Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Stromatolites as biosignatures and paleoenvironmental records: experiments with modern mats and examples from the Eocene Green River Formation
(USC Thesis Other)
Stromatolites as biosignatures and paleoenvironmental records: experiments with modern mats and examples from the Eocene Green River Formation
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Stromatolites as biosignatures and paleoenvironmental records: experiments with modern mats and examples from the Eocene Green River Formation by Carie M. Frantz Approved August 16, 2013 by: Frank A. Corsetti, Committee Chair Department of Earth Science University of Southern California Kenneth H. Nealson Department of Earth Science University of Southern California Moh El-Naggar, Outside Member Department of Physics University of Southern California Copyright 2013 Carie M. Frantz Stromatolites as biosignatures and paleoenvironmental records: experiments with modern mats and examples from the Eocene Green River Formation by Carie M. Frantz _____________________________________________ A dissertation presented to the Faculty of the USC Graduate School University of Southern California In partial fulfillment of the requirements for the degree Doctor of Philosophy in Earth Sciences December 2013 i Table of Contents List of figures vi List of tables xi Acknowledgements xiii Introduction 1 Figures 4 References 5 Chapter 1: Grain trapping and binding by filamentous photosynthetic cyanobacterial (Coleofasciculus chtonoplastes) and algal (Chaetomorpha) mats 6 Abstract 6 Introduction 8 Stromatolites 8 The stromatolite grain size conundrum 9 Grain trapping and binding by microbial communities to understand grain size distributions in stromatolites 11 Methods 13 Model organisms 13 Experimental setup 15 Scanning Electron Microscopy (SEM) 20 Results 21 Cyanobacteria 21 Algae 25 Discussion 28 Microbial mats trap brains beyond the abiotic angle of repose: implications for the interpretation of stromatolite biogenicity 28 Trapping and binding by filamentous cyanobacterial mats 29 Trapping and binding by Chaetomorpha filamentous algal mats 31 ii General observation: Size matters! 32 Implications for the interpretation of modern and ancient stromatolites 33 Conclusions 35 Acknowledgements 36 Figures 38 Tables 70 References 79 Chapter 2: Dramatic local environmental change during the Early Eocene Climatic Optimum detected using high resolution chemical analyses of Green River Formation Stromatolites 85 Abstract 85 Introduction 87 The Green River Formation 87 Paleoclimate during the Early Eocene Climatic Optimum 89 Stromatolites as environmental indicators 91 Geological setting 92 Methods 94 Stromatolite collection and processing 94 ICP-MS elemental analyses 96 Standard stable isotopes 99 Clumped isotopes 100 Deep-ultraviolet (deep-UV) native fluorescence spectroscopy 103 Results 104 Petrography and mineralogy 104 ICP-MS elemental results 106 Stable isotopes 107 Clumped isotopes 108 Deep-UV native fluorescence spectroscopy 108 Discussion 109 Evaluating biogenicity in the different microfabrics 109 iii Evidence for minimal diagenesis 110 Stromatolite chemistry and consistency across layers 111 Chemistry, temperature, and microfabric linked to lake volume changes 112 Evaporation models 115 Micrite and fans formed in different environments: significance for stromatolites as near-shore deposits 127 Shifting shorelines and climate 128 Stromatolite growth rate constraints 129 Lake water temperatures: evidence for thermal stratification? 130 Hypothesis: lake level impacts biology, biology (or lack thereof) determines microfabric 133 Conclusions 135 Acknowledgements 136 Figures 138 Tables 189 References 202 Chapter 3: Transient periods of basin closure associated with changes in the microfabric of stromatolites from the Lower LaClede Bed of the Laney Member of the Eocene Green River Formation 212 Abstract 212 Introduction 213 Geological setting 213 Paleoclimate 214 Depositional enviornments 214 Methods 215 Sample collection and processing 215 Chemical analyses 217 iv Results 218 Petrography and mineralogy 218 ICP-MS elemental results 219 Stable isotopes 220 Discussion 222 Stromatolites formed in a saline lake 222 Basin hydrology: open, balanced, or closed? 222 Application of lake volume models during closed intervals 224 Lamination counts and growth rate estimates 229 Conclusions 230 Acknowledgements 231 Figures 232 Tables 252 References 259 Appendix A: Supplemental stromatolite images 262 Stromatolite collection 263 Preparation of stromatolite slabs and billets 264 Thin section photomosaics 274 Appendix B: Supplemental ICP-MS results 281 Full table of results 282 Sources of contamination 290 Dilution checks 295 Appendix C: Supplemental standard stable isotope results 297 Boar’s Tusk stromatolites 298 LaClede stromatolites 301 v Appendix D: Detailed clumped isotope methods and results 304 Methods 304 Results 307 References 308 Appendix E: Matlab code 309 Code descriptions 309 Code 311 vi List of Figures Introduction Figure 1. Stromatolites with similar shape but different origins 6 Chapter 1 Figure 1. Distribution of grain size in stromatolites and thrombolites over time 38 Figure 2. Photomicrograph of a thin section of a LaClede stromatolite 39 Figure 3. Catalina Harbor mudflat cyanobacterial mat sampling site 41 Figure 4. Mature Catalina Harbor cyanobacterial mats 43 Figure 5. Transmitted light microscope images from filament samples from the Catalina Harbor cyanobacterial mats 44 Figure 6. Collection of the Chaetomorpha algal mats 45 Figure 7. Transmitted light photomicrograph mosaic of a filament from the Chaetomorpha algal mats 46 Figure 8. Experimental setup for trapping and binding experiments 47 Figure 9. Setup for flow experiments 49 Figure 10. Fine, medium, and coarse grains trapped as mass fraction of grains delivered by cyanobacterial mats of varying maturity compared with a glass slide control 50 Figure 11. Trapping and binding of grains by developed cyanobacterial mats 52 Figure 12. Growth of cyanobacterial filaments onto medium sized grains over time 54 Figure 13. Flow experiments conducted using Catalina Harbor cyanobacterial mats showing the loss of grains under flow conditions 55 Figure 14. Fraction of grains remaining on cyanobacterial mats during different phases of the flow experiments 57 Figure 15. Results of the experiment testing binding of medium grains by a cyanobacterial mat under flow conditions 58 Figure 16. Scanning electron microscopy images of grains trapped by Catalina Harbor microbial mats 60 Figure 17. A cyanobacterial filament bundle binding a medium-sized grain after 2 days of mat incubation under flow conditions 61 vii Figure 18. Trapping of fine, medium, and coarse grains by Chaetomorpha mats of different densities 62 Figure 19. Trapping of grains of various sizes after two days of mat incubation by Chaetomorpha mats 64 Figure 20. Retention of fine and medium grains over various incubation periods by Chaetomorpha mats 65 Figure 21. SEM images of algal mats 67 Figure 22. Cartoon of hypothesized grain size distribution in laminae of stromatolites formed under different conditions by cyanobacterial vs. Chaetomorpha-like algal mats 69 Chapter 2 Figure 1. Cenozoic CO 2 and temperature estimates 138 Figure 2. The Rife Bed stromatolites in outcrop 140 Figure 3. Map of stromatolite sampling locations 141 Figure 4. Map and stratigraphic representation of the Boar’s Tusk outcrop 142 Figure 5. The Boar’s Tusk outcrop showing the exposed Tipton Shale Member, the bounding stromatolite horizon in the Rife Bed, and the Wilkins Peak Member 143 Figure 6. Microdrilled locations of the BT08 stromatolite 144 Figure 7. Closeup of the BT08 stromatolite after re-drilling for elemental analysis 146 Figure 8. Map of microdrilled locations of the BT12-CF-1 and BT12-CF-2 stromatolites 148 Figure 9. Map of microdrilled locations of the BT12-CF-4 stromatolite 150 Figure 10. Drill holes for the BTM2 (micrite) and BTF2 (fan) clumped isotope runs. 152 Figure 11. Thin section image of a segment of a Rife Bed stromatolite showing the digitate columns and the three dominant microfabrics: calcite fans, micrite, and mixed 153 Figure 12. Photomicrograph of a Boar’s Tusk stromatolite following staining with Alizarin Red S 154 Figure 13. Photomicrographs of a calcite fan layer showing sharp crystal termination boundaries, and of a micrite layer showing hanging grains 155 Figure 14. Cathodoluminescence of a section of the BT08 stromatolite 156 Figure 15. Molar fraction of magnesium in carbonate (XMg) plotted by fabric for the Boar’s Tusk stromatolites 157 viii Figure 16. Correlation of stromatolite laminae across a 200 m section of outcrop 158 Figure 17. Correlation of laminations in the BT12-CF-1 and BT12-CF-2 stromatolites 160 Figure 18. Stromatolites drilled for chemical analyses arranged in stratigraphic order 161 Figure 19. Chemical results for the BT08 stromatolite samples plotted by layer 163 Figure 20. Chemical results for the BT12-CF-1 and BT12-CF-2 stromatolite samples plotted by layer 165 Figure 21. Chemical results for the BT12-CF-4 stromatolite samples plotted by layer 167 Figure 22. Correlation of oxygen and carbon stable isotopes for the Boar’s Tusk stromatolites 169 Figure 23. Comparison of clumped isotope results with results from standard stable isotope measurements from the BT08 stromatolite 170 Figure 24. Results of deep-UV native fluorescence spectroscopy 171 Figure 25. Crossplots of all elements measured vs. į 18 O for all stromatolites plotted by fabric type 172 Figure 26. Crossplot of Crossplot of ɷ 18 O and sodium measurements from the Boar’s Tusk stromatolites 174 Figure 27. Covariant carbon and oxygen stable isotopes in carbonates from modern closed lakes compared with the Rife Bed stromatolites 175 Figure 28. Water temperature change required to produce the carbonate į 18 O shift observed from BT12-CF-4 calcite fan layers 10-12 176 Figure 29. Oxygen isotope model sensitivity curves showing maximum lake depth vs. freshwater input value for all temperature scenarios for the Boar’s Tusk stromatolites 177 Figure 30. Isotope model of lake volume and depth change by layer for the Boar’s Tusk stromatolites showing the model’s sensitivity to freshwater input and temperature scenarios assumed 179 Figure 31. Comparison of the ion, conservative isotope, and intermediate isotope model results for the Boar’s Tusk stromatolites 181 Figure 32. Lake depth calculations sorted by microfabric for the Boar’s Tusk stromatolites 182 Figure 33. Historic lake level changes in the Great Salt Lake, Utah 183 Figure 34. Schematic of Great Salt Lake shoreline shifts showing lake levels plotted on an elevation profile of the Great Salt Lake Basin 184 ix Figure 35. Comparison of the ion, conservative isotope, and intermediate isotope model results by layer shown with their corresponding microfabrics and clumped isotope temperatures 185 Figure 36. Light penetration and water temperature at two different sites in the Great Salt Lake 187 Figure 37. Cartoon illustrating formation hypotheses for the different stromatolite microfabrics 188 Chapter 3 Figure 1. Map showing the location of the outcrop from which samples were collected 232 Figure 2. Typical lacustrine depositional cycles in the Laney Member 233 Figure 3. Stratigraphy of the LaClede site 234 Figure 4. Map of microdrilled locations of the LC12-3 stromatolite 235 Figure 5. Map of microdrilled locations of the LC12-E stromatolite 237 Figure 6. Map of microdrilled locations of the LC12-E2 stromatolite 238 Figure 7. Chemical results for the LC12-3 stromatolite samples plotted by layer 239 Figure 8. Chemical results for the LC12-E stromatolite samples plotted by layer 240 Figure 9. Comparison of chemical data measured from all Green River Formation stromatolites analyzed in this dissertation 241 Figure 10. Fraction of magnesium in carbonate (XMg) in the LaClede stromatolites plotted by layer 242 Figure 11. Map of microdrilled locations for the LC12-E and LC12-E2 showing layer correlations (LC12-E combined ) 243 Figure 12. Oxygen isotope values measured for the LaClede stromatolites 245 Figure 13. Crossplots of oxygen and carbon stable isotopes from the LaClede stromatolites 246 Figure 14. Oxygen isotope model sensitivity curves showing maximum lake depth vs. freshwater input value for all temperature scenarios for periods of closure in the LaClede stromatolites 248 Figure 15. Lake volume model results for periods of closure in the LaClede stromatolites displaying closed 249 Figure 16. Plots of sodium concentration, ɷ 18 O, and calculated lake depth shown with the five different microfabric regions of the LC12-E stromatolite 251 x Appendix A Figure A1. Boar’s Tusk stromatolite beds and sample locations 262 Figure A2. BT12-CF-1 stromatolite sectioning 264 Figure A3. BT12-CF-2 stromatolite sectioning 266 Figure A4. BT12-CF-4 stromatolite sectioning 268 Figure A5. LC12-3 stromatolite sectioning 270 Figure A6. LC12-E stromatolite sectioning 272 Figure A7. Photomicrograph mosaic of a thin section from BT08 273 Figure A8. Transmitted light photomicrograph mosaics of thin sections BT12-CF- 1Ba-c from the BT12-CF-1 stromatolite 274 Figure A9. Cross-polarized light photomicrograph mosaics of thin sections BT12-CF- 1Ba-c from the BT12-CF-1 stromatolite 276 Figure A10. Transmitted light photomicrograph mosaic of thin section BT12-CF-1Bd from the BT12-CF-1 stromatolite 278 Figure A11. Cross-polarized light photomicrograph mosaic of thin section BT12-CF- 1Bd from the BT12-CF-1 stromatolite 279 Figure A12. Photomicrograph mosaics of thin sections from the LC12-3 stromatolite 280 Figure A13. Photomicrograph mosaics of thin sections from the LC12-E stromatolite 282 Appendix B Figure B1. Microdrill bit contamination test results 293 Figure B2. Concentration of select elements measured using different drill bits as a test of contamination by drill bits 294 Figure B3. Water soluble vs. carbonate-bound fraction of key elements in the BT08 stromatolite 295 Figure B4. Fraction of element present in the water-soluble fraction 296 Figure B5. Dilution check of elements measured via ICP-MS 298 xi List of Tables Chapter 1 70 Table 1. Grain sizes used in this study 70 Table 2. Grains trapped by cyanobacterial mats of varying maturity compared with a glass slide control 71 Table 3. Grains trapped, bound on contact, and bound by well-developed by cyanobacterial mats after 12 hours of incubation 72 Table 4. Results of flow experiments with cyanobacterial mats 73 Table 5. Grains trapped immediately upon contact with Chaetomorpha algal mats 74 Table 6. Grains trapped by dense Chaetomorpha algal mats following two days of incubation 75 Table 7. Grains trapped by thin Chaetomorpha algal mats after different periods of incubation 76 Table 8. Grains trapped vs. bound by thin Chaetomorpha algal mats 77 Table 9. Medium-sized grains trapped vs. bound by dense Chaetomorpha mats under flow conditions 78 Chapter 2 Table 1. Natural isotopic abundances and calculated mass correction factors for the isotopes measured in this study 189 Table 2. Microfabric mineralogy determined using X-ray powder diffraction compared with calculated XMg values from elemental results 190 Table 3. Elemental measurements for the Boar’s Tusk stromatolites 191 Table 4. Standard stable isotope measurements from the Boar’s Tusk stromatolites 193 Table 5. Clumped isotope results from the BT08 stromatolite fabrics 195 Table 6. Comparison of clumped isotope results with results from standard stable isotope measurements from the BT08 stromatolite 196 Table 7. Chemical measurement results for all stromatolites by fabric 197 Table 8. Summary of assumptions used in the lake volume models 198 Table 9. Temperature scenarios tested in the oxygen isotope volume model 199 Table 10. Lake volume model results for all stromatolites 200 xii Chapter 3 Table 1. Descriptions of the five different microfabric regions in the LC12-E and LC12-E2 stromatolites 252 Table 2. Elemental measurements for the LaClede stromatolites 253 Table 3. Stable isotope measurements for the LaClede stromatolites 254 Table 4. Assessment of basin closure using plots of į 13 C vs. į 18 O for different microfabric regions of the LaClede stromatolites 256 Table 5. Lake volume model results for LaClede stromatolite sequences during periods of closure 257 Appendix B Table B. ICP-MS results for all Green River Formation stromatolites measured 285 Appendix C Table C1. All standard stable isotope measurements from the Boar’s Tusk stromatolites 301 Table C2. All standard stable isotope measurements from the LaClede stromatolites 304 Appendix D Table D. Full table of clumped isotope results measured from the BT08 stromatolite 310 xiii Acknowledgements I attribute my survival of these past six years, and eventual success in the form of this finished document, to the incredible support network that I was fortunate to have. Working with and learning from so many amazing, brilliant, interesting, and good people during my time as a Ph.D. student was one of the greatest privileges of my life. In particular, I had a team of mentors who “raised” me as a scientist in the same way that “it takes a village to raise a child”. First and foremost, the two ‘academic parents’ who produced this Geobiologist: Ken Nealson, with his irresistible enthusiasm, made the decision to come to USC an easy one. From there, he gave me complete freedom to pursue my interests and crazy ideas, encouraged me to attend the summer courses (the Agouron Geobiology course and Microbial Diversity at MBL) that molded my path and changed my life, and his boundless positivity was often a source of energy when I most needed it. I have Ken to thank for infecting me with his deep love of the microbial world, and I owe the ‘bio’ half of my new title of Geobiologist to him. Frank Corsetti, my “Geo parent” and the Earth Science Department’s guardian angel of grad students, was my lighthouse from Day 1. Despite my stubbornness and free-bird tendencies, Frank managed to keep me mostly in line, no easy task, and also managed to turn me from a ‘chemist/microbiologist’ into a ‘stromatolitoligist’ in two short years. He should probably be awarded an honorary doctorate in psychology, or at least the title of Jedi Master for his skill in that regard. Frank was my advisor, but also a coach, teacher, mentor, role model, cheerleader, Yoda, and constant friend, and I am deeply grateful for his support and guidance, both scientific xiv and in life. Also, the fact that this thesis mentions the word ‘stromatolite’ more than 500 times is entirely his fault. Some other key members of the science village include: Jörg Overmann, who adopted me while I was in Germany and taught me the ways of phototrophs; Gijs Kuenen, whose scientific feedback and friendship I am blessed to have benefitted from; Will Berelson who continues to call me on my sh*t better than anyone I know; my Geobio course family (Ann, Amber, John, Russell, Alex, Hope, Vicky, and others); my MBL course family (Dan, Steve, Victoria, Roland, and others); and my USC family (too many to mention, but especially my academic siblings and cousins in the Nealson, Corsetti, and Berelson labs). I would not have survived the past six years without my friends, an incredible group of A-class humans. They let me sob on their shoulders when my heart was broken and when hope had left me. They dragged me out for a stiff drink when nothing was working; brought me flowers, food, and bandages when I was recovering from surgery; brought me movies when I had pneumonia; took me hiking when I was crabby; cheered with me in my little victories; and fed, housed, and even clothed me in various times of need. Thank you for bringing so many smiles and so much humor to my life! And finally, a big thank you to my family: my dad who taught me to love science, my mom who gave me fire, my sister (BFF), my grandma the great supporter of advanced education, and my aunts and uncles who have tutored, housed, fed, and encouraged me my whole life. 1 Introduction Stromatolites, defined descriptively by Semikhatov et al. (1979) as “attached, laminated, lifhified, sedimentary growth structures, accretionary away from a point or limited surface of initiation”, are considered important structures for understanding the evolution of life on earth and potentially elsewhere. From their first detailed description by Kalkowsky (1908, who ascribed the stromatolites and oolites of the Harz Mountains to the activities of “lower plants”), stromatolites have been interpreted as biogenic structures—that is, evidence of life—usually formed by the trapping and binding of grains by cyanobacteria. The assumption of a biogenic origin of stromatolites and link between stromatolites and the past presence of cyanobacteria is ubiquitous: it is found, for example, in both geology and biology textbooks (e.g., Brock Biology of Microorganisms, 13 th edition, 2012), encyclopedias (e.g., the Encyclopedia Britannica and Wikipedia, accessed August 2013), science museum displays, and point of interest information stands worldwide. However, it is well-established that stromatolites can form abiogenically (e.g., Grotzinger and Rothman, 1996; Grotzinger and Knoll, 1999; Riding, 2008; McLoughlin et al., 2008), and trapping and binding by cyanobacteria is only one of several potential biological mechanisms for their formation. The discrepancy between reality and the generally held assumptions about stromatolites stems largely from an overreliance on modern marine “living stromatolite” systems such as Shark Bay, Australia and Highborne Cay in the Bahamas as “analog systems”, despite dramatic textural differences between modern marine and ancient stromatolites (e.g., Figure 1; Riding, 2011). Furthermore, the presence of complex communities in modern stromatolites— 2 including organisms that only recently evolved—suggests that ancient stromatolites formed via different mechanism(s) than modern marine stromatolites. In reality, the mechanisms—both biotic and abiotic—by which different types of stromatolites form are not very well understood. The problems interpreting the meaning of stromatolites are not limited to establishing their biogenicity (or lack thereof). The full range of environments under which stromatolites can form is also not well understood. As an example, stromatolites are widely assumed to be a shallow facies (e.g., Grotzinger and Knoll, 1999), despite evidence for deep-water stromatolites (e.g., Hoffman, 1974), another assumption perhaps skewed by an overemphasis on modern marine “analog” systems in the study of stromatolites. This dissertation refines and redefines the current understanding of stromatolites in several ways: In the first chapter, experiments with modern cyanobacterial and algal mats demonstrate the link between the physical properties of mat-forming organisms and the size of grains that the mats are able to trap and bind, as well as the angle at which grains are trapped. Results of the mat experiments suggest that the evolution of stromatolite grain size distributions can be linked to the evolution of filamentous forms of bacteria and, later, to the evolution of photosynthetic eukaryotes and establishment of mixed photosynthetic communities. The second and third chapters of this thesis demonstrate the utility of stromatolites as high-resolution records of environmental change. Stromatolites from the Green River Formation capture the dynamics of environmental change during the Eocene Climatic Optimum when global temperatures and CO 2 levels were the highest in the Cenozoic. The stromatolites record dramatic fluctuations of lake chemistry, volume, extent, and even hydrology, with the lake 3 expanding and contracting several times within the record of the 10-30 cm thickness of the stromatolites. These fluctuations imply a degree of climate and hydrologic variability on the scale of years to decades that was previously thought to occur only on orbital timescales (tens of thousands to hundreds of thousands of years). An additional key finding of this work is that the stromatolites we investigated did not grow in a single environment (e.g., shallow photic zone waters), but instead grew in a range of depths (possibly experiencing as much as >10 meters of lake depth change), water temperatures, and distances from shore. The implication of this finding is that stromatolites are not necessarily a shallow near-shore facies marker, or a marker of one specific environment of growth. While they can form in shallow, near-shore environments, the stromatolites of this study seem to have formed, at times, tens of kilometers from shore in much deeper waters. This thesis highlights the need for continued experimental study to understand the specific mechanisms responsible for forming the wide range of stromatolite fabrics observed in the fossil record. It also demonstrates the potential usefulness of stromatolites as fine-scale records of the environment in which they formed, as well as of the types of biological communities that contributed (or did not contribute) to their growth. 4 FIGURES Figure 1. Stromatolites with similar shape but different origins (Figure 7 from Grotzinger and Knoll, 1999). A Modern stromatolites from Shark Bay,Western Australia. Scale bar is 40 cm. B Cross-section of a Shark Bay stromatolite, which was formed by microbial trapping and binding. Knife is 7.5 cm long. C Neoarchean stromatolites from the Campbellrand Subgroup, South Africa. Hammer is 35 cm long. D Cross section of the Campbellrand stromatolites, which consist of abiogenically precipitated crystal fans. Scale bar is 20 cm. 5 REFERENCES Grotzinger, J.P., and Knoll, A.H., 1999, Stromatolites in Precambrian carbonates: evolutionary mileposts or environmental dipsticks?: Annual review of earth and planetary sciences, v. 27, no. 1, p. 313–58, doi: 10.1146/annurev.earth.27.1.313. Grotzinger, J.P., and Rothman, D.H., 1996, An abiotic model for stromatolite morphogenesis: Nature, v. 383, no. 6599, p. 423–425, doi: 10.1038/383423a0. Hoffman, P., 1974, Shallow and deepwater stromatolites in lower Proterozoic platform-to-basin facies change, Great Slave Lake, Canada: American Associated Petroleum Geologists Bulletin, v. 58, no. 5, p. 856–867. Kalkowsky, E., 1908, Oolith und Stromatolith im norddeutschen Buntsandstein: Zeitschrift der Deutschen Geologischen Gesellschaft, v. 60, p. 68–125. Madigan, M.T., Martinko, J.M., Stahl, D.A., and Clark, D.P., 2012, Brock Biology of Microorganisms: Pearson, San Francisco. McLoughlin, N., Wilson, L.A., and Brasier, M.D., 2008, Growth of synthetic stromatolites and wrinkle structures in the absence of microbes - implications for the early fossil record: Geobiology, v. 6, no. 2, p. 95–105, doi: 10.1111/j.1472-4669.2007.00141.x. Riding, R., 2008, Abiogenic, microbial and hybrid authigenic carbonate crusts: components of Precambrian stromatolites: Geologia Croatica, v. 61, no. 2-3, p. 73–103, doi: 10.4154/GC.2008.10. Riding, R., 2011, The nature of stromatolites: 3,500 million years of history and a century of research, in Reitner, J., Quéric, N.-V., and Arp, G. eds., Advances in Stromatolite Geobiology, Springer Berlin Heidelberg, Berlin, Heidelberg, p. 29–74. Semikhatov, M.A., Gebelein, C.D., Cloud, P.E., Awramik, S.M., and Benmore, W.C., 1979, Stromatolite morphogenesis—progress and problems: Canadian Journal of Earth Sciences, v. 16, no. 5, p. 992–1015, doi: 10.1139/e79-088. 6 Chapter 1: Grain trapping and binding by filamentous photosynthetic cyanobacterial (Coleofasciculus chtonoplastes) and algal (Chaetomorpha) mats ABSTRACT In general, Archean and Proterozoic stromatolites are fine-grained whereas most modern marine examples are comparatively coarse-grained. Given that the modern marine forms are commonly studied as analogues to the ancient forms, it is important to understand the processes responsible for the textural differences. Cyanobacteria are typically considered the dominant stromatolite builders through time, but it is well known that many modern marine stromatolites also contain a eukaryotic component (various algae, diatoms, etc.). Thus, we conducted experiments to test the grain trapping and binding capabilities of filamentous cyanobacterial mats (dominated by the trichome-forming Coleofasciculus chtonoplastes) versus algal mats (Chaetomorpha) in order to better understand the grain-size trends in stromatolites through time. The mats were cut into coupons, inclined at angles from 0-75° in saltwater tanks to approximate the angle of lamina observed in typical stromatolites, and grains of various sizes (fine sand, coarse sand, and fine pebbles) were delivered to their surface. We measured both trapping and binding as a function of mat properties such as filament length and mesh density. Experiments were done under very low flow and moderate flow conditions. The cyanobacterial mats were able to trap fine grains consistently at all angles. At angles beyond the angle of repose, medium and coarse grains were not trapped as efficiently, but some 7 (although very few) coarse grains were trapped even at high angles depending on the maturity of the mat and filament bundle length. Dense algal mats trapped medium and coarse grains efficiently at all angles, but were poor at trapping fine grains. The cyanobacteria bound the grains over time, regardless of angle, over time by physically wrapping them. The algae, in contrast, were unable to bind grains and tended to shed trapped grains over time. When flow was added to the experiment, trapping was significantly reduced in the cyanobacterial experiments, but not the algal experiments. Our experiments suggest that the presence of grains beyond the abiotic angle of repose can be considered a biosignature in ancient stromatolites where biogenicity is in question. Although we cannot conclude that all fine-grained stromatolites were formed by cyanobacteria, our results suggest that stromatolites where coarse grains are present at high angles at much the same frequency as at low angles (e.g., most modern marine stromatolites) may require a filamentous eukaryotic component in order to efficiently trap coarse grains beyond the angle of repose, and give insight into the evolution of stromatolite microfabrics through time. 8 INTRODUCTION Stromatolites Stromatolites are classically defined as laminated organosedimentary structures formed by microorganisms (Kalkowsky, 1908; Awramik, 1971; Walter, 1976; Riding, 2011). As such, stromatolites have long been the subject of research, speculation, and debate because of their interpretation as some of the oldest fossils and evidence for the presence of oxygenic phototrophs (especially cyanobacteria) prior to the Great Oxidation Event (ca. 2.4 Ga). However, stromatolites for which a microbial origin has been conclusively demonstrated are more the exception than the rule (e.g., Grotzinger and Knoll, 1999). In order to avoid the need to demonstrate a biological origin before identifying a structure as a stromatolite, an alternate descriptive definition was proposed by Semikhatov et al. (1979): “an attached, laminated, lithified sedimentary growth structure, accretionary away from a point or limited surface of initiation”. Stromatolites as defined using Semikhatov’s descriptive definition form by any of the following mechanisms (or combinations thereof): 1) Abiotic mineral precipitation and/or sedimentation (e.g., Grotzinger and Rothman, 1996; Pope et al., 2000; McLoughlin et al., 2008). 2) Microbial micrite precipitation by metabolic activities that favor carbonate precipitation or by the calcification of cyanobacterial sheaths (e.g., Pentecost and Riding, 1986; Altermann et al., 2006). 3) Trapping and binding of sediments by sticky or filamentous microbial mats (e.g., Gebelein, 1969). 9 The different growth mechanisms are thought to result in different microfabrics (e.g., Riding, 2008; 2011). Abiogenic mineral precipitation generally forms laterally continuous, isopachous, sparry crusts. Microbial micrite precipitation or trapping of micrite results in fine- grained, often uneven, wavy and/or peloidal textures. Finally, trapping and binding of coarser allochthonous grains by microbial mats produces more crudely-layered, often thrombolitic, agglutinated stromatolites (Pope et al., 2000; Riding, 2008; Riding, 2011). The stromatolite grain size conundrum Brief history of stromatolite microstructure The microfabric and morphology of stromatolites has not been constant over the past 3.5 billion years of Earth’s history (Walter et al., 1980; Riding, 2000; Riding, 2011; Figure 1), setting up a grain size conundrum for stromatolite research: modern marine forms are coarse grained, whereas nearly all ancient examples are fine grained. The earliest stromatolite are dominated by precipitated/sparry microstructures; the oldest stromatolites (those >3.2 Ga) display generally isopachous layering and lack universally convincing microfossils, an observation that has led to considerable debate regarding their biogenicity (see Lowe 1994 and subsequent comment/reply correspondence, Van Kranendonk et al., 2003; Schopf et al., 2007; Allwood et al., 2009). By the Proterozoic, stromatolite microfabrics consisted of sparry, fine-grained, or hybrid textures (a mix of sparry and fine-grained crusts as defined by Riding). Most Proterozoic stromatolites are finely-laminated and can occur as elaborate macro-scale structures diverse and distinctive enough to have been assigned Linnaean-style form names (Krylov, 1976; Walter et al., 1992; Awramik and Sprinkle, 1999; Riding, 2011). Despite the presence of coarse grains in 10 the environments in which they formed, Proterozoic stromatolites are predominantly fine- grained, highlighting one aspect of the “grain size conundrum”. A good example of coarse grain rejection is found in the finely laminated stromatolites of the LaClede Bed, Green River Formation, discussed elsewhere in this dissertation (Figure 2). The late Neoproterozoic/early Phanerozoic is marked by the appearance of a new form of microbialites—thrombolites (like stromatolites, but with a clotted instead of laminated meso- structure)—which has led to speculation that thrombolites are the bioturbated version of stromatolites (Aitken, 1967; Walter and Heys, 1985; Riding, 2000). However, although evidence of burrowing is occasionally present in thrombolites, the fabrics appear to originate from the calcification of photosynthetic microbial communities (Kennard and James, 1986); thus, they may be better explained by the increased presence of algal or mixed photosynthetic communities (including calcifying microorganisms) in high-energy environments versus stromatolite formation by cyanobacterial trapping and binding in lower-energy environments (Aitken, 1967; Feldmann and McKenzie, 1997). True coarse-grained stromatolites appear quite late in the history of stromatolites, with the earliest documented course-grained stromatolites occurring in the Miocene (Feldmann and McKenzie, 1997). Modern marine stromatolites: terrible analogues for the ancient True finely-laminated stromatolites, texturally similar to the vast majority of ancient stromatolites, are very uncommon in modern systems. So-called “modern analogue” marine stromatolites, such as those found in Shark Bay and the Bahamas (Logan, 1961; Dravis, 1983; Dill et al., 1986; Awramik, 1988; Feldmann and McKenzie, 1998; Macintyre et al., 2000; Reid et al., 2000), are coarse-grained, poorly laminated if laminated at all, and typically thrombolitic. 11 Therefore, modern marine stromatolites represent a class of stromatolite that did not appear until very late in the Phanerozoic. The modern microbial mat systems that are responsible for building the modern marine coarse-grained and thrombolitic stromatolites include diatoms, other algae, and a host of other organisms in addition to cyanobacteria; many of these organisms did not evolve until relatively late in the Earth’s history (Fischer, 1965; Monty, 1973; Falkowski et al., 2004; Bengtson et al., 2009), which may partially explain the relatively late appearance of coarse-grained and thrombolitic stromatolite forms. Thus, the modern marine stromatolites most commonly used as analogues for ancient stromatolites are hardly analogues at all. While they are interesting geobiologic systems in their own right, we need to turn to other analogues (e.g., finely-laminated stromatolites found in alkaline lakes and hot springs; Osborne et al., 1982; Jones et al., 1997; Berelson et al., 2011; Petryshyn et al., 2012; Pepe-Ranney et al., 2012; Mata et al., 2012) and laboratory experiments to better understand the formation of ancient stromatolites. Grain trapping and binding by microbial communities to understand grain size distributions in stromatolites In stromatolite research, trapping is defined as the localization of sediment by baffling or adhesion of irregular (vs. smooth or flat) mat surfaces, whereas binding is defined as grain incorporation into the mat by the overgrowth of microbes that may or may not be the same as the organisms responsible for trapping the grains (Gebelein, 1969; Riding, 2000). Biogenic stromatolites are built by some combination of sediment trapping and binding and mineralization, and the microbial community responsible will leave its mark in the microstructure in some way. Research on modern marine stromatolites has shown that the ability 12 of microbial communities to trap and bind sediment depends on physical characteristics of the microbes such as size and orientation as well as “stickiness” (influenced by EPS production and perhaps other physical or chemical properties; Gebelein, 1969; Golubic, 1976; Monty, 1976). The eventual morphology of stromatolites formed through trapping and binding and/or calcification depends on many factors both biological (the species present and their physical properties such as size, arrangement, EPS production, degree of nutrient limitation, presence of grazers and bioturbators, etc.) and environmental (size of delivered grains, flow, wave energy, carbonate saturation, nutrient availability, etc.; Grotzinger and Knoll, 1999; Dupraz et al., 2006). The size of particles trapped depends on the grain size of supplied sediment as well as properties of the mat (Ginsburg, 1957; Gebelein, 1969). For example, Gebelein 1969 did a thorough study of grain trapping by stromatolites around Bermuda and found that the cyanobacterial mats there (a mix of the sheathed filamentous bundles, an Oscillatoria, and an unidentified biofilm-forming coccoid cyanobacterium) predominantly trapped grains <500µm in diameter with the fraction of grains <80µm in diameter five times higher in the mats than in adjacent sediments, suggesting a preferential trapping of fine grains. Recalling the conundrum that most ancient stromatolites are fine grained despite the presence of coarser grains in their environment, what exactly can microbial mats trap and bind under various conditions? Experiments investigating the relationship between specific properties of microbial mats and their grain-trapping abilities in controlled environments have not been reported. In their seminal paper on the state of Precambrian stromatolite research, Grotzinger and Knoll (1999) called for more experimental research to elucidate the links between organisms and stromatolite morphology. 13 In that spirit, this study investigated the grain trapping and binding ability of two distinctly different types of filamentous mat: a cyanobacterial type and a much larger green algal type. We studied the impact of filament size and length on the trapping of grains of different sizes at different incline angles (including angles well beyond the abiotic angle of repose, the “angle with the horizontal at which loose granular material will stand when piled or dumped” (Carrigy 1970, p.148)) with and without flow conditions. We found that the size of grains trapped by filaments is directly related to the length of the filament, suggesting a strong impact of the evolution of filamentous forms (and larger eukaryotic filament forms) on stromatolite grain size. By understanding what certain types of microbial mats can trap, we can better understand the meaning of the microstructure in ancient stromatolites. METHODS Model organisms Cyanobacterial mats We collected mats containing cyanobacteria and other oxygenic phototrophs from a low- energy subtidal mudflat on the western edge of Catalina Harbor on Catalina Island, California (33.430°N, 118.505°W) (Figure 3). The mats were dark green from the filamentous cyanobacteria on the surface. The mat field was pitted with the burrows of small crabs spaced at an average distance of ~30cm, but appeared otherwise free of the effects of bioturbation or grazing. After resting undisturbed for several days in seawater tanks in the lab, the mats “fluffed out” with cyanobacterial filament bundles protruding as much as 1 cm from the sediment and forming a green lawn of filaments (mats used in experiments had filament lengths of 1-10 mm) 14 (Figure 4). Upon exposure to light, bubbles formed on, and were released by, the filaments, suggesting that the mats were actively photosynthesizing in the seawater tank and during the experiments. The protruding filamentous portion of the cyanobacterial mats was dominated by filamentous bundles of Coleofasciculus chtonoplastes (previously Microcoleus chtonoplastes; Siegesmund et al., 2008), a cosmopolitan mat-building cyanobacterium commonly found in marine environments (Stal, 2012). Coleofasciculus trichome bundles (referred to broadly as “filaments” in the remainder of this text; for example “filament length” refers to the length of the trichome bundle protruding from the mat surface) are surrounded by polysaccharide sheaths that can become calcified. In addition, Coleofasciculus mats have been implicated in the trapping of sediments in marine mats (e.g., Gunatilaka 1975). Consequently, these organisms have been noted elsewhere as stromatolite-builders (Golubic, 1976; Semikhatov et al., 1979), although they were not mineralized in this environment. While other morphotypes of cyanobacteria and diatoms were present in the mats, they made up only a minor fraction of the biomass and were not responsible for the formation of the macroscopic filaments that protruded out of the mat surface when experiments were conducted (Figure 5). Algae We collected filamentous green algae visually identified as Chaetomorpha linum (Abbott and George Jacob Iollenberg, 1993; Guiry, 2013) from the seawater table drainage outflow of the USC Wrigley Institute for Environmental Science at Big Fisherman Cove, Catalina Island, CA (Figure 6). The algae formed large (~0.5 m 2 ) floating rafts of interwoven filaments (referred to as algal mats in this study) in the drainage stream. Individual uniseriate algal filaments were 100 15 µm in diameter and up to 10 cm long (Figure 7) with bundles of filaments growing much longer. The authors of this paper are aware of at least one locality where dense mats dominated by algae of the genus Chaetomorpha were found to trap small amounts of sediments (Gunatilaka, 1975), although the Chaetomorpha described in that study differed in morphology and mat character from the ones used in this study and may have been a different species. Experimental setup Overview The experiments involved inclining coupons of photosynthetic mats (the Coleofasciculus cyanobacterial type and the Chaetomorpha green algal type) at various angles in a water tank, carefully delivering a known amount of grains of various defined sizes to them, and collecting samples at several time points to determine the amount of material trapped and to assess the active binding of grains. The angles used here mimic the angles observed on a typical Proterozoic marine stromatolite, ranging from horizontal at the tip of the column or dome to nearly vertical on the sides. Lithification is very important for stromatolite formation, but the focus of this experiment was to test the trapping and binding abilities of certain filamentous microorganisms, especially their ability to trap and bind grains beyond the angle of repose. Thus, lithification was not part of this experiment. EPS produced by a mat may also play a role in trapping and/or binding grains, although the evidence available is somewhat anecdotal, at best. The model organisms we chose lacked copious EPS, thus reducing the number of variables significantly. Future experiments could involve mats producing abundant EPS. 16 Tanks The tanks used were large (1 x 0.6 meter x 0.3 meter deep) acrylic tanks with a drain and water inlet to permit flushing of water with filtered seawater from the Fishermans Cove Marine Reserve on Catalina Island, California (Figure 8A). Except during experiments testing the impact of flow, flow rates in the tank were very low (~25·10 -6 m 3 /s) and inlet hoses were positioned away from mats to minimize turbulence near the mats while supplying the mats with an influx of water-soluble nutrients and facilitate gas exchange in the tanks. Algal tanks were additionally bubbled with diffuse air using aquarium airstones in order to prevent excess oxygen buildup and to supply the algae with dissolved CO 2 for photosynthesis. Tanks received natural daylight from windows in the lab as well as light from overhead fluorescent bulbs that we turned off at night. Algal mats received supplemental light up to ~10% of PAR at noon from small fluorescent lights placed above the tanks that we also turned off at night to simulate a normal diurnal cycle. Quantifying grain trapping Procedure. In order to quantify the amount of grains trapped by photosynthetic mats, mats were cut into coupons of known area and distributed into petri dishes for the experiments. The cyanobacterial mats were not buoyant and rested in the bottom of the petri dishes, and appeared to photosynthesize during the experiment. Due to the buoyant nature of the algal filaments, algal mat coupons were bound in their petri dishes using a narrow band of lab tape and the petri dishes were weighted to keep them secured at set angles in the seawater tanks (Figure 8C). Even after the addition of nutrients, supplemental light, and diffused air, the algae did not appear to actively photosynthesize or grow. While it is possible that these algae were not alive or 17 died during the experiments, they appeared to be actively growing in the drainage area from which they were collected. A predetermined amount (1 g when using large 7x7 cm coupons, 0.5 g when using smaller 5x4 coupons) of grains of a defined size were then carefully delivered to mat coupons placed in pre-weighed Petri dishes. Grains that were not trapped by the mats would roll off or fall through the mats and collect in the bottom of the petri dish. At the end of the experiments, dishes were dried, mats removed, and dishes with the collected un-trapped grains were weighed. Grain binding was tested by inverting and very gently shaking submerged mats to release unbound grains and collecting and weighing any grains that were dislodged before or during mat inversion; grains bound to the mats were determined by subtracting the measured mass from the mass of grains delivered. Sedimentation controls. In order to assess the mass of external sediment (sediment not sourced from delivered grains) that accumulated in the petri dishes under various conditions, control dishes with mats that had not received delivered grains were weighed. The mass of external sediment was negligible except in the case of fresh/immature cyanobacterial mats. Where possible, the mass of this external sediment was estimated from the control dishes and used the estimates to correct measured masses of un-trapped grains for grain trapping experiments. In some cases (many of the cyanobacterial experiments except where explicitly stated) the mass of external sediment was variable, but relatively small; in these cases, we did not attempt corrections and the values of trapped grains calculated therefore represent underestimates. Where trapping and/or binding results are presented as a percentage of the mass 18 of grains delivered, values less than 0% and greater than 100% result from uncertainties in the mass of external sediment. Grain sizes. The grain size types “fine”, “medium”, and “coarse” used in this study are defined in Table 1. Flow experiments A somewhat different setup was used to investigate the impact of flow on grain trapping and binding than for the quantitative grain trapping experiments described above. Cyanobacterial mats. First, a weighed amount of medium sized grains were delivered evenly across the surface of large ~24 x 10 x 2 cm filamentous cyanobacterial mat slabs. Mat slabs were then draped over 11 cm Pyrex beakers placed in a tray, effectively inclining different portions of the mats at increasingly steep angles, mimicking the typical domed shape of many ancient stromatolites. Any grains that rolled off when the mats were inclined were collected from the tray and weighed. Next, a stream of water of measured flow rate was positioned in the tank, either directly onto the mats (Figure 9A) or in gentler cross-flow conditions (Figure 9B) that increased circulation in the tank while approximating laminar flow, which is thought to be required for stromatolite formation (Feldmann and McKenzie, 1997). Flow conditions used in this study were realistic and within the range of flow rates measured, for example, by Gebelein (1969) for environments in which stromatolites form in the Bahamas. Grains that were dislodged under flow were again collected and weighed. Mats that remained intact were left under these conditions for two days, with fallen grains collected and weighed each day. Mats were then inverted to release any unbound grains, which were again collected and weighed. From these masses of lost grains in addition to photographs of the mat taken at 19 each step, the fraction of grains remaining on the mat at different phases during the experiment and at different angle ranges could be determined. Due to a temporary problem with seawater pressure in the lab (and a subsequently flooded lab), flow was turned off for several of the experiments for a period of 12 hours at which point the tanks drained, inducing a subaerial exposure condition. Mats that “survived” this period without breaking apart were again submerged and the experiments continued under the previously-used flow conditions. Four successful flow experiments are described in detail here: Mat A. A poorly-developed (filament length <1mm) 1-day old mat subjected to low (0.067 L/s) direct water flow across the top of the mat. After 30 hours under flow conditions, flow was shut off and the tank dried out, but the mat remained intact. The tank was refilled after 12 hours of subaerial exposure and flow was turned back on for 4 hours before ending the experiment. Mat B. A developed 3-day old mat (filament length 1 – 5 mm) subjected to low (0.141 L/s) direct water flow across the top of the mat. After ~10 minutes under flow conditions, the mat broke apart and the grain collection experiment ended. However the mat was kept in the tank under flow conditions for 2 days, at which point it was dissected and analyzed for trapping vs. binding of grains at different incline angles (Figure 15A-B). Mat C. A developed 2-day old mat (filament length 2 – 3 mm) subjected to low-volume cross-flow conditions (hoses not pointing directly at the mat but inducing water circulation in the tank) with two hoses producing flow rates of 0.089 and 0.074 L/s. After 10 hours under flow conditions, flow shut off and the tank dried out but the mat remained intact. The tank was refilled 20 after 12 hours of subaerial exposure and flow was resumed and the experiment continued overnight. Mat D. A developed 2-day old mat (filament length <1 – 4mm, filaments were significantly less developed in a large patch on one side than in the rest of the mat) subjected to high-volume cross-flow conditions with two hoses producing flow rates of 0.40 and 0.42 L/s for 10 hours. Algal mats. Two large ~30x20 cm, ~5 cm thick Chaetomorpha mats were collected, placed them on trays, inclined at an angle of 30° in seawater tanks, and submerged in water. Given the setup, tanks were not deep enough to attempt higher incline angles. The mats were then subjected to direct water flow at a moderate rate of 0.2 L/s (Figure 9C-D). ~5 g of medium- sized grains were delivered to the mats under flow conditions. Mats were left under flow conditions for 2 days, at which point flow was terminated and trays with the mats gently laid flat. One mat was removed and the grains that the mat had not trapped remained on the tray and were weighed to determine the mass of trapped grains. The second mat was inverted over the tray to collect grains that were not bound in the mat; released grains were weighed to determine the mass of bound grains. Scanning Electron Microscopy (SEM) A tabletop environmental scanning electron microscope (Hitachi TM-1000) was used to analyze samples collected at regular intervals during the trapping and binding experiments. Samples were air- or oven-dried prior to analysis. 21 RESULTS Cyanobacteria Trapping of grains Trapping experiments involved delivering grains to cyanobacterial mats inclined at different angles, measuring the mass of grains that rolled off, and subtracting this measured mass from the mass of grains delivered to determine the grains trapped. The results of grain trapping experiments for cyanobacterial mats of different degrees of maturity/filament lengths are compiled in Table 2 and shown in Figure 10. Glass slide controls (absent of any microbial material) retained fine grains up to incline angles of 30-45° and medium and coarse grains only up to 15°. Freshly-collected mats with no visible filament protrusion retained (trapped) fine and medium grains at greater incline angles (as high as 60° for fine grains and 30° for medium grains), while coarse grains were only trapped at angles up to 15°, the same as glass slide controls. More mature mats with filament lengths up to 2 mm trapped medium and coarse grains at higher incline angles than the undeveloped mats or the glass slide control; at least 40% of the grains (by mass) that were delivered to them were trapped at angles as high as 75° for fine grains, 45° for medium grains, and 30° for coarse grains. No significant difference in trapping ability was observed between one day old mats with filament lengths <1 mm and three day old mats with filament lengths from <1-2 mm. However, mature mats with filament lengths of ~1 cm that we collected and tested five months earlier (February 2013 vs. July 2013) for the trapping and binding experiment described in the section that follows trapped a greater fraction of fine and medium grains (coarse grains were not tested) at high angles, trapping ~all of fine grains at 22 angles of 60° (greater angles were not tested) and >50% of medium grains at angles as high as 75° (Table 3). Binding of grains Developed mats (filament length = ~1 cm) bound some grains by the definition of this study (i.e., the grains remained attached to the mat when the mat was inverted) within seconds of the grains contacting the mats (Table 3). The mats bound a significant fraction (30-70%) of trapped fine grains immediately, with mats inclined at lower angles appearing to bind a greater fraction of fine grains than mats inclined at higher angles (Figure 11A). In contrast, medium grains were not immediately bound by the mats (Figure 11B). We did not test coarse grains, assuming that their even greater mass and lower surface area to mass ratios would cause them to also not be bound on contact. Incubating the mats in the tanks overnight (12 hours) following delivery of grains greatly increased the fraction of bound grains, with grains bound by the mats approximately equal to the fraction of grains trapped by similarly-developed mats (Table 3, Figure 11; measurements for grains trapped and bound on contact presented in the table were done in February 2013 vs. July 2012 for grains bound after 12 hours). The mats bound coarse grains at high angles at significantly lower rates than medium and fine grains (Figure 11C). Mat filaments “grew” (probably more accurately, glided) out of the mats and around the grains, physically wrapping them within days following grain delivery (Figure 12). In the overnight binding experiment, grains were visibly greenish with fresh cyanobacterial growth/wrapping when mats were inverted to test for binding. 23 Flow experiments Four experiments were conducted covering a range of mat types, flow speeds, and flow directions: (A) a poorly-developed 1-day old mat under low direct water flow across the top of the mat, (B) a developed 3-day old mat under low direct flow, (C) a developed 2-day old mat under low cross flow, and (D) a developed 2-day old mat under high direct flow. Results are summarized in Table 4 and illustrated in Figure 14 and Figure 15. The poorly-developed mat A lost nearly 50% of the grains delivered to it immediately upon draping, suggesting that many grains were not trapped, especially at high angles. Of those grains remaining, most were dislodged once direct flow was applied. Grains continued to shed over time while exposed to direct flow until very few grains remained attached. However, of the grains that remained attached after two days, most remained bound when the mat was inverted (Figure 14A). In contrast, the well-developed mats B-D lost few grains when the mats were draped (Figure 14B-D), and nearly all of the grains lost during the draping of these more mature mats were lost from poorly-developed patches of these mats. Most initially trapped grains remained attached to the mats under all flow conditions tested, however a greater portion of grains were lost under direct flow than under cross flow, and greater flow rates caused the loss of a greater fraction of grains. Mat B broke apart after exposure to flow conditions and the measurement of lost grains was halted, but the mat was used to monitor the binding of grains over time (Figure 15). Mat C was observed over extended exposure to flow conditions; as with mat A nearly all of the grains that remained attached to mat C after two days under flow conditions were physically 24 bound such that inverting the mat resulted in a release of only a few unbound grains (Figure 14C). The cyanobacterial mats exposed to flow conditions visibly bound grains (Figure 15C). The majority of grains that remained bound to the mats when the mats were inverted after two days of incubation under moderate flow conditions could still be released under strong direct flow conditions (Figure 15D-E). Grains trapped at very high incline angles were more likely to be retained by mats when exposed to high flow than grains trapped at lower incline angles (Figure 15E). Scanning electron microscopy SEM images of the cyanobacterial mats show cyanobacterial filament bundles that appear to wrap and physically bind delivered grains. However, images of mats immediately following grain delivery (t0) do not look significantly different than images of mats following two days of incubation following grain delivery (t2) after the filaments had a chance to grow around the grains (Figure 16). This suggests that, at least in some cases, the filaments lying on top of grains in SEM images are an artifact of removing the mats from the tank and drying them for analysis rather than an accurate representation of how the filaments bind the grains. However, the images do show how individual filaments or a net of filaments growing around and/or onto grains could physically bind them to the mat (e.g., Figure 16, which is an image of a grain that was visibly bound by filaments prior to preparation for SEM imaging). Indeed, long-term visible monitoring of mats revealed that filaments do grow around grains, binding them over time (Figure 12). SEM in addition to normal light microscopy also showed the bundled nature of the filaments and the dominance of these bundled filaments in the mat structure (Figure 5, Figure 16). 25 Algae Trapping of grains The algal mats used in different tests had visible differences in the tightness of filament weave of which the mats were constructed. These differences in mat “density” (with density measured as g/cm 2 , the mass of the mat divided by the surface area over which grains were delivered) were a major factor in the trapping ability of the algal mats (Table 5, Figure 18A). For medium and coarse sized grains, the ability of the mats to trap grains scaled with mat density (Figure 18B). For fine grains, however, mat trapping ability did not appear to scale with mat density, with mats trapping ~15-35% of grains delivered regardless of density (Figure 18B). Chaetomorpha green algal mats trap grains regardless of incline angle (Table 6 and Table 7, Figure 19). While dense mats trapped medium and coarse grains at rates >70% and thin mats trapped coarse grains at rates from 0-35%, fine grains tended to slip through the weave of the mats resulting in a significantly lower trapped fraction at several angles. It should be noted that “trapped” fine grains were likely overestimated due to drying effects: fine grains tended to stick together upon drying and fine grains that were not bound to the filament in water became bound to the filaments upon drying (e.g., Figure 21A-B). Binding of grains In contrast to the cyanobacterial mats, which bound trapped grains over time, no increase in grain retention over time was observed; indeed, in some cases the algal mats shed grains over time (Table 7, Figure 20). Mats collected 14 days following grain deposition had retained fewer than half of the medium grains compared with mats collected immediately following grain deposition at most incline angles. Trapping of fine grains remained low over time except for the 26 sample measured after 14 days which had values much higher than those measured previously at two angles, probably again a result of drying effects as mats would not be expected to acquire grains over time. The algal mats did not bind grains of any size to a significant extent. Unlike with the cyanobacterial mats, we did not observe an increase in binding by mats left to incubate with delivered, trapped grains for two days (Table 8, Figure 20C). Flow experiments Results of the algal mat flow experiments are summarized in Table 9. The thick, dense mats tested trapped the majority of medium grains delivered and approximately half of the delivered grains remained adhered to the mat when it was inverted (“bound” by the definition of this study, although not in the sense of cyanobacterial wrapping of grains over time). Scanning electron microscopy Despite measurable “trapping” of fine grains by the algal mats, we saw very few of fine grains associated with algal filaments in SEM. One such example is Figure 21A-B, where a fine grain is suspended between two algal filaments and several other fine grains appear to be adhered without actually touching any filaments. Because the spacing between algal filaments was often greater than the diameter of fine grains, it is easy to see how fine grains would not be easily trapped by these mats. Those fine grains remaining attached could be due to the drying artifacts discussed in previous sections (and see an example of “trapped” grains not actually in contact with a filament in Figure 21A-B). Medium-sized grains, however, were large enough to be trapped by the algal mesh, as in Figure 21C, M, and P. Interestingly, a very small grain was 27 observed attached to an algal filament in a sample where no grains had been added (Figure 21D), suggesting that these algae may trap grains in their environment. The wrinkled nature of the filaments under SEM appears to be an artifact of the drying and/or vacuum process as undisturbed filaments observed under light microscopy are cylindrical and unwrinkled (Figure 6). Some filaments (a few can be seen in Figure 21J) did not wrinkle, but the majority did. This makes it difficult to determine actual environmental relationships between grains and filaments; for example the trapped grain in Figure 21D appears to be caught between two wrinkles, where these wrinkles almost certainly did not exist prior to the drying/vacuum process. We observed a layer of diatoms (~20 !m) and other unidentified substances encrusting filaments that we used for the thin algal mat trapping time series (done in February 2013; Figure 21E-H). This was not the case for filaments used for the other experiments (done in July 2012). We did not observe diatoms in SEM scans of dried water samples from the seawater tanks from February 2013 suggesting that they were genuinely associated with the filaments and not a product of drying diatom-rich seawater onto the filaments. There was no significant increase in encrustation observed over two weeks of observation (Figure 21I-L). Grains added to the mats appeared in some cases to develop films (possibly bacterial biofilms) on them over time. (Figure 21M-O). This was not the case for all grains (Figure 21P) and the makeup of the films is undetermined. Brown-colored debris accumulated in the seawater tanks hosting the algal mats over time. Large quantities of sponge spicules were present in the debris and were also observed in samples 28 that had been incubating in the tanks. Crusts and crusty pellets of unknown origin were also common in the debris, presumably sourced from the seawater used to fill the tanks. DISCUSSION Microbial mats trap grains beyond the abiotic angle of repose: implications for the interpretation of stromatolite biogenicity Establishing the biogenicity of stromatolites is critical to their interpretation, but is often thwarted by lack of conclusive evidence (Buick et al., 1981; Lowe, 1994; Grotzinger and Knoll, 1999; Corsetti and Storrie-Lombardi, 2003; Van Kranendonk et al., 2003; Allwood et al., 2009). Unambiguous microfossils are rarely found in ancient stromatolites, and even when they are, their role in the construction of stromatolites cannot be assumed. Other indicators of biogenicity include isotopes, trace elements, and microfabrics, but all of these can be, and frankly usually are, altered by diagenesis. The presence of grains in stromatolite laminae at angles greater than the abiotic angle of repose (i.e., on the steep sides of stromatolite domes) could be used as an additional indication of biogenicity. The angle of repose is defined in geology as “the angle (with the horizontal) at which loose granular material will stand when piled or dumped” (Carrigy 1970, p.148). This angle differs for different sediment types, but is typically 45° or less in abiotic systems Carrigy, 1970; Glover, 1995. Glass slide controls were used in this study to approximate the abiotic angle of repose, which was found to be between 30-45° for fine grains and 15-30° for medium and coarse grains. The presence of living organisms, whether in the form of a sticky biofilm on the sediment bed, filaments that prevent the rolling of grains by baffling, or the actual overgrowth of grains by 29 biofilms or filaments can significantly increase the angle of repose (Meadows et al., 1994). This concept was used by Petryshyn et al. (manuscript in prep) as the basis for a novel biogenicity test utilizing magnetic susceptibility to quantify the amount of sediment trapped at high angles in stromatolite laminae. The results presented above conclusively demonstrate that microbial mats can trap—and in some cases bind—grains beyond the abiotic angle of repose. In addition, the differences found in this study in the size of grain and angle at which different types of mats are able to trap and bind grains suggests that properties of the stromatolite-building microbial communities may be inferred from the distribution of grains present in stromatolite laminae. Trapping and binding by filamentous cyanobacterial mats Cyanobacterial mats trap fine grains best The trapping of grains by the filamentous cyanobacterial mats in this study was dependent on (1) mat incline angle, (2) grain size, and (3) mat maturity (filament length). The developed cyanobacterial mats of this study trapped even some coarse grains beyond the abiotic angle of repose on contact, a feature that differentiates the biotic system from the abiotic glass slide control. Thus, the presence of coarse grains beyond the angle of repose can be cautiously considered a biosignature, as noted above. While the mats trapped nearly all grains delivered at low angles, they trapped significantly fewer grains at greater incline angles, not unlike ancient predominantly fine-grained stromatolites. The fraction of grains the mats trapped at high angles was greatest for fine grains and smallest for coarse grains. This preferential trapping of fine grains by cyanobacterial mats is also observed in modern marine microbialite- forming mats Gebelein, 1969. 30 In contrast to developed mats, fresh, undeveloped mats appear to behave more like abiotic surfaces. Undeveloped mats held significantly more fine grains at high angles than the glass slide controls, but did not hold medium grains angles >30°, and did not hold coarse grains at all beyond what was held by the glass slide control. When flow was added to the experiment—a condition that likely approaches real world conditions somewhat better than the static experiment—the mats trapped significantly less grains versus the static experiments. Flow experiment A lost nearly 50% of the medium grains delivered to it immediately upon draping (primarily from the steep sides of the mat). Of those grains remaining, most fell off once direct flow was applied. Grains continued to fall off over time while exposed to direct flow until very few grains remained attached. This suggests that any trapping of grains beyond the angle of repose or under strong flow conditions or coarse grains at high angles requires the presence of filaments of sufficient length/maturity. Cyanobacterial mats bind grains over time In all cases observed, the majority of grains trapped by the cyanobacterial mats became bound over time (compare Figure 11A-B values of grains trapped to values of grains bound following overnight incubation). However, flow conditions tended to release grains that were bound by the definition used in this study, i.e., grains that remain attached to the mats when the mats were inverted; the majority of grains that remained bound when the mats were inverted after two days of tank incubation could still be released under strong direct flow conditions. For medium and coarse grains, binding appears to take time, with the percentage of medium sized grains that are bound immediately upon contact with the cyanobacterial mats at roughly zero versus ~40-80% after overnight incubation (Figure 10B). However, approximately 31 40-70% of fine grains that are trapped by the cyanobacterial mats are also immediately bound (Figure 11A). This could be due to electrostatic charge holding the low-mass fine grains—which have a much higher surface area, and therefore surface charge, to mass ratio than medium grains—to the surface of the cyanobacterial filaments. Trapping and binding by Chaetomorpha filamentous algal mats We observed the following key trends in the algal mats: (1) trapping ability was independent of incline angle for all grain sizes, (2) mats trapped larger grains more readily than smaller grains at most angles, and (3) the fraction of grains retained did not increase with time (i.e., no binding occurred). The Chaetomorpha mats are essentially a net of twisted filaments with voids between filaments that grains—especially finer grains—can slip through. This is in stark contrast to the cyanobacterial mats which form a relatively uniform vertical lawn of filaments with a solid base. This explains the lack of a strong influence of incline angle on the trapping of grains by these mats. Smaller grains slipped through these voids more readily than larger grains. In an additional contrast to cyanobacterial mats, which tended to physically bind grains over time by growing around them, algal mats collected after 21 days of incubation following grain deposition had retained fewer than half of the medium grains as were retained by mats collected immediately following grain deposition at most incline angles. This could be due to slight perturbations in the water shifting grains such that they fell through the mats over time with more time resulting in more grains released in this way, or to physical changes to the algal filaments (e.g., degradation) that in some way influences the stickiness of the filaments. This trend was much more pronounced for medium sized grains than for fine grains. This implies that 32 the binding of grains does not occur in Chaetomorpha, making these organisms unlikely stromatolite builders in the absence of other “binding” organisms. However, as a model organism in our experiments, they demonstrate the abilities of larger filamentous eukaryotic algae to trap larger grains versus their smaller, cyanobacterial cousins. This observation of a lack of binding was additionally supported by the observation that when Chaetomorpha mats with trapped grains were inverted, they lost nearly all of their grains, even after 2 days of incubation in the tanks (Figure 20C). This is in contrast to the cyanobacterial mats, which bound nearly all of their trapped grains after an overnight incubation in the tanks. Medium grains were “bound” by the definition of this study in thick, dense mats during the flow experiment (Table 9). However, in the Chaetomorpha mats “bound” grains are simply caught in the thick mesh of filaments vs. actively bound by the filaments as in the cyanobacterial mats. These results do, however, suggest that at least in very thick, dense Chaetomorpha mats, grains of sufficient size become caught and held in the mats even under flow conditions. General observation: Size matters! For both algal and cyanobacterial mats, the size of grains trapped appears to be correlated to filament length and spacing. The cyanobacterial mats consist of closely-spaced linear bundles of filaments ~30 µm in diameter that protrude up to 1cm upward from the solid mat surface. Algal mat filaments are significantly larger with a diameter of ~100 µm and a length of several cm (with bundles of filaments growing much longer) that in dense mats pack closely but still have voids between bundles up to 1 cm in diameter. These geometrical properties alone impact the size of grains that can be trapped. 33 Very large grains are not easily held at high angles by the fine cyanobacterial filaments, although longer filaments hold larger grains more readily. This finding echoes a statement by Riding (2000, p.183) that trapping is “facilitated where mats have irregular surface topography, such as that formed by relatively large microbes with abundant erect filaments…In contrast, smooth mats or films with little surface topography trap only very fine grains or none at all.” The eukaryotic filamentous green algal mats (Chaetomorpha) that were tested were more effective at trapping larger grains, trapping grains as large as 2 mm in diameter at angles up to 75° at a rate greater than 80%. However, these mats were less effective at trapping finer grains, particularly in thin mats, because fine grains slipped through the larger mesh size formed by the algal mats. Implications for the interpretation modern and ancient stromatolites Our results are perhaps intuitive and to be expected: smaller organisms can only trap smaller particles, and larger organisms can trap larger particles. However, our results are significant when compared to stromatolite microfabrics through time. In this study, fine grains ( <200 !m) were readily trapped and bound well beyond the abiotic angle of repose by even immature filamentous cyanobacterial mats and not by the Chaetomorpha-type algal mats. Medium grains were also trapped and bound beyond the angle of repose by developed cyanobacterial mats, but retention of the grains under strong flow conditions was low. The algal mats, meanwhile, had a far greater ability to trap coarse grains, and trapped them independent of incline angle, but in this case were not adept at binding particles of any size. Given these results, it is tempting to conclude that fine-grained stromatolites are formed by filamentous cyanobacteria while coarse-grained stromatolites require the presence of larger 34 filamentous algae or other eukaryotic component. While a better understanding of the mechanisms responsible for trapping grains of various sizes is required, this study lends support to the hypothesis that the sudden appearance of coarse-grained stromatolites and thrombolites in the Phanerozoic relates to the evolution of eukaryotic filamentous phototrophs (or at least of larger and potentially mesh-forming filamentous microbial forms). However, our results also suggest some larger algae, while they may trap larger grains, may not be adept at binding them. Hence, if coarse grains are present in abundance in a stromatolite, the stromatolite building community likely also contained active binder organisms (e.g., cyanobacteria), precisely the type of community seen in modern marine stromatolites. Thus, our results lend insight into the composition of the microbiota that built stromatolites through time and the type of properties the microbial communities need to construct various stromatolites, coarse or fine. In addition, assuming that grains are lithified in the same proportions as they are bound, the results of this study suggest the following hypotheses with respect to stromatolite grain size distribution: 1) Stromatolites that formed in still water by developed filamentous cyanobacterial mats would be expected to consist of a relatively even blanket of fine grains with larger grains concentrated at lower incline angles, but still possibly present at high angles (e.g., Figure 22D). Under heavy flow conditions grain distribution would be less angle-dependent. 2) Stromatolites formed by undeveloped filamentous cyanobacterial mats would be expected to look much like abiotic surfaces, perhaps inducing chemical precipitation but holding few grains larger than fine grains beyond the angle of repose (e.g., Figure 22C). Under high enough flow conditions few grains would be expected at any angles (e.g., Figure 22A). 35 3) Because Chaetomorpha mats did not appear to actually bind grains in the traditional sense of microbial overgrowth and at least in thin mats grain loss over time was observed, these mats may not be good stromatolite-building candidates in the absence of a community including binding microorganisms (e.g., cyanobacteria). However, dense mats did retain medium grains that became trapped in the mesh of filaments and held these grains under direct flow conditions as well as when the mat was inverted. It is therefore conceivable that these mats could accumulate grains that could become lithified. In this case, Chaetomorpha-like mats consisting of coarse filamentous meshes vs. the fine vertical filaments of the cyanobacterial mats would be expected to be dominated by larger grains, with the dominant grain size dependent on void space size. Little difference in grain distribution would be expected at low vs. high incline angles (Figure 22E-H). CONCLUSIONS This study demonstrates the ability of filamentous cyanobacterial mats to trap and bind grains even at high angles and highlights several key differences between filamentous cyanobacterial mats and filamentous algal rafts in the trapping and binding of grains. The cyanobacterial mats readily trapped and bound fine—and in some cases larger—grains well beyond the abiotic angle of repose. The algal mats, in contrast, trapped only large grains well. Specific relationships between filament length, incline angle, water flow rate, and the size of trapped and bound grains additionally imply that grain size distributions in stromatolites could be suggestive of particular mat communities and environments. 36 This study was limited in the organisms used and the conditions under which grains were delivered and mats were incubated. The temptation to generalize these results to all cyanobacterial mats or all algae or all filamentous organisms should be resisted. Additional studies with non-filamentous organisms, organisms producing copious extracellular polysaccharides (EPS), and other algal forms (including diatomaceous mats) are needed in order to compare the trapping and binding of grains of various sizes by other mat-building organisms. In addition, this study ignored the impact of mat calcification and stromatolite lithification, which is critical to the formation of both modern and ancient stromatolites. However, this study does strongly suggest that the differences in grain sizes in Phanerozoic stromatolites compared with Proterozoic stromatolites could be explained by differences in the organisms that formed them. The sudden appearance of coarse-grained stromatolites in the Phanerozoic may well be related to the evolution of filamentous forms that could trap larger grains than filamentous cyanobacterial mats are capable of. ACKNOWLEDGEMENTS A portion of the experimental work was carried out by the 2012 International Geobiology Course (in particular the following students of the course: Amanda Achberger, Flavia Boidi, Caitlin Cox, Verónica Durán, and Daniel Mills), and students and faculty from the 2011 and 2013 courses are acknowledged for their insights. The 2012 course was supported by the USC Wrigley Institute for Environmental Studies and the Colorado School of Mines as well as grants from the Agouron Institute, the Gordon and Betty Moore Foundation, the NASA Astrobiology Institute, and the National Science Foundation. 37 The authors are grateful to the wonderful staff of the USC Wrigley Institute for Environmental Studies on Catalina Island, especially Ann Close, Amber Brown, Lauren Czarnecki, and Kellie Spafford for facilitating the experimental work presented in this chapter. In addition, SEM work was done in the lab of Karla Heidelberg who provided helpful tips and insight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igure 2. A) Photomicrograph of a thin section of a LaClede stromatolite (discussed in detail in a separate chapter of this thesis). B) Close-up image of the highlighted region of A. In both images, arrows point to coarse grains that were deposited in intercolumn spaces (demonstrating the presence of coarse grains in the formation environment) but that are absent in actual stromatolite laminae. ! "# ! ! $%&'()!*! !"## !$$"%# ! " ! 41 Figure 3. Catalina Harbor mudflat cyanobacterial mat sampling site. A Satellite view of Catalina Harbor showing the sampling location (blue marker). Inset shows location (blue marker) mapped on a map of Catalina Island. The green marker in the inset denotes the location of the USC Wrigley Institute for Environmental Science where the algal mats were collected. Maps modified from Google Maps. B Photograph of the sampling location in Catalina Harbor. ! "# ! ! $%&'()!*! !"# $%%# ! "#$%&'(#') *&+&,-.&$ %&'(#' *&+&,-.&$ /),&.0 &'()*'#+,-./),01'(-0. 1 2 43 Figure 4. Mature Catalina Harbor cyanobacterial mats "fluffed out" after several days in a seawater tank. A-B show mats collected and used for experiments during July 2012 as part of the 2012 International Geobiology Course. C shows a mat collected and used for experiments in February 2013 by the author of this thesis. 1cm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µm 47 Figure 8. Experimental setup for trapping and binding experiments. A Seawater tanks lit by overhead fluorescent lights with mat coupons in petri plates inclined at various angles. B Undeveloped cyanobacterial mat coupon in a petri dish inclined at 45°. C Developed cyanobacterial mat coupons in petri dishes inclined at angles between 45-60°. D Green algal mat coupons in petri plates inclined at angles between 15-30°. Algal coupon dishes were weighted with airstones to keep them from floating and algae were restrained to their plates using a thin belt of lab tape across the center of the dish. ! ! "# ! ! ! $%&'()!#! ! " # $ ! ! "# ! ! $%&'()! #*! +),'-! ./(! .0/1! )2-)(%3)4,5*! !! 6784/98:,)(%80 ! 38,! .0/1! )2-)(%3)4, ! 5),! '-! ./(! ;%():,! .0/1*! "! 6784/98:,)(%80 ! 38,! .0/1! )2-)(%3)4, ! 5),! '-! ./(! :(/55! .0/1*! #! </-! =%)1! /.! 80&80 ! 38,! .0/1! )2-)(%3)4,*! $!+%;)!=%)1!/.!80&80!38,!.0/1!)2-)(%3)4,* ! ! !"#$ !"#$ !"#$ !"#$ !"#$ %"&'( )*'+#,'-.(/0' ! " # $ 50 Figure 10. A Fine, B medium, and C coarse grains trapped as mass fraction of grains delivered by cyanobacterial mats of varying maturity compared with a glass slide control. Key: open gray markers = glass slides (glass); orange markers = undeveloped fresh mat, filament length = 0 mm (f0); light green markers = mat after 1 day of growth, filament length <1 mm (f1); dark green markers = mat after 3 days of growth, filament length <1-2 mm (f2); blue green markers = mature mat, filament length ~1 cm (f10), not measured for coarse grains. Error bars represent standard deviations from duplicate measurements. ! ! "# ! ! ! $%&'()!#*! !"#$%&'("#))*+ ,- , ,- . ,- / ,- 0 ,- 1 2- , , 2, ., 3, /, 4, 0, 5, 1, !"#$%&'("#))*+ ,- , ,- . ,- / ,- 0 ,- 1 2- , , 2, ., 3, /, 4, 0, 5, 1, 6%7 8 $ %*' #%98 *' : +*9" **& ; !"#$%&'("#))*+ ,- , ,- . ,- / ,- 0 ,- 1 2- , , 2, ., 3, /, 4, 0, 5, 1, A Fine grains B Medium grains 6%7 8 $ %*' #%98 *' : +*9" **& ; 6%7 8 $ %*' #%98 *' : +*9" **& ; C Coarse grains 98 #& & <, <2 <. <2, 52 Figure 11. Trapping and binding of grains by developed (filament length ~1 cm) cyanobacterial mats. A Fine grains trapped, bound on contact, and bound after 12 hours. B Medium grains trapped, bound on contact, and bound after 12 hours. C Fine, medium, and coarse grains bound after 12 hours. Key: circles = fine grains; triangles = medium grains; squares = coarse grains; dotted line, open symbols = grains trapped; dashed line, filled symbols = grains bound on contact (0 hrs after grain delivery); solid line, filled symbols = grains bound after 12 hours following grain delivery. Error bars represent standard deviations from duplicate measurements. ! ! "# ! ! ! $%&'()!**! A Fine grains B Medium grains !" ! !" # !" $ !" % !" & '" ! ! '! #! (! $! )! %! *! &! + , -. /0 1 , 23 -. /. /4 5/6 7 . /21 -/47 21 8 924, 220 : !" ! !" # !" $ !" % !" & '" ! ! '! #! (! $! )! %! *! &! + , -. /0 1 , 23 -. /. /4 5/6 7 . /21 -/47 21 8 924, 220 : C Binding after 12 hours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igure 13. Flow experiments conducted using Catalina Harbor cyanobacterial mats showing the loss of grains under flow conditions. White arrows indicate the outlet of water hoses used to induce flow conditions. Images show a series beginning with the naked mat (mat), the delivery of grains to the mat (+grains), draping of the mat onto a beaker (draped), onset of flow (flow t0), after 1 and 2 days of incubation under flow conditions (flow t1 and flow t2, respectively), the mat again laid flat (flat t2), and the mat after being tilted to >90° to release unbound grains (inverted). A Mat A experiment using a poorly-developed mat under low direct water flow. B Mat B experiment using a developed mat under low direct water flow. After t0 was collected the mat broke apart. The broken mat was left in the tank for an additional 2 days and samples were collected to assess the binding of grains under flow conditions (see Figure 14). C Mat C experiment using a developed mat under low-volume cross-flow conditions. D Mat D experiment using a developed mat under high-volume cross-flow conditions. After t0 was collected the mat broke apart and the experiment was ended. ! "# ! ! $%&'()!*+! !"#$ $%& '()%*+, -)%./- 0123"&4 0123"&5 0123"&6 *+7/)&/- !".22)189-/7/12./-"$%&:"123"-*)/#&"0123 $%& '()%*+, -)%./- 0123"&4 ;)2</+"&5 #"-/7/12./-"$%&:"=*(="#)2,,"0123 +2".=2&2 01%&"&6 $%& '()%*+, -)%./- 0123"&4 0123"&5 0123"&6 *+7/)&/- 01%&"&6 $"-/7/12./-"$%&:"123"#)2,,"0123 %"-/7/12./-"$%&:"123"-*)/#&"0123 $%& '()%*+, -)%./- 0123"&4 ;)2</+"&5 ! ! "# ! ! $%&'()! *+,! $(-./%01! 02! &(-%13! ()4-%1%1& ! 01! .5-106-./)(%-7 ! 4-/3! 8'(%1&! 8%22)() 1/! 9:-3)3! 02! /:)! 270;! )<9)(%4)1/3, ! =)//)(3 ! .0(()39018! /0! /:)! )<9)(%4)1/3 ! 3:0;1! %1! /:)! 7)//)()8 ! 3'62%&'()3! %1! $%&'()! *>,! ?@1%/%-7A ! -()! /:)! &(-%13! %1%/%-775 ! 8)7%B)()8 ! /0! /:)! 4-/C! ?8(-9)8A! -()! &(-%13! ()4-%1%1&! -2/)(! 8(-9%1&! /:)! 4-/! 01/0! /:)! 6)-D)(C! ?270;! /EA! -()! &(-%13! ()4-%1%1& ! -2/)(! /'(1%1&! 01! 270;! %1! /:)! /-1D3C! ?270;! /*A! -18! ?270;! />A! -()! &(-%13! ()4-%1%1& ! -2/)(! /:)! 4-/! ;-3! )<903)8! /0! 270;! 20(! *! -18! >! 8-53! ()39)./%B)75C ! -18! ?%1B)(/)8A ! %3! /:)! 4-33! 02! &(-%13! /:-/! ()4-%1)8 ! -//-.:)8 ! F60'18G! /0! /:)! 4-/! ;:)1! %/! ;- 3! %1B)(/)8 ! -2/)(! >! 8-53! 02! %1.'6-/%01 ! %1! /:)! /-1D3, ! H()-D3! %18%.-/) ! *>! :0'(! 9)(%083! 02! 3'6-)(%-7!)<903'()!;:)1!270;!;-3!/'(1)8!022!-18!/-1D3!8(%)8!0'/,! ! !"! !"# !"$ !"% !"& !"' !"( !") !"* !"+ #"! ,-,.,/0 12/341 50678.! 50678.# 50678.$ ,-942.41 :2/,-;824</,-,-= A undeveloped mat, low direct ow B developed mat, low direct ow C developed mat, low cross ow D developed mat, high cross ow subaerial exposure period 58 Figure 15. Results of the experiment testing binding of medium grains by a cyanobacterial mat under flow conditions using a three-day old developed mat that broke during initial flow tests (Flow Experiment B) but was left under low direct flow conditions for two days. A The draped, broken mat showing the relative positions and angles of numbered slices used for analyses. B After two days under flow conditions the mat from A was gently laid flat and cut into 5x7 slices for microscopy (M), binding tests by mat inversion (B1 and B2), and binding tests under close direct flow (F1 and F2). Vertical slices 1-7 are representative of different incline angles as shown in A. C Close-up images of mat slices M1 – M7 showing grains bound by filaments (white arrows). A Binding results by slice, where slice numbers correspond to the slices cut shown in A and B. Red squares represent the percentage of grains that remained bound to the mat slice when the mat was inverted. Blue triangles represent the percentage of grains that remained bound to the mat slice when the mat was held at 90° at a distance of 10 cm facing a hose producing 0.14 L/s of water. B Results by the angle at which the mat was inclined during incubation. ! "# ! ! $%&'()!*"! ! " # $ % & ' ! " # $ %& ' ( )" *! *" )! ! " %+,, (! (" (# ($ (% (& (' # -.- -." -.$ -.& -./ !.- ! "# $%& ' *01 2 3 4 56+ 57 + 801 4 69 + 0: ,1 4 64 68 ; < 4 2 : + 6=,>: 0 #- $- %- &- '- /- ? 62< 4 6:+ 168< :+ @ A:80 ::9B $ % C43D9355A+46E:09456 C43D9355A+2<59:+A40:23+7<5C 66° 72° 65° 39° 61° 61° 61° Slice number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igure 18. Trapping of fine, medium, and coarse grains by Chaetomorpha mats of different densities inclined at 0°. A Mats used for the experiment showing visible density variation. B Grain trapping results, where mat density (thickness) is approximated by dividing the mass of the mat by the cross-sectional area of the mat to which the grains were delivered. Equations and R 2 values of best-fit trend lines for each grain size are shown to the right of the plot. Standard deviations of measured percentage of grains trapped at a given angle normalized by mat density is on the order of 10% (data not shown), so the trapping percentages measured here for fine grains are not considered significantly different from one another. ! "# ! ! ! $%&'()!*+! !"#"$% & ' () " * " + & ,% - . " #"+& /0 ! " # " 1 1 & , 2) "$"+&3/" - . " #"1& ++ ! "#" '& '%) " *"+& 1, - . " #"+& 0% +&+ +&1 +&3 +&% +&, +&2 +&( +&' +&0 +&/ 1&+ + +& +3 +& +, +& +( +& +0 +& 1 456789":56;;<= > 6 :"=< 897 :! " ?@ A BC 3 D E7 8< C<=7F C BG 65 9< E78< C<=7FC BG659< C6:"=<897:! ! " 64 Figure 19. Trapping of grains of various sizes after two days of mat incubation by Chaetomorpha green algal mats inclined at angles between 0-75°. Key: circles = fine grains; triangles = medium grains; squares = coarse grains; dark green = dense mats; light green = thin mats. Example images of dense and thin mats used for experiments are shown. Error bars for the dense mats represent the standard deviation for two replicate samples. Thin mats measurements represent single samples. Grains trapped 0 10 20 30 40 50 60 70 80 Incline angle (degrees) 0.0 0.2 0.4 0.6 0.8 1.0 fine medium coarse dense thin 65 Figure 20. Retention of A fine and B medium grains over various incubation periods by thin Chaetomorpha green algal mats inclined at angles between 0-75°. C Trapping vs. binding of grains of different sizes by thin Chaetomorpha algal mats inclined at 45° for 0 or 2 days. Lighter line colors represent longer periods of incubation following grain delivery. Error bars where present represent the standard deviation for two replicate samples, otherwise only one sample was measured. ! ! "" ! ! ! #$%&'(!)*! A Fine grains B Medium grains ! "! #! $! %! &! '! (! )! * + , - . / 0 1+,2234 5.6 7 - .30 ,.87 30 9 438+ 33/ : ! "! #! $! %! &! '! (! )! 5.6 7 - .30 ,.87 30 9 438+ 33/ : !; ! !; # !; % !; ' !; ) "; ! 00!04,</ 00#04,</ 00(04,</ " %04,</ # "04,</ * + , - . / 01+,2234 !; ! !; # !; % !; ' !; ) "; ! *+,-./0+31,-.34 !;! !;# !;% !;' !;) ";! =-.3 >34-?> 6@,+/3 * + , -. 0 /-A 3 1+,223401! 1+,223401# B @ ? . 401! B @ ? . 401# C Trapping vs. binding at 45° 67 Figure 21. SEM images of algal mats. A-B Trapped fine grains at 30° after 0 days of incubation. Note that several “trapped” grains do not appear to be touching an algal filament. Image B is a close-up of the inset in A. C Medium grain trapped in an algal filament at 75° after 0 days of incubation. C Grain trapped in an algal filament at 0° after 0 days of incubation in an algal control where no grains were added, suggesting that the filaments also trap grains in their environment. E-H Encrustation of algal filaments after 0 days of incubation by diatoms and other substances at various magnifications. I-L Encrustation of algal filaments over several different incubation periods. M-P Biofilms on delivered grains after incubation periods of 2 (M- N) and 14 (O-P) days. ! "# ! ! $%&'()!*+! A-B Grain trapping !"## $%%"&# !"## !%%"&# !"## !%%"&# '%"&# $%"&# !"## !"## !"## !"## !"## !"## !"## !"## C 0 days 75° medium D 0 days 0° no grain control E-H 0 days Filament encrustation I 0 days Filament encrustation over time J 1 days K 2 days L 14 days M 2 days N 2 days O 14 days P 14 days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able 1. Grain sizes used in this study. ! " # $% &' ( ) * & ! " # $ % & +$ ,* & - * %' . /" ' 0&+ 1 # 2 * &* 34$ 5 # 2 * %' &6 - * %' . /" ' 07&8 9 : : ; & < $ % * & => ? :@@&A B & < $ % *&+ #% C & B *C $ 4 B & @D => ? :D @&B B & 1 / #" + *& E & 5 *" ( & 1 / #" + *& + #% C & 1/ # " +* & : D@ ? =D >&B B & 5 *" ( &< $ % *&E &< $ % *&) *F F 2 *+ & 71 Table 2. Grains trapped (as mass fraction of grains delivered) by cyanobacterial mats of varying maturity compared with a glass slide control. Averages (avg) and standard deviations (!) are for duplicate measurements. ! 2# + + &+ 2 $C * & 1 / % ' " / 2 + & G" *+ 0 &B #' + & < $ 2 #B *% ' &2 *% H' 0 & I &@ &B B & 8 &C # ( &/ 2C &B # ' + & < $ 2 #B *% ' &2 *% H' 0 & J 8 &B B & > &C # ( &/ 2C &B # '+ & < $ 2 #B *% ' &2 *% H' 0 & & J8 ? : &B B & K % 1 2 $ % * & #% H 2 *& & ! "# $ % $ ! "# $ % $ ! "# $ % $ ! "# $ % $ G$ % *&H " #$ % + & & & & & & & @ L & @D MN & @D @= & & & & & & & 8N L & @D =O & @D @= & 8D @P & @D 8@ & @D MO & @D @P & @D M9 & @D @O & >@ L & @D NM & @D @> & @D O@ & @D 8: & @D OO & @D 8> & @D O> & @D PP & PN L & @D @9 & @D :N & @D PO & @D 8: & @D M> & @D :@ & 8D @9 & @D 8P & =@ L & ? @D @: & @D @N & @D =: & @D >> & @D M= & @D @N & @D N= & @D >> & MN L & @D @M & % E# & @D 8N & @D 8: & @D P9 & @D 8= & ? @D @P & @D PN & & & & & & & & & & Q *C $ 4 B &H" #$ % + & & & & & & & @ L & @D 9> & @D @P & & & & & & & 8N L & @D 99 & @D @8 & @D ON & @D 8P & @D O= & @D @> & @D 99 & @D 88 & >@ L & @D 8@ & @D 8P & @D 98 & @D P8 & @D 98 & @D @: & @D =O & @D >P & PN L & @D @@ & @D @@ & @D 8> & @D @O & @D O> & @D @M & @D N8 & @D :9 & =@ L & @D @@ & @D @@ & ? @D 8O & @D @8 & @D 8P & @D :8 & ? @D :P & @D 8> & MN L & @D @@ & @D @@ & ? @D 8O & @D @O & ? @D 89 & @D @> & ? @D 8@ & @D @O & & & & & & & & & & R/# " + * &H " # $ %+ & & & & & & & @ L & @D 9M & @D @: & & & & & & & 8N L & @D 9> & @D @P & @D 9> & @D N> & @D 98 & @D :@ & @D OM & @D @8 & >@ L & @D @@ & @D @@ & @D @@ & @D 8: & @D == & @D 8= & @D =N & @D >= & PN L & @D @@ & @D @@ & @D @> & @D 8@ & @D :M & @D 8@ & @D :M & @D @> & =@ L & @D @@ & @D @@ & ? @D 8P & @D 8@ & @D @: & @D @P & ? @D @> & @D @9 & MN L & @D @@ & @D @@ & ? @D @: & @D @M & ? @D :N & @D >N & ? @D :N & @D @P & 72 Table 3. Grains trapped, bound on contact (expressed as a fraction of grains delivered and as a fraction of grains trapped), and bound by well-developed (filament length ~1cm) cyanobacterial mats after 12 hours of incubation with delivered grains (*measurements done 7 months prior to other measurements presented in this table: July 2012 vs. February 2013) as mass fraction of grains delivered. Averages (avg) and standard deviations (!) are for duplicate measurements. S" #) ) * C & T/4%C&/%&1 /%' # 1 ' & G " # 1 ' $ /%&/< &' " # ))* C& H " # $%+ &F/4%C&/%& 1 / % ' #1 ' & T/4%C&# < ' * " &8 : & 0/4" + U & K % 1 2 $ % * & # % H 2 * & ! "# $ % $ ! "# $ % $ ! "# $ % $ ! "# $ % $ G $ % *& H" #$ % + & & & & & & & @L & 8D @9 & @D 8P & @D M: & @D 8M & @D == & @ D8 O & @D O9 & @D @8 & 8NL & & & & & & & @D OO & @D @> & >@L & @D ON & @D :: & @D N@ & @D 8P & @D N9 & @D :: & @D M= & @D @9 & PNL & & & & & & & @D O: & @D @@ & =@L & 8D 88 & @D 8P & @D >@ & @D PM & @D :M & @D P: & @D M@ & @D :@ & MNL & & & & & & & @D =8 & @D :8 & & & & & & & & & & Q *C $ 4 B &H" #$ % + & & & & & & & @L & 8D 8: & @D 8= & @D :8 & @D :: & @D 8O & @D 89 & @D M9 & @D 8: & 8NL & 8D @@ & @D 8P & ? @D @: & @D 8P & @ & @ & @D O= & @D @: & >@L & 8D @P & @D 8P & ? @D @> & @D 8O & @ & @ & @D OP & @D @N & PNL & @D >= & @D =9 & @D @O & @D 8P & @D :: & @D NM & @D O@ & @D @@ & =@L & @D =P & @D 8O & ? @D @N & @D 8P & @ & @ & @D PM & @D :: & MNL & @D N> & @D 8P & @D @M & @D 8P & @D 8> & @D := & @D >8 & @D 88 & & & & & & & & & & R/# " + * &H " # $ %+ & & & & & & & @L & & & & & & & @D 98 & % E# & 8NL & & & & & & & @ DO > & @D @9 & >@L & & & & & & & @D O> & @D @O & PNL & & & & & & & @D M@ & @D 8O & =@L & & & & & & & @D >P & @D 89 & MNL & & & & & & & @D @= & @D 88 & 73 Table 4. Results of flow experiments with cyanobacterial mats showing the fraction of grains still attached to the mats calculated from the measured mass of lost grains after draping the mats onto the beakers, initiating flow conditions, one and two days under flow conditions, and inverting the mat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able 5. Grains trapped immediately upon contact with Chaetomorpha algal mats of different densities inclined at 0° as mass fraction of grains delivered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able 6. Grains remaining on (trapped by) dense Chaetomorpha algal mats as mass fraction of grains delivered following two days of incubation with delivered grains. Averages (avg) and standard deviations (!) are for duplicate measurements. S" #) ) * C &F ( &C * % + * &B #' & K % 1 2 $ % * & # % H 2 * & ! "# $ % $ G$ % *&H " #$ % + & & & @L & @D :> & @D 8= & 8NL & @D PO & @D @N & >@L & @D MO & @D @: & PNL & @D P= & @D :@ & =@L & @D O@ & @D := & MNL & @D 9= & @D O@ & & & & Q *C $ 4 B &H" #$ % + & & & @L & @D MO & @D @P & 8NL & @D 99 & @D @@ & >@L & @D M= & @D @N & PNL & @D 9@ & @D @N & =@L & @D 98 & @D @8 & MNL & @D MM & @D 8= & & & & R/# " + * &H " # $ %+ & & & @L & @D M: & @D @N & 8NL & @D 9N & @D 8M & >@L & @D O> & @D @: & PNL & @D 9> & @D @= & =@L & @D ON & @D @= & MNL & @D 98 & @D 8P & 76 Table 7. Grains remaining on (trapped by) thin Chaetomorpha algal mats as mass fraction of grains delivered after different periods of incubation with the delivered grains: 0 days (trapped on contact), 7 days, 14 days, and 21 days. Where a standard deviation (!) value is present, averages (avg) are for duplicate samples, where no standard deviation is present the single measured value at that angle is presented. @&C #( + & :&C #( + & M&C #( + & 8P&C #( + & :8&C #( + & K % 1 2 $ % * & # % H 2 * & ! "# $ % $ ! "# $ % $ ! "# $ % $ ! "# $ % $ ! "# $ % $ G$ % *&H " #$ % + & & & & & & & & & & @L & @D @O & % E# & @D :@ & % E# & @D :@ & @D P9 & @D :8 & % E# & & & 8NL & @D 8= & % E# & & & & & & & & & :NL & @D @> & % E# & @D @M & % E# & @D @: & @D @O & @D :8 & % E# & & & P@L & @D @N & % E# & & & & & & & & & N@L & @D 88 & % E# & @D 8P & % E# & @D @@ & @D @= & @D M= & % E# & & & =@L & @D 8P & % E# & & & & & & & & & & & & & & & & & & & & Q *C $ 4 B &H" #$ % + & & & & & & & & & & @L & @D N> & % E# & @D >> & % E# & @D :O & @D 8= & @D @= & % E# & @D 8: & @D @O & 8NL & @D >P & % E# & @D 8N & % E# & @D @9 & @D @N & @D 8: & % E# & @D 8N & @D @P & :NL & @D 8@ & % E# & @D 8= & % E# & @D 89 & @D 88 & @D 8> & % E# & @D :8 & @D :P & P@L & @D P: & % E# & @D @8 & % E# & @D 8P & @D @N & @D @9 & % E# & @D 8> & @D :8 & N@L & @D >= & % E# & @D :O & % E# & @D >@ & @D 8@ & @D :@ & % E# & @D 8> & @D 8: & =@L & @D N@ & % E# & @D :M & % E# & @D 89 & @D @M & @D :@ & % E# & @D PP & @D 8P & 77 Table 8. Grains trapped vs. bound by thin Chaetomorpha algal mats as mass fraction of grains delivered on contact (0 days) and following two days of incubation following grain delivery. Averages (avg) and standard deviations (!) are for duplicate measurements. & & & S" #) ) * C & & T/4%C & & & & @&C #( + & :&C #( + & & @&C #( + & :&C #( + & ! " # $% &+ $, * & W % H 2* & $ ! "# $ % $ ! "# $ % $ $ ! "# $ % $ ! "# $ % $ < $ % * & PNL & & @D @M & @D 88 & @D @P & @D 8: & & @D @N & @D @@ & ? @D @= & @D @M & & & & & & & & & & & & & B *C $ 4 B & PNL & & @D 8P & @D @M & @D 8M & @D @N & & ? @D @: & @D @N & @D @: & @D @8 & & & & & & & & & & & & & 1/ # " +* & PNL & & @D >: & @D @= & @D >O & @D @8 & & @D 8@ & @D @: & ? @D @8 & @D 8@ & 78 Table 9. Medium-sized grains trapped vs. bound by dense Chaetomorpha mats under flow conditions. Q #' & W % H 2* & G 2/. &' ( )* & G 2 / . & "# '* & 6 VE+; & ! " # $ % +& C* 2 $ 5 * " * C & ! " # $ % +& 1 / 2 2 *1 ' *C & G "# 1 '$ / % & " * B # $ %$ %H & R/%C $ ' $ / % & 8 & >@L & C$" * 1 ' & @ D: & ND 8@= & @D =9@ & @D O= & ' " # ))* C & : & >@L & C$" * 1 ' & @ D: & ND N>> & :D 9@8 & @D PO & F/4%C & 79 REFERENCES Abbott, I.A., and George Jacob Iollenberg, 1993, Marine Algae of California: Stanford University Press, Palo Alto, CA. Aitken, J., 1967, Classification and environmental significance of cryptalgal limestones and dolomites, with illustrations from the Cambrian and Ordovician of southwestern Alberta: Journal of Sedimentary Research, v. 37, no. 4, p. 1163–1178, doi: 10.1306/74D7185C- 2B21-11D7-8648000102C1865D. Allwood, A.C., Grotzinger, J.P., Knoll, A.H., Burch, I.W., Anderson, M.S., Coleman, M.L., and Kanik, I., 2009, Controls on development and diversity of Early Archean stromatolites: Proceedings of the National Academy of Sciences of the United States of America, v. 106, no. 24, p. 9548–55, doi: 10.1073/pnas.0903323106. Altermann, W., Kazmierczak, J., Oren, A., and Wright, D., 2006, Cyanobacterial calcification and its rock-building potential during 3.5 billion years of Earth history: Geobiology, v. 4, no. 3, p. 147–166, doi: 10.1111/j.1472-4669.2006.00076.x. Awramik, S.M., 1971, Precambrian columnar stromatolite diversity: reflection of metazoan appearance: Science (New York, N.Y.), v. 174, no. 4011, p. 825–827, doi: 10.1126/science.174.4011.825. Awramik, S.M., 1988, Role of Algal Eukaryotes in Subtidal Columnar Stromatolite Formation: Proceedings of the National Academy of Sciences, v. 85, no. 5, p. 1327–1329, doi: 10.1073/pnas.85.5.1327. Awramik, S.M., and Sprinkle, J., 1999, Proterozoic stromatolites: The first marine evolutionary biota: Historical Biology, v. 13, no. 4, p. 241–253, doi: 10.1080/08912969909386584. Bengtson, S., Belivanova, V., Rasmussen, B., and Whitehouse, M., 2009, The controversial “Cambrian” fossils of the Vindhyan are real but more than a billion years older: Proceedings of the National Academy of Sciences of the United States of America, v. 106, no. 19, p. 7729–34, doi: 10.1073/pnas.0812460106. Berelson, W.M., Corsetti, F.A., Pepe-Ranney, C., Hammond, D.E., Beaumont, W., and Spear, J.R., 2011, Hot spring siliceous stromatolites from Yellowstone National Park: assessing growth rate and laminae formation: Geobiology, v. 9, no. 5, p. 411–424, doi: 10.1111/j.1472-4669.2011.00288.x. 80 Buick, R., Dunlop, J.S.R., and Groves, D.I., 1981, Stromatolite recognition in ancient rocks: an appraisal of irregularly laminated structures in an Early Archaean chert-barite unit from North Pole, Western Australia: Alcheringa, v. 5, no. 3, p. 161–181. Carrigy, M.A., 1970, Experiments on the angles of repose of granular materials: Sedimentology, v. 14, no. 474, p. 147–158. Corsetti, F.A., and Storrie-Lombardi, M.C., 2003, Lossless Compression of Stromatolite Images: A Biogenicity Index?: Astrobiology, v. 3, no. 4, p. 649–655, doi: doi:10.1089/153110703322735980. Dill, R.F., Shinn, E.A., Jones, A.T., Kelly, K., and Steinen, R.P., 1986, Giant subtidal stromatolites forming in normal salinity waters: Nature, v. 324, no. 6092, p. 55–58, doi: 10.1038/324055a0. Dravis, J.J., 1983, Hardened subtidal stromatolites, Bahamas: Science, v. 219, no. 4583, p. 385– 6, doi: 10.1126/science.219.4583.385. Dupraz, C., Pattisina, R., and Verrecchia, E.P., 2006, Translation of energy into morphology: Simulation of stromatolite morphospace using a stochastic model: Sedimentary Geology, v. 185, no. 3-4, p. 185–203, doi: 10.1016/j.sedgeo.2005.12.012. Falkowski, P.G., Katz, M.E., Knoll, A.H., Quigg, A., Raven, J.A., Schofield, O., and Taylor, F., 2004, The evolution of modern eukaryotic phytoplankton: Science, v. 305, no. 5682, p. 354–60, doi: 10.1126/science.1095964. Feldmann, M., and McKenzie, J., 1997, Messinian stromatolite-thrombolite associations, Santa Pola, SE Spain: an analogue for the Palaeozoic?: Sedimentology, v. 44, no. 5, p. 893–914, doi: 10.1046/j.1365-3091.1997.d01-53.x. Feldmann, M., and McKenzie, J.A., 1998, Stromatolite-thrombolite associations in a modern environment, Lee Stocking Island, Bahamas: Palaios, v. 13, no. 2, p. 201–212, doi: 10.1043/ 0883-1351(1998)013<0201:SAIAME>2.0.CO;2. Fischer, A.G., 1965, Fossils, early life, and atmospheric history: Proceedings of the National Academy of Sciences, v. 53, no. 6, p. 1205–1215. Gebelein, C.D., 1969, Distribution, morphology, and accretion rate of recent subtidal algal stromatolites, Bermuda: SEPM Journal of Sedimentary Research, v. Vol. 39, no. 1, p. 49– 69, doi: 10.1306/74D71BE0-2B21-11D7-8648000102C1865D. 81 Ginsburg, R., 1957, Early diagenesis and lithification of shallow-water carbonate sediments in south Florida, in Regional aspects of carbonate deposition: a symposium. No. 5. Society of Economic Paleontologists and Mineralogists, p. 80–98. Glover, T.J., 1995, Pocket Ref: Sequoia Publishing, Littleton, CO. Golubic, S., 1976, Taxonomy of extant stromatolite-building cyanophytes: Developments in Sedimentology, v. 20, p. 127–140, doi: 10.1016/S0070-4571(08)71133-6. Grotzinger, J.P., and Knoll, A.H., 1999, Stromatolites in Precambrian carbonates: evolutionary mileposts or environmental dipsticks?: Annual review of earth and planetary sciences, v. 27, no. 1, p. 313–58, doi: 10.1146/annurev.earth.27.1.313. Grotzinger, J.P., and Rothman, D.H., 1996, An abiotic model for stromatolite morphogenesis: Nature, v. 383, no. 6599, p. 423–425, doi: 10.1038/383423a0. Guiry, M.D., 2013, Chaetomorpha linum (O.F.Müller) Kützing (M. D. Guiry & G. M. Guiry, Eds.): AlgaeBase: http://www.algaebase.org/search/species/detail/?species_id=222 (accessed February 2013). Gunatilaka, A., 1975, Some aspects of the biology and sedimentology of laminated algal mats from Mannar Lagoon, Northwest Ceylon: Sedimentary Geology, v. 14, no. 4, p. 275–300, doi: 10.1016/0037-0738(75)90003-2. Jones, B., Renaut, R.W., and Rosen, M.R., 1997, Biogenicity of silica precipitation around geysers and hot-spring vents, North Island, New Zealand: Journal of Sedimentary Research, v. 67, no. 1, p. 88–104. Kalkowsky, E., 1908, Oolith und Stromatolith im norddeutschen Buntsandstein: Zeitschrift der Deutschen Geologischen Gesellschaft, v. 60, p. 68–125. Kennard, J.M., and James, N.P., 1986, Thrombolites and stromatolites: two distinct types of microbial structures: Palaios, v. 1, no. 5, p. 492–503, doi: 10.2307/3514631. Van Kranendonk, M.J., Webb, G.E., and Kamber, B.S., 2003, Geological and trace element evidence for a marine sedimentary environment of deposition and biogenicity of 3.45 Ga stromatolitic carbonates in the Pilbara Craton, and support for a reducing Archaean ocean: Geobiology, v. 1, no. 2, p. 91–108, doi: 10.1046/j.1472-4669.2003.00014.x. Krylov, I.N.I., 1976, Approaches to the classification of stromatolites: Developments in Sedimentology, v. 20, p. 31–43, doi: 10.1016/S0070-4571(08)71127-0. 82 Logan, B.W., 1961, Cryptozoon and associate stromatolites from the Recent, Shark bay, Western Australia: The Journal of Geology, v. 69, no. 5, p. 517–533. Lowe, D.R., 1994, Abiological origin of described stromatolites older than 3.2 Ga: Geology, v. 22, no. 2, p. 387–90, doi: 10.1130/0091-7613(1994)022<0387:AOODSO>2.3.CO;2. Macintyre, I.G., Prufert-Bebout, L., and Reid, R.P., 2000, The role of endolithic cyanobacteria in the formation of lithified laminae in Bahamian stromatolites: Sedimentology, v. 47, no. 5, p. 915–921, doi: 10.1046/j.1365-3091.2000.00327.x. Mata, S.A., Harwood, C.L., Corsetti, F.A., Stork, N.J., Eilers, K., Berelson, W.M., and Spear, J.R., 2012, Influence of gas production and filament orientation on stromatolite microfabric: Palaios, v. 27, no. 4, p. 206–219, doi: 10.2110/palo.2011.p11-088r. McLoughlin, N., Wilson, L.A., and Brasier, M.D., 2008, Growth of synthetic stromatolites and wrinkle structures in the absence of microbes - implications for the early fossil record: Geobiology, v. 6, no. 2, p. 95–105, doi: 10.1111/j.1472-4669.2007.00141.x. Meadows, A., Meadows, P.S., Wood, D.M., and Murray, J.M., 1994, Microbiological effects on slope stability: an experimental analysis: Sedimentology, v. 41, no. 3, p. 423–435, doi: 10.1111/j.1365-3091.1994.tb02004.x. Monty, C., 1973, Precambrian background and Phanerozoic history of stromatolitic communities, an overview: Annales de la Société Géologique de Belgique, v. 96, p. 585– 624. Monty, C.L. V., 1976, The origin and development of cryptalgal fabrics: Developments in sedimentology, v. 20, p. 193–249, doi: 10.1016/S0070-4571(08)71137-3. Osborne, R.H., Licari, G.R., and Link, M.H., 1982, Modern lacustrine stromatolites, Walker Lake, Nevada: Sedimentary Geology, v. 32, no. 1-2, p. 39–61, doi: 10.1016/0037- 0738(82)90013-6. Pentecost, A., and Riding, R., 1986, Calcification in cyanobacteria, in Leadbeater, B.S. and Riding, R. eds., Biomineralization in lower plants and animals, Clarendon Press, Oxford, p. 73–90. Pepe-Ranney, C., Berelson, W.M., Corsetti, F.A., Treants, M., and Spear, J.R., 2012, Cyanobacterial construction of hot spring siliceous stromatolites in Yellowstone National Park: Environmental microbiology, v. 14, no. 5, p. 1182–97, doi: 10.1111/j.1462- 2920.2012.02698.x. 83 Petryshyn, V.A., Corsetti, F.A., Berelson, W.M., Beaumont, W., and Lund, S.P., 2012, Stromatolite lamination frequency, Walker Lake, Nevada: Implications for stromatolites as biosignatures: Geology, v. 40, no. 6, p. 499–502, doi: 10.1130/G32675.1. Pope, M.C., Grotzinger, J.P., and Schreiber, B.C., 2000, Evaporitic subtidal stromatolites produced by in situ precipitation: textures, facies associations, and temporal significance: Journal of Sedimentary Research, v. 70, no. 5, p. 1139–1151, doi: 10.1306/062099701139. Reid, R.P., Visscher, P.T., Decho, A.W., Stolz, J.F., Bebout, B.M., Dupraz, C., Macintyre, I.G., Paerl, H.W., Pinckney, J.L., Prufert-Bebout, L., Steppe, T.F., and DesMarais, D.J., 2000, The role of microbes in accretion, lamination and early lithification of modern marine stromatolites: Nature, v. 406, no. 6799, p. 989–92, doi: 10.1038/35023158. Riding, R., 2008, Abiogenic, microbial and hybrid authigenic carbonate crusts: components of Precambrian stromatolites: Geologia Croatica, v. 61, no. 2-3, p. 73–103, doi: 10.4154/GC.2008.10. Riding, R., 2000, Microbial carbonates: the geological record of calcified bacterial-algal mats and biofilms: Sedimentology, v. 47, no. s1, p. 179–214, doi: 10.1046/j.1365- 3091.2000.00003.x. Riding, R., 2011, The nature of stromatolites: 3,500 million years of history and a century of research, in Reitner, J., Quéric, N.-V., and Arp, G. eds., Advances in Stromatolite Geobiology, Springer Berlin Heidelberg, Berlin, Heidelberg, p. 29–74. Schopf, J.W., Kudryavtsev, A.B., Czaja, A.D., and Tripathi, A.B., 2007, Evidence of Archean life: Stromatolites and microfossils: Precambrian Research, v. 158, p. 141–155. Semikhatov, M.A., Gebelein, C.D., Cloud, P.E., Awramik, S.M., and Benmore, W.C., 1979, Stromatolite morphogenesis—progress and problems: Canadian Journal of Earth Sciences, v. 16, no. 5, p. 992–1015, doi: 10.1139/e79-088. Siegesmund, M.A., Johansen, J.R., Karsten, U., and Friedl, T., 2008, Coleofasciculus gen. nov. (Cyanobacteria): morphological and molecular criteria for revision of the genus Microcoleus gomont: Journal of Phycology, v. 44, no. 6, p. 1572–1585, doi: 10.1111/j.1529-8817.2008.00604.x. Stal, L.J., 2012, Cyanobacterial Mats and Stromatolites, in Whitton, B.A. ed., Ecology of Cyanobacteria II, Springer Netherlands, Dordrecht, p. 65–125. Walter, M.R., 1976, Stromatolites (M. R. Walter, Ed.): Elsevier, New York. 84 Walter, M.R., Buick, R., and Dunlop, J.S.R., 1980, Stromatolites 3,400–3,500 Myr old from the North Pole area, Western Australia: Nature, v. 284, no. 5755, p. 443–445, doi: 10.1038/284443a0. Walter, M.R., Grotzinger, J.P., and Schopf, J.W., 1992, Proterozoic Stromatolites, in Schopf, J.W. and Klein, C. eds., The proterozoic biosphere: a multidisciplinary study, Cambridge University Press, New York, p. 253–260. Walter, M.R., and Heys, G.R., 1985, Links between the rise of the metazoa and the decline of stromatolites: Precambrian Research, v. 29, no. 1-3, p. 149–174, doi: 10.1016/0301- 9268(85)90066-X. Wentworth, C.K., 1922, A Scale of Grade and Class Terms for Clastic Sediments: The Journal of Geology, v. 30, no. 5, p. 377–392. 85 Chapter 2: Dramatic local environmental change during the Early Eocene Climatic Optimum detected using high resolution chemical analyses of Green River Formation stromatolites ABSTRACT The Eocene Green River Formation represents a system of lakes that covered parts of what is now Wyoming, Colorado, and Utah, and captures the Early Eocene Climatic Optimum (EECO, 52-50 Ma), a period of very high global temperatures representing the warmest period of the Cenozoic Era that is associated with very high levels of atmospheric CO 2 . Lakes, especially closed basin lakes, can respond dramatically to environmental change because of their sensitivity to precipitation and evaporation. In this study, I 1) use stromatolites from the Rife Bed of the Green River Formation as fine-scale records of terrestrial paleoenvironmental change during a global hothouse climate, and 2) investigate how the environmental dynamics within the lake system affect the growth of stromatolites. The stromatolites we studied are composed of branching microdigitate columns laminated on the 10-100 µm scale. Laminae are grouped in cm-scale layers that alternate between micritic, calcite fan, and mixed microstructures. The micrite layers contain evidence for a biogenic origin and are enriched in 18 O, Na, and Mg/Ca. The fan layers, in contrast, appear to have formed abiogenically and are relatively depleted in 18 O, Na, and Mg/Ca compared to the micrite layers. The į 13 C and į 18 O strongly co-vary, suggesting the stromatolites formed in a closed basin lake (e.g., Talbot, 1990). In addition, clumped isotope analyses provide the first 86 quantitative values for water temperatures in lake water from the Green River Formation. The different microfabrics are associated with significantly different lake water temperatures: ~35°C for micrite and ~28°C for fan layers. Given that the stromatolites grew in a closed basin lake, we can link changes in the į 18 O (via a Rayleigh distillation model) and sodium ion concentration (assuming it behaves conservatively) to periods of evaporation or recharge, and thus model changes in lake volume/level during stromatolite growth. The two models converge upon similar results that suggest that dramatic lake volume changes occurred many times during the accretion of the ~30 cm thick stromatolite, with lake level rising and falling as much as 8 meters. Because of broad, flat bathymetry of the lake, such lake volume changes would have been accompanied by shoreline shifts on the order of tens of kilometers while the stromatolites were actively growing, throwing the traditional interpretation of stromatolites as shoreline facies into question. As a reality check, we note that the modern Great Salt Lake, a similarly broad and flat lake, experienced similar shoreline shifts over several years in the 1980’s. The calculated lake volume—and consequently lake depth—changes also provide insight into the formation of the stromatolites studied. The micrite microfabric formed when the lake was shallow and warm, whereas the fan microfabric formed in cooler waters when the lake was deeper, possibly below a thermocline. I hypothesize that the alternation between biogenic and abiogenic microfabrics present in the Rife Bed stromatolites are the result of dramatic changes in lake level influencing the microbiology and chemistry of the waters in which the stromatolites were forming, potentially due to the at least intermittent existence of thermal and chemical stratification in the lake. 87 INTRODUCTION The Green River Formation The Eocene Green River Formation represents the remains of a large system of intermittently interconnected intermountain lakes that covered much of southwestern Wyoming, northwestern Colorado, and northeastern Utah. The formation contains a record of lacustrine deposition in several different sub-basins over 10 million years during the first half of the Eocene. The largest of these lakes, paleolake Gosiute (52.5-48.5 Ma; Smith et al., 2008), is the focus of this study and is represented by the Green River Formation deposits of the Greater Green River Basin: the Tipton Shale Member (interpreted as a fresh, overfilled to balanced fill, saline lake), the Wilkins Peak Member (an evaporative, closed basin system), and the Laney Member (the return to balanced saline to overfilled fresh system). The Green River Formation is one of the most extensive and best-characterized paleolake systems in the world (a Google Scholar search for “Green River Formation” in July 2013 turned up over 7,000 research papers and conference abstracts), largely because of significant economic interest as one of the world’s largest known oil shale deposits (estimated at 1.5 trillion U.S. barrels, Dyni, 2006; compared to 2.4 billion barrels of crude oil produced by the U.S. in 2012 and a world oil consumption of 32 billion barrels in 2011, U.S. Energy Information Administration, 2012). It also contains some of the world’s largest deposits of trona and nahcolite, and is a potential analogue system for the Brazilian pre-salt oil deposits (Awramik and Buchheim, 2012). The formation preserves extensive fossil beds including the fish fossil Lagerstätte for which it is famous, as well as fossils that provide a snapshot of the emergence of modern mammals (Gazin, 1965). The formation is 88 also famous for recording varves (yearly sedimentary laminations) in its extensive oils shales, interpreted by others as a record of orbital forcing (Milaknovitch cycles) as well as annual and ENSO-style cycles (Ripepe et al., 1991; Fischer and Roberts, 1991; Carroll et al., 2003; Machlus et al., 2008). In addition, the Green River Formation records the Early Eocene Climatic Optimum (EECO) from 52-50 Ma, the warmest period of the Cenozoic. The EECO was associated with high levels of atmospheric CO 2 (Figure 1) and has been compared to future climate projections (Zachos et al., 2008), and thus provides a benchmark for study to understand future environmental change. As a very broad system with relatively flat bathymetry (depositional slopes of ~20 cm per km), the ancient lake was very sensitive to input and evaporation. As the most recent hothouse climate period, the EECO may be the best available analog for understanding the impact of current global climate change and for testing the predictions of climate models. Many workers have investigated the history of the lake system (e.g., Bradley, 1929; Bradley, 1964b; Picard and High, 1968; Trudell et al., 1970; Eugster and Surdam, 1973; Eugster and Hardie, 1975; Surdam and Stanley, 1979; Smoot, 1983; Fischer and Roberts, 1991; Roehler, 1991; Roehler, 1992; Roehler, 1993; Clyde et al., 2001; Rhodes et al., 2002; Keighley et al., 2003; Carroll et al., 2003; Pietras et al., 2003; Pietras and Carroll, 2006; Smith et al., 2008; Aswasereelert et al., 2012). Abundant fish, plant, and invertebrate fossils have been used to constrain ages of the members (Buchheim and Surdam, 1977; Swain, 1999; Grande, 2001; Leggitt and Cushman, 2001); and 40 Ar/ 39 Ar dating of volcanic ash beds provide chronometric 89 constraints for much of the formation (Carroll et al., 2003; Pietras et al., 2003; Smith et al., 2008). Stromatolites are present in the Green River Formation and are typically associated with facies interpreted as paleoshoreline systems (e.g., Buchheim and Surdam, 1977; Roehler, 1993). The actual duration of stromatolite lamination is not well understood; the formation of stromatolite laminae in other systems are considered diurnal (Golubi ü and Focke, 1978; Vanyo and Awramik, 1982; Berelson et al., 2011 ), seasonal (Jones, 1981), annual (e.g., Grotzinger and Knoll, 1999 and references therein), or even multi-annual (Petryshyn et al., 2012). Regardless of the precise timing, Green River Formation stromatolites existed within the lifetime of the shoreline system, and their laminated nature provides a unique opportunity to study environmental change in the basin on a potentially highly resolved temporal scale. This study focuses on the petrography and geochemistry of a stromatolite from a period when the lake system transitioned from a balance-filled, freshwater lake to an underfilled (closed), saline lake (Pietras et al., 2003). Our work permits estimates of lake environmental change at a critical period in the lake’s evolution on a fine temporal scale and gives insight into the dynamics of the climate during the EECO, which may better inform future climate projections. Furthermore, the study elucidates how the changing lake environments affected the growth of the stromatolites, impacting how we view stromatolites as environmental indicators in deep time. Paleoclimate during the Early Eocene Climatic Optimum The deposits at the Boar’s Tusk outcrop record the warmest period of the Cenozoic, the EECO from 52-50 Ma. The stromatolites of this study formed in the middle of the EECO ca. 90 51.3 Ma (Smith et al., 2008). Using several proxies, atmospheric CO 2 levels are estimated to have been >1000 ppm (Figure 1; Yapp, 2004; Lowenstein and Demicco, 2006; Beerling and Royer, 2011) compared with 400 ppm during the writing of this thesis (2013) and a projected value of 1800 ppm in 2400 due to anthropogenic CO 2 release (Zachos et al., 2008). Studies of leaf margins and areas from Green River Basin paleofloral collections indicate a mean annual temperature of 23±4 ͼ C at ~52.5 Ma (Niland Tongue, Wasatch Formation) and 20±2ͼ C at ~50 Ma (Little Mountain, Wilkins Peak Member, Green River Formation), and a mean annual precipitation of 100 cm and 77 cm, respectively. These paleoflora results bracket the time during which the Rife Bed stromatolites were forming and were part of a general trend of dramatic decrease in precipitation levels with relatively stable temperatures from the mid- Wasatchian to early Bridgerian (Wilf, 2000). The Tipton Shale and Farson Sandstone were deposited when the region was hot and humid (subtropical), during which time the fauna of the lake consisted of mammals, alligators, flamingos, and other subtropical flora and fauna (Roehler, 1991). A decrease in precipitation and a transition to the warm and arid conditions, indicated by the paleofloral studies, may have contributed to the drying out of the region represented by the evaporitic Wilkins Peak member. On a finer timescale, į 18 O values of deep-sea foraminerifera in marine Eocene sediments indicate that global temperatures did not change significantly during the period represented in the Boar’s Tusk outcrop (Zachos et al., 2001; Pearson et al., 2007). The climate of North American continental regions is generally thought to have been equable, with only mild seasonality. The huge lake system that the Green River Formation represents would have additionally buffered local seasonal temperature swings (Sloan, 1994). However, seasonal climate variations may have influenced the lake system. Lowenstein and 91 Demicco (2006) note that the Dead Sea of Jordan and Israel experiences a similar mean annual temperature as the Green River Formation at the time of stromatolite deposition, and Dead Sea surface water temperatures vary seasonally from 21-36°C. Likewise the Salton Sea, a shallow, closed-basin, evaporitic lake in an arid environment (Southern California, USA) with a similar mean annual temperature, has a mean water column temperature that varies seasonally from 13-34°C and varies greatly in extent and depth due to seasonal evaporation (Hely, 1964; Watts et al., 2001). Depending on their lamination frequency, stromatolites could be used to resolve environmental variability on seasonal (or finer) timescales. Stromatolites as environmental indicators Stromatolites are laminated, lithified, sedimentary structures that accrete from a point or initiation surface (Semikhatov et al., 1979) and can form by the microbial trapping and binding of sediments or by in-situ precipitation of minerals (this can be abiogenic or mediated by microbial activities). Layering in stromatolites, such as is present in Green River stromatolites, is caused by episodic accretion related to intermittent periods of mineral precipitation, microbial growth (if present), and sedimentation (Braga et al., 1995). Major environmental changes resulting from climate variations can be reflected in stromatolite growth and morphology as well as in the chemistry of individual stromatolite laminae. For example, carbonate-associated sulfate measured in different stromatolitic laminae of a tufa from the shoreline of Walker Lake, Nevada was found to have recorded changes in the lake’s volume over a period of ~1000 years (Berelson et al., 2008). Carbon isotopes measured in laminae of a modern siliceous stromatolite growing in a hot spring in Yellowstone National Park recorded changes in community composition apparently driven by periodic drops in the depth of 92 the pool in which they were forming (Berelson et al., 2011). The microstructure of stromatolites from the Laney Member of the Green River Formation has been used to infer annual to decadal fluctuations in the balance of precipitation and evaporation (Seard et al., 2013), but this has never been explicitly demonstrated or quantified. This study investigates stromatolites from the Rife Bed as fine-scale indicators of past lake chemistry and environment in order to better understand the evolution and variability of the Green River Lake system during the Early Eocene Climatic Optimum. Geological setting Stromatolites for this study were collected from the Boar’s Tusk outcrop (41.97 ͼ N, 109.25ͼ W) on the eastern side of White Mountain, approximately 41 km north of Rock Springs, Wyoming (location shown in Figure 2A, maps in Figure 3C and Figure 4A, also described by Bradley, 1929 and by Roehler, 1991 as the type locality for the Farson Sandstone). The outcrop is named for a distinctive nearby volcanic tower, the Boar’s Tusk, located 3 km to the west of the outcrop (Figure 2A). The site is within the Bridger Basin of the greater Green River Basin. The paleolake is commonly termed “Lake Gosiute”. The stratigraphic succession at the Boar’s Tusk site includes, from base to top, the Farson Sandstone and the Rife Bed of the Tipton Member, the Wilkins Peak Member (including an incursion from the Cathedral Bluffs Tongue of the Wasatch Formation), and a partial section of the overlying Laney Member (Figure 4B, following the nomenclature of Roehler, 1991). The Farson Sandstone is interpreted as a freshwater deltaic system based on faunal assemblages and sedimentary structures (Roehler, 1991). Rife Bed, in contrast, contains stromatolites (the subject of this study), oolites, mudstones, and oil shales (some with evaporite 93 pseudomorphs), and is interpreted to mark the lake’s abrupt transition to a saline system (Roehler, 1993). The presence of the stromatolites in association with oolite and cross-bedded sandstones suggest formation adjacent to an ancient shoreline complex (see Figure 30 in Roehler, 1993). The overlying Wilkins Peak Member is thought to represent a playa/mudflat system surrounding a hypersaline lake (Eugster and Surdam, 1973; Eugster and Hardie, 1975; Surdam and Wolfbauer, 1975; Lundell and Surdam, 1975), indicating further desiccation of the system. The lake was probably alkaline (pH 8-10; Smith and Robb, 1973) based on the sodium carbonate evaporite mineral assemblage and likely periodically stratified; anoxic bottom waters would explain the general lack of bottom sediment bioturbation (Boyer, 1982). Pietras et al. (2003) proposed that the cause of this abrupt transition from a balance-filled to underfilled evaporitic lake was the uplift of the Wind River Mountains along the continental fault to the north diverting a major source (or sources) of freshwater. The stromatolite-rich unit in the Rife Bed forms a ~30 cm-thick, resistant bench at the base of the Rife Bed of the Tipton Member of the Green River Formation (Roehler, 1991), near the contact with the overlying Wilkins Peak Member (Figure 5). The Rife Tuff is superjacent to the stromatolite unit in the Rife Bed at the Boar’s Tusk locality and was dated at 51.30±0.30 Ma (Smith et al., 2008), providing an approximate depositional age for the stromatolite. An overlying ash bed in the Wilkins Peak Member (Firehole Tuff, Smith et al., 2003) is dated at 50.70 Ma, further constraining the age of the Rife Bed stromatolites. Chronographically, the age of the stromatolites represents the middle of the Early Eocene Climatic Optimum (Zachos et al., 2001). Stromatolites in the bed are discontinuously exposed and the thickness of the stromatolite 94 bed pinches and swells between ~20 cm and ~1 m when traced laterally for 1 km—the extent of this investigation. Stromatolites are present throughout the Green River Formation and are typically associated with paleoshoreline systems consisting of thin oolite and intraformational breccias (Buchheim and Surdam, 1977; Roehler, 1993). The Rife Bed stromatolites represent what was likely part of an extensive area of stromatolites growing along the shallow rim of Lake Gosiute just outboard of a beach/ooid shoal system (see Figure 30 in Roehler, 1993 for an illustration), and thus were in a location sensitive to small changes in lake level and chemistry (e.g., Schultz et al., 2004; Buchheim et al., 2007; Buchheim et al., 2009). METHODS Stromatolite collection and processing Stromatolite collection With the help of students from the 2008, 2012 and 2013 International Geobiology Courses (the author took the course in 2008, and TA’d the course in 2012 and 2013), stromatolites were collected from along one kilometer of the Rife Bed stromatolite horizon (collection locations of stromatolites used for chemical analyses are shown in Figure 3). In order to minimize modern environmental contamination/weathering, fresh, unweathered samples were collected from the outcrop. Stromatolites were sectioned into vertical slabs using a water-cooled rock saw and stored in sterile, combusted aluminum foil to prevent organic contamination during storage. Large format (2x3 inch) thin sections were prepared and standard petrographic and cathodoluminescent 95 studies were conducted. The mineralogy of identified microfabrics was confirmed using powder- diffraction X-Ray crystallography and staining with Alazarin Red S. Four stromatolite samples were chosen for detailed analysis: BT08, BT12-CF-1, BT12- CF-2, and BT12-CF-4. Sample BT08 displayed an excellent succession of layers, so it was chosen for detailed analysis. Samples BT12-CF-1 and 2 were collected from within about a meter of one another and were used to test the uniformity of fabrics and geochemical signals between closely-spaced stromatolites. Sample BT12-CF-4 also displayed an interesting succession of layers. Although collected from different locations, the stromatolites formed in vertical succession (see discussion section), from bottom to top: BT08, BT12-CF-1 & 2, and BT12-CF-4. Microdrilling The Boar’s Tusk stromatolites are characterized by cm-scale, light and dark banding patterns (referred to as layers in the rest of this paper in order to differentiate them from the finer scale laminations present within the stromatolites) that reflect different microfabrics. These layers provided targets for chemical analyses. The layers were microdrilled for subsequent chemical analyses using a Carpenter Microsystems CM-2 microdrill fitted with a 1 mm carbide bit, taking care to drill only the targeted microfabric (avoiding detrital infill and cements). While microdrilling was limited to no more than 2 mm depth, contamination of microdrilled fabrics by detrital infill, cements, or other non-targeted fabrics due to the unseen three-dimensional structure of the stromatolite slabs cannot be entirely ruled out. However, different fabrics typically had significantly different powder colors; samples that were visibly contaminated were 96 excluded from analyses. In addition, the elemental contamination from the drill bits was tested and was negligible (see Appendix B). Several miligrams of powder were collected from each drilled site, with at least three sites drilled as replicate samples from each microdrilled layer (Figures 6-9). ICP-MS elemental analyses Processing of the BT08 samples Elemental analyses were conducted via ICP-MS in the lab of Dr. Pedro Marenco, Bryn Mawr College. For the elemental analyses of the BT08 stromatolite, 1.8-2.8 mg microdrilled powder samples were spun down in microcentrifuge tubes to collect powder in the bottom of the tube. The samples were dissolved by adding 1.5 mL 0.22M trace metal grade nitric acid, spun down to collect acid in the bottom of the tubes, and agitated on an orbital shaker for 60 minutes at 200 rpm. The samples were pelleted by centrifugation for 1 minute at maximum speed and the supernatant was decanted into 15 mL Falcon tubes. Samples reacted overnight with 7.5 mL of 633 ppb BaCl in 0.22M nitric acid in order to precipitate out any sulfate originally bound in the carbonates (carbonate associated sulfate, CAS) as BaSO 4 . BaSO 4 formed was pelleted out by centrifuging samples for 10 minutes at 1350 g, and the supernatant was transferred to fresh falcon tubes. Due to the high amounts of barium discovered to be present in the carbonates prior to reaction with BaCl, measuring CAS using this method was not possible. Barium results reported in text, tables, and figures for the BT08 stromatolites are measured values minus the amount of barium added to the samples for the CAS procedure. 97 Processing of other stromatolite samples In addition to the BT08 stromatolite, three other stromatolites were analyzed: BT12-CF- 1, BT12-CF-2, and BT12-CF-4. As a result of the intrinsic barium and the inability to accurately measure CAS in the BT08 stromatolite, the reaction with BaCl was not repeated for the other stromatolites. Instead, the following protocol was used: Several milligrams of powder (sufficient to run both elemental and subsequent stable isotope analyses) were drilled from layers of all of the collected stromatolites using carbide bits and the collected powder was then stored in wax-coated weigh paper. 1.5-3.5 mg of powder was transferred from each drill site into 15 mL Falcon tubes. The powder was reacted with 9 mL 0.22M nitric acid under agitation on an orbital shaker at 200 rpm for 1 hour. The insolubles were pelleted from the samples by centrifuging them at 1350 g rpm for 10 minutes. The supernatent was decanted and transferred to fresh Falcon tubes for elemental analysis. ICP-MS measurements Dissolved carbonate samples were analyzed using an Agilent 7500 series ICP-MS with an octopole reaction cell using helium gas to minimize interferences (this was accomplished by the author at Bryn Mawr College in the lab of Dr. Pedro Marenco). The following masses were measured: 23(Na), 24(Mg), 27(Al), 43(Ca), 44(Ca), 55(Mn), 56(Fe), 57(Fe), 60(Ni), 63(Cu), 66(Zn), 86(Sr), 87(Sr), 88(Sr), 137(Ba), 138(Ba), and 238(U). Resulting values were compared to standards of known concentration that were run every hour to calibrate the response of the instrument and check for machine drift. In addition, blanks that underwent the same procedure as the powdered samples were run after every 3 samples. Because they gave the best instrument reads compared with other isotopes of the same element, we used values measured for 44 Ca, 56 Fe, 98 88 Sr, and 138 Ba to determine elemental abundances for Ca, Fe, Sr, and Ba, respectively. Measurements that fell outside the calibrated measurement range for a given mass were discarded except where analyses with diluted samples (see Appendix B) indicated that measurement of the element in question was linear well beyond the calibrated range. Correcting for insoluble material The insoluble fraction of different stromatolite microfabrics was determined later, using additional microdrilled powder. Averages for different microfabrics (and their corresponding standard deviations) were used for mass corrections. 1-4 mg of microdrilled powder was dissolved in 1.5mL Eppendorf microcentrifuge tubes by adding 1.5mL of 2% nitric acid and agitating samples on an orbital shaker at 200 rpm for 1 hour. The remaining insoluble material was pelleted out by centrifuging the samples at max speed. 1.0mL of supernatant was removed from the samples and the remaining liquid was evaporated out by drying samples overnight (17.5 hours) in an oven at 56°C. Dry pellets in tubes were cooled to room temperature and weighed, subtracting out the initial mass of the tube and correcting for scale drift and tube mass loss in the oven using blank samples. Calculating elemental concentrations from ICP-MS results Calibrated ICP-MS results for each elemental isotope measured gave the concentration of that isotope in the liquid sample run. Standard natural isotope abundances were used to correct for the total concentration of the element in the samples. These concentrations needed to be corrected for sample mass, soluble fraction of the sample, and sample dilution to get the abundance of that isotope in the soluble (carbonate) fraction of the sample. 99 sol sample I liquid measured I carbonate fm F mE E ][ ] [ (1) Where [E]carbonate is the calculated concentration of the element in the soluble (carbonate) fraction of the sample; [ I E] measured is the isotope concentration in the dissolved sample measured via ICP-MS; m liquid is the total mass of the dissolved (liquid) sample; II E I AM M F is the mass correction factor based on the natural abundance of the isotope measured (A I ) and the molar masses of the element (M E ) and isotope (M I ), see Table 1 for values used; m sample is the mass of the powdered sample; and f sol is the soluble (carbonate) fraction of the powdered sample. The elemental measurements of Mg and Ca were used to determine the relative magnesium content in the carbonates: >@ > @> @ Ca Mg Mg XMg (2) When XMg<0.04, the mineralogy is low-Mg calcite. When 0.12<XMg<0.28 the mineralogy is generally considered high-Mg calcite, and when XMg 0.5, the mineralogy is dolomite. Standard stable isotopes Because the BT08 samples were exhausted for the elemental analysis described above, the stromatolite was redrilled using a carbide bit, with new drill holes directly adjacent to the original drill holes so the isotope results could be correlated to the elemental results for the same sites (Figure 6). For all other stromatolites, elemental and stable isotope analyses were performed on aliquots of the powdered material from a single microdrilled site. 100 Standard stable isotopes were measured with the help of Miguel Rincon in the lab of Lowell Stott at the University of Southern California. For the isotope measurements, we added approximately 20 µg of powder from each sample to clean glass scintillation vials and ran them on a micromass, dual-inlet IsoPrime MassSpec mass spectrometer alongside limestone internal standards calibrated against VPDB for 18 O and 13 C. Samples were run using a timed reaction with phosphoric acid at 90°C for 20 minutes to release CO 2 from the carbonates in the samples. This reaction is sufficient to dissolve dolomite as well as calcite. For the runs done for this study, the instrument had an internal precision of 0.1‰ for both 18 O and 13 C. The standard deviation of the standards for all runs completed over a three-week period were 0.049‰ and 0.084‰ for 13 C and 18 O, respectively, and the maximum daily standard deviation for the standards were 0.07‰ and 0.15‰, respectively. Of the 190 samples run, 11 were sample replicates (different aliquots of powder from the same sample), which deviated by 0.09‰ ( į 13 C) and 0.3‰ ( į 18 O) from original sample measurements. For samples for which sample replicates were measured, the value presented for a sample is the mean of sample replicates. Clumped isotopes Carbonate clumped isotope paleothermometry Clumped isotope thermometry is a novel technique based on the measurement of the rare “heavy” mass 47 isotopic species (isotopologues) of CO 2 produced during acid digestion of carbonates. “Clumped” refers to the occurrence of bonds between two heavy isotopes in a molecule, e.g., the bond between 18 O and 13 C in mass 47 18 O= 13 C= 16 O. A stochastic distribution of isotopes amongst all isotopologues of CO 2 contains 47.5 ppb for the three isotopologues with mass 47, vs. 98.4% of the “standard” mass 44 isotopologue. The abundance of the mass 47 101 isotopologue is reported in ǻ 47 notation, which indicates the per mil difference between the measured abundance of mass 47 CO 2 in a sample and the expected stochastic abundance of mass 47 CO 2 : 1000 1 *45 45 1 *46 46 1 *47 47 47 » ¼ º « ¬ ª ¸ ¹ · ¨ © § ¸ ¹ · ¨ © § ¸ ¹ · ¨ © § ' R R R R R R (3) where R47, R46, and R45 are the abundance ratios of the heavy isotopologues to the standard mass 44 isotopologue, e.g.,: > @ > @ 2 44 2 47 47 CO CO R (4) and R47*, R46*, and R45* are the abundance ratios of heavy isotopologues in a standard with a stochastic isotopic distribution, generally a CO 2 gas of known composition heated to sufficient temperature to produce a stochastic isotopic distribution. For a review on the clumped isotopes method, see Eiler, 2007. In carbonate minerals, the abundance of mass 47 isotopologues is temperature-dependent (Schauble et al., 2006). When carbonate precipitation proceeds at equilibrium, lower temperatures favor the formation of the clumped mass 47 isotopologue, while progressively higher temperatures produce isotopologue abundances that approach a stochastic value. These properties of clumped isotopes led to the development of the carbonate clumped isotope paleothermometer (e.g., Ghosh et al., 2006, 2007; Eiler, 2007; Tripati et al., 2010). Samples Victoria Petryshyn (UCLA) microdrilled two layers (fan layers 8-8.5 and micrite layer 9) of the BT08 stromatolite for clumped isotope analysis (Figure 10). Powdered carbonate material 102 was combined from multiple drill holes to achieve the sample sizes required (approximately 40 mg of sample is required to analyze a sample in triplicate). Measurement We determined G 13 C, G 18 2 , ' 47 and ' 48 in CO 2 derived from the phosphoric acid digestion of 8-12 mg of carbonate samples (described in detail in Appendix D and in Passey et al., 2010) using a Thermo Scientific MAT 253 gas-source mass spectrometer using published configuration and methods in the lab of Dr. Aradhna Tripati at UCLA (Ghosh et al., 2006). We ensured that mass 44 voltages remained stable at 16 V throughout the course of each analysis. Typical precision was 0.005-0.009‰, equivalent to about 1-2 qC, consistent with other studies (Eiler, 2007; Huntington et al., 2009; Tripati et al., 2010; Eagle et al., 2010; Thiagarajan et al., 2011). Carbonate standards were prepared and analyzed in the same manner as the samples between every 3-4 samples; in addition, 1000 qC and 25 qC equilibrated gas standards were run each day. Samples were screened for the presence of contaminating molecules (such as hydrocarbons and sulfur compounds) using mass 48 anomalies. ' 48 values were calculated in the same way as ' 47 values, by references to the ' 48 stochastic distribution as defined by the analysis of heated gases. We considered samples with large measured deviations (>1‰) from the ' 48 heated gas line potentially indicative of the presence of contaminants and excluded them from further analysis. In total, we conducted 12 analyses of samples microdrilled from four locations—two each from two different BT08 stromatolite layers. 103 Calculations to derive clumped isotope values and their errors Sample reactions were carried out at 90 qC, so we applied a published (Passey et al., 2010), empirically derived acid digestion fractionation correction of 0.08‰ for ' 47 values to allow comparison to the published calcite line calibration in which samples were reacted at 25 qC (Ghosh et al., 2006). Data that have been collected thus far for aragonite, dolomite, and calcite indicate no discernible mineral-specific acid digestion effects. Errors in reported ' 47 values and calculated temperatures include the propagated uncertainty in heated gas determination and in sample measurement (Huntington et al., 2009). We collected the majority of data for this study before the proposition of an “absolute reference frame” for clumped isotope studies of CO 2 based on the analysis of water equilibrated CO 2 gases. We therefore present temperature data both relative to the stochastic value (i.e., the nomenclature used in most previous studies) and in the absolute reference frame (Dennis et al., 2011), while results in figures and described in the main text are reported in the absolute reference frame. These determinations are described in more detail in Appendix D. Deep-ultraviolet (deep-UV) native fluorescence spectroscopy Deep-UV native fluorescence permits the in situ detection of organic compounds while minimizing the interfering fluorescence by minerals that plagues other spectroscopic methods such as infrared and Raman spectroscopy (Bhartia et al., 2008). Species that fluoresce strongly in the UV under deep-UV excitation are primarily pi-bonded organic compounds. A deep-UV raster-scanning system was used that combines a 224 nm laser with a custom detector (Targeted UV Chemical Sensor, Photon Systems, Inc., Covina, California) that measures fluorescence emission in the ultraviolet simultaneously at 280, 300, 320, 340, 360, and 380 nm (described in 104 Bhartia et al., 2010) to highlight regions of organic material in a cleaned, polished slab of a stromatolite from the Boar’s Tusk. To clean the polished stromatolite of surface contaminants, the samples were sanded using 12 !m (yellow) Al-oxide lapping film and ddH 2 O, sonicated in ethanol for 10 minutes, and rinsed with acetone. This effectively removed surface features that were visible when the sample (which had not been handled in a sterile manner) was scanned prior to cleaning. The sample was then scanned using a 200 !m laser spot size with a 30% raster overlap. RESULTS Petrography and mineralogy The stromatolites are composed of microdigitate (“finger-like”) columns, where each individual column is approximately 1-2 mm in width and 5-15 mm in height (Figure 11). The digitae are narrow at their base and slightly wider (up to 1.5 times) at their tops. Some columns branched as they grew upwards. The microdigitae, in turn, constitute the major constructional element in larger domes (up to 1 m wide, ~30 cm high). The small columns grew normal to their underlying surface on the domes. Interstitial spaces between the columns are filled with siliciclastic sediment, ooids, and carbonate cement. Thin section photomosaic images of the stromatolites analyzed in this study are in Appendix A. The cm-scale layers visible in the Boar’s Tusk stromatolites represent alternations of three distinct microstructures: calcite fans (spar), micrite, and a mixture of the two. The specific mineralogy of the microstructures was determined using x-ray powder diffraction (XRD; Table 2) and Alizarin Red staining (Figure 12) as well as traditional petrography. 105 The fan layers consist of radiating low-Mg calcite fans that are clearly visible under cross-polarized light (Figure 11). The fans have well-developed crystal terminations and lack evidence of surrounding sediment compression (Figure 13A), suggesting that the calcite fans are a primary fabric of this stromatolite (not a diagenetic fabric). In addition, cathodoluminescence shows that the fans are only weakly luminescent and have similar luminescence to adjacent micrite layers, while both nonluminescent and brightly luminescent secondary cements are present in intercolumn spaces (Figure 14). XRD analysis confirmed that fans are composed of calcite, with no evidence for magnesium incorporation (Table 2). The micritic microstructure is composed primarily of dolomite and high-Mg calcite, and contains a small amount of very fine siliciclastic grains that commonly define laminations. Some siliciclastic grains appear to be trapped in the micrite beyond the angle of repose on the sides of the columns (Figure 13B). Identification of the micrite composition as dolomite was confirmed by staining (Figure 12), XRD (Table 2), and elemental composition. While the XMg (the molar fraction of Mg) calculations using the Mg and Ca concentrations measured by ICP-MS assume that all of the measured Mg and Ca were from carbonate (vs. present as other acid soluble material), correcting Mg and Ca values for the approximate percentage of total Mg and Ca present in carbonate (vs. in salts) did not significantly change calculated XMg values. Measured elemental values provided additional support to the XRD and staining classification of fans as primarily low-Mg calcite and micrite as dolomite or very high-Mg calcite (Figure 15). A third microstructure consists of a sub-mm intercalation of the calcite fan and micritic microfacies, termed “mixed” throughout this paper (Figure 11). Individual microdrilled layers 106 were classified as fans, micrite, or mixed fabrics and microdrilling was carefully carried out so as to avoid contamination of non-targeted fabrics. Intercolumn spaces were often partially filled with detrital infill, especially ooids, and closed with secondary cements, some of which were brightly luminescent under cathodoluminescence in contrast to the primary micrite and fan fabrics (Figure 14). Cements were generally absent in the columns and were therefore relatively easy to avoid while microdrilling. Layering patterns were traceable across adjacent stromatolites and even recognizable in stromatolites separated by several hundred meters (Figure 16); therefore the Rife Bed stromatolite bench appears to represent a stromatolite-forming event as opposed to a time- transgressive system or one strongly influenced by local conditions (e.g., groundwater). Stromatolites BT12-CF-1 and 2 were collected ~1 m apart from the same stromatolite head and showed identical lamination patterns (Figure 17). Stromatolites collected from the different locations record very similar layering patterns (Figure 16). While micritic layers varied in thickness at the different locations where stromatolites were collected, fan layers were quite uniform. Distinctive color variations in the fan layers made particularly useful markers for correlating one stromatolite to the other, such that it was straightforward to place stromatolites from different locations in a relative sequence (Figure 18). ICP-MS elemental results Results for ICP-MS element abundance measurements for all drilled samples are presented in Appendix B. Averages and standard deviations for measurements from each stromatolite layer after outliers were removed (as discussed below) are presented in Table 3 and 107 Figures 19-21. In general, replicate samples drilled from the same stromatolite layer gave very similar values (i.e., had low relative standard deviations) for all elements measured. Several clear outlier points (noted in the table in Appendix B) were removed from subsequent analyses. Four of the 122 samples measured gave elemental results that were consistently shifted from values from other samples for the layers from which they were drilled (BT08 sample 15, BT12-CF-2 samples 2, 3 and 4, and BT12-CF-4a Sample 18) and were considered suspect (possibly due to drilling into unseen textural elements in the third dimension, as discussed previously) and removed from subsequent analyses. The fan microfabric consistently recorded lower sodium and magnesium versus the micrite. No correlation between fabric/position in the stromatolite was noted for aluminum. Stable isotopes Stable isotope measurements for all drilled samples are presented in Appendix C. Averages and standard deviations for measurements from each stromatolite layer after outliers were removed are presented in Table 4 and Figures 19-21. The range of values measured (vs. VPDB) for all of the Boar’s Tusk stromatolites were 0.5- 5‰ for į 13 C and -7 to 0‰ for į 18 O, with the exception of two clear outlier points (one from BT08 layer 7 and one from BT12-CF-1 layer -1) which may have been contaminated during re-drilling for isotopes and were excluded from subsequent analyses. Values for į 13 C and į 18 O were strongly correlated; the plot of į 13 C vs. į 18 O had a best-fit line with an R 2 value of 0.89 and a slope of 0.64 (Figure 22). Furthermore, a clear trend is apparent between the microfabrics: fan layers consistently recorded lower į 13 C and į 18 O, whereas micrite recorded more enriched isotopic compositions (discussed in detail in a subsequent section). 108 Clumped isotopes Clumped isotope measurements gave a mean temperature of 28.5±1.0°C for the formation of BT08 fan layers 8-8.5 (which were combined for clumped isotope analysis) and 35.0±0.4°C for the formation of the immediately superjacent micrite layer 9 (Table 5, full table of results in Appendix D). In addition, the mean į 18 O and į 13 C in carbonate were -4.6±0.1‰ and 2.6±1.3‰ (VPDB), respectively, for fan layers 8-8.5; and -1.5±0.2‰ and 4.3±0.1‰ (VPDB), respectively, for micrite layer 9. With the exception of the į 18 O values for micrite layer 9, the į 18 O and į 13 C values we measured using clumped isotope analysis are not significantly different from the values that we measured using standard stable isotope measurements (Table 6, Figure 23). Deep-UV native fluorescence spectroscopy The deep-UV native fluorescence scan of the Boar’s Tusk stromatolite highlighted conspicuous UV-fluorescent regions in the scanned stromatolite (Figure 24A). Spectral features of the rim of the scanned slab were distinct from the cleaned carbonate in the interior. The spectra of micrite and fans were nearly identical, with the exception of spectrally-distinct spots (called the “pink spots” due to their color in the false-color representation of the detected fluorescence) that were present in the micrite layers and especially concentrated at the base of a micrite layer superjacent to a fan layer (Figure 24B). The spots had particularly high intensities in the 340-380 nm range. With the exception of spectra measured from the rim, which were similar to spectra measured for benzene derivatives, none of the spectra measured matched spectra of compounds previously reported using deep-UV native fluorescence spectroscopy (Bhartia et al., 2008, Figure 24D). However, the location of the “pink spot” peaks and general 109 shape of the spectrum are suggestive of large, extensively conjugated organic molecules (Rohit Bhartia, personal communication). DISCUSSION Evaluating biogenicity in the different microfabrics Direct evidence for biogenicity was absent in the fan layers; the fans appear to have formed abiogenically as chemical precipitates from lake water, consistent with Riding’s interpretation of sparry microfabrics in stromatolites as being abiogenic in origin (Riding, 2008; Riding, 2011). Here, we intentionally use the term abiogenic to mean that the stromatolites were not actively built by microorganisms. Microorganisms were almost certainly present in the waters in which the stromatolite formed, with some perhaps even living on the fans, so the system was not abiotic. However, it is difficult to envision a system in which continuous mm- scale calcite crystals would form by uninterrupted precipitation in a microbial mat or biofilm (in some stromatolites calcite fan crystals >2mm are present with continuous extinction patterns under cross-polarized light, see Figure 11). However, several lines of evidence suggest that the micritic layers formed with microbial influence. First, grains trapped at steep angles (angles beyond the angle of repose) are present in micritic laminae (Figure 13B); as discussed in a different chapter of this thesis, grains trapped beyond the angle of repose is an indication of the presence of a filamentous microbial mat during stromatolite growth. Second, the cloudy amber-colored material (Figure 13B) visible in the micrite layers (and absent in fan layers) is likely disseminated organic material. Third, deep-UV native fluorescence spectroscopy highlighted the presence of material that is present in micritic 110 layers and absent in fan layers that is spectrally distinct from the carbonate that surrounds it and is likely organic in origin (Figure 24). While the presence of organic material alone is not proof of a microbial origin for the micrite, it is an expected result if a microbial mat was present during micrite formation. Finally, dolomite formation at low temperatures has never been demonstrated in controlled, sterile environments but does occur under microbial mediation (Vasconcelos et al., 1995; Land, 1998; Warthmann et al., 2000). Thus, the two different microfabrics appear to have formed by very distinct mechanisms: the fans by abiogenic precipitation from lake water, and the micrite by microbially-mediated precipitation and the trapping and binding of grains by filamentous organisms. Why would microbial mats have been active builders during some episodes of stromatolite growth but not others? The chemistry of these different microstructures provides critical insight into the environmental factors that may have driven these different growth mechanisms. Evidence for minimal diagenesis The sedimentary rocks of this formation have not been deeply buried and the microfacies targeted were not subjected to significant diagenesis; thus, we consider the geochemical signals to be primary. Deposits in the Wasatch and Fort Union Formations underlying the Green River Formation in the area of this study (and thus more deeply buried than the Green River Formation) have vitrinite reflectance indices of <0.55% (Pawlewicz and Finn, 2002) which indicates that the deposits at the Boar’s Tusk never reached temperatures in excess of 60°C associated with the oil window, so burial-related heating was minimal. Mineral replacement could be an issue, especially since the elemental measurement results are for all water-soluble ions and could include secondary evaporites. However, secondary evaporites were not visible in 111 thin section. In addition, the microfabrics targeted would be relatively impermeable to water, making secondary ion replacement unlikely. Stromatolite chemistry and consistency across layers All of the stromatolites collected recorded similar ranges in elemental composition, but variability between layers, even super- and subjacent layers, was often quite high (Figures 19-21). Samples drilled from different points across a layer, both from within the same stromatolite and from different stromatolites collected from the same stromatolite head (as in BT12-CF-1 and BT12-CF-2, which were collected from approximately 1 m apart) gave relatively consistent elemental results. Outliers tended to be within the range of adjacent layers, suggesting that most outliers could be explained by microdrilling into adjacent layers (or infill or cements in interstitial spaces) due to the unseen 3D structure of the stromatolites. A few samples produced results that were consistently shifted from other samples for the same layer; these samples were considered suspect, possibly misdrilled, and excluded from subsequent analyses as previously discussed. Because layers were composed of many microlaminae smaller than the size of the drill, each analysis is necessarily an average across those microlaminae. While most stromatolites were microdrilled by drilling several small holes in a defined area within a layer to obtain sufficient sample, the BT12-CF-4 stromatolite was drilled in such a way as to minimize issues of vertical heterogeneity. Rather than drilling individual points, samples were drilled from paths along vertical transects up an entire layer. These samples gave similar standard deviations between triplicate samples and had similar variations between layers as measured in the other 112 stromatolites, suggesting that the impact of vertical heterogeneity on our measurements was small. Chemistry, temperature, and microfabric linked to lake volume changes Stromatolite chemistry consistent with phases of evaporation Evidence for evaporation in a closed basin. In a closed system, evaporation concentrates conserved ions (e.g. Na and Mg) as well as 18 O and 13 C, and these evaporation indicators co-vary in the Boar’s Tusk stromatolites (Figures 25-26). Previously published facies analyses (e.g., Carroll and Bohacs, 1999; Pietras et al., 2003) suggest that the Rife Bed represents the transitioning of Lake Gosiute to a closed system. The elemental and isotopic patterns of covariance suggest that, by the time the Boar’s Tusk stromatolites formed, the basin was hydrologically closed and the elemental and isotopic changes observed within the stromatolite are consistent with lake volume changes. All of the stromatolites in this study exhibit a strong covariation of į 18 O and į 13 C, as well as a large (>6‰) spread in į 18 O fractionation values; both features are characteristic of closed- basin lakes where evaporation significantly impacts the į 18 O and į 13 C signals (Figure 27; Talbot, 1990). In addition, the slope of the isotope covariation line (slope = 0.66 for į 13 C vs. į 18 O for all samples) is consistent with relatively low groundwater input; high influence of groundwater would buffer changes in į 18 O due to evaporation and result in a steeper slope. The isotope values themselves provide clues about the environment. The minimum mean water į 18 O value for a layer of -3.4‰ (SMOW) for BT08 layer 7.5 is a maximum value for the inflow waters feeding the lake. In addition, the relatively high į 13 C values (compared to the values in the Talbot 1990 study, Figure 27) suggest an arid catchment (a wet climate would support more vegetation and/or 113 a higher proportion of C3 vs. C4 plants, the degradation of which in the catchment would drive the carbon isotopes of source waters more negative). Alternate hypotheses considered. We argue that the stromatolites of this study formed in lake water and, therefore, differences in their chemistry in some way reflect changes in lake water chemistry. Alternative hypotheses include groundwater discharge into the lake fostering carbonate precipitation (e.g., Surdam and Stanley (1979) noted tufa-tower like microbialites associated with springs elsewhere in the Green River Formation), and the potential for the micrite within the stromatolites to be allochthonous (the fans, however, clearly formed in place within the lake). Arguments against the groundwater hypothesis are apparent. For example, sodium and magnesium (as well as several other elements) co-vary with į 18 O (Figure 25); elements such as sodium and magnesium that are conserved in closed systems at moderate concentrations are concentrated in an evaporating body of water, as is 18 O, so one would expect sodium or magnesium and 18 O to deviate from covariance if groundwater was important. In addition, the slope with which carbonate į 13 C and į 18 O co-vary suggests that the waters from which the carbonates formed were strongly influenced by evaporation (cf., Talbot, 1990), which would not be the case for spring waters (this is discussed in more detail in the following section). Previous authors have suggested that the large amount of dolomicrite in the Green River system could originate from evaporative pumping in the mudflats that surrounded the lake (Wolfbauer and Surdam, 1974; Mason and Surdam, 1992); dolomicrite produced in the mudflats could wash into the lake, and could perhaps become trapped and bound into stromatolites by microbial mats. While the mudflat hypothesis for micrite formation may be applicable to other 114 facies, the consistency of the geochemistry along micritic layers within the stromatolites argues against an allochthonous origin (much higher variability along layers could be expected if the micrite originated from multiple sources and washed into the lake system). For each stromatolite layer measured, three samples were taken from different locations along the layer in order to assess variability within a layer. The variability of key indicators of evaporation, Na and į 18 O, were not greater in micrite layers than in fan layers. While this does not rule out the possibility of an allochthonous source for the micrite, we found no evidence to support this alternate hypothesis. Fan and micrite layers are chemically distinct, formed at different lake depths Significant differences were found for nearly every measurement type between calcite fan and micrite samples (Table 7) including both į 18 O and į 13 C as measured by standard isotope analysis, the temperature of formation measured, and all of the elements measured except barium (although when the barium-spiked BT08 samples were excluded from analyses, a significant difference between the fan and micrite samples from the remaining stromatolites was found). In addition, two evaporation indicators, sodium concentration and į 18 O (Figure 26), suggest that micrite formed when the lake volumes were low (evaporated) while fans formed when lake volumes were high (freshened). When taken together, the elemental, isotopic, and temperature data suggest that the fans formed in deeper, cooler waters when the lake was volumetrically expanded, and the micrite formed in shallower, warmer waters when the lake volume was reduced. The magnitude of potential volume changes is considered in the next section. 115 Evaporation Models Above, we demonstrated that the stromatolite microfabrics and geochemistry indicate alternating phases of expansion (fans) and contraction (micrite) of the closed basin lake system, involving some magnitude of lake volume change, and corresponding lake level elevation fluctuation. In order to estimate lake volume and depth changes over the course of stromatolite growth, two simple models were constructed using (1) sodium and (2) į 18 O as tracers of evaporation and freshwater recharge in Lake Gosiute. Both systems respond to evaporation and freshwater recharge in different ways. Assumptions Assumptions used in the lake volume models are summarized in Table 8 and discussed in detail below. Lake area. The area of the lake at the time of stromatolite deposition is constrained by the sedimentological record of Lake Gosiute. One lower bound for the extent of the lake is the areal extent of the stromatolite-rimmed Rife Bed, which was mapped at 18,130 km 2 by Roeher (1992). Lake shape. Lake Gosiute is assumed to have been quite shallow considering its large extent (not unlike modern shallow evaporitic lakes such as the Great Salt Lake). An approximate lake shape was determined by Stanley and Surdam (1978) using Gilbert-type delta foresets around the edges of Lake Gosiute during deposition of the Laney Member. They stated that the maximum depth of the lake at that time must not have been much more than 25 m. The maximum areal extent of the lake during deposition of the Laney member was 40,000 km 2 , with shorelines ~120 km from basin center. Assuming a conical shape of the lake (also used by 116 Doebbert et al., 2010 in their Lake Gosiute isotope mass balance model), a relationship of 1 km of shoreline shift for every 0.2 m of depth increase is apparent. This relationship supports the shoreline gradient of ~0.2-0.4 m/km suggested by Bradley (1964a). Approximating the surface area (18,130 km 2 ) of the lake as a circle and the 0.2 m/km lake slope discussed above gives a lake depth at basin center of 15 m. Additional constraints on lake depth are provided by 15 m thick delta foreset beds noted by Pietras et al. (2003) for the Farson Sandstone (coeval with deposition of the Tipton Shale) at Whitehorse Creek, which would have been approximately 80 km from the center of Lake Gosiute (see Figure 52 of Roehler, 1993). Using the 0.2 m/km depth relationship above suggests that the Whitehorse Creek site was ~16 m shallower than the lake center, giving a total lake depth in excess of 30 m around the time that the stromatolite was forming. In this study, we use 15 m as the minimum lake depth and assume calculated lake depths up to 30 m to be within the range of reasonable values. The stromatolites did not form subaerally, so we assume they were at least slightly submerged during their growth (at least 0 meters water depth). While it is conceivable that the stromatolites could have accreted carbonate from spray from the lake, stromatolites still would not have formed a large distance above lake surface, especially since this would also put them a significant lateral distance from the edge of the lake. A consequence of this assumption is that the depth of water above the stromatolites is equal to the depth of the lake at basin center minus the 15 meter assumed minimum lake depth at basin center during stromatolite growth. 117 Lake volume. Using the conical lake shape assumption, the 15 m minimum lake depth, and 0.2 m/km lake slope constraints discussed above, the minimum lake volume used in our model is 88.4 km 3 . Layers included vs. excluded. Calcite fan, micrite, and mixed layers were included in our lake volume model calculations. Mixed layers, like single-fabric layers, consist of many individual laminae. Drilled samples represent an averaging of the chemistry of individual lamininae, and consequently a time average across the fan and micrite phases of the mixed layers. Ooids, while they did form in the lake, are potentially subject to considerable lateral transport and may not reflect lake chemistry at the time that they were incorporated into stromatolites. Therefore, the ooid layer present in BT12-CF-4 (layer 8) was excluded from lake volume calculations (this was the only ooid layer within the stromatolites studied). The stromatolites collected are from different horizons representing different periods in the growth of the Rife Bed stromatolite bench. The stromatolites analyzed in this study stack on top of one another and represent a nearly continuous record of the accretion of the stromatolite bench (Figure 18). Because of this, and because chemical heterogeneity was observed across layers, the stromatolites BT08, BT12-CF-1, and BT12-CF-4 (stacked in that order) were treated as a continuous record in our modeling. In the model results, calculated lake volumes are presented with respect to “stacked layers”, which numbers the layers from bottom to top as 1-10 for BT08 layers 1-10, 11-22 for BT12-CF-1 layers -1-10, and 23-34 for BT12-CF-4 layers 1-12. Additional model-specific assumptions are discussed in the detailed model descriptions below. 118 Description Sodium model. In a closed hydrologic system, sodium behaves as a conserved ion until concentrations become high enough to trigger sodium-bearing evaporite precipitation. In addition, sodium is incorporated into calcite in a linear relationship to source-water sodium concentration (Okumura and Kitano, 1986 for concentrations up to 1 g/L). Evaporation concentrates sodium; during periods of evaporation, volume change can be calculated from the change in sodium concentration: >@ >@ 1 0 01 Na Na VV (5) Freshwater input, in contrast, would dilute sodium concentration in the lake. We assume that the sodium concentration of the input waters was much less than the sodium concentration in the lake (a reasonable assumption; alkaline, evaporitic lakes can have sodium concentrations several orders of magnitude greater than freshwater). Consequently, volume change is approximated by Equation 5. The model assumes that sodium behaves conservatively over the range of measured values. Non-conservative behavior at high sodium concentrations would remove sodium from the system in the form of evaporites (resulting in an artificially low sodium measurement, which would result in an underestimation of lake volume changes) or result in disproportionate incorporation into carbonate. Depending on ion concentrations, this could result in an artificially high sodium measurement, which would overestimate lake volume changes. Oxygen isotope model. The į 18 O values of carbonates are directly related to the į 18 O value of the water from which they formed: 119 1000 1000 water carbonate carbonate water D G G (6) The fractionation factor Į carbonate-water is a function of mineralogy and temperature. The empirically-based equations used to determine Į carbonate-water for calcite and dolomite, the two dominant mineralogies of the Boar’s Tusk stromatolite microfabrics, are: 26 . 0 10 73 . 2 ln 1000 26 T water dolomite D from Vasconcelos et al., 2005 1 (7) 89 . 2 10 78 . 2 ln 1000 26 T water calcite D from Friedman and O’Neil, 1977 (8) where T is the temperature in Kelvin. For sequences of a single mineralogy, į 18 O carbonate values varied as much as 1.7‰ (from BT12-CF-4 layers 10-12; BT08 layers 7-8.5 see a similar increase in į 18 O). If temperature differences alone are to account for the change in į 18 O, this would require a temperature change of 7°C in the source water (Figure 28). Clumped isotope analyses suggest that water temperature varied by as much as ~6°C between the formation of adjacent layers (BT08 layers 8-9), so this is possible, but not likely: the dramatic variations in elemental composition and the correlation of į 18 O and į 13C values are more consistent with the contribution of evaporation in changing the į 18 O of lake water, suggesting that temperature is not the primary control on į 18 O in the system. For our simple oxygen isotope model, we treat evaporative and freshwater recharge phases separately (i.e., where layers become successively isotopically heavier vs. lighter). More elaborate isotope models have been used to understand isotope values measured from the Green 1 Based on microbially-mediated low-temperature dolomite precipitation, which was deemed a more realistic formula for this study than dolomite fractionation formulae based on high-temperature dolomite precipitation. 120 River Formation (e.g., Doebbert et al., 2010), but the number of unknowable parameters (e.g., rates of evaporation, precipitation, inflow, etc.) increases the uncertainty in the model. Here, we are interested in the direction and estimates of the magnitude of lake volume/level changes, for which this model suffices. Isotope model part I: Evaporative phases. An evaporating body of water will fractionate the isotopes in water because isotopically heavier water does not enter the gas phase as easily as lighter water; thus, as water in a closed system evaporates it will become increasingly isotopically heavy. Rayleigh fractionation describes the relationship between the isotope ratio R = [ 18 O]/[ 16 O] of an evaporating body of water, the initial isotope ratio of the body of water before evaporation R o , and the fraction of the initial volume of water remaining f: 1 0 D f RR (9) where Į is the temperature-dependent fractionation factor for the vapor-liquid phase transition of water, or 1/ Į liquid-vapor: 38 26 13 10 5041 . 3 10 6664 . 1 10 7123 . 6 685 . 7 ln 1000 T T T vapor liquid D (10) from Horita and Wesolowski, 1994 Solving Equation 9 for the end volume V f gives: 1 1 0 0 ¸ ¸ ¹ · ¨ ¨ © § D R R VV f (11) where V 0 is the initial volume of the body of water. The relative final and initial isotope ratios R and R 0 are related to their corresponding water isotope fractionation values: 121 1000 1000 00 G G R R (12) Equation 11 was used to calculate lake volume when į water 1 > į water 0 , i.e., when oxygen isotope values were heavier going from one layer to the one above it. Isotope model part II: Freshwater recharge phases. When isotope ratios in the lake become lighter, there must be a source of light isotope filling into the lake and diluting the heavy isotope signal. This could be precipitation, runoff, groundwater inflow, or all of the above. In a closed system, this is described by the mixing equation: ¸ ¸ ¹ · ¨ ¨ © § i i VV GG GG 0 0 (13) where the final lake volume i VV V 0 , V i is the volume of the freshwater input, and į i is the mean oxygen isotope fractionation value of the freshwater input. If į i is fixed to some assumed value, V f can be solved for using measured values of į 0 and į f and an assumed V 0 . Equation 13 was used to calculate lake volume when įwater 1 < įwater 0 , i.e., when oxygen isotope fractionation became lighter going from one layer to the one above it. Model limitations. This model assumes that evaporation and freshwater recharge occur in separate phases, i.e., that only evaporation (with no freshwater recharge) occurs during phases when isotope ratios become heavier and only freshwater recharge (with no evaporation) occurs when isotope ratios become lighter. Any freshwater input during “evaporative” phases would result in an underestimation of the total amount of evaporation that would have occurred to drive isotope values to the levels measured. Likewise, any evaporation occurring during “freshening” 122 phases would result in an underestimation of the total freshwater input needed to drive isotope values to the levels measured. In addition, the Rayleigh fractionation model for evaporation is only accurate for a system in which evaporated water is removed from contact with lake water so that there can be no back-reaction. In natural systems, back reactions do occur that can be quantified if atmospheric vapor į 18 O, humidity, and wind speed are known, which they are not for this system. However, given the arid conditions that prevailed during deposition of the subsequent Wilkins Peak Member, we can assume that humidity was low. Wave ripples at depth in the lake suggest that wind speeds were relatively high, similar to what they are today (e.g., Stanley and Surdam, 1978). Both of these factors would serve to minimize back-reactions. Assuming an open system represented by Equation 9 provides a minimum bound for volume changes in the system; any back-reactions would require greater volume changes to produce the same water isotopic changes. Freshwater isotope assumptions and choice of į i . The isotope model is quite sensitive to į i , the į 18 O value assumed for the freshwater input (Figure 29, Figure 30A). Thus, it is important to choose an appropriate į i for the model. The sensitivity analysis reveals that į i ~-11‰ produces the most conservative lake volume changes. This value is similar to modern surface water values in the region, is consistent with a mid-continental high-altitude freshwater source, and is well within the range of values predicted by previous studies for precipitation (Dettman and Lohmann, 2000) and river waters (c.f. Fricke and Wing, 2004; Doebbert et al., 2010) in the region. We therefore use -11‰ to provide a conservative estimate of lake change. Fricke and Wing (2004) suggest that river water entering the Green River Basin was closer to -3.9‰ using 123 oxygen isotopes from the teeth of suspected river-dwelling Eocene mammals and assuming equivalent fractionation values to modern mammals. Our isotope model becomes somewhat unstable at į i >-8‰, requiring extremely large freshwater input volumes to sufficiently freshen the lake. An intermediate į i of -7‰ approaches the volume predictions of the ion model and is presented as a less conservative/intermediate case for the isotope model results. Temperature assumptions. Because of the temperature dependence of both the water- carbonate and water-vapor fractionation factors, the isotope model is also sensitive to temperature assumptions. There are two potentially distinct temperatures of concern: (1) the temperature of the water in which the carbonate of a given layer formed and (2) the surface water temperature from which evaporation occurred. The model was tested using seven distinct potential temperature scenarios summarized in Table 9. Temperatures used in the scenarios were based on clumped isotope temperature measurements for BT08 calcite fan layers 8-8.5 (28°C) and dolomicrite layer 9 (35°C). Fixed temperature scenarios used a single temperature for all carbonate points (T min = 28°C or T max = 35°C). Given the large temperature differences measured for adjacent layers using clumped isotopes, we consider these the least realistic scenarios. The scaled temperature scenarios assign each į 18 O measurement a temperature based on either the sodium content or fraction of magnesium in carbonate, >@ > @> @ Ca Mg Mg XMg , measured for the sample as indicators of evaporation. These scenarios assume that temperature scaled approximately linearly with these evaporation indicators: 124 C XX XX C CT fan micrite fan $ $ $ 28 * ) 28 35 ( (14) where X is the Na or XMg value for the point in question and X fan and X micrite are the average values measured for BT08 calcite fan layers 8-8.5 and dolomicrite layer 9, respectively. The seasonal temperature scenario assigns a temperature to each į 18 O measurement based on its mineralogy (calcite fan vs. micrite), assuming that all calcite fans formed in cooler (28°C) waters while all micrite formed in warm (35°C) waters, as could be the case if carbonate mineralogy was controlled by seasonal lake water temperature swings (and/or concurrent increases in photosynthetically active radiation). The thermocline scenario assumes that micrite formed in a warm (35°C) epilimnion while fans formed in the cooler (28°C) hypolimnion of a thermally stratified lake. The only difference between the seasonal scenario and the thermocline temperature scenario is that in the seasonal scenario surface waters are considered equal to the carbonate formation water values (i.e., a thermally well-mixed lake), while in the thermocline scenario surface waters are equal to the 35°C water temperature in which the micrite formed (which we propose formed in shallow waters). The effect of these different temperature scenarios on model results is summarized in Figure 29 and Figure 30B. At freshwater input oxygen isotope fractionation values close to the į i = -11‰ value used as a conservative lake change estimate based on the previously discussed sensitivity analysis, the different temperature scenarios give similar model results, with the smallest lake volume and depth changes obtained using the fixed temperature scenarios 125 (especially fixed T min ) and greater depth changes with the variable scenarios (especially the thermocline model). Model Results Ion model. The results of the lake volume models are summarized in Figure 31 and Table 10. The sodium ion model suggests dramatic lake volume changes, with a maximum volume calculated for BT08 layer 7.5 of ~440 km 3 and a maximum lake depth at basin center of ~26 m (which is still well within the 30 m maximum lake depth considered reasonable), or a depth change (in this case, evaporation from BT08 layer 7.5 to BT08 layer 9) of ~10 m. Assuming the 0.2 m/km lake slope discussed above, this would imply a shoreline shift during lake highstand of 50 km or more from the Boar’s Tusk site. In the NW direction where the lake was shallowly-sloping (vs. to the N or S where the lake slope would have been greater due to the bordering Wind River and Uinta mountains), this would still be well within the range of recorded synchronous floodplain mudstone and sandstone deposits (Roehler, 1993). Isotope model. Using the oxygen isotope model, the smallest change in lake volume was calculated using the fixed T min = 28°C temperature scenario and a freshwater input value į i = -11‰, which gave a maximum lake volume, again in BT08 layer 7.5, of 135 km 3 . This corresponds to a maximum depth of ~17 m at basin center (equivalent to a 2 m water column depth over the stromatolite bed), and a depth change of ~2 m (between BT08 7.5 to BT08 layer 10) and correlating shoreline shift of ~10 km. This would put the lake highstand well within the range of mapped synchronous mudstone mudflat and dolomite (previously interpreted as forming in a playa-lake system) deposits (Roehler, 1993). For comparison, changing the freshwater input value to į i = -7‰ and using the thermocline temperature scenario gave a depth change of ~5 m 126 for the same layers, or a shoreline shift of ~25 km, which is intermediate to the values calculated using the ion model and still gives a lake highstand within the range of mudflat/playa and mudstone and dolomite deposits mapped by Roehler (1993). As discussed in the previous section, using different assumptions for the freshwater input isotope fractionation value changes these results significantly, with values >-7 " approaching and exceeding the values calculated using the sodium model (Figure 29). Reasons for discrepancies between the isotope and elemental models As discussed above, the isotope model is very sensitive to the assumed freshwater input į 18 O and temperature scenario, and much of the differences observed between the isotope and ion models displayed, e.g., in Figure 31, could be explained by our choice in these parameters. Additional divergence may be explained by a difference in how the ion and isotope models treat chemical differences between adjacent layers. The isotope model requires knowledge of the isotope value and lake volume of the underlying layer. The ion model avoids this by pinning the minimum lake volume to the layer with the maximum sodium concentration and calculating the lake volume for all other layers from all stromatolites based on the difference in concentration between the layer and question and the maximum concentration. Although samples were drilled in close succession in all of the stromatolites, they do not represent a complete record of values; missing a set of laminations that record a particularly low or high lake stand would cause subsequent values to be over- or underestimated. Because of this artifact, we consider the model results for the BT08 (lowermost) stromatolite to be the most accurate. 127 Because the ion model is less sensitive than the isotope models to factors that are unknown in this system, the ion model is used as the baseline for the discussion, with the isotope model results presented as a counterpoint. Micrite and fans formed in different environments: significance for stromatolites as near- shore deposits When results are sorted by microfabric, clear differences are apparent between micrite and fan layers of the stromatolites (Figure 32), with micrite forming at shallower lake levels than fan layers. The hypsometry of the lake system was broad and flat, with a presumed slope of ~0.2 m/km. Thus, a small change in lake level would force a relatively significant lateral migration of the shoreline facies. Thus, the fans (on average) formed at much greater distances from lake shoreline than the micrite. For example, the aforementioned depth changes between layers 7.5 to 10 were between 2m (conservative isotope model), 5m (intermediate isotope model) and 8 m (ion model), corresponding to shoreline shifts of 10, 25, and 40 km, respectively. Even the most conservative estimate shifts the shoreline by 10 km. This observation could have profound implications in the interpretation of stromatolites as a shoreline facies in the Green River Formation and elsewhere: stromatolites may not simply represent one “stromatolite facies”. Rather, the stromatolites may have experienced many different environments during the course of their formation. Are such extreme shoreline shifts geologically reasonable? The Great Salt Lake basin is similarly broad and flat, and presents a reasonable modern analogue for the hypsometry of Lake Gosiute. For example, in the early 1980’s, the lake level of the Great Salt Lake rose by 1.6 m over a period of 9 months (September 1982-June 1983) and continued to rise an additional 2.1 m 128 to a highstand in June 1986 (Figure 33. This depth change resulted in a >30% increase in lake surface area and shoreline shift in excess of 30 km in some locations (Figure 34 2 ), causing destructive flooding, which resulted in hundreds of millions of dollars in damage to developed areas in the lake basin (e.g. Arnow and Stephens, 1990; Atwood, 1994). Fluctuations in lake extent were even larger in the prehistoric past when the lake was deep enough to spill into what are now the salt flats (e.g. the Bonneville Salt Flats) surrounding the greater lake basin, with shoreline shifts over 150 km. Thus, it is not surprising that Lake Gosiute exhibited similarly large shoreline shifts. Shifting shorelines and climate Furthermore, the significant volume changes over the formation time of the stromatolite suggest that a dynamic climate may have been present during the Eocene Climatic Optimum. To be clear, we do not know the timing of stromatolite formation, but from the lamination counts and evaporation rate estimates discussed above, as well as from analogy with stromatolites forming in similar systems in the modern (e.g., Golubi ü and Focke, 1978; Berelson et al., 2011; Petryshyn et al., 2012), it is likely to be on the order of days to years per lamination. Thus, the climate and the hydography of the basin appear to have varied significantly on the scale of years to hundreds of years, adding to our understanding of one of the hottest times in the past 60 million years. 2 See also the following animation produced by the Utah Reclamation, Mitigation, and Conservation Commission showing the expansion and contraction of the Great Salt Lake over a 20-year period as viewed by satellite imagery: http://ut.water.usgs.gov/greatsaltlake/elevations/elevationmovie.html (accessed July 27, 2013). 129 Stromatolite growth rate constraints We cannot measure the growth rate of the Rife Bed stromatolites directly, but we can make certain assumptions about evaporation rates that allow us to estimate growth rates of certain stromatolite microfabrics. For example, using the evaporation sequence recorded in BT08 calcite fan layers 7.5 to 8.5 (1 cm), the models predict a depth decrease of 1-2 m. Assuming an evaporation rate of 0.5-1.5 m/year (cf. Morrill et al., 2001) gives a growth period of ~0.5-4 years for the precipitation of the fan. This section of stromatolites is 1 cm thick, giving a stromatolite growth rate of ~0.5-1 cm/year. Lamination counts revealed ~50-150 microlaminae, giving a lamination frequency of ~10-300 laminations per year (depending on the evaporation rate assumed). A similar fan sequence in BT12-CF-1 (layers 2-4, ~70 microlaminae) gave an evaporation period of <3 years and stromatolite growth rates on the order of 1 cm/year or a lamination frequency of >20 laminations per year. These growth rates are significantly faster than rates previously reported for stromatolites from Holocene alkaline lake systems (e.g., Petryshyn et al., 2012) but on the same order as rates reported for microbialites found in the Laney Member of the Green River Formation that nucleated on caddisfly casings (Leggitt and Cushman, 2001). Also, unless evaporation rates were much lower than the range we suggest here, this implies that the microlaminations that compose the calcite fan layers occur multiple (or many) times a year and are not annual or longer-term growth laminations. One potential mechanism could be occasional storm events that intensified weathering in the catchment basin and contributed to pulses of Ca 2+ to the system. The large changes in isotope and elemental composition between certain layers/microfabrics (e.g., stacked layers 8.5 and 9) raise a different issue with growth rates. By 130 the same logic (0.5-1.0 m of evaporation potential per year), it would take 8 to 16 years to transition between fan layer 8.5 and micrite layer 9. Thus, the stromatolite did not grow at one continuous rate during its formation. Did the micrite simply grow very slowly vs. the underlying fan, or did it cease growth during the transition? The answer is unclear, but it seems geologically reasonable that fan-forming conditions ceased before micrite-forming conditions restarted, perhaps due to changes in chemistry, or perhaps even changes in the energy of the formation environment (and associated magnitude of detrital input). Lake water temperatures: evidence for thermal stratification? Clumped isotope analysis from BT08 stromatolite fan layers 8-8.5 and adjacent micrite layer 9 reveal a ~6°C difference in water temperatures. The temperature shifts are accompanied by the previously discussed dramatic shifts in elemental composition, conserved ion concentration, and stable isotope fractionation values, which are consistent with a major evaporation event. The fan layers 8-8.5, with their clumped isotope temperature of ~28°C, formed in deeper waters. The micrite layer 9 formed in shallower, potentially near-surface waters that according to the clumped isotope paleothermometer were quite warm: ~35°C (Figure 35). This shift could be explained by: (1) lake water temperature changes driven by seasonal variability (with fans forming in cooler months and micrite forming in warmer months), (2) lake water temperature changes driven by longer-term air temperature variability, or (3) the existence of a thermocline in Lake Gosiute, with micrite forming in the epilimnion and fans forming in the hypolimnion of a stratified lake. Although high altitude lake water temperatures can display large seasonal variability (Figure 36B), we reject the seasonal temperature hypothesis based on the clear geochemical 131 signals for evaporation amounts that are unlikely to have occurred over the course of one season, as discussed in the previous section. The ion model predicts a net loss (by evaporation) of 8 m of water between the adjacent fan and micrite layers from which the clumped isotope temperatures were measured. This would require extremely high rates of evaporation (and extremely low precipitation) to occur over one season. In addition, assemblages of plant and animal fossils from the Green River Formation and elsewhere suggest that the Early Eocene Climatic Optimum was a time of relatively low seasonal temperature variability (Wing and Greenwood, 1993), and the presence of the large lake system would have further reduced seasonal temperature swings in the region (Sloan, 1994). We also argue that the changes in mean annual temperatures that would be required to produce the lake water temperature variations seen over longer periods are too large to be likely. Lake water temperatures during carbonate formation do not directly reflect air temperatures. Modern relationships between seasonal lake water temperatures and mean annual air temperatures compiled by Hren and Sheldon (2012) would suggest that if micrite was forming in shallow surface waters, MAAT would have been ~25-28°C (depending on the season during which micrite layers were actively formed), on the higher range of what paleobotanical evidence predicts (Wilf, 2000). Likewise, if the adjacent calcite fans formed in shallow surface waters (which our models suggest was not the case) or in thermally well-mixed waters (i.e., in the absence of thermal stratification), MAAT at the time of their formation would have been ~19-24°C (again, depending on the assumed active growth season). This could suggest a significantly higher MAAT during formation of the micrite layer compared with the fan layer directly beneath it, implying dramatic MAAT variability during the Eocene, which is in stark 132 contrast to the equable climate during the Eocene predicted by the paleobotanical and modeling studies discussed above. Alternately, the presence of thermal stratification in the lake (or the lack of complete thermal mixing) would completely explain the temperature differences without requiring MAAT variation. Saline lakes can become chemically and thermally stratified due to the lack of mixing between overlying freshwater and underlying saltwater. The large, shallow nature of Lake Gosiute would have made the formation of a permanent thermocline unlikely; wave ripples that formed at the bottom of the lake during deposition of the Laney Member (a period when the lake may have been deeper than during formation of the Rife Bed of this study) suggest that the lake was at least occasionally well-mixed during storms and strong winds not unlike those that occur in the modern location of the Green River Formation (Stanley and Surdam, 1978). However, intermittent stratification is likely to have occurred due to the saline nature of Lake Gosiute, especially during periods of greater lake depth. Indeed, there is evidence that oil shale deposition in the Green River Formation occurred when the lake was stratified (Bradley, 1929; Desborough, 1978; Boyer, 1982). Due to these considerations, we consider the existence of a thermocline a likely scenario that would explain the temperature variability observed between adjacent fan and micrite fabrics if the ion model depths or the less-conservative isotope models are used. However, if the most conservative assumptions for the isotope model are used, the minimal depth difference calculated between the layers where clumped isotope measurements were done would argue against the presence of a thermocline. 133 Hypothesis: lake level impacts biology, biology (or lack thereof) determines microfabric Results of both the ion and isotope models combined with the clumped isotope measurements for carbonate formation suggest that micrite formed in warm, shallow waters while fans formed in cooler, deeper waters (Figures 35 and 37). The micritic layers appear to be biogenic in origin. According to the ion model, micrite did not form at calculated lake depths deeper than 21 m at basin center (Figure 32), which corresponds to a water depth of 6 m at the location of the stromatolites. Fans, meanwhile, formed abiogenically under a water column depth at the location of the stromatolites of >4 m (most at depths >7 m). Is there something about these several meters of difference that could control the combination of microbial community and water chemistry responsible for the formation of these stromatolites? If the micritic layers are formed by photosynthetic communities (e.g. cyanobacteria, algae, or the mixed communities responsible for the building of most modern stromatolites), these communities could be absent at deeper water levels if there is insufficient light penetrating at depth to support photosynthesis. Light penetration in modern lakes is extremely variable and depends on such factors as concentrations of light-absorbing minerals and humic material, water turbidity, and especially the presence of pigmented microorganisms (algae and other phototrophs; e.g. Cristofor et al., 1994). In the Great Salt Lake, the Secchi Depth (a measure of water clarity—and by extension, light penetration) of lake water is highly variable, from <0.5 m year-round in a highly productive eutrophic subbasin (Farmington Bay) and varies in the greater basin from <0.5 m in the fall up to ~7 m in summer months when algal populations are reduced by brine shrimp grazing (Figure 36A, Wurtsbaugh et al., 2012). 134 Although the extensive oil shale deposits in the Green River Formation indicate that the lake was at least seasonally very productive (which would reduce the depth of light penetration), there is little evidence of high productivity in the Rife Bed stromatolites. The Rife Bed stromatolites are relatively poor in organic material versus the oil shale present in the Green River Formation. Thus, light may have penetrated well beyond the depths at which the apparently biogenic micrite ceased forming in the Rife Bed stromatolites. In addition, the grains trapped beyond the angle of repose that were observed in the Boar’s Tusk stromatolites did not exceed 200 !m (e.g. Figure 13B). In reference to the previous chapter of this thesis discussing trapping and binding of grains by microbial mats, this would be consistent with the presence of a microbial mat capable of binding grains but does not require the presence of long filament bundles; therefore, the mats responsible for micrite precipitation were not necessarily photosynthetic (although they could have been). An alternate hypothesis is that lake chemistry influenced the [micro]biology of the lake, which in turn impacted stromatolite microfabric. The climate-driven variations in lake level observed in the Great Salt Lake are associated with major shifts in the biology of the lake due to changes in salinity, with periods of low salinity (and high lake level) favoring cyanobacteria (vs. more halophilic organisms), which have been associated with the formation of carbonate bioherms in the lake (e.g. Arnow and Stephens, 1990). In addition, as discussed previously, we consider it likely that the lake was at least intermittently thermally (and chemically) stratified. Stratification may have resulted in conditions (e.g. anoxia) in the hypolimnion (i.e. at depth in the water column) unfavorable to the growth of the organisms responsible for forming the micrite in these stromatolites. 135 CONCLUSIONS This study yielded novel information about Lake Gosiute during a critical period in its development. For the first time, absolute values for lake water temperatures have been measured for the Green River Formation using the clumped isotope paleothermometer: 28.5°C during the formation of calcite fans in a deeper lake and 35°C during the formation of dolomicrite in a shallower lake. There is suggestive evidence from this study to support the idea first set forth by Bradley (1929) that Lake Gosiute was at least occasionally thermally stratified (which, as Boyer, 1982, argued, does not invalidate the well-established model of the lake as a playa-lake system). The position of the stromatolites within these stratified layers would likely have influenced their chemistry, biology, and mineralogy. Another important conclusion drawn from the results of this study is the significant changes in lake volume, depth, and extent—as well as lake chemistry—that occurred over the scale of stromatolite growth. The results suggest Lake Gosiute was highly sensitive to changes in the balance of evaporation to precipitation on the timescale of stromatolite growth, much like analogous modern systems such as the Great Salt Lake. Stromatolites are typically thought to represent a near-shore environment, however, our model results suggest active stromatolite growth in the calcite fan layers in deep waters while several kilometers from shore. Thus, the traditional view of stromatolites representing a single depositional environment needs 136 reconsideration, especially in lake systems where vast shoreline changes are possible over a short period of time. Finally, we have demonstrated the utility of stromatolites, as finely laminated structures that record the chemistry of the waters in which they formed, for paleoclimate reconstructions, particularly in systems where post-depositional alteration has been minimal. ACKNOWLEDGEMENTS I acknowledge the support of Lowell Stott and Miguel Rincon at the University of Southern California for assistance with the standard stable isotope measurements. Clumped isotope measurements were performed by Victoria Petryshyn in the lab of Aradhna Tripati at UCLA. Elemental analyses were done in collaboration with Pedro Marenco at Bryn Mawr College. X-ray diffraction analyses were done with the assistance of Anthony Kampf at the Natural History Museum of Los Angeles. Deep-UV native fluorescence spectroscopy was done in collaboration with Rohit Bhartia (NASA JPL) and Everett Salas (Photon Systems). This study began as a project in the 2008 International Geobiology Course, and additional data and stromatolite samples came from the 2012 and 2013 Geobiology Course groups, which received support from the USC Wrigley Institute for Environmental Studies and the Colorado School of Mines as well as grants from the Agouron Institute, the Gordon and Betty Moore Foundation, the NASA Astrobiology Institute, and the National Science Foundation. Will Berelson is acknowledged for his always-insightful comments and demands that greatly improved this work. In addition, I am grateful for the thorough reviews of a previous version of this manuscript by Kathleen Benison and Linda Kah. Frank Corsetti read and provided detailed feedback on at least 137 a hundred versions of this chapter over the course of five years and is largely responsible for preventing my throwing my laptop and all of the sample tubes (but not the stromatolites—I always loves the stromatolites!) into the nearest fountain when the going was rough. 138 FIGURES Figure 1. Cenozoic CO 2 (top, red) and temperature (bottom, green) estimates. The time period represented by Green River Formation deposits are highlighted in green, with the Early Eocene Climatic Optimum (EECO) highlighted. The upper CO 2 plot is modified from Beerling and Royer (2011) and depicts the range of CO 2 values allowed by several different proxies. The lower temperature plot is modified from Zachos et al. (2008). 139 ! Figure 1 1500 2000 1000 500 0 Eocene Green River Formation Paleocene Oligocene Miocene Plio Atmospheric CO 2 (ppm) Nacolite Trona 0 1 2 3 4 5 4 0 8 12 Antarctic ice sheets EECO N. Hem. ice sheets ? ? Ice-free temperature (°C) Age (millions of years ago) į 18 O (‰) 0 20 10 30 40 50 60 140 ! Figure 2. A The Rife Bed stromatolite bench (highlighted with white arrows) at the Boar’s Tusk locality with the 2013 International Geobiology Course participants providing scale and the Boar's Tusk volcanic tower for which the outcrop is named in the left background. Photo credit: Ann Close. B Rife bed stromatolite head at the outcrop with the distinctive cm-scale layering visible. A B 141 ! Figure 3. Map of stromatolite sampling locations. A Satellite image of the Boar’s Tusk outcrop with the locations where stromatolites were collected mapped as blue points. The orientation of North is the same for all maps. B Topographic map of the Boar’s Tusk outcrop showing the locations where stromatolites were collected. C Contextual map of Southwestern Wyoming showing the location of the Boar’s Tusk outcrop. Satellite image from Google Earth, maps from Google Maps. 50m BT08 BT12-CF-4 BT12-CF-1&2 20 km Boar’s Tusk Outcrop Rock Springs Uplift Rock Springs WY UT CO 500m to the Boar’s Tusk White Mountain North Sampling Sites 3k to Rock Springs A B C N 142 ! Figure 4. A Map showing the location of the Boar’s Tusk outcrop in the context of the mapped extent of the Rife Bed (Tipton Shale Member, Green River Formation) and surrounding depositional environments at the time when the stromatolites in this study were forming near the shoreline of Paleolake Gosiute. Modified from Roehler, 1993.B Stratigraphic representation of the Boar's Tusk Outcrop showing the location of the stromatolite bed (marked with *) and the transitioning of paleolake Gosiute from overfilled to balanced-filled to underfilled and back (Roehler, 1991; Pietras and Carroll, 2006). 43º 110º 108º Extent of Rife Bed (Tipton Shale Member) WYOMING COLORADO UTAH Lacustrine oil shale Pre-Cretaceous basement Mudflat mudstone Floodplain mud- & sandstone Shoreline sandstone Pediment arkose Stromatolites Rock Springs Uplift 0 40 80 km 41º A B 30 m Scheggs Bed Farson Sandstone Rife Bed Wilkins Peak Member Tipton Member Green River Formation Laney Member saline lacustrine saline lacustrine fresh lacustrine deltaic fresh lacustrine saline lacustrine and mudflat (playa) Underfilled Balanced Overfilled 109º 42º 107º * Boar’s Tusk 143 ! Figure 5. The Boar’s Tusk outcrop showing the exposed Tipton Shale Member, the bounding stromatolite horizon in the Rife Bed, and the Wilkins Peak Member. Tipton Member Wilkins Peak Member Farson Sandstone stromatolite bed Rife Bed 144 Figure 6. A BT08 stromatolite showing face drilled for elements and stable isotopes. B Map of microdrilled locations (outlined numbers correspond to sample numbers). Layers are outlined and labeled with their number and fabric: fan (f), mixed (x) or micrite (m). Holes not labeled in this image were drilled previously for other studies. 145 ! Figure 6 A 1 1 2 2 3 3 57 57 56 56 55 55 54 54 48 48 44 44 42 42 41 41 40 40 39 39 37 37 36 36 35 35 34 34 33 33 32 32 31 31 30 30 29 29 28 28 27 27 25 25 26 26 24 24 22 22 21 21 20 20 19 19 18 18 17 17 16 16 15 15 14 14 13 13 12 12 11 11 10 10 9 9 8 8 7 7 6 6 5 5 4 4 10 (x) 9 (m) 8.5 (f) 8 (f) 7.5 (f) 7 (f) 6 (x) 5 (m) 4 (f) 3 (x) 2 (x) 1 (m) B 1cm 38 38 146 Figure 7. A Closeup of the BT08 stromatolite after microdrilling for the elemental analysis with a few drill sites labeled with their corresponding sample numbers. B Same area after re-drilling for stable isotope analysis with the same locations labeled. Stable isotope drill holes approximately correspond to the elemental holes. 147 ! Figure 7 1cm 39 38 8 4 2 16 39 38 8 4 2 16 A B 148 Figure 8. A Map of microdrilled locations (outlined numbers correspond to sample numbers) from the BT12-CF-1 stromatolite. B Microdrill site map for the BT12-CF-2 stromatolite. Both stromatolites are shown to the same scale. 149 ! Figure 8 11 11 1 1 2 2 3 3 4 4 5 5 6 6 7 7 9 9 8 8 10 10 13 13 14 14 15 15 16 16 17 17 18 18 19 19 21 21 20 20 22 22 23 23 24 24 25 25 26 26 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 22 22 12 12 2 cm A. BT12-CF-1 B. BT12-CF-2 1 1 3 3 2 2 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 150 Figure 9. Stromatolites BT12-CF-4b (top) and BT12-CF-4a (bottom) showing faces drilled for elements and stable isotopes with microdrilled paths labeled (outlined numbers). An ooid layer that was microdrilled (samples 22-24) crumbled away before the stromatolite could be scanned. 151 ! Figure 9 19 19 2 cm 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 16 16 17 17 18 18 14 14 15 15 25 25 22-24 22-24 20 20 21 21 Ooid layer (missing) 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 35 35 34 34 36 36 152 ! Figure 10. Drill holes for the BTM2 (micrite) and BTF2 (fan) clumped isotope runs. BTM1 and BTF1 were drilled on the same layers but no images were taken. 200µm BTM2 BTF2 153 ! Figure 11. Thin section image of a segment of a Rife Bed stromatolite showing the digitate columns and the three dominant microfabrics: calcite fans (Fan), micrite (Micrite), and mixed (Mixed) under A transmitted and B cross-polarized light. Micrite Mixed Fan 2.0 mm A B 154 ! Figure 12. A Photomicrograph of an unstained Boar’s Tusk stromatolite thin section under transmitted light. B Alizarin Red S stain of the same section showing differential staining of calcite fans (stained) and dolomicrite (unstained). mixed micrite mixed fan mixed micrite mixed fan A B 155 ! Figure 13. A Photomicrograph of a calcite fan layer showing sharp crystal termination boundaries.B Photomicrograph of a micrite layer showing hanging grains. 500µm B 50µm A 156 ! Figure 14. Cathodoluminescence (left) of a section of the BT08 stromatolite (shown in plain light at right) showing relative luminescence of calcite fans, micrite, and cement. cement calcite fan micrite 157 ! Figure 15. Molar fraction of magnesium in carbonate (XMg) plotted by fabric for the BT08 stromatolite (BT08) and for all points for the BT12-CF-1, BT12-CF-2, BT12-CF-4, and BT08 stromatolites (All BT Stroms). Boxes denote the 25-75 percentile range, red bars denote the median value, whiskers denote the full range of values not considered outliers, and red + denote statistical outliers. The dotted line at XMg=0.5 denotes the value above which the mineralogy is classified as dolomite. 0 0.2 0.4 0.6 0.8 1.0 fan mixed micrite BT08 XMg fan mixed micrite ooid All BT Stroms dolomite calcite 158 Figure 16. Correlation of stromatolite laminae across a 200 m section of outcrop. A-C stromatolites collected from different locations in the Rife Bed of the Boar's Tusk. All stromatolites are shown to the same scale. D Satellite image (Google Maps) showing collection sites with arrows pointing to the stromatolites collected at each site. 159 ! Figure 16 100 m 5 cm A B C D N 160 ! Figure 17. Correlation of laminations in the A BT12-CF-1 and B BT12-CF-2 stromatolites. Arrows highlight particularly distinctive laminae. A BT12-CF-1 B BT12-CF-2 1 cm 161 Figure 18. Stromatolites drilled for chemical analyses arranged in stratigraphic order with distinct layering patterns used to arrange them shown on a larger stromatolite collected subsequently. All stromatolites are shown at the same scale. 162 ! Figure 18 5 cm BT12-CF-4b BT12-CF-4a BT12-CF-1 BT08 BT13-5-A&B 163 Figure 19. Chemical results for the BT08 stromatolite samples plotted by layer. Red points are points that were considered outliers for the purpose of subsequent analyses. Blue lines show the outlier-excluded average for each layer with error bars representing standard deviations for each layer. 164 ! Figure 19 0 0.5 1 1 2 3 4 5 6 7 8 9 10 Na (%) Layer 0 20 40 1 2 3 4 5 6 7 8 9 10 Mg (%) 20 40 60 1 2 3 4 5 6 7 8 9 10 Ca (%) 0 0.2 0.4 1 2 3 4 5 6 7 8 9 10 Al (%) Layer 0 0.2 0.4 1 2 3 4 5 6 7 8 9 10 Mn (%) 0 5 10 1 2 3 4 5 6 7 8 9 10 Fe (%) 50 100 150 1 2 3 4 5 6 7 8 9 10 Ni (ppm) Layer 0 50 100 1 2 3 4 5 6 7 8 9 10 Cu (ppm) 0 100 200 1 2 3 4 5 6 7 8 9 10 Zn (ppm) 0 0.5 1 1 2 3 4 5 6 7 8 9 10 Sr (%) Layer 0 0.2 0.4 1 2 3 4 5 6 7 8 9 10 Ba (%) 0 20 40 1 2 3 4 5 6 7 8 9 10 U (ppm) 0 0.5 1 1 2 3 4 5 6 7 8 9 10 XMg (mol Mg/Mg+Ca) Layer -5 0 5 1 2 3 4 5 6 7 8 9 10 į 13 C (‰ VPDB) -10 -5 0 1 2 3 4 5 6 7 8 9 10 į 18 O (‰ VPDB) 165 Figure 20. Chemical results for the BT12-CF-1 (black points) and BT12-CF-2 (green x) stromatolite samples plotted by layer. Red points and x markers are points (from BT12-CF-1 and BT12-CF-2, respectively) that were considered outliers for the purpose of subsequent analyses. Blue lines show the outlier-excluded average of BT12-CF-1 samples for each layer with error bars representing standard deviations for each layer. 166 ! Figure 20 0 1 2 0 20 40 20 40 60 -0.5 0 0.5 0 0.2 0.4 0 5 50 100 150 0 20 40 0 20 40 0.5 1 1.5 0 0.5 1 0 20 40 0 0.5 0 2 4 -10 -5 0 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 -1 0 1 2 3 4 5 6 7 8 9 10 0.25 2.5 Na (%) Layer Mg (%) Ca (%) Al (%) Layer Mn (%) Fe (%) Ni (ppm) Layer C u (ppm) Zn (ppm) Sr (%) Layer Ba (%) U (ppm) XMg (mol Mg/Mg+C a) Layer ! 13 C (‰ V P DB) ! 18 O (‰ V P DB) 167 Figure 21. Chemical results for the BT12-CF-4 stromatolite samples plotted by layer. Red points are points that were considered outliers for the purpose of subsequent analyses, with the red points in Layer 6 from one sample that was excluded on the basis that it was an outlier in many measurements. Blue lines show the outlier-excluded average for each layer with error bars representing standard deviations for each layer. Layer 8, which shows high variability in most measurements, was composed of ooids. 168 ! Figure 21 0 0.5 1 0 20 40 0 50 100 0 0.5 0 0.2 0.4 0 2 4 0 50 100 0 100 200 0 20 40 0 0.5 1 0 0.2 0.4 0 10 20 0 0.5 1 0 5 -10 -5 0 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Na (%) Layer Mg (%) Ca (%) Al (%) Layer Mn (%) Fe (%) Ni (ppm) Layer C u (ppm) Zn (ppm) Sr (%) Layer Ba (%) U (ppm) XMg (mol Mg/Mg+C a) Layer ! 13 C (‰ V P DB) ! 18 O (‰ V P DB) 169 ! Figure 22. Correlation of oxygen and carbon stable isotopes with best fit lines through the data for the Boar's Tusk stromatolites BT08 (red), BT12-CF-1 (BT1, dark green), BT12-CF-2 (BT2, light green), and BT12-CF-4 (BT4, blue). The best fit line for all Boar’s Tusk data points (BT all) is shown in black. í í í í í í í 0 0.0 &DUERQDWHį O (‰ VPDB) &DUERQDWHį C (‰ VPDB) All BT \ [5 %7 \ [5 %7&) \ [5 %7&) \ [5 %7&) \ [5 170 ! Figure 23. Comparison of clumped isotope results with results from standard stable isotope measurements from the BT08 stromatolite. fs: fan layer 8-8.5, standard isotopes; fc: fan layer 8- 8.5, clumped isotopes; ms: micrite layer 9, standard isotopes; mc: micrite layer 9, clumped isotopes. Boxes denote the 25-75 percentile range, red bars denote the median value, whiskers denote the full range of values not considered outliers, and red + denote outlier points. 1.5 2.0 2.5 3.0 3.5 4.0 4.5 A į 13 C į 13 C (‰ VPDB) į 18 O (‰ VPDB) std. clump -6.0 -5.0 -4.0 -3.0 -2.0 -1.0 B į 18 O std. clump fans micrite fans micrite std. clump std. clump 171 ! Figure 24. Results of deep-UV native fluorescence spectroscopy. A False-color image depicting results of the six-channel UV fluorescence detection scan of a Boar’s Tusk stromatolite with key features highlighted: “pink spots”, locations that had anomalously high intensities in the 340- 380nm range; “rim”, locations around the outer rim of the stromatolite; and micrite and fans, carbonate material of a given microfabric. B Map of the locations of the “pink spots” visible in A on a photomicrograph mosaic of the scanned stromatolite. C Normalized spectral signatures of the features noted in A. Error bars represent the standard deviation of measurements of multiple sites where the feature was noted. D Normalized deep-UV native fluorescence spectroscopic signatures of several classes of organic compounds (Rohit Bhartia, unpublished data). 1 cm 280 300 320 340 360 380 Wavelength (nm) pink spots rim micrite fans pink spots rim micrite fans A B C 280 300 320 340 360 380 Wavelength (nm) 260 D 172 Figure 25. Crossplots of all elements measured vs. į 18 O for all stromatolites plotted by fabric type: fans (blue), mixed (purple), micrite (red), and ooids (green). Crossplots showing sodium and magnesium, both tracers of evaporation in closed systems, are highlighted. 173 ! Figure 25 -8 -6 -4 -2 0 0.0 0.2 0.4 0.6 0.8 Na (%) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) į 18 O (‰ VPDB) -8 -6 -4 -2 0 0.0 0.1 0.2 0.3 0.4 Mn (%) -8 -6 -4 -2 0 0 20 40 60 80 Zn (ppm) -8 -6 -4 -2 0 0.2 0.4 0.6 0.8 1.0 Sr (%) -8 -6 -4 -2 0 0 10 20 30 40 Mg (%) -8 -6 -4 -2 0 0 2 4 6 8 10 Fe (%) -8 -6 -4 -2 0 0.1 0.2 0.3 0.4 0.5 Ba (%) -8 -6 -4 -2 0 10 20 30 40 50 60 Ca (%) -8 -6 -4 -2 0 0.0 0.1 0.2 0.3 0.4 0.5 Al (%) -8 -6 -4 -2 0 0 10 20 30 U (ppm) -8 -6 -4 -2 0 20 40 60 80 100 120 Ni (ppm) -8 -6 -4 -2 0 0 50 100 150 Cu (ppm) fans mixed micrite ooids 174 ! Figure 26. Crossplot of ɷ 18 O and sodium measurements from the BT08, BT12-CF-1, and BT12- CF-4 stromatolites with measurements from different microfabric types (fans = blue, mixed = purple, micrite = red, ooids = green points) highlighted. 0.2 0.4 0.6 0.8 -7 -6 -5 -4 -3 -2 -1 0 Na (%) į 18 O (‰ V PDB ) fans micrite HY DSRU D WLRQ fans mixed micrite ooids 175 ! Figure 27. Covariant carbon and oxygen stable isotopes in carbonates from modern closed lakes (1-6) compared with the Rife Bed stromatolites of this study (7, red points and trendline). r=regression coefficient; n=number of samples. Plot modified from Talbot 1990. 7 Closed lakes 1. Turkana (r=0.86, n=34) 2. Great Salt Lake (r=0.87, n=27) 3. Van (r=0.81, n=58) 4. Natron-Magadi (r=0.84, n=19) 5. Bosumtwi (r=0.97, n=12) 6. Rukwa (r=0.95, n=4) 7. Boar’s Tusk (r=0.94, n=112) į 18 O į 13 C 10.0 5.0 5.0 10.0 -5.0 -10.0 -5.0 -10.0 1 2 3 4 5 6 176 ! Figure 28. Water temperature change required to produce the carbonate į 18 O shift from -5.5 to - 3.8‰ observed from BT12-CF-4 calcite fan layers 10-12 as a function of the starting water temperature (temperature during formation of layer 10 at -5.5‰) ignoring all other sources of potential water į 18 O change. 10 15 20 25 30 35 40 -9.5 -9.0 -8.5 -8.0 -7.5 -7.0 T start ǻ7 UHTXLUHGWRSURGXFHį 18 2VKLIW 177 Figure 29. Maximum lake depth (at basin center) calculated using the oxygen isotope model showing the model's sensitivity to both the assumed freshwater input value į i and the temperatures assumed for lake surface water and the waters from which carbonates formed. A Results calculated for the BT08 stromatolite only. B Results calculated with the BT08, BT12- CF-1, and BT12-CF-4 stromatolites treated as a single continuous (stacked) stromatolite. The gray bar shows the maximum lake depth (with a 1 standard deviation range) calculated using the sodium ion model. The minimum lake depth assumed in both cases was 15 m, so maximum depth changes calculated for each set of parameters is equal to the maximum calculated lake depth minus 15 m. 178 ! Figure 29 -20 -18 -16 -14 -12 -10 -8 -6 -4 16 18 20 22 24 26 28 30 max calculated lake depth (m) į i (‰ VSMOW) ion model fixed Tmin fixed Tmax fixed Tmid scaled TNa scaled TXMg seasonal thermocline Temperature Scenario -20 -18 -16 -14 -12 -10 -8 -6 -4 16 18 20 22 24 26 28 30 max calculated lake depth (m) į i (‰ VSMOW) ion model A BT08 B All BT 179 Figure 30. Isotope model of lake volume and depth change for the Boar’s Tusk stromatolites showing the model’s sensitivity to key assumptions. A Model results using different assumptions for freshwater input į i using the T min temperature scenario (carbonate formation and surface water temperatures = 28°C for all points). The range of possible į i values are plotted from -20‰ (dark blue) to -3‰ (red) with the value giving the lowest volume and depth changes ( į i =-11‰) highlighted in black. For the sake of readability, error bars are not shown. B Model results for the seven different temperature scenarios assuming a freshwater input į i = -11‰. Error bars represent the standard deviation for volumes and depths calculated for replicate data points. In both plots, the calculated lake depth at basin center is presented. 180 ! Figure 30 -20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 į i assumed (‰) Calculated lake volume (km 3 ) Stacked layer 0 5 10 15 20 25 30 35 Calculated lake depth (m) 100 200 300 400 500 16 20 22 18 24 26 28 30 fixed T min fixed T max fixed T mid scaled T Na scaled T XMg seasonal thermocline Calculated lake volume (km 3 ) Stacked layer 0 5 10 15 20 25 30 35 Calculated lake depth (m) 100 120 140 160 180 14 15 16 17 18 19 20 Temperature scenario A Isotope model results using different assumptions for į i , T min scenario B Isotope model results for different temperature scenarios, į i = -11‰ 181 ! Figure 31. Comparison of the ion (red), conservative isotope (blue; uses T min temperature scenario and į i = -11‰) and intermediate isotope (green; uses thermocline temperature scenario and į i = -7‰) model results for the Boar’s Tusk stromatolites BT08, BT12-CF-1, and BT12-CF- 4 (stacked in that order). Error bars represent the standard deviation for volumes and depths calculated for replicate data points. Both the calculated lake depth at basin center and the corresponding depth of the stromatolites in the water column are presented. 14 16 18 20 22 24 26 28 0246 8 10 12 Calculated lake depth (m) Stromatolite depth (m) 100 200 300 400 500 0 5 10 15 20 25 30 35 Calculated lake volume (km 3 ) Stacked layer Isotope model T min į i = -11‰ ,VRWRSHPRGHOWKHUPRFOLQHį i = -7‰ Ion model 182 ! Figure 32. Lake depth calculations sorted by microfabric for BT08 only (top) and for the combined data for BT08, BT12-CF-1, and BT12-CF-2 (bottom) showing results for the ion model (left), the conservative oxygen isotope model using the fixed T min with į i = -11‰ (center), and the intermediate isotope model using the thermocline temperature scenario with į i = -7‰ (right). Calculated depth (m) Calculated depth (m) fans micrite Ion Model fans micrite Fabric Intermediate Isotope Model fans micrite fans micrite fans micrite fans micrite Conservative Isotope Model A BT08 B All BT Stromatolites 14 16 18 20 22 24 26 14 16 18 20 22 24 26 Ion Model Fabric Intermediate Isotope Model Conservative Isotope Model 183 ! Figure 33. Historic lake level changes in the Great Salt Lake, Utah. Data from the U.S. Geological Survey. Great Salt Lake 1278 1280 1282 1284 1286 0 1 2 3 4 5 6 7 8 9 recent pre-historic highstand ca. 1700 AD historic highstand historic lowstand Calendar Year Lake Level Elevation (m) Relative Lake Level Change (m) 1840 1860 1880 1900 1920 1940 1960 1980 2000 2020 184 ! Figure 34. Schematic of Great Salt Lake shoreline shifts showing lake levels plotted on an elevation profile of the Great Salt Lake Basin. 1280 1282 1284 Elevation (m) 1286 1276 1278 1288 1290 J une 1986: 1283.7 m (rec ent highs tand) s horeline 0 km 25 km 50 km 75 km 100 km 125 km 150 km 175 km 1280 1282 1284 Elevation (m) 1286 1276 1278 1288 1290 S ept.1982: 1280.0 m s horeline 0 km 25 km 50 km 75 km 100 km 125 km 150 km 175 km 1280 1282 1284 Elevation (m) 1286 1276 1278 1288 1290 J une 1983: 1281.6 m s horeline 0 km 25 km 50 km 75 km 100 km 125 km 150 km 175 km 1280 1282 1284 Elevation (m) 1286 1276 1278 1288 1290 J une 1963: 1277.5 m (rec ent lows tand) s horeline 0 km 25 km 50 km 75 km 100 km 125 km 150 km 175 km 185 Figure 35. Comparison of the ion (red), conservative isotope (blue; uses T min temperature scenario and į i = -11‰) and intermediate isotope (green; uses thermocline temperature scenario and į i = -7‰) model results by stacked layer shown with their corresponding microfabrics and the clumped isotope temperatures measured for layers 8-8.5 and 9. 186 ! Figure 35 31 32 33 34 Layer Fabric Fan Fan Fan Fan 21 22 26 23 28 29 30 27 24 25 Ooid Fan Micrite Fan Mixed Mixed Micrite Fan Micrite Fan 11 12 16 13 18 19 20 17 14 15 Fan Micrite Fan Fan Fan Fan Fan Fan Fan Fan 1 2 6 3 8 9 10 7 4 5 Mixed Micrite Fan Fan Mixed Micrite Fan Mixed Mixed Micrite Ion model Isotope model 14 16 18 20 22 24 26 28 0246 8 10 12 Calculated lake depth (m) Stromatolite depth (m) 35.0±0.4°C 28.5±1.0°C 187 ! Figure 36. A Light penetration and B water temperature at two different sites in the Great Salt Lake: Farmington Bay, a restricted, eutrophic sub-basin that comprises the SE arm of the lake; and Gilbert Bay, the larger, more saline portion of the southern half of the lake. Figure modified from Wurtsbaugh et al., 2012. Light Penetration Great Salt Lake 1 2 3 4 5 6 7 Farmington Bay Gilbert Bay Month (2009) Secchi Depth (m) Temperature C April May June July Aug. Sept. Oct. Temperature Great Salt Lake 5 10 15 20 25 30 Month (2009) April May June July Aug. Sept. Oct. A B 188 ! Figure 37. Cartoon illustrating formation hypotheses for the different stromatolite microfabrics. A Fan layers form abiogenically in deep water, potentially below a thermo/chemocline and/or outside the photic zone. B Micrite layers form by the activities of a microbial mat in warmer, shallow waters. C The deepening of the lake (and consequently the water column above the stromatolites) again favors abiogenic fan precipitation. A B C 189 TABLES Table 1. Natural isotopic abundances and calculated mass correction factors for the isotopes measured in this study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able 2. Microfabric mineralogy determined using X-ray powder diffraction compared with calculated XMg values from elemental results. G+)%(& 7+2164 J?- JͲ1+ K %#L3&1 36MM1+4$6#* 1&"0($ HN=A O+K&1 A<E 7 + * =<=B 5+(46$&P *# &Q63&*4& # M ? - 6 * 4 # 1 % # 1 + $ 6 # * HN=A O+K&1 B ?6416 $ & =<D9 R#(#)6$& HN=A O+K&1 ; ?6416 $ & =<EE R # ( # ) 6 $ & < C336$6#*+( )6*# 1 %&+S" 6 *36 4 + $& % # $ & * $ 6 + ( "6(6464(+"$6 4 4#*$+)6*+ $ 6#*T %&+S" M#0*3 363 !" # )+$4U %+$$&1*" M#1 U+(6$&P $1#*+P # 1 *+U4#(6$& < 191 Table 3. Elemental measurements from the Boar's Tusk stromatolites, averaged by stromatolite (Strom.) layer and presented with the stacked layer number (the layer when stromatolites are stacked in stratigraphic sequence) and corresponding microfabric (Fabric key: F = fan, X = mixed, M = micrite, O = ooid). Average values (avg) are the mean of individual drill holes from a layer; standard deviations ( ı ) are presented where multiple measurements were averaged. G$+4S & 3 (+K&1 : + ,V/ ?- ,V/ C( ,V/ 5+ ,V/ ?* ,V/ 7& ,V/ : 6 , % % )/ 50 , %%)/ F* , % % )/ G1 ,V/ H+ ,V/ I , %%)/ O+K&1 7+2164 $% & ʍ $% & ʍ $% & ʍ $% & ʍ $% & ʍ $% & ʍ $% & ʍ $% & ʍ $% & ʍ $% & ʍ $% & ʍ $% & ʍ H N=A ; ; ? =< E@A =< =9> 8E< @ 8 <8 =< 8EA =< =>; 9>< = 9 <A =< 889 =< =;= >< ;= =< DE DB< > @ <D 9;< = > <D D;< A 8 <= =< @EA =< ;;> =< 88B =< =88 8D< B ; <= 8 8 J =< 9@> =< =9@ ;9< B 8 <@ =< ;EB =< =D@ 9E< 8 8 <9 =< 8;= =< =E9 9< ;= =< 9B @>< ; = <A 8E< = ; <B 9A< E 9 <D =< D@B =< =9= =< 8>= =< =;@ 8>< = 8 <; 9 9 J =< 9;E =< =9; ;=< E ; <> =< ;E; =< =ED 9D< E 8 <9 =< 8A@ =< =A; 9< =@ =< EB @>< > E <E 8@< @ @ <8 >E< D @ <B =< D=D =< =EE =< ;B= =< =98 ;E< @ ; <E > > 7 =< 88@ =< =9= 9 <A = <B =< =AB =< =9E 9B< 9 ; <8 =< ;;D =< ==D ;< A@ =< D9 @9< ; = <@ 8=< 9 ;;< 8 9D< A E <9 =< D== =< =8= =< ;DE =< =;> ;;< B = <A E E ? =< >E; =< =>E ;D< 9 ; <= =< 8== =< =E@ >=< 8 8 <B =< 89A =< =D8 E< 8; =< >D A@< D A <8 9@< D 8 <@ >E< @ ; <B =< DB8 =< =D@ =< 89E =< =9> 8=< ; 8 <= D D J =< 9=@ =< =>A @ <> ; <8 =< ;>A =< ==8 >;< A = <@ =< 88= =< =@> 8< DA =< ;A B=< > 9 <= 9D< A 9 <9 >B< D > <@ =< D@8 =< =9E =< ;AA =< =9E ;D< ; ; <@ @ @ 7 =< ;A> =< =;8 = <A = <; =< =@; =< ==B >9< ; ; <; =< 8=B =< =>= =< >@ =< =@ A=< ; = <@ ;=< @ ; <9 8;< 9 ; <= =< E@E =< =>= =< ;9A =< =;E A <E = <8 @ <E @ <E 7 =< ;ED =< =99 ; <8 = <E =< ;=9 =< =8E >=< E = <B =< ;DE =< =99 =< E8 =< =A @D< E ; <= ;=< > > <D 8=< 8 E< = =< E99 =< =9E =< ;99 =< =;D ;=< > ; <A A A 7 =< ;D; =< =>9 ; <D = <> =< =@; =< =;E >;< 8 ; <; =< ;>D =< =;9 =< DA =< 8> @B< > = <8 ;=< B = <E 9@< > ;B< E =< EAA =< =9D =< ;EB =< =;A B <; ; <8 A <E A <E 7 =< 8;; =< =;B 8 <E ; <= =< =E= =< =>> >=< ; ; <= =< ;@9 =< =89 =< D9 =< 8= AA< D ;D< D ;8< 8 =< 8 9;< 8 8 <@ =< D8> =< =>= =< ;B8 =< =99 B <8 ; <B B B ? =< @EA =< =;; 98< 9 9 <E =< ;@8 =< =9A 9=< B > <8 =< 8B> =< =8= D< B9 ;< DA @B< B ;9< ; 9A< = ; <D @E< @ 8@< @ =< DE9 =< =@9 =< ;B@ =< =9; 8@< A ; <A ;= ;= J =< 9A@ =< =;> ;=< D = <> =< ;;A =< =9= 9B< > ; <9 =< ;DE =< ==> ;< BE =< =D @9< ; ; <= ;@< B ; <@ 9@< = > <8 =< D=9 =< =89 =< ;8E =< ==8 ;8< 8 = <8 H N;8 Ͳ 57 Ͳ ; Ͳ ; ;; 7 =< ;@9 =< =;A 8 <= = <D =< ;98 =< =@; >>< E ; <; =< ;>B =< =;9 =< @B =< ;D @>< 9 = <@ ;8< 9 = <A ;>< 8 > <= =< DDD =< =EE =< 8E= =< =9A ;E< 8 ; <9 = ;8 7 =< 8D= =< =;= > <8 = <9 =< ;E= =< =9A >8< = 8 <@ =< ;A8 =< =;B ;< 9; =< => DB< = > <8 ;E< 9 9 <B ;>< B 9 <D =< DD9 =< =99 =< 8EE =< =;= ;E< = = <A ; ;9 7 =< 89E =< =;A ; <8 = <D =< =>8 =< =88 >A< > 9 <9 =< ;@A =< =;> =< D= =< ;> A;< ; E <D @ <= ; <9 B <= = <@ =< @A9 =< =>@ =< 9;B =< =;A ;>< > ; <; 8 ;> 7 =< 8;9 =< =;@ ; <8 = <> =< =AB =< =D; E=< A 8 <= =< 8=@ =< =ED =< @9 =< ;D AD< B > <B @ <= ; <A ;;< E E <B =< @@D =< =88 =< 98; =< =;> 8;< ; = <D 9 ;E 7 =< 8;B =< =8A ; <> = <> =< ;>A =< =AE >B< B E <8 =< 8=E =< =99 =< A@ =< 8E A>< A B <8 D <B 8 <= ;D< 9 = <B =< @EA =< =@B =< 8AA =< ==A 8A< 8 = <9 > ;D 7 =< 99> =< =9E E <A ; <> =< =@A =< =AD >A< A 9 <> =< ;AD =< ==A ;< B; =< D; AE< > 8 <8 B <= > <9 ;@< B A <= =< A8D =< =E8 =< 9>8 =< =;D ;A< D ; <B 192 E ;@ 7 =< 8=; =< ==> 8 <; = <; =< ;=8 =< =8E E>< 9 9 <@ =< ;BE =< =>; =< BD =< => B9< @ @ <= @ <B = <@ ;>< @ 8 <@ =< AD= =< =E9 =< 99B =< =8> 8=< D ; <9 D ;A 7 =< 98E =< =;B D <@ ; <D =< =E9 =< =8; >9< B 9 <B =< ;@@ =< =;> ;< A; =< 8B @E< = E <; @ <; = <8 ;9< A 9 <> =< @DD =< =>8 =< 9;= =< =;D ;8< B = <@ @ ;B ? =< 8BA =< =;= A <> ; <8 =< 8D@ =< =AA 9@< D 9 <E =< 8;B =< =;8 8< 8@ =< ;E DA< = E <E ;A< A 9 <E ;B< @ 9 <> =< D8> =< =9> =< 8E@ =< =;B ;9< E ; <> A 8= 7 =< 8>@ =< =99 ; <8 =< > =< =ED =< =;A >@< E 9 <9 =< ;99 =< =;> =< @; =< ;; @D< D E <@ @ <= = <; A <8 9 <@ =< @DA =< =E; =< 8AA =< =8> ;8< D = <9 B 8; 7 =< 9BA =< =EB > <; = <9 =< =D8 =< =;@ >A< = 9 <@ =< ;E@ =< ==A ;< 8B =< =E @@< ; D <9 B <> ; <= B <= 8 <8 =< A;B =< =D8 =< 99E =< =9= B <D = <D ;= 88 ? =< 9DE =< =88 ;8< 8 ; <D =< ;@; =< =9A 99< > 8 <> =< ;@B =< =;> 8< 8; =< 8@ E>< B 9 <= ;;< 8 ; <> ;E< 8 ; <A =< DE> =< =D8 =< 8AE =< =9D ;9< > = <E H N;8 Ͳ 57 Ͳ 8 9 ;E 7 =< 8B9 ; <> =< ;9D >E< E =< 8=E =< B= @8< D > <E D <> =< DAE =< 8D= 88< 9 > ;D 7 =< 8@E =< =E8 ; <; = <> =< ;8; =< =>D >;< 8 D <A =< ;A@ =< =E= =< DB =< ;> D@< 8 ;;< 9 9 <@ ; <= 89< ; ;B< 9 =< D=D =< =BA =< 88A =< =9D ;B< @ ; <; D ;A 7 =< >>9 @ <; =< =>A >>< B =< ;@B 8< 8@ @8< 8 @ <9 A <A =< @9D =< 8A; 88< E @ ;B ? =< 8DD =< =8; > <8 ; <= =< =@D =< =98 9>< ; 8 <B =< ;88 =< =;= ;< ;D =< 8= EE< A > <D E <8 = <E D <B ; <8 =< ED> =< =>9 =< 88E =< =;> A <D = <D H N;8 Ͳ 57 Ͳ > ; 89 7 =< 8;; =< =;B ; <A ; <8 =< =9E =< ==E >B< B E <B =< ;=8 =< ==A =< ED =< ;B @;< D E <D ;=< 9 E <A D <B 8 <; =< DD9 =< =9> =< 8=8 =< =;E ;8< A = <B 8 8> ? =< 9@A =< =>E ;;< = 8 <= =< =ED =< =9D 9A< = = <@ =< ;;8 =< =;= 8< =@ =< 9; >D< B B <A D=< B ;B< @ ;=< E 8 <@ =< E9; =< ;88 =< 8;@ =< =>@ ;8< 8 ; <E 9 8E J =< 8D8 =< =8D E <; ;< 9 =< =B; =< =;B 9D< > 9 <= =< ;;8 =< =;@ ;< 8> =< 88 E8< 8 > <E A9< > 98< A ;=< E 9 <A =< EE= =< =@9 =< ;B8 =< =8@ ;8< E 8 <B > 8D J =< ;B8 =< =9E ; <; = <E =< =B8 =< =EE 9@< ; E <> =< ;;8 =< =;9 =< E> =< ;> E>< > A <9 88< ; ;E< 8 @ <D 8 <; =< ED@ =< =@A =< ;B> =< =8> B <A 8 <> E 8@ 7 =< 89E =< =E9 = <@ = <; =< =8A =< =;= >9< ; 9 <@ =< ;9= =< =;B =< >; =< =8 D9< A D <> E <> ; <> D <8 9 <= =< D>@ =< =E@ =< 8;B =< =;> A <; = <E D 8A ? =< EAD =< =>9 88< 8 = <; =< ;9E =< =9= 8>< @ = <@ =< ;BA =< =;; 9< ;= =< 9> 9B< ; = <@ @ <@ 9 <= ;D< = ; <A =< >@; =< =;; =< ;AE =< ==E ;E< A ; <= @ 8B J =< 8@> =< =;= > <= = <A =< ;;8 =< =;9 9B< E 9 <8 =< ;DE =< =;D ;< =E =< ;D D=< B E <8 D <A = <9 A <D = <; =< D== =< =EE =< 8EE =< =EE ;=< B 8 <= A 9= W =< 8;E =< ;;E 9 <B 9 <E =< 88A =< ;DB 89< B A <8 =< ;8D =< =D9 ;< ;; =< BD 9B< B ;E< D B <B > <A ;9< > @ <E =< 9DB =< ;>= =< 88@ =< =B@ @ <> 8 <B B 9; 7 =< 888 =< =8> = <@ = <8 =< =E; =< =8A >>< = 8 <> =< ;9= =< ==B =< >9 =< =D DA< E 9 <@ > <8 = <8 E <9 ; <8 =< D9= =< =8E =< 8>; =< ==B ;9< D 8 <@ ;= 98 7 =< 8;@ =< =88 = <B = <8 =< =B> =< =;E >E< > > <= =< 8=> =< ==> =< E; =< =; @=< A D <E E< 8 = <E D <8 = <9 =< D8B =< =D8 =< 89> =< =8; ;E< ; = <B ;; 99 7 =< ;EB =< =;9 ; <; = <> =< =88 =< ==A >>< ; = <D =< ;8B =< ==9 =< >D =< =B DA< E ; <@ > <= ; <; > <A = <E =< DD; =< ==B =< 89A =< ==A ;E< B ; <; ;8 9> 7 =< 8>@ =< ==D 8 <= = <9 =< =>> =< ==@ >8< 9 8 <> =< ;8; =< ==E =< DD =< =@ DD< A 8 <E 9 <A = <@ D <8 = <D =< DEB =< =>= =< 8DE =< ==8 @ <A = <> 193 Table 4. Standard stable isotope measurements from the Boar's Tusk stromatolites, averaged by stromatolite layer and presented with the stacked layer number (the layer when stromatolites are stacked in stratigraphic sequence) and corresponding microfabric (Fabric key: F = fan, X = mixed, M = micrite, O = ooid). Average values (avg) are the mean of individual drill holes from a layer; standard deviations ( ı ) are presented where multiple measurements were averaged. ɷ ; 9 5 ,X YZRH/ ɷ ; A W ,X Y Z R H / O+K&1 G$+4S&3 O+K&1 7+2164 $%& ʍ $%& ʍ HN=A ; ; ? ><=> =<8@ Ͳ;<A =<8 8 8 J 9<>8 =<8A Ͳ9<= =<> 9 9 J ;<>> =<D= ͲE<A ;<; > > 7 8<;= =<8D Ͳ><8 =<9 E E ? 8<A; =<>B Ͳ><; ;<9 D D J 8<AD =<;8 Ͳ9<9 =<; @ @ 7 ;<D> =<8D ͲE<9 =<E @<E @<E 7 ;<8= =<88 ͲD<; =<> A A 7 ;<DA =<9B ͲE<= =<A A<E A<E 7 ;<@; =<98 Ͳ><D =<9 B B ? ><9B =<;B Ͳ;<; =<9 ;= ; = J 9<@= =<== Ͳ8<B =<; HN;8 Ͳ57 Ͳ; Ͳ; ; ; 7 8<;@ =<=B Ͳ><; =<8 = ; 8 7 ;<B@ =<8= Ͳ><9 =<> ; ; 9 7 ;<D> =<;= Ͳ><> =<; 8 ; > 7 ;<D= =<8= Ͳ><@ =<D 9 ; E 7 ;<@> =<8; Ͳ><E =<D > ; D 7 8<98 =<88 Ͳ><= =<8 E ; @ 7 ;<B> =<8A Ͳ><9 =<D D ; A 7 8<@E =<;8 Ͳ9<> =<; @ ; B ? 8<@D =<;D Ͳ8<B =<8 A 8 = 7 8<>8 =<;= Ͳ9<A =<; B 8 ; 7 8<>D =<;9 Ͳ9<@ =<; ;= 8 8 ? 9<;D =<=A Ͳ8<E =<; 194 HN;8 Ͳ57 Ͳ8 9 ; E 7 ;<>@ Ͳ E<8 > ; D 7 8<E8 =<=; Ͳ><= =<; D ; A 7 8<>; Ͳ9<B @ ; B ? 9<E> =<8> Ͳ;<B =<9 HN;8 Ͳ57 Ͳ> ; 8 9 7 ;<DD =<99 ͲE<9 ;<= 8 8 > ? 9<;D =<=9 Ͳ8<@ =<; 9 8 E J 8<=8 =<8E Ͳ><9 =<@ > 8 D J ;<B9 =<=; Ͳ><> =<; E 8 @ 7 ;<A> =<;; Ͳ><E =<; D 8 A ? ><9E =<8; Ͳ;<E =<9 @ 8 B J 8<>8 =<8A Ͳ9<D =<8 A 9 = W 8<8; =<E@ Ͳ><8 ;<9 B 9 ; 7 ;<9; =<8A ͲE<; =<@ ;= 9 8 7 ;<88 =<8D ͲE<E =<> ;; 9 9 7 ;<EE =<8D Ͳ><> =<9 ;8 9 > 7 8<8= =<DD Ͳ9<A ;<; 195 Table 5. Clumped isotope results for the BT08 stromatolite fabrics. Each row represents a different run of a split of the sample listed. Temperatures presented here were calculated by comparison to the stochastic value using the method described by Ghosh et al. (2006) as well as using the absolute reference frame described by Dennis et al. (2011). Samples in gray text were excluded from analyses due to suspicions of drilling into a secondary cement phase (Sample BTF1c) or ǻ 48 values with large internal errors indicating contamination with organic material during measurement (Samples BTF1a and BTF2b). G+)%(& ɷ ; 9 5 X ,Z RH / ɷ ; A W X ,Z RH / ȴ > @ X ,W[ / ȴ > @ ͲN \5 ,] U# "U/ ȴ > @ ͲN \5 ,R&**6" / O+K&1 A ͲA<E ,M +*/ H N 7 ; + 8<==@Ͳ ><D99Ͳ =<=BB 8A<E 98<8 H N 7 ; 2 ><E>8Ͳ ><D>>Ͳ =<=E@ 8E<; 8A<A H N 7 ; 4 ;<BD;Ͳ ><D8=Ͳ =<=;= 9B<D >=<D H N 7 ; 3 ;<B8BͲ ><D;E ^=<=9E 8A<E 8B<@ H N 7 8 + 8<=9BͲ ><E=9Ͳ =<=A9 8E<= 8A<D H N 7 8 2 ><EE9Ͳ ><D8AͲ =<=9; ;A<B 88<D H N 7 8 4 8<=9>Ͳ ><E89 ^=<=E; 8E<B 8@<; O+K&1 B ,)6416 $&/ HN?;+ ><98=Ͳ ;<DA8Ͳ =<=>B 9;<9 9E<= HN?;2 ><9A8Ͳ ;<>EB ^=<;9B 99<@ 9><A HN?8+ ><8E;Ͳ ;<>=@ ^=<;9> 9><@ 9E<A HN?82 ><8D8Ͳ ;<>== ^=<;9A 99<@ 9><A HN?84 ><8;=Ͳ ;<DB=Ͳ =<=>B 9;<; 9><A 196 Table 6. Comparison of clumped isotope results with results from standard stable isotope measurements of layers from the BT08 stromatolite. A statistically significant difference between standard isotope and clumped isotope measurements using a two-sample t-test with a 95% confidence interval assuming a two-end tail and unequal variances 1 was only found for the į 18 O values measured for micrite. G$+*3+13 "$+2(& 6 " # $ # % & " 5(0)%&3 6"#$#%&" G6-< 36MM< O+K&1 7+2164ɷ ; 9 5 ,X /ɷ ; A W ,X /ɷ ; 9 5 ,X /ɷ ; A W ,X / 5.W A ͲA<E M+* ^;<8EA_^8<=8 =Ͳ E<ABD_ Ͳ><99 ^;<B8B_^><E> 8Ͳ ><D>>_ Ͳ><E= 9 : . : B )6416$& ^><;DA_^><E9 EͲ ;<>=9_ Ͳ=<A9 = ^><8;=_^><9A 8Ͳ ;<DB=_ Ͳ;<>= = :.` 1 Calculated using the Matlab function WWHVW from the Statistics Toolbox: WWHVW[\ ERWK XQHTXDO 197 Table 7. Chemical measurement results for all stromatolites by fabric. Statistically significant measurement differences between fan and micrite fabrics (Sig. diff.) were tested using a two- sample t-test with a 95% confidence interval assuming a two-end tail and unequal variances. 7+2164a 7 + * ?6b& 3 ?6416 $ & W#63 G6-< 36MM< ?&+"01&)&*$ $%& ʍ $%& ʍ $%& ʍ $%& ʍ ' ( & ) & * $ + ( )&+"01&)& * $" ,!5ZͲ? G / : + V =<8> =<=A =<9= =<=@ =<>E =<;D =<8; =<;; `&" ? - V 8<E 8<@ @<E ><> ;D<B B<@ 9<B 9<E `&" C( V =<=A =<=D =<;8 =<=> =<;@ =<=A =<89 =<;@ `&" 5+ V >><B ><@ 9A<= 9<9 9><8 E<= 89<B A<8 `&" ? * V =<;D =<=> =<;A =<=@ =<8= =<=D =<;9 =<=D `&" 7& V =<B =<D ;<B ;<= 9<E 8<; ;<; ;<= `&" : 6 %%) @ D B D @ ; 9 D 8 ; D > = ; D `&" 50 %%) A<E ><B 9;<A 8@<> 8E<B 8=<> B<B ><A `&" F * %%) ; > ; = 8 @ ; A 9 = 8 @ ; 9 @ `&" G1 V =<DB =<;= =<D; =<=D =<D9 =<;= =<9@ =<;> `&" H + V =<8E =<=@ =<8= =<=E =<89 =<=> =<89 =<;= : # I %%) ;9<B ><B ;><> ><A ;@<A @<; @<> 8<B `&" J?- =<=A =<=@ =<89 =<;8 =<>8 =<;D =<;B =<;= `&" G$+*3+13 "$+2(& 6"#$#%&" ɷ ; 9 5 X ;<B =<E 8<E =<A 9<E =<@ 8<8 =<D `&" ɷ ; A W XͲ ><E =<A Ͳ><= ;<= Ͳ8<9 ;<= Ͳ><8 ;<9 `&" 5(0)%&3 6 " # $ # % & " ,H N=A (+K&1" A Ͳ B # * ( K / ɷ ; 9 5 X 8<D ;<9 ><9 =<; :#c ɷ ; A W XͲ ><D =<; Ͳ;<E =<; `&" ȴ > @ XͲ =<=; =<=@ =<=D =<; :#c N \5 8A<E ;<; 9E<= =<> `&" *A significant difference was calculated when the confidence interval was lowered to 90%. 198 Table 8. Summary of assumptions used in the lake volume models for this study. ? # 3 & ( Z+1+)&$&1 C""0)%$6#* d&M&1&*4&.e 0"$6M64+$6#* H#$U )6*6) 0) (+S& +1&+ C )6* f ; AP= = = S) 8 ?+%%& 3 & b$& *$ # M $U& d6M& H&3 ,d#&U(&1P ;BB9/ H#$U )6*6) 0) (+S& 3&%$U 3 )6* g ; E ) I"6*- C )6* +*3 + (+S& "(#%& # M =<8 ).S) ,H 1+3(&KP ;BD >2/ H#$U (+S& 3& %$U.Q # (0)& 1&(+$6#*"U6% 3 7 3 10 5.2 d V S Y )6* g AA <> S ) 9 5 # * 6 4 + ( (+S& ,+("# 0"&3 2K R # & 2 2 & 1 $ &$ +(<P 8 = ; = / L6$U "(#%& # M =<8 ). S) H#$U (+S& UK 31#(#-K 4(#"&3 2+"6* G$1#*-ɷ ;9 5. ɷ ; A W 4#11&(+$6#* L6$U U6 -U ɷ ; A W Q+16+$6#* ,N+( 2# $P ; B B = / !"#$#%& (+S& L+$&1 "01M+4& $&)% &1+$01& Q+16+2(& 8 A Ͳ9E\5 G&Q&1+( )#3& (" M#1 (+S& L+ $&1 "01M+4& $&)% &1+$01& L&1& $ & " $ & 3 !"#$#%& (+S& L+$&1 $&)% &1+$01& Q+16+2(& 8 A Ͳ9E\5 G&Q&1+( )#3& (" M#1 4+12#*+ $ & M#1)+$6#* L+$&1 $&)%&1+$01&" L&1& $&"$& 3 !"#$#%& )6*6) 0) M1&"UL+$&1ɷ ;A W ɷ 6 Ͳ )6* gͲ 8 = X ,YG ? Wh / G*#L)&($ ,:#116" &$ +(<P ; B B E / !"#$#%& )+b6)0 ) M1&"UL+$&1ɷ ;A W ɷ 6 Ͳ )+b gͲ 9<>X ,YG ? Wh / ?6*6 )0) ) & + * L+$&1 6"#$#%& Q+(0& 3&$&1 )6*&3 M1#) + "$1#)+ $#(6$& (+K&1 ,H N=A (+K&1 @<E/ 199 Table 9. Temperature scenarios tested in the oxygen isotope volume model. G4&*+16# N "0 1M+4 & N 4+1 2 # * +$& R&"416%$6#* M6b&3 N )6* 8A\5 8A\5 5+12#*+$&" M#1)&3 6* + $U&1)+((K L&((Ͳ )6b&3 (+S& +$ + M6b&3 (#L $& )%&1+$01 & < M6b&3 N )+b 9E\5 9E\5 5+12#*+$&" M#1)&3 6* + $U&1)+((K L&((Ͳ )6b&3 (+S& +$ + M6b&3 U6- U $& )%&1+$01 &< M6b&3 N )63 9E\5 9=\5 5+12#*+$&" M#1)&3 +$ + M6b&3 )#3&1+$& $&)%&1+$01&< G01M+4& L + $&1" L&1& L+1 ) &1< "4+(&3 N : + 8 A Ͳ9E\5 ,"4+( &3 L6$U : + 4#*$&* $/ 8 A Ͳ9E\5 ,"4+( &3 L6$U : + 4#*$&* $/ 5+12#*+$&" M#1)&3 6* + $U&1)+((K L&((Ͳ )6b&3 (+S& $U+$ L+1 ) &3 ,+*3 &Q+%#1+$&3/ + * 3 4##(&3 3 0& $# "&+"#*+( # 1 4 ( 6 ) + $ & 4U+* -& < "4+(&3 N J?- 8 A Ͳ9E\5 ,"4+( &3 L6$U J ? - 4#*$&* $/ 8 A Ͳ9E\5 ,"4+( &3 L6$U J ? - 4#*$&* $/ 5+12#*+$&" M#1)&3 6* + $U&1)+((K L&((Ͳ )6b&3 (+S& $U+$ L+1 ) &3 ,+*3 &Q+%#1+$&3/ + * 3 4##(&3 3 0& $# "&+"#*+( # 1 4 ( 6 ) + $ & 4U+* -& < "&+"#*+( 8A\5 , M + * " / 9=\5 ,)6 b&3 / 9E\5 ,)6 4 16$& / 8A\5 , M + * " / 9=\5 ,)6 b&3 / 9E\5 ,)6 4 16$& / 5+12#*+$&" M#1)&3 6* + $U&1)+((K L&((Ͳ )6b&3 (+S&< ?6* &1+(# -K 6" ( 6 * S & 3 $# (+S& $&) % & 1+ $ 0 1&< $U&1)#4(6*& 9E\5 8A\5 , M + * " / 9=\5 ,)6 b&3 / 9E\5 ,)6 4 16$& / 5+12#*+$&" M#1)&3 6* + $U&1)+((K "$1+$6M6&3 (+S& L6$U ) 6 4 1 6 $ & M # 1)6*- 6* $U& L + 1) & % 6(6)*6#* + * 3 4+(46$& M+*" M # 1)6*- 6* $U& 4##( UK%#(6 ) * 6#*< 200 Table 10. Lake volume model results for all stromatolites using key models. The maximum change ǻ = maximum – minimum calculated value for all measured layers of a given stromatolite. ǻ all is the maximum change over all stromatolites. 5+(40(+$&3 (+S& Q#(0)& ,S) 9 / 5+(40(+$& 3 (+S& 3&%$U ,)/ ?#3&(a !#* !"#$#%& M6b&3 N ) 6 * ɷ 6 g Ͳ ;;X !"#$#%& $U& 1 )#4(6*& ɷ 6 g Ͳ ;;X !#* !"#$#%& M6b&3 N ) 6 * ɷ 6 g Ͳ @X !"#$#%& $U& 1 )#4(6*& ɷ 6 g Ͳ @X O+K&1 G$+4S&3 (+K& 1 7+2164 $%& ʍ $%& ʍ $%& ʍ $%& ʍ $%& ʍ $%& ʍ HN=A ; ; ? ;;D @ ;=B 9 B8 8 ;D<> =<9 ;D<; =<; ;E<8 =<; 8 8 J ;A= ;A B> > AA > ;B<= =<D ;E<9 =<8 ;E<= =<8 9 9 J 8;> 88 ;8A ;D ;E; 9E 8=<; =<@ ;@<= =<@ ;@<B ;<> > > 7 8BA 9A ;=@ 9 ;9; > 88<E ;<= ;D<= =<8 ;@<; =<8 E E ? ;>B ;D ;=@ ;E ;;; ;D ;@<B =<D ;D<= =<A ;D<8 =<A D D J 888 9A BA 8 ;;E 9 8=<9 ;<; ;E<E =<; ;D<> =<8 @ @ 7 9DD 89 ;88 A ;@A 88 8><; =<E ;D<@ =<> ;A<B =<A @<E @<E 7 >>; A9 ;9E @ 88= 8E 8E<D ;<@ ;@<9 =<9 8=<9 =<A A A 7 >9E B@ ;;B ;; ;B9 ;B 8E<> 8<= ;D<D =<E ;B<E =<D A<E A<E 7 9;B 9= ;;> > ;AE @ 89<= =<@ ;D<9 =<8 ;B<8 =<9 B B ? AA ; ;=9 > ;>; E ;E<= =<; ;E<A =<8 ;@<E =<8 ;= ;= J ;@9 D BE ; ;>D ; ;A<A =<8 ;E<> =<= ;@<@ =<= ȴ !"# $" %& && &!# #' &()* &)$ #)( ()" ")! &)( HN;8 Ͳ57 Ͳ ; Ͳ ; ;; 7 9B; >> ;=@ 8 ;AD D 8><D =<B ;D<= =<; ;B<8 =<8 = ;8 7 8EA ;= ;=B E ;BE ;E 8;<> =<9 ;D<; =<8 ;B<E =<E ; ;9 7 8AD 88 ;;; ; 8== > 88<8 =<D ;D<8 =<; ;B<@ =<; 8 ;> 7 9;D 8D ;;> A 8;> 8E 88<B =<D ;D<9 =<> 8=<; =<A 9 ;E 7 9=A 9D ;;8 @ 8;9 ;B 88<@ =<B ;D<8 =<9 8=<; =<D > ;D 7 8=8 89 ;=D 9 8=; E ;B<@ =<@ ;D<= =<; ;B<@ =<8 E ;@ 7 998 @ ;=B A 8;8 8E 89<9 =<8 ;D<; =<> 8=<; =<A D ;A 7 8=D ;8 BB ; ;B; 9 ;B<B =<> ;E<D =<; ;B<> =<; @ ;B ? 88E A B> 8 ;E9 > 8=<E =<8 ;E<9 =<; ;A<= =<8 A 8= 7 8@> 9E ;=8 ; 8;8 E 8;<A ;<= ;E<@ =<; 8=<; =<; B 8; 7 ;@; 8@ ;=; 8 8;; > ;A<@ ;<= ;E<@ =<; 8=<= =<; ;= 88 ? ;A> ;; AA ; ;EE 8 ;B<8 =<> ;E<= =<; ;A<; =<; ȴ ##( +( #* ' *& #' ")' &)' &)! ()% #)& ()' 201 HN;8 Ͳ57 Ͳ > ; 89 7 9;B 8B ;;B ;D 9;@ AD 89<= =<@ ;D<D =<@ 88<A 8<= 8 8> ? ;@B 88 AB ; ;BA 9 ;B<= =<A ;E<; =<; ;B<D =<; 9 8E J 8EA 8D ;=E A 8B= 9A 8;<> =<@ ;E<B =<> 88<9 ;<= > 8D J 9ED D> ;=D ; 8B> 9 89<A ;<> ;E<B =<= 88<> =<; E 8@ 7 8BE D9 ;=@ ; 988 E 88<9 ;<D ;D<= =<= 89<; =<; D 8A ? ;;E A ;=9 9 8D9 A ;D<9 =<> ;E<A =<8 8;<D =<8 @ 8B J 8>> B B@ 9 8A8 ;; 8;<; =<9 ;E<E =<; 88<; =<9 B 9; 7 9=E 99 ;;D B >=A D> 88<D =<A ;D<> =<> 8><B ;<9 ;= 98 7 9;; 9; ;8; @ >E= >E 88<A =<A ;D<@ =<9 8E<A =<B ;; 99 7 >89 9E ;=@ 9 9B> ;8 8E<9 =<@ ;D<= =<8 8><@ =<9 ;8 9> 7 8@; D ;=; ;8 9@9 >B 8;<A =<8 ;E<D =<D 8><8 ;<; ȴ !(' %! !# $ #"! %$ $)' &)& &)* ()% *)# &)( ȴ ,-- !"# $" %+ ' !*# %' &()* &)$ #)! ()% &()$ &)& 202 REFERENCES Arnow, T., and Stephens, D.W., 1990, Hydrologic characteristics of the Great Salt Lake, Utah, 1847-1986: U.S. Geological Survey Water-Supply Paper, v. 2332, p. 40. Aswasereelert, W., Meyers, S.R., Carroll, A.R., Peters, S., Smith, M.E., and Feigl, K., 2012, Basin-scale cyclostratigraphy of the Green River Formation, Wyoming: Geological Society of America Bulletin, v. 125, no. 1-2, p. 216–228, doi: 10.1130/B30541.1. Atwood, G., 1994, Geomorphology applied to flooding problems of closed-basin lakes... specifically Great Salt Lake, Utah: Geomorphology, v. 10, no. 1-4, p. 197–219, doi: http://dx.doi.org/10.1016/0169-555X(94)90017-5. Awramik, S.M., and Buchheim, H.P., 2012, The quest for microbialite analogs to the South Atlantic Pre-Salt carbonate hydrocarbon reservoirs of Africa and South America, in Houston Geological Society Joint General and International Dinner Meeting, Houston, TX, p. 25–28. Beerling, D.J., and Royer, D.L., 2011, Convergent Cenozoic CO 2 history: Nature Geoscience, v. 4, no. 7, p. 418–420, doi: 10.1038/ngeo1186. Berelson, W.M., Corsetti, F.A., Johnson, B., Vo, T., and Der, C., 2008, Carbonate-associated sulfate as a proxy for lake level fluctuations: a proof of concept for Walker Lake, Nevada: Journal of Paleolimnology, v. 42, no. 1, p. 25–36, doi: 10.1007/s10933-008-9245-z. Berelson, W.M., Corsetti, F.A., Pepe-Ranney, C., Hammond, D.E., Beaumont, W., and Spear, J.R., 2011, Hot spring siliceous stromatolites from Yellowstone National Park: assessing growth rate and laminae formation: Geobiology, v. 9, no. 5, p. 411–424, doi: 10.1111/j.1472-4669.2011.00288.x. Bhartia, R., Hug, W.F., Salas, E.C., Reid, R.D., Sijapati, K.K., Tsapin, A., Abbey, W., Nealson, K.H., Lane, A.L., and Conrad, P.G., 2008, Classification of organic and biological materials with deep ultraviolet excitation: Applied Spectroscopy, v. 62, no. 10, p. 1070–1077. Bhartia, R., Salas, E.C., Hug, W.F., Reid, R.D., Lane, A.L., Edwards, K.J., and Nealson, K.H., 2010, Label-free bacterial imaging with deep-UV-laser-induced native fluorescence: Applied and Environmental Microbiology, v. 76, no. 21, p. 7231–7237. Boyer, B., 1982, Green River laminites: Does the playa-lake model really invalidate the stratified-lake model?: Geology, v. 10, no. 6, p. 321–324, doi: 10.1130/0091-7613(1982) 10<321:GRLDTP>2.0.CO;2. 203 Bradley, W.H., 1964a, Geology of Green River Formation and associated Eocene rocks in southwestern Wyoming and adjacent parts of Colorado and Utah: U.S. Geological Survey Professional Paper 496-A, 84 p. Bradley, W.H., 1964b, The geology of the Green River Formation and associated Eocene rocks in southwestern Wyoming and adjacent parts of Colorado and Utah: Washington, D.C., U.S. Geological Survey, 86 p. Bradley, W.H., 1929, The varves and climate of the Green River epoch: U.S. Geological Survey Professional Paper 158-E, 110 p. Braga, J.C., Martin, J.M., and Riding, R., 1995, Controls on microbial dome fabric development along a carbonate-siliciclastic shelf-basin transect, Miocene, SE Spain: Palaios, v. 10, no. 4, p. 347–361. Buchheim, H.P., Awramik, S.M., and Leggitt, V.L., 2009, Paleoenviornmental factors controlling microbialite bioherm deposition and distribution in the Green River Formation, in Geological Society of America Abstracts with Programs, Vol. 41, No. 7, Portland, OR, p. 511. Buchheim, H.P., Awramik, S.M., and Schultz, C., 2007, Stromatolites as indicators of past climate and chemistry, in Geological Society of America Abstracts with Programs Vol. 39 No. 6, Denver, CO, p. 194. Buchheim, H.P., and Surdam, R.C., 1977, Fossil catfish and the depositional environment of the Green River Formation, Wyoming: Geology, v. 5, no. 4, p. 196, doi: 10.1130/0091- 7613(1977)5<196:FCATDE>2.0.CO;2. Carroll, A.R., and Bohacs, K.M., 1999, Stratigraphic classification of ancient lakes: Balancing tectonic and climatic controls: Geology, v. 27, no. 2, p. 99–102, doi: 10.1130/0091- 7613(1999)027<0099:SCOALB>2.3.CO;2. Carroll, A.R., Singer, B.S., and Smith, M.E., 2003, 10 k.y. depositional cyclicity in the early Eocene: Stratigraphic and 40Ar/39Ar evidence from the lacustrine Green River Formation: Geology, v. 31, no. 7, p. 593–596, doi: 10.1130/0091-7613(2003)031<0593:KDCITE> 2.0.CO;2. Clyde, W.C., Sheldon, N., Koch, P., Gunnell, G., and Bartels, W., 2001, Linking the Wasatchian/Bridgerian boundary to the Cenozoic Global Climate Optimum: new magnetostratigraphic and isotopic results from South Pass, Wyoming: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 167, no. 1-2, p. 175–199, doi: 10.1016/S0031- 0182(00)00238-8. 204 Cristofor, S., Vadineanu, A., Ignat, G., and Ciubuc, C., 1994, Factors affecting light penetration in shallow lakes: Developments in Hydrobiology, v. 94, p. 493–498. Dennis, K.J., Affek, H.P., Passey, B.H., Schrag, D.P., and Eiler, J.M., 2011, Defining an absolute reference frame for “clumped” isotope studies of CO 2 : Geochimica et Cosmochimica Acta, v. 75, no. 22, p. 7117–7131, doi: 10.1016/j.gca.2011.09.025. Desborough, G.A., 1978, A biogenic-chemical stratified lake model for the origin of oil shale of the Green River Formation: An alternative to the playa-lake model: Geological Society of America Bulletin, v. 89, no. 7, p. 961–971, doi: 10.1130/0016- 7606(1978)89<961:abslmf>2.0.co;2. Dettman, D.L., and Lohmann, K.C., 2000, Oxygen isotope evidence for high-altitude snow in the Laramide Rocky Mountains of North America during the Late Cretaceous and Paleogene: Geology, v. 28, no. 3, p. 243, doi: 10.1130/0091-7613(2000)28<243:OIEFHS>2.0.CO;2. Doebbert, A.C., Carroll, A.R., Mulch, A., Chetel, L.M., and Chamberlain, C.P., 2010, Geomorphic controls on lacustrine isotopic compositions: Evidence from the Laney Member, Green River Formation, Wyoming: Geological Society of America Bulletin, v. 122, no. 1-2, p. 236–252, doi: 10.1130/b26522.1. Dyni, J.R., 2006, Geology and resources of some world oil-shale deposits: U.S. Geological Scientific Investigations Report 2005–5294, 42 p. Eagle, R.A., Schauble, E.A., Tripati, A.K., Tütken, T., Hulbert, R.C., and Eiler, J.M., 2010, Body temperatures of modern and extinct vertebrates from 13 C- 18 O bond abundances in bioapatite: Proceedings of the National Academy of Sciences of the United States of America, v. 107, no. 23, p. 10377–82, doi: 10.1073/pnas.0911115107. Eiler, J.M., 2007, “Clumped-isotope” geochemistry—The study of naturally-occurring, multiply- substituted isotopologues: Earth and Planetary Science Letters, v. 262, no. 3-4, p. 309–327, doi: 10.1016/j.epsl.2007.08.020. Eugster, H.P., and Hardie, L.A., 1975, Sedimentation in an ancient playa-lake complex: The Wilkins Peak Member of the Green River Formation: Geological Society of America Bulletin, v. 86, p. 319–334. Eugster, H.P., and Surdam, R.C., 1973, Depositional environment of the Green River Formation of Wyoming: A preliminary report: Geological Society of America Bulletin, v. 84, no. 4, p. 1115–1120, doi: 10.1130/0016-7606(1973)84<1115:deotgr>2.0.co;2. 205 Fischer, A.G., and Roberts, L.T., 1991, Cyclicity in the Green River Formation (lacustrine Eocene) of Wyoming: Journal of Sedimentary Petrology, v. 61, no. 7, p. 1146–1154, doi: 10.1306/D4267852-2B26-11D7-8648000102C1865D. Fricke, H.C., and Wing, S.L., 2004, Oxygen isotope and paleobotanical estimates of temperature and ɷ 18 O-latitude gradients over North America during the early Eocene: American Journal of Science, v. 304, no. 7, p. 612–635, doi: 10.2475/ajs.304.7.612. Friedman, I., and O’Neil, J.R., 1977, Compilation of stable isotope fractionation factors of geochemical interest, in Fleischer, M. ed., Data of Geochemistry, United States Government Printing Office, Washington, p. 12. Gazin, C.L., 1965, Early Eocene mammalian faunas and their enviornment in the vicinity of the Rock Springs Uplift, Wyoming, in De Voto, R. and Bitter, R. eds., Sedimentation of Late Cretaceous and Tertiary Outcrops, Rock Springs Uplift; 19th Annual Field Conference Guidebook, Wyoming Geological Association, Casper, WY, p. 171–180. Ghosh, P., Adkins, J., Affek, H.P., Balta, B., Guo, W., Schauble, E.A., Schrag, D., and Eiler, J.M., 2006, 13 C– 18 O bonds in carbonate minerals: A new kind of paleothermometer: Geochimica et Cosmochimica Acta, v. 70, no. 6, p. 1439–1456, doi: 10.1016/j.gca.2005.11.014. Golubi ü , S., and Focke, J.W., 1978, Phormidium Hendersonii Howe: Identity and significance of a modern stromatolite building microorganism: Journal of Sedimentary Petrology, v. Vol. 48, no. 3, p. 751–764, doi: 10.1306/212F7559-2B24-11D7-8648000102C1865D. Grande, L., 2001, An updated review of the fish faunas from the Green River Formation, the world’s most productive freshwater Lagerstätten, in Topics in Geobiology Vol. 18: Eocene Biodiversity, Springer, New York, p. 1–38. Grotzinger, J.P., and Knoll, A.H., 1999, Stromatolites in Precambrian carbonates: evolutionary mileposts or environmental dipsticks?: Annual review of earth and planetary sciences, v. 27, no. 1, p. 313–58, doi: 10.1146/annurev.earth.27.1.313. Hely, A.G., 1964, Precipitation, runoff and water loss in the lower Colorado River-Salton Sea area: U.S. Geological Survey, 17 p. Horita, J., and Wesolowski, D.J., 1994, Liquid-vapor fractionation of oxygen and hydrogen isotopes of water from the freezing to the critical temperature: Geochimica et Cosmochimica Acta, v. 58, no. 16, p. 3425–3437, doi: 10.1016/0016-7037(94)90096-5. 206 Hren, M.T., and Sheldon, N.D., 2012, Temporal variations in lake water temperature: Paleoenvironmental implications of lake carbonate į 18 O and temperature records: Earth and Planetary Science Letters, v. 337-338, p. 77–84, doi: 10.1016/j.epsl.2012.05.019. Huntington, K.W., Eiler, J.M., Affek, H.P., Guo, W., Bonifacie, M., Yeung, L.Y., Thiagarajan, N., Passey, B., Tripati, A.K., Daëron, M., and Came, R., 2009, Methods and limitations of “clumped” CO 2 isotope ( ȴ 47 ) analysis by gas-source isotope ratio mass spectrometry: Journal of mass spectrometry, v. 44, no. 9, p. 1318–29, doi: 10.1002/jms.1614. Jones, C.B., 1981, Periodicities in stromatolite lamination from the Early Proterozoic Hearne Formation, Great Slave Lake, Canada: Palaeontology, v. 24, no. 2, p. 231–250. Keighley, D., Flint, S., Howell, J., and Moscariello, A., 2003, Sequence stratigraphy in lacustrine basins: a model for part of the Green River Formation (Eocene), Southwest Uinta Basin, Utah, U.S.A.: Journal of Sedimentary Research, v. 73, no. 6, p. 987–1006, doi: 10.1306/ 050103730987. Land, L.S., 1998, Failure to precipitate dolomite at 25°C from dilute solution despite 1000-fold oversaturation after 32 years: Aquatic Geochemistry, v. 4, no. 3-4, p. 361–368, doi: 10.1023/A:1009688315854. Leggitt, V.L., and Cushman, R.A., 2001, Complex caddisfly-dominated bioherms from the Eocene Green River Formation: Sedimentary Geology, v. 145, no. 3-4, p. 377–396, doi: 10.1016/S0037-0738(01)00155-5. Lowenstein, T.K., and Demicco, R. V, 2006, Elevated Eocene atmospheric CO 2 and its subsequent decline: Science, v. 313, no. 5795, p. 1928, doi: 10.1126/science.1129555. Lundell, L.L., and Surdam, R.C., 1975, Playa-lake deposition: Green River Formation, Piceance Creek Basin, Colorado: Science, v. 62, no. 1593, p. 34, doi: 10.1126/science.62.1593.34. Machlus, M.L., Olsen, P.E., Christie-Blick, N., and Hemming, S.R., 2008, Spectral analysis of the lower Eocene Wilkins Peak Member, Green River Formation, Wyoming: Support for Milankovitch cyclicity: Earth and Planetary Science Letters, v. 268, no. 1-2, p. 64–75. Mason, G.M., and Surdam, R.C., 1992, Carbonate mineral distribution and isotope-fractionation: An approach to depositional environment interpretation, Green River Formation, Wyoming, USA: Chemical Geology: Isotope Geoscience section, v. 101, no. 3-4, p. 311–321, doi: http://dx.doi.org.libproxy.usc.edu/10.1016/0009-2541(92)90010-3. Morrill, C., Small, E.E., and Sloan, L.C., 2001, Modeling orbital forcing of lake level change: Lake Gosiute (Eocene), North America: Global and Planetary Change, v. 29, no. 1-2, p. 57– 76, doi: 10.1016/S0921-8181(00)00084-9. 207 Norris, R.D., Hole, W., Corfield, R.M., and Cartlidge, J.E., 1995, Skiing in the Eocene Uinta Mountains? Isotopic evidence in the Green River Formation for snow melt and large mountains: Geology, v. 24, no. 5, p. 403–406, doi: 10.1130/0091-7613(1996) 024<0403:SITEUM>2.3.CO;2. Okumura, M., and Kitano, Y., 1986, Coprecipitation of alkali metal ions with calcium carbonate: Geochimica et Cosmochimica Acta, v. 50, no. 1, p. 49–58, doi: 10.1016/0016- 7037(86)90047-5. Passey, B.H., Levin, N.E., Cerling, T.E., Brown, F.H., and Eiler, J.M., 2010, High-temperature environments of human evolution in East Africa based on bond ordering in paleosol carbonates: Proceedings of the National Academy of Sciences of the United States of America, v. 107, no. 25, p. 11245–9, doi: 10.1073/pnas.1001824107. Pawlewicz, M., and Finn, T., 2002, Vitrinite reflectance data for the Greater Green River basin, southwestern Wyoming, northwestern Colorado, and northeastern Utah: U.S. Geological Survey Open-File Report 02-339, 16 p. Pearson, P.N., van Dongen, B.E., Nicholas, C.J., Pancost, R.D., Schouten, S., Singano, J.M., and Wade, B.S., 2007, Stable warm tropical climate through the Eocene Epoch: Geology, v. 35, no. 3, p. 211, doi: 10.1130/G23175A.1. Petryshyn, V.A., Corsetti, F.A., Berelson, W.M., Beaumont, W., and Lund, S.P., 2012, Stromatolite lamination frequency, Walker Lake, Nevada: Implications for stromatolites as biosignatures: Geology, v. 40, no. 6, p. 499–502, doi: 10.1130/G32675.1. Picard, M.D., and High, L.R.J., 1968, Sedimentary cycles in the Green River Formation (Eocene), Uinta Basin, Utah: SEPM Journal of Sedimentary Research, v. Vol. 38, no. 2, p. 378–383, doi: 10.1306/74D7199C-2B21-11D7-8648000102C1865D. Pietras, J.T., and Carroll, A.R., 2006, High-Resolution stratigraphy of an underfilled lake basin: Wilkins Peak Member, Eocene Green River Formation, Wyoming, U.S.A.: Journal of Sedimentary Research, v. 76, no. 11, p. 1197–1214, doi: 10.2110/jsr.2006.096. Pietras, J.T., Carroll, A.R., and Rhodes, M.K., 2003, Lake basin response to tectonic drainage diversion: Eocene Green River Formation, Wyoming: Journal of Paleolimnology, v. 30, no. 2, p. 115–125, doi: 10.1023/A:1025518015341. Rhodes, M.K., Carroll, A.R., Pietras, J.T., Beard, B.L., and Johnson, C.M., 2002, Strontium isotope record of paleohydrology and continental weathering, Eocene Green River Formation, Wyoming: Geology, v. 30, no. 2, p. 167, doi: 10.1130/0091- 7613(2002)030<0167:SIROPA>2.0.CO;2. 208 Riding, R., 2008, Abiogenic, microbial and hybrid authigenic carbonate crusts: components of Precambrian stromatolites: Geologia Croatica, v. 61, no. 2-3, p. 73–103, doi: 10.4154/GC.2008.10. Riding, R., 2011, The nature of stromatolites: 3,500 million years of history and a century of research, in Reitner, J., Quéric, N.-V., and Arp, G. eds., Advances in Stromatolite Geobiology, Springer Berlin Heidelberg, Berlin, Heidelberg, p. 29–74. Ripepe, M., Roberts, L.T., and Fischer, A.G., 1991, ENSO and sunspot cycles in varved Eocene oil shales from image analysis: Journal of Sedimentary Petrology, v. 61, no. 7, p. 1155– 1163, doi: 10.1306/D4267857-2B26-11D7-8648000102C1865D. Roehler, H.W., 1991, Revised stratigraphic nomenclature for the Wasatch and Green River formations of Eocene age, Wyoming, Utah, and Colorado: U.S. Geological Survey Professional Paper 1506-B, 38 p. Roehler, H.W., 1992, Correlation, composition, areal distribution, and thickness of Eocene stratigraphic units, Greater Green River Basin, Wyoming, Utah, and Colorado: U.S. Geological Survey Professional Paper 1506-E, 49 p. Roehler, H.W., 1993, Eocene climates, depositional environments, and geography, Greater Green River Basin, Wyoming, Utah, and Colorado: U.S. Geological Survey Professional Paper 1506-F, 74 p. Schauble, E.A., Ghosh, P., and Eiler, J.M., 2006, Preferential formation of 13C–18O bonds in carbonate minerals, estimated using first-principles lattice dynamics: Geochimica et Cosmochimica Acta, v. 70, no. 10, p. 2510–2529, doi: 10.1016/j.gca.2006.02.011. Schultz, C., Buchheim, H.P., and Awramik, S.M., 2004, A high resolution archive of lake dynamics preserved in the stromatolites of the Laney Member of the Green River Formation (Eocene), in Geological Society of America Abstracts with Programs, Vol. 36, No. 5, Denver, CO, p. 285. Seard, C., Camoin, G., Rouchy, J.-M., and Virgone, A., 2013, Composition, structure and evolution of a lacustrine carbonate margin dominated by microbialites: Case study from the Green River formation (Eocene; Wyoming, USA): Palaeogeography, Palaeoclimatology, Palaeoecology, v. 381-382, p. 128-144, doi: 10.1016/j.palaeo.2013.04.023. Semikhatov, M.A., Gebelein, C.D., Cloud, P.E., Awramik, S.M., and Benmore, W.C., 1979, Stromatolite morphogenesis—progress and problems: Canadian Journal of Earth Sciences, v. 16, no. 5, p. 992–1015, doi: 10.1139/e79-088. 209 Sloan, L.C., 1994, Equable climates during the early Eocene: Significance of regional paleogeography for North American climate: Geology, v. 22, no. 10, p. 881–884, doi: 10.1130/0091-7613(1994)022<0881:ECDTEE>2.3.CO;2. Smith, M.E., Carroll, A.R., and Singer, B.S., 2008, Synoptic reconstruction of a major ancient lake system: Eocene Green River Formation, western United States: Geological Society of America Bulletin, v. 120, no. 1/2, p. 54–84, doi: 10.1130/B26073.1. Smith, J.W., and Robb, W., 1973, Aragonite and the genesis of carbonates in Mahogany Zone oil shales of Colorado’s Green River formation: Bureau of Mines Report of Investigations 7727, 24 p. Smoot, J.P., 1983, Depositional subenvironments in an arid closed basin; the Wilkins Peak Member of the Green River Formation (Eocene), Wyoming, U.S.A.: Sedimentology, v. 30, no. 6, p. 801–827, doi: 10.1111/j.1365-3091.1983.tb00712.x. Stanley, K., and Surdam, R.C., 1978, Sedimentation on the front of Eocene Gilbert-type deltas, Washakie Basin, Wyoming: Journal of Sedimentary Petrology, v. 48, no. 2, p. 557–573. Surdam, R.C., and Stanley, K.O., 1979, Lacustrine sedimentation during the culminating phase of Eocene Lake Gosiute, Wyoming (Green River Formation): Geological Society of America Bulletin, v. 90, p. 93–110. Surdam, R.C., and Wolfbauer, C.A., 1975, Green River Formation, Wyoming: A playa-lake complex: Geological Society of America Bulletin, v. 86, p. 335–345. Swain, F.M., 1999, Fossil non-marine ostracoda of the United States: Developments in Paleontology and Stratigraphy, v. 16, p. 16–21. Talbot, M.R., 1990, A review of the palaeohydrological interpretation of carbon and oxygen isotopic ratios in primary lacustrine carbonates: Chemical Geology: Isotope Geoscience section, v. 80, no. 4, p. 261–279. Thiagarajan, N., Adkins, J., and Eiler, J.M., 2011, Carbonate clumped isotope thermometry of deep-sea corals and implications for vital effects: Geochimica et Cosmochimica Acta, v. 75, no. 16, p. 4416–4425, doi: 10.1016/j.gca.2011.05.004. Tripati, A.K., Eagle, R.A., Thiagarajan, N., Gagnon, A.C., Bauch, H., Halloran, P.R., and Eiler, J.M., 2010, 13 C– 18 O isotope signatures and “clumped isotope” thermometry in foraminifera and coccoliths: Geochimica et Cosmochimica Acta, v. 74, no. 20, p. 5697–5717, doi: 10.1016/j.gca.2010.07.006. 210 Trudell, L.G., Beard, T.N., and Smith, J.W., 1970, Green River Formation lithology and oil shale correlations in the Piceance Creek Basin, Colorado: U.S. Bureau of Mines Report of Investigations, 226 p. U.S. Energy Information Administration, 2012, Independent Statistics: www.eia.gov (accessed March 2013). Vanyo, J.P., and Awramik, S.M., 1982, Length of day and obliquity of the ecliptic 850 MA ago: Preliminary results of a stromatolite growth model: Geophysical Research Letters, v. 9, no. 10, p. 1125–1128, doi: 10.1029/GL009i010p01125. Vasconcelos, C., McKenzie, J.A., Bernasconi, S.M., Grujic, D., and Tiens, A.J., 1995, Microbial mediation as a possible mechanism for natural dolomite formation at low temperatures: Nature, v. 377, no. 6546, p. 220–222, doi: 10.1038/377220a0. Vasconcelos, C., McKenzie, J.A., Warthmann, R., and Bernasconi, S.M., 2005, Calibration of the į 18O paleothermometer for dolomite precipitated in microbial cultures and natural environments: Geology, v. 33, no. 4, p. 317, doi: 10.1130/G20992.1. Warthmann, R., van Lith, Y., Vasconcelos, C., McKenzie, J.A., and Karpoff, A.M., 2000, Bacterially induced dolomite precipitation in anoxic culture experiments: Geology, v. 28, no. 12, p. 1091, doi: 10.1130/0091-7613(2000)28<1091:BIDPIA>2.0.CO;2. Watts, J.M., Swan, B.K., Tiffany, M.A., and Hurlbert, S.H., 2001, Thermal, mixing, and oxygen regimes of the Salton Sea, California, 1997–1999: Hydrobiologia, v. 466, no. 1-3, p. 159– 176, doi: 10.1023/A:1014599719989. Wilf, P., 2000, Late Paleocene-early Eocene climate changes in southwestern Wyoming: Paleobotanical analysis: Geological Society of America Bulletin, v. 112, no. 2, p. 292–307, doi: 10.1130/0016-7606(2000)112<292:lpecci>2.0.co;2. Wing, S.L., and Greenwood, D.R., 1993, Fossils and fossil climate: the case for equable continental interiors in the Eocene: Philosophical Transactions of the Royal Society B: Biological Sciences, v. 341, no. 1297, p. 243–252, doi: 10.1098/rstb.1993.0109. Wolfbauer, C.A., and Surdam, R.C., 1974, Origin of nonmarine dolomite in Eocene Lake Gosiute, Green River Basin, Wyoming: Geological Society of America Bulletin, v. 85, no. 11, p. 1733–1740, doi: 10.1130/0016-7606(1974)85<1733:OONDIE>2.0.CO;2. Wurtsbaugh, W.A., Marcarelli, A.M., and Boyer, G.L., 2012, Eutrophication and Metal Concentrations in Three Bays of the Great Salt Lake (USA): Watershed Sciences Faculty Publications, Paper 550. 211 Yapp, C.J., 2004, Fe(CO 3 )OH in goethite from a mid-latitude North American Oxisol: estimate of atmospheric CO 2 concentration in the Early Eocene “climatic optimum:” Geochimica et Cosmochimica Acta, v. 68, no. 5, p. 935–947, doi: 10.1016/j.gca.2003.09.002. 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. 279–83, doi: 10.1038/nature06588. Zachos, J.C., Pagani, M., Sloan, L., Thomas, E., and Billups, K., 2001, Trends, rhythms, and aberrations in global climate 65 Ma to present: Science, v. 292, no. 5517, p. 686–93, doi: 10.1126/science.1059412. 212 Chapter 3: Transient periods of basin closure associated with changes in the microfabric of stromatolites from the Lower LaClede Bed of the Laney Member of the Eocene Green River Formation ABSTRACT Major changes in basin hydrology, including transient periods of basin closure during a time when the basin is generally considered to have been balanced-filled, are captured in both the chemistry and microfabric of stromatolites from the Lower LaClede Bed of the Laney Member of the Eocene Green River Formation (Sweetwater County, Wyoming). Our high-resolution sampling reveals significant changes in 18 O and elemental composition on the cm-scale, providing a finer scale resolution to the filling and evaporation of Lake Gosuite during the waning Eocene Climatic Optimum, and is complementary to broader scale studies on the hydrology of the basin (e.g., Doebbert et al, 2010). Our results potentially indicate that several filling and evaporation stages representing depth changes of ~3m occurred over the course of a few cm of stromatolite accretion, potentially indicating the magnitude of short-term climate change during the Eocene Climatic Optimum, the period with the highest temperatures and atmospheric CO 2 levels in the Cenozoic. In addition, periods of basin filling are often marked by sudden changes in stromatolite microfabric. 213 INTRODUCTION Geological setting The Green River Formation The Green River Formation was introduced in the previous chapter as a large system of interconnected lakes that existed during the Eocene in Western North America. The formation is of particular economic interest because of the abundance of oil shale and evaporite deposits it contains. The Green River Formation also captures the Early Eocene Climatic Optimum, a period of high global temperatures (the warmest period of the Cenozoic) and atmospheric CO 2 levels (Zachos et al., 2008). Green River sediments responded to major fluctuations in the extent and chemistry of the lake system that appear to reflect a combination of tectonic (e.g. Pietras et al., 2003; Doebbert et al., 2010) and climatic (Bradley, 1929; Fischer and Roberts, 1991) influences on lake fill. The largest of the system’s lakes, paleolake Gosiute, filled a synclinal depression formed by downwarping that accompanied the uplift of the bordering Wind River and Uinta Mountains during the Laramide Orogeny in the Greater Green River Basin (Surdam and Wolfbauer, 1975; Dickinson et al., 1988). The LaClede Bed of the Laney Member The Laney Member records the Green River Formation’s final stages in the Greater Green River Basin. It rests above the evaporative playa-lake deposits recorded in the Wilkins Peak Member and, in some basin locations, above the floodplain deposits of the Cathedral Bluffs Tongue of the Wasatch Formation. The stromatolites of this study were collected from the LaClede Bed, which is the lower oil shale-bearing unit of the Laney Member (Roehler, 1973). The LaClede Bed represents a 214 period between ~49.5-48 Ma (Smith et al., 2008) when Lake Gosiute had transitioned from the closed system (underfilled, lacking a water outlet) represented by the Wilkins Peak Member to a balance-filled lake (water input and output approximately balanced) characterized by the fluctuating profundal facies of the Lower LaClede Bed, and finally to an overfilled system (where water influx exceeded accommodation and spilled into neighboring basins) represented by the Upper LaClede Bed (Carroll and Bohacs, 1999). Lake Gosiute reached its greatest extent during the deposition of the LaClede Bed (~40,000 km 3 ; Roehler, 1993). The LaClede Bed is overlain by the Sand Butte Bed, which contains the volcaniclastic sandstones deposited in deltas that eventually filled the lake in the end stages of the Green River Formation (Surdam & Stanley 1980, Smith 2008). Paleoclimate During deposition of the LaClede Bed, global climate began to cool from the peak of the Early Eocene Climatic Optimum (Zachos et al., 2001). Study of paleofloral assemblages from the region (the Little Mountain flora) indicates a warm, subtropical, and moderately seasonal environment with mean annual temperatures of ~20°C (using leaf margin and area analyses) somewhat prior to deposition of the Laney Member (Wilf, 2000). Depositional environments The Lower LaClede Bed consists of several meter-scale cycles of flooding and desiccation; typical sequences are marked by ostracods, stromatolites and/or oolites and pisoid grainstones followed by organic rich microlaminated marl (carbonate with a significant detrital content) that increases in Mg concentration as the lake became shallower, and end with mud cracked grain- and mudstones, occasionally containing evaporite minerals and Magadi-style 215 cherts (Roehler 1973, Rhodes et al., 2002). During flooding periods, occasional spillover into neighboring Lake Uinta probably occurred (Smith et al., 2008). The upper LaClede Bed represents the lake’s transition from the balance-filled saline lake of the Lower LaClede to a permanently overfilled freshwater basin; shallow-water and mudflat facies are absent in the upper LaClede Bed. The transition from the Lower to Upper LaClede (the so-called “fill-to-spill” transition) is marked by a limestone bench consisting of a stromatolite horizon, ostracod grainstones, and flat-pebble conglomerates accompanied by a sudden reduction in dolomicrite and a decrease in oxygen isotopes attributed by Doebbert et al. (2010) to the capture of a river with high-altitude source-waters. Stromatolites and other microbialites have been described in the LaClede Bed, including the caddisfly-dominated bioherms described by Leggitt et al. (Leggitt and Cushman, 2001; Leggitt and Loewen, 2002) and have been used to imply annual to multi-decadal climate variations in the region (Seard et al., 2013). METHODS Sample collection and processing Site Description The stromatolites described in this study were collected from a location on the western end of the Delaney Rim approximately 65 km southwest of Rock Springs in Sweetwater County, Wyoming (Figure 1). The outcrop (41.43 ͼ N, 108.44ͼ W), accessible by BLM road 4409, is exposed along a gully that drains into Bitter Creek and is 2 km west of the Bitter Creek site described by Stanley and Surdam (1978). The site is in the northwestern part of the Washakie Basin of the greater Green River Basin. The section studied is from the Lower LaClede bed; the 216 section displays several repeating depositional cycles, described by Surdam and Stanley (1979), from bottom to top, as (Figure 2): 1. Thin shallow facies (generally a meter or less in thickness), including stromatolites, ostracod grainstones, pisolites/oolites, microbial intraclasts, and flat-pebble conglomerates; 2. Deep facies, including well laminated, organic rich marl (so-called oil shale), several meters in thickness; contains fossil fish in the lower portion; 3. A return to shallow dolomicrite facies and evidence of emergence, including mud cracks, evaporite minerals/casts, and chert. At the study site, two prominent stromatolite beds (Figure 3) were noted and collected for further study. The stromatolites, some of which nucleated from a mud-cracked surface, likely formed during the initial transgression of the lake system following a period of desiccation. Stromatolite collection and processing Stromatolites were collected from both of the two exposed stromatolite beds; the smaller LC12-3 stromatolite from the lower stromatolite bed and the very large LC12-E stromatolite from the upper bed (Figure 3). The stromatolites were sectioned into vertical slabs using a water- cooled rock saw (See Appendix A) and large format (2x3 inch) thin sections were prepared. Microdrilling Microdrilling was performed as described in the previous chapter. Both LaClede stromatolites display very distinct, laterally continuous mm-scale banding and sub-mm scale (~100-1000µm) laminations that were targeted for microdrilling and chemical analysis. In one stromatolite (LC12-3), ~1 cm long transects were drilled perpendicular to the laminations in 217 order to capture a record of average chemistry through time (Figure 4). In the other stromatolite (LC12-E) ten individual laminae were targeted for analysis (Figure 5). In both cases three transects were drilled in order to assess chemical heterogeneity along layers. Following initial analyses, a second slab cut from the LC12-E stromatolite (referred to here as LC12-E2) was microdrilled at greater resolution ( 1 mm drill holes every ~5 mm) in order to obtain a more detailed record of chemical variability within the stromatolite (Figure 6). Chemical analyses ICP-MS and standard stable isotope measurements were done on microdrilled samples of the LaClede stromatolites LC12-3 and LC12-E as described in the previous chapter for the Boar’s Tusk stromatolites. Stable isotope measurements for LC12-E2 were measured separately; the standard deviations of standards run during the measurement of these samples were 0.055‰ and 0.093‰ for į 13 C and į 18 O, respectively (vs. 0.049‰ and 0.084‰ for the runs in which all other stromatolite samples were measured). Of the 34 LC12-E2 samples run, 12 were sample replicates (different aliquots of powder from the same sample), which deviated by 0.06‰ ( į 13 C) and 0.2‰ ( į 18 O) from original sample measurements. Sample replicates from the Boar’s Tusk and LC12-3 and LC12-E runs deviated by 0.09‰ ( į 13 C) and 0.3‰ ( į 18 O). For samples for which sample replicates were measured, the value presented for a sample is the mean of sample replicates. All isotope values are presented vs. the VPDB standard unless otherwise specified. 218 RESULTS Petrography and mineralogy The LaClede stromatolites are composed of laminated micrite with a coarser detrital fraction composed of ooids, ostracods, and siliciclastic detritus. When cut, the stromatolites smelled distinctly of hydrocarbons. Laminations ranged in size from ~20-1000µm and were defined by differences in color and texture. Laminations in the LaClede Bed stromatolites varied from chalky buff to green-gray to reddish-brown in color. Ooids and other coarser grains are commonly present in the intra-columnar area, in microtopographic pits within the stromatolite, and in some rare cases within laminae. Most of the laminae in the LC12-3 stromatolite were wavy or clotty in nature. Striking circular structures, ~200 !m in diameter, are present throughout LC12-3, and in some cases are the dominant structure within laminae. Most are filled with an unidentified, clear mineral. LC12-3 appears to be significantly altered by diagenesis, including widespread replacement by silica and the aforementioned clear mineral. Most of the laminae in the LC12-E and E2 stromatolites were smooth (vs. clotty in LC12-3). The LC12-E stromatolite had very little visible evidence of diagenesis. One prominent silica- replaced horizon is present (discussed, below). In general, the laminae themselves are very fine, but their character and bundling changes up section, revealing five distinct microfacies visible at the hand sample scale, which allow the stromatolite to be subdivided into five regions of interest, summarized in Table 1. . 219 ICP-MS elemental results Insoluble fraction A large fraction (68% for material drilled from LC12-3, 74% for LC12-E) of powdered samples combined from several carbonate layers of both stromatolites did not dissolve during a 60 minute reaction with 2% nitric acid. The fraction of insoluble material in the carbonate layers is similar to what was measured for the Boar’s Tusk stromatolites discussed in the previous chapter (50-85%). The measured fraction of insoluble material was used to correct elemental results for the actual amount of carbonate dissolved as discussed in the previous chapter. ICP-MS Results Results for ICP-MS element abundance measurements for all drilled samples are presented in Appendix B. Averages and standard deviations for measurements from each stromatolite layer after outliers were removed presented in Table 2 and Figures 7-8. LC12-3 elemental abundances had relatively high standard deviations for replicate samples drilled from the same stromatolite layer for many elements, most notably aluminum. Five of the samples from LC12-13 gave elemental results that were consistently shifted from the adjacent samples, were considered suspect (LC12-3 samples 6, 13, 16, 19, and 22), and removed from subsequent analyses. LC12-E recorded much less variability within layers. Several clear outlier points (noted in the table in Appendix B) were removed from subsequent analyses. LC12-E samples 4, 11, and 20 values were consistently shifted from adjacent samples, were considered suspect (possibly misdrilled or influenced by diagenesis), and removed from subsequent analyses. 220 The LaClede stromatolites were chemically different from one another and from the Boar’s Tusk stromatolites (discussed in the previous chapter of this thesis) for several elements (Figure 9). Both LaClede stromatolites had higher concentrations of Na and Mg (both conserved ions) and lower concentrations of Cu and U than the Boar’s Tusk stromatolites. LC12-3 had generally higher concentrations of Mg and Fe, and lower concentrations of Mn, Zn, and Ba than LC12-E. In addition, nearly all of the samples measured from LC12-3 had Mg and Ca concentrations suggestive of dolomite (some contained Mg molar ratios beyond 50%), whereas nearly all of the LC12-E measurements represented calcite (Figure 10). Stable isotopes Stable isotope measurements for all drilled samples are presented in Appendix C and shown in Figures 7-8. Averages and standard deviations for measurements from each stromatolite layer for both stromatolite slabs are presented in Table 3. Most of the stable isotope values measured for the LaClede stromatolites fell in the range of 0.0 to 4.0‰ and -6.0 to -0.5‰ VPDB for į 13 C and į 18 O, respectively, similar to values measured from carbonates elsewhere in the Lower LaClede by Doebbert et al. (2010), and similar to the range measured in the Boar’s Tusk stromatolites of the previous chapter. Because LC12-E and LC12-E2 represent different slabs of the same stromatolite and both were microdrilled for stable isotope measurements, results from LC12-E and LC12-E2 were combined for this analysis. The laminae can be correlated from one sample to the other, allowing the construction of one continuous higher-resolution dataset from the two samples: LC12- E combined ; Figure 11. 221 Evaluation of outliers: One sample from LC12-3 layer 3 had a į 13 C value of -1.9‰, almost 2‰ lighter than any other value measured in either the Boar’s Tusk or LaClede stromatolites. The same sample, however, gave a į 18 O value well within the normal range (-4.9‰) and the spread in į 13 C values for that layer were quite large, so the value was kept in subsequent analyses. In addition, several extremely light ( į 18 O <-10‰) values were measured for LC12-3 layer 6 (all three samples from the layer drilled several cm apart from one another gave similarly light values); these were also kept in subsequent analyses. Samples A4 and A22 from LC12-E record extremely light oxygen isotopic values ( į 18 O <-10‰). Doebbert et al. (2010) measured į 18 O values in the Upper LaClede Bed as low as -10.5‰ (average for the Upper LaClede = -8.5‰) and argued that these low signals were primary by comparison to į 18 O values as low as -11‰ measured from aragonitic bivalves in a stratigraphically equivalent section in Manila, Utah (Morrill and Koch, 2002). The Upper LaClede Bed, however, represents a permanently hydrologically open system vs. the balanced (intermittently closed) system represented by the Lower LaClede, we expect values from the Lower LaClede Bed to be heavier; indeed, no values measured by Doebbert in the Lower LaClede Bed were lower than -6 ". Sample A22 does not differ an extreme amount from the value from A23 directly above it, and is therefore kept in subsequent analyses. A4, however, differs by >6 " from samples directly above and below and is considered an outlier. The coherence of closely spaced samples is generally taken as an indicator of a near primary signal in chemostratigraphic studies, and extreme outliers are commonly excluded from analyses (e.g., Kaufman et al., 1993). The oxygen isotope values used for analyses are shown in Figure 12. 222 DISCUSSION Stromatolites formed in a saline lake Both LaClede stromatolites had higher average levels of the conserved ions Na and Mg than the Boar’s Tusk (BT) stromatolites (with the LaClede stromatolites reaching values in excess of 1 # compared with <0.4% in the Boar’s Tusk stromatolites), supporting the idea of Lake Gosiute’s transition from a freshwater lake during deposition of the Tipton Member (the end of which is marked by the BT stromatolites) to an increasingly saline lake during deposition of the Wilkins Peak Member. The Lower LaClede records a period before the lake transitioned back to permanently freshwater conditions when a gradual decrease in accommodation space in the basins hosting Lake Gosiute (Surdam and Stanley, 1980) and the apparent capture of a large freshwater source (Doebbert et al., 2010) led to the lake becoming overfilled and spilling into Lake Uinta to the South. The high sodium content of these stromatolites suggests that the lake was more saline during stromatolite deposition in the Lower LaClede Bed than during stromatolite deposition during the Rife Bed of the previous chapter, but became fresh enough for freshwater fish to re-appear during the deposition of the oil shales (Surdam and Stanley, 1980; pers. obs., 2012, 2013). Basin hydrology: open, balanced, or closed? į 13 C vs. į 18 O Talbot (1990) argues that strong covariation (r>0.7) between į 13 C and į 18 O is the hallmark of a closed basin lake, as discussed in the previous chapter. When considered as a complete dataset, į 13 C and į 18 O were not strongly correlated in the LaClede stromatolites 223 (Figure 13A), with a best-fit slope of į 13 C vs. į 18 O for all points giving an R 2 value <0.01 (vs. 0.89 in the Boar’s Tusk stromatolites), implying that the LaClede system was not a continuously closed basin. In order to evaluate the possibility that sequences within the stromatolites might have formed during temporary closure of the basin, a Matlab script (Appendix E, ILQG6HTXHQFHV ) was written to identify all sequences of three or more consecutive layers that displayed a covariation of į 13 C and į 18 O with a correlation coefficient R 0.7. Using the Matlab script, six sequences were identified in the LC12-3 stromatolite between layers 1-5, while the upper half of the stromatolite did not display covariance (Figure 13B, Table 4). The same analysis of the LC12-E combined stromatolite identified 188 unique sequences displaying potentially closed system behavior. Of these, several of the sequences with the highest R values were limited to particular microfabric regions identified for the stromatolite. When basin closure was assessed for the five microfabric regions of the LC12-E stromatolite (Table 1), clear patterns indicating periods basin closure and periods when the basin was open or overfilled were apparent (Figure 13C, Table 4). Stromatolites record major transitions in basin hydrology Previous studies have considered the Lower LaClede Bed to record the transition from the underlying evaporitic or terrestrial units to a balanced filled saline lake, where input essentially equals evaporation. Facies alone suggest the system may have been closed at times, given the repeating shallowing upward cycles of laminated oil shale to mudcracked, evaporite- bearing dolomicrite, with the stromatolites occupying a position during the initial trangressive phase (e.g., Surdam and Stanley, 1979). 224 Our analysis of į 13 C vs. į 18 O in the stromatolite reveals when the basin may have been closed or balanced/open. For example, LC12-3 records some covariation during its inception, but does not record obvious covariation in į 13 C vs. į 18 O in the remaining structure, suggesting it may indeed have been deposited initially during a time of basin closure followed by balanced fill (the alternative being that it is diagenetically altered and does not record primary isotopic signatures). LC12-E combined , which we believe to be more robust with respect to diagenetic alteration, clearly reveals episodes of į 13 C vs. į 18 O co-variance in the middle of the structure, which could represent basin closure, also followed by periods of non-covariance in the upper layers, perhaps representing a transition to balanced fill. However, the lower ~2 cm of the stromatolite do not display sufficiently strong coviariance to indicate basin closure. Thus, the pattern of basin closure followed by balance fill is apparent in both stromatolites. The presence or absence of covariance corresponds to previously identified changes in the microfabric of the stromatolite in both stromatolites. Thus, the LaClede stromatolites appeared amenable to the application of the previously-developed models for lake volume change during times of closed basin behavior, in order to assess the magnitude of lake volume change during their deposition. Application of lake volume models during closed intervals Assumptions Lake area. The minimum area of the lake at the time of stromatolite deposition is assumed to be the 25,000 km 2 , the approximate mapped area of the stromatolite-bound Lower LaClede bed (Figure 70 of Roehler, 1993); the maximum areal extent of the LaClede bed (during the upper LaClede) is ~40,000 km 2 (Roehler, 1993). 225 Lake depth. 15-25 m thick Gilbert-style delta foreset sequences described by Stanley and Surdam (1978) for the Laney Member at a site just south of the location from which we collected our stromatolites provide constraints on lake depth. We use a minimum lake depth of 18 m for our lake volume calculations, which uses the lake slope of 0.2 m/km discussed in the previous chapter. Note that Doebbert et al. (2010) use 50 m in their isotope mass balance model of the lake during deposition of the Laney Member. Lake volume. As in the previous chapter, we use a conical lake volume model. Using the above assumptions for minimum lake area and depth, this gives a minimum lake volume of 153 km 3 . This minimum lake volume as well as the 18 m minimum lake depth are applied to each sequence of both stromatolites. This provides a minimum estimate for lake volume and depth changes recorded in the stromatolite. Carbonate formation temperature. The oxygen isotope model is sensitive to temperature. For both stromatolites, we assumed that the stromatolites formed at 20°C, the approximate regional temperature based on Wilf’s paleobotanical assessment for the Little Mountain flora from the upper Wilkins Peak and lower Laney Member (Wilf, 2000). We assumed the same value for surface water temperatures. Isotope model assumptions. The assumptions of the oxygen isotope model are discussed in the previous chapter. The model requires an assumed freshwater input į 18 O value. We use -14‰ (VSMOW), a value that gives a lower bound on lake volume and depth change estimates (Figure 14). This value is lighter than the value used for the model in the previous chapter (-11‰), which, if used, would require significantly greater lake volume increases to sufficiently freshen the lake to the isotope levels seen in the LaClede stromatolites. This value is similar to 226 values Doebbert (2010) used for riverine input to explain changes in isotope levels seen in the LaClede Beds. Additional assumptions. Both the ion and isotope models assume that the chemistry derived from the stromatolites is primary in nature. We consider the results from the LC12-E stromatolite to be more reliable than those from the LC12-3 stromatolite due to the higher degree of diagenesis visible in the latter stromatolite, particularly in the upper half of the stromatolite (layers 5 and above). Model results The results of the lake volume models are summarized in Figure 15 and Table 5. Where sodium ion data were available, the results of the sodium ion model are very similar to results using the isotope model. Overall, the LC12-3 ion results were shifted to higher values from the isotope model results and showed somewhat greater variation, whereas the LC12-E combined ion and isotope results were very similar, falling within a standard deviation range of one another. However, because ion data were not available for most layers, results for the ion model in LC12- E combined are not considered as robust as the isotope model. LC12-3. The ion model for the LC12-3 sequence had high associated errors and showed no significant variation in lake depth during the period assessed as representing a closed system. The isotope model did show a small but significant change in depth from ~18 m in layer 1 to ~20 m in layer 3. Thus the LC12-3 stromatolite may have recorded ~2 m of lake depth change, although signals could have been obscured by diagenesis. LC12-12E. The LC12-E sequences showed slightly larger lake changes, and the changes were greater than likely explained by measurement error (i.e., depth or volume changes are much 227 greater than typical layer replicate standard deviation values). In the previous description of the co-variance of į 13 C vs. į 18 O, we noted two sections in the lower portion of the stromatolite that displayed co-variance (microfabric region 2-3 and microfabric region 4) and two sections that did not display co-variance (microfabric region 1, poor covariance, and microfabric region 5, no covariance). Treating the two covarying sections separately reveals: 1. Microfabric region 2-3 records one clear cycle of expansion and contraction, with a magnitude of ~3 meters of depth change. Lake level rose through microfabric region 2 (coarse, wavy lamination), then fell through microfabric 3 (finely-laminated, smooth laminations). At the end of the lake level fall recorded in microfabric 3 there is a layer of diagenetic chert, which is indicative of very shallow, nearly emergent conditions (cf., Surdam and Stanley, 1979). 2. Microfabric region 4 records the expansion of the lake following the chert layer, with a similar depth change of ~3 meters. Microfacies are banded, but less distinctly than the previous segment. 3. Microfabric region 5 reveals no covariation, and thus likely represents a more balance filled system, as one would expect if the basin filled to capacity. Like the other segments, it is concomitant with a fabric change in the stromatolite to darker laminations, perhaps representing the transition into a more productive, organic rich lake. The isotope model was also run assuming basin closure over the entire lower ! of the stromatolite (microfabric regions 1-4; R=0.78). Model results were nearly identical to the results found when the microfabric regions were treated separately. Although the results over 228 microfabric region 1 are not considered robust (because stable isotope covariation is poor over this region), there is suggestive evidence that this section of the stromatolite also records a deepening of the lake (as is displayed as a dashed red line in Figure 15). It is striking that microfabric changes occur at the beginning of every major deepening sequence recorded in these stromatolites (Figure 16). Model implications: stromatolites record major lake transitions For both stromatolites, the greatest lake depth calculated was ~21 m. This approaches the maximum thickness of Gilbert-type delta foresets measured by Stanley and Surdam (1978), which, they argued, implies that the depth of Lake Gosiute during formation of the Laney Member did not much exceed 25 m. Therefore, our model results are consistent with the LC12-E stromatolite (and potentially the LC12-3 stromatolite as well) recording a “fill-to-spill” transition marked by the chemical transition to non-covarying stable isotopes and a marked, abrupt change in stromatolite microfacies. The shoreline shifts that would accompany such depth changes are dramatic. Previous authors have estimated a 0.2 m/km lake slope, which would imply shoreline shifts in excess of 14 km. This is reasonable, considering the ~30 km of mapped mudflat deposits beyond the stromatolite bed during deposition of the Lower LaClede member (see e.g. Roehler, 1993 Figure 70). As discussed in the previous chapter, frequent large shoreline shifts are not unusual in shallow lakes. The model indicates LC12-E combined records two, or perhaps three, “fits and starts”, representing the filling and evaporation of ~3 m of water depth before the basin transitioned into an open system. Our models, using data obtained on a mm-scale within a stromatolite, capture basin variations on a much finer timescale than was attempted by previous modeling efforts and 229 provides information that is complementary to previous longer-scale models. For example, Doebbert et al. (2010) analyzed samples at meter scale sampling intervals which indicated large scale changes in the basin caused by the capture of a high altitude freshwater source. We measured similar isotope variability on a mm-scale in a single stromatolite and attribute the changes seen to cycles of lake evaporation and freshening, indicating that the history of Lake Gosiute was more complex than previously thought and changed on a timescale much faster than previously detected. Stated more broadly, the “noise” observed in coarser-scale isotope records can contain, as in this case, information about high frequency environmental variability. Lamination counts and growth rate estimates There are approximately 250 laminations up the ~11 cm of the LC12-E stromatolite. From the microdrilled lamination at 3.9 cm to the lamination at 4.9 cm (corresponding to a depth decrease of 2.8 m over microfabric region 4) there are 24 laminations or light/dark couplets. Using an estimate of evaporation rate of 0.5-1.5 m per year (cf. Morrill et al., 2001), this implies that the 24 laminations formed over a period of ~2-6 years, giving a growth rate of ~4-12 laminations (or ~0.2-0.5 cm) per year. This lamination period does not follow an obvious forcing, although stromatolites that do not follow a typical lamination period are known (c.f., Berelson et al., 2011; Petryshyn et al., 2012). However, annual laminations may be more parsimonious with the hypothesis that these stromatolites formed seasonal layers of precipitated or trapped-and-bound micrite, especially given the strong presumed-annual varving (Surdam and Stanley, 1979) found in the overlying oil shale. If this were the case, the 3 meters of evaporation, represented by 24 laminations, would have occurred over ~24 years, well within the realm of possibility observed in modern lakes. Net evaporation rates <0.2 m per year would be required if 230 the couplets were annual, which may imply a somewhat humid environment. This is consistent with prior assertions that the refilling of the basin represented by the transition of the Wilkins Peak to the Laney member occurred during a transition of regional climate from arid to more humid (Bradley, 1964; Wilf, 2000). Our results give an indication of the magnitude of short term climate change (e.g., ~3 m of lake level change over ~24 years) experienced by Lake Gosiute during the waning stage of the Eocene Climatic Optimum. CONCLUSIONS This study analyzed stromatolites from two horizons of the Lower LaClede bed of the Laney Member of the Green River Formation, a period when the lake had recently transitioned from the closed evaporative basin of the Wilkins Peak member to a balanced-filled system. The chemistry of the upper of the two stromatolites suggested significant changes in lake depth and extent, with lake depth changes of ~3 m estimated over a sequence of laminations spanning <1 cm. As in the previous chapter, where depth changes (and corresponding dramatic shoreline shifts) were accompanied by distinct changes in microstructure and mineralogy, subtle changes in the microstructure of the stromatolite laminae were observed in the LaClede stromatolites that occurred with large shifts in basin hydrology. Isotope correlation in this stromatolite implies that the Lower LaClede system was not consistently closed (in agreement with the extensive literature suggesting that the lake was balance-filled at the time); however sequences are present in consecutive stromatolite layers that are suggestive of periods of lake closure that lasted several years or even decades (if our estimates of stromatolite growth rates are correct). As in the prior 231 chapter, this study demonstrates the utility of stromatolites as tools for high-resolution paleoenvironmental reconstructions. ACKNOWLEDGEMENTS Stable isotope work on the LC12-E2 stromatolite slab was done by Max Wagner for his undergraduate senior thesis and were run by Miguel Rincon at USC. ICP-MS work was done with Pedro Marenco at Bryn Mawr. Stratigraphy of the studied section was done and stromatolite samples were collected by the 2012 and 2013 International Geobiology Courses, which received support from the USC Wrigley Institute for Environmental Studies and the Colorado School of Mines as well as grants from the Agouron Institute, the Gordon and Betty Moore Foundation, the NASA Astrobiology Institute, and the National Science Foundation. We are particularly grateful to John Spear’s dedication in collecting and sectioning the LC12-E stromatolite, affectionately dubbed “John Spear’s Loaf of Bread”. 232 FIGURES Figure 1. A) Map showing location of the outcrop from which samples were collected. B) Satellite image of the outcrop. Modified from Google Maps. 20 km WY UT CO Delaney Rim Outcrop BLM 4409 Rd Rock Springs Rock Springs Uplift Outcrop Washakie Basin A B N 250 m 233 Figure 2. Surdam and Stanley, 1979, Figure 9: Typical lacustrine depositional cycles in the Laney Member, showing A) carbonate mineralogy [dolomite/(dolomite + calcite)] and B) stratification sequence and depositional structures. mud cracks DO/(Ca + DO) 0.0 1.0 mud cracks Dolomicrite Oil shale Limestone A B 234 Figure 3. Stratigraphy of the LaClede site, measured at five different locations, showing the two stromatolite horizons (starred) from which the stromatolite samples LC12-E, LC12-E2, and LC12-3 were collected. N S 100 m shale limestone silt/sandstone Key nodular limestone ostracods mud cracks stromatolite concretions/carbonate lenses ooids & flat-pebble conglomerate ash chert 5 m 2 3 4 5 Lat: 41.43488 Lon: -108.44001 Lat: 41.43457 Lon: -108.44005 * * * * * * * gypsum (probably secondary) * 1 * * * LC12-E, LC12-E2 LC12-3 235 Figure 4. LC12-3-C stromatolite slab. A) Photograph showing bands microdrilled for ICP-MS and standard stable isotope measurements (1-30) and material drilled for clumped isotope measurements (C-1 and C-2). B) High-resolution scan. 236 Figure 4 A B 237 Figure 5. LC12-E-B (LC12-E) stromatolite slab showing the location of holes microdrilled for ICP-MS and standard stable isotope measurements (1-30). Scale divisions are mm. 238 Figure 6. LC12-E2 stromatolite slab showing holes drilled by Max Wagner for standard stable isotope analysis. Two sets of holes were drilled for most layers in order to obtain sufficient material for isotope analyses. 1 cm 239 Figure 7. Chemical results (concentration by mass) for the LC12-3 stromatolite samples plotted by layer. Red points are points that were considered outliers for the purpose of subsequent analyses. Blue lines show the outlier-excluded average for each layer with error bars representing standard deviations for each layer. 02 1 2 3 4 5 6 7 8 9 10 Na (%) Layer 0 100 200 1 2 3 4 5 6 7 8 9 10 Mg (%) 0 50 100 1 2 3 4 5 6 7 8 9 10 Ca (%) 0 0.2 0.4 1 2 3 4 5 6 7 8 9 10 Al (%) Layer 0.1 0.3 0.5 1 2 3 4 5 6 7 8 9 10 Mn (%) 0 10 1 2 3 4 5 6 7 8 9 10 Fe (%) 0 200 400 1 2 3 4 5 6 7 8 9 10 Ni (ppm) Layer 0 20 1 2 3 4 5 6 7 8 9 10 Cu (ppm) 0 20 40 1 2 3 4 5 6 7 8 9 10 Zn (ppm) 0 2 1 2 3 4 5 6 7 8 9 10 Sr (%) Layer 0 0.2 0.4 1 2 3 4 5 6 7 8 9 10 Ba (%) 0 40 1 2 3 4 5 6 7 8 9 10 U (ppm) 0 1 2 3 4 5 6 7 8 9 10 XMg (mol Mg/Mg+Ca) Layer -5 0 5 1 2 3 4 5 6 7 8 9 10 į 13 C (‰ VPDB) -20 -10 0 1 2 3 4 5 6 7 8 9 10 į 18 O (‰ VPDB) 4 1 1 5 10 20 0.5 240 Figure 8. Chemical results for the LC12-E stromatolite samples plotted by layer. Red points are points that were considered outliers for the purpose of subsequent analyses. Blue lines show the outlier-excluded average for each layer with error bars representing standard deviations for each layer. 0 1 2 1 2 3 4 5 6 7 8 9 10 Na (%) Layer 0 20 40 1 2 3 4 5 6 7 8 9 10 Mg (%) Layer 0 50 100 1 2 3 4 5 6 7 8 9 10 Ca (%) Layer 0 0.2 0.4 1 2 3 4 5 6 7 8 9 10 Al (%) Layer 0 0.2 0.4 1 2 3 4 5 6 7 8 9 10 Mn (%) Layer 0 1 2 1 2 3 4 5 6 7 8 9 10 Fe (%) Layer 0 100 200 1 2 3 4 5 6 7 8 9 10 Ni (ppb) Layer 0 10 20 1 2 3 4 5 6 7 8 9 10 Cu (ppb) Layer 0 50 100 1 2 3 4 5 6 7 8 9 10 Zn (ppb) Layer 0 0.5 1 1 2 3 4 5 6 7 8 9 10 Sr (%) Layer 0 0.5 1 1 2 3 4 5 6 7 8 9 10 Ba (%) Layer 0 10 20 1 2 3 4 5 6 7 8 9 10 U (ppb) Layer 0 0.5 1 1 2 3 4 5 6 7 8 9 10 XMg (mol Mg/Mg+Ca) Layer 0 2 4 1 2 3 4 5 6 7 8 9 10 d 13 C (permil vs. VPDB) Layer -10 -5 0 1 2 3 4 5 6 7 8 9 10 d 18 O (permil vs. VPDB) Layer 241 Figure 9. Comparison of chemical data measured (with outlier points deleted) from five Green River stromatolites analyzed in this thesis. From the Boar’s Tusk location discussed in the previous chapter (BT): BT08, BT12-CF-1, and BT12-CF-4; from the LaClede bed: LC12-3 (LC3) and LC12-E (LCE). Boxes denote the 25-75 percentile range for each stromatolite, red bars denote the median value, dots denote points outside the 25-75 percentile range of values. 0.2 0.6 1.0 1.4 BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE BT LC3 LCE Na (%) 20 40 60 80 Mg (%) 20 40 60 Ca (%) 0 0.2 0.4 Al (%) 0.1 0.2 0.3 Mn (%) 0 2 4 6 8 Fe (%) 20 60 100 Ni (ppm) 0 50 100 Cu (ppm) 0 50 100 Zn (ppm) 0.2 0.4 0.6 0.8 1.0 Sr (%) 0.0 0.2 0.4 0.6 Ba (%) 0 10 20 30 U (ppm) 0 0.2 0.4 0.6 XMg [Mg/(Mg+Ca)] -2 0 2 4 į 13 C (‰ VPDB) -10 -8 -6 -4 -2 į 18 O (‰ VPDB) 242 Figure 10. Fraction of magnesium in carbonate (XMg) in the LaClede stromatolites LC12-3 and LC12-E. The dotted black line at 0.5 indicates the XMg value above which the mineralogy of the carbonate is considered dolomite. 0 0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 8 9 10 XMg (mol Mg/Mg+Ca) Layer LC12-3 LC12-E calcite dolomite 243 Figure 11. LC12-E and LC12-E2 (inset) stromatolites showing the correlation of drilled layers. Five microfabric regions (MR1-5) are shown bounded by white dotted lines. LC12-E layers drilled are shown as solid black lines with the layers numbered in black text on white circles. LC12-E2 layers drilled are shown in the inset image as well as represented as dotted black lines on the LC12-E stromatolite with the drill hole numbers in white text on black circles. LC12- E combined layers are numbered (as approximate distance from the bottom of the LC12-E stromatolite) as black text on grey backgrounds; normal font represents layers drilled only in LC12-E, italicized text represents layers drilled only in LC12-E2, and bold text represents layers drilled in both stromatolites. 244 Figure 11 1 4 5 6 7 8 9 10 11 12 24 1 2 3 23 22 13 14 19 20 21 18 17 16 15 8 9 10 3 6 5 4 7 2 1.0 1.4 1.6 1.9 2.0 2.2 2.5 3.0 3.5 3.9 4.2 4.6 4.8 4.9 5.2 5.4 5.5 6.1 6.3 7.0 7.3 7.7 8.1 8.3 8.6 9.0 9.3 9.6 9.9 LC12-E2 LC12-E MR1 MR2 MR3 MR4 MR5 Microfabric regions LC12-E Layers Rel. Layers LC12-E2 Drill Hole Numbers 245 Figure 12. Oxygen isotope values measured for the LaClede stromatolites A LC12-3 and B LC12-E (blue dots) and LC12-E2 (red dots) showing the outlier point excluded from analyses (red x). Black lines show the average for each layer with error bars representing standard deviation for each layer. -15 -10 -5 0 1 2 3 4 5 6 7 8 9 10 į 18 O (‰ vs. VPDB) į 18 O (‰ vs. VPDB) Layer A LC12-3B -15 -10 -5 0 1 2 3 4 5 6 7 8 9 10 B LC12-E combined x Combined layer 246 Figure 13. Crossplots of oxygen and carbon stable isotopes. A All data from the LaClede stromatolites with best fit lines through the data for LC12-3 (blue), LC12-E combined (red), and all LaClede data points (black). B Isotope data from the LC12-3 stromatolite with best fit lines through the data for the lower four micritic layers (bottom, blue) and the upper layers (top, red). C Isotope data from the LC12-E combined stromatolite with best fit lines through data from the different microfabric regions: Microfabric 1 (blue), Microfabric 2 (green) and 3 (yellow) with a trendline for the two fabrics combined (yellow-green), Microfabric 4 (orange), and Microfabric 5 (red). 247 Figure 13 y = 0.80x + 3.22 R = 0.85 y = -0.02x + 1.23 R = 0.043 -2 -1 0 1 2 3 4 5 -12 -10 -8 -6 -4 -2 0 į 13 C (‰ VPDB) į 18 O (‰ VPDB) B LC12-3 A All LaClede bottom top y = 0.37x + 4.73 R = 0.59 y = 0.28x + 3.28 R = 0.87 y = 0.17x + 3.61 R = 0.48 y = 0.54x + 4.67 R = 0.85 -2 -1 0 1 2 3 4 5 -12 -10 -8 -6 -4 -2 0 į 18 O (‰ VPDB) C LC12-E Microfabric 1 Microfabric 2 Microfabric 3 Microfabric 4 Microfabric 5 y = 0.34 + 3.91 R = 0.61 y = 0.06x + 1.36 R = 0.12 -2 -1 0 1 2 3 4 5 -12 -10 -8 -6 -4 -2 0 į 13 C (‰ VPDB) į 18 O (‰ VPDB) LC12-E LC12-3 y = -0.02 + 1.57 R = 0.036 All 248 Figure 14. Sensitivity of the oxygen isotope model results to the assumed freshwater input į 18 O value ( į i ). LC12-3 uses the sequence for layers 1-4. LC12-E MF1-4 uses the sequence of layers represented by microfabric regions 1-4 for LC12-E combined . Likewise LC12-E MF2-3 and LC12-E MF4 use the sequences of layers represented by microfabric regions 2-3 and 4, respectively. Error bars represent the standard deviations of values calculated from individual drill sites in a single layer; because replicate measurements were not done for most LC12-E2 samples, standard deviations could not be calculated for most LC12-E combined layers. -20 -18 -16 -14 -12 -10 -8 15 20 25 30 35 į i (‰VSMOW) max lake depth calculated (m) LC12-3 Layers 1-4 LC12-E MF1-4 LC12-E MF2-3 LC12-E MF4 249 Figure 15. Lake volume model results for sequences in the LaClede stromatolites displaying closed basin covariance. A Ion and isotope model results for the LC12-3 layer 1-4 sequence. B Isotope and ion model results for the LC12-E combined sequences: microfabric region 2-3 (green), microfabric region 4 (blue), and microfabric regions 1-4 treated together (dotted red line) with black points representing the average volume calculated using the ion model for microfabric regions 1-4. Error bars represent the standard deviations of values calculated from individual drill sites in a single layer; because replicate measurements were not done for most LC12-E2 samples, standard deviations could not be calculated for most LC12-E combined layers. All calculations use a minimum lake volume of 154 km 3 , a minimum lake depth of 18 m, and a freshwater input į i = -14‰. 250 Figure 15 100 150 200 250 300 350 400 1 2 3 4 Calculated lake volume (km 3 ) Layer 16 18 19 17 20 21 22 23 24 1 2 3 4 Calculated lake depth (m) Calculated lake volume (km 3 ) Calculated lake depth (m) Layer 100 150 200 250 300 350 400 16 18 17 19 20 21 22 23 24 MF1-4 MF2-3 MF4 ion A. LC12-3 B. LC12-E combined 2 3 1 4 5 6 7 8 9 10 Layer 2 3 1 4 5 6 7 8 9 10 Layer isotope ion 251 Figure 16. Plots of sodium concentration ([Na]), į 18 O, and calculated lake depth (using the calculation for microfabrics 1-4 treated together) from the LC12-E combined stromatolite shown with the five different microfabric regions. 1 2 3 4 5 1 cm Hydrology Lake Volume Model Open ? chert Closed 1 Closed 1 Closed 2 Open Microfabric [Na] į 18 O Observations Interpretations 17 18 19 20 21 22 Calculated lake depth (m) į 18 O (‰ VPDB) Na (%) -10 -8 -6 -4 -2 0 0.2 0.4 0.6 0.8 1.0 252 TABLES Table 1. Descriptions of the five different microfabric regions in the LC12-E and LC12-E2 stromatolites (shown in Figure 11) with layers drilled from each region noted. $%&'( ) *+'% & ,-.%(/ 0-1&'%23%(/ 4567 Ͳ87 9'%:: ;(:-1 4567 Ͳ8 :*<-'1 6 =-'< )%/-:< :* >%/*3- 9? 1>((3; 9(>% &*: :*>%/*3%(/1 @6 Ͳ@A 7 7 4-11 9%13%/ &3: < :*>%/*3-9? B * C<? * / 9 1(>-B;*3 &(*' 1-' 3;*/ 3 ;- 2'- & -9%/. :*>%/* 3%(/1 @D Ͳ@E /F* G )%/-:< :*>% /*3-9? +H 3 B%3; 2 ' ( > % / - / 3 */9 *:3-'/*3%/. &'-*> */9 +H )) &(:('-9 +* /91 ( / 3;- I7 JJ >%&'(/ Ͳ1&*:- @K Ͳ@67 G ͲL M%:%&* Ͳ'-2:* & -9 ; ( ' % N ( / A L O%/-:< :* >%/* 3-9 * /9 +*/9 -9? +H 3 :-11 9 % 1 3 % / & 3 : < +* /9-9 3;*/ @K Ͳ@67 P 4*>%/*- * 2 2 - * ' 3( %/ &('2 ('* 3- &(*'1-' >*3-'%*: )'( > +(33(> 3( 3(2P @6G Ͳ@6E D ͲE A 0*'Q-' * / 9 )%/-' .'*%/ -9 %/ *22-*'*/&-? B%3; 1 3' ( /. +*/9%/. ( / 3;- IGJJ >% &'( / 1&*:-P @6K Ͳ@7L K Ͳ6 J 253 Table 2. Elemental measurements for the LaClede stromatolites, averaged by layer. Average values (avg) are the mean of individual drill holes from a layer; standard deviations ( ı ) are presented where multiple measurements were averaged. R * S# T $. S# T @: S# T 5* S# T $/ S# T O- S# T R % S22>T 5H S 22>T U/ S22> T M' S# T V* S# T W S22> T 4*<-' !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ 4567 Ͳ G 6 J P L X D J P 6 G 7 G L P 7 K P J J P J D L J P J L 7 7 K P K D P A J P 6 7 D J P J G X 6 P G L J P L D A J P 7 6 J P K A P J 6 P D A P X 6 P J J P L 7 D J P 6 6 X J P J K D J P J 7 E E P G G P 6 7 J P L L A J P 6 G 7 7 D P L G P 7 J P J D E J P J A K 7 K P A 6 P D J P J X X J P J J G J P K A J P 7 A A J P 6 G P E 7 P X J P D 6 A P D 6 7 P D J P G K 7 J P J E A J P J D D J P J J 6 L P D 6 P X G J P A J G J P G J L 7 G P X 6 E P D J P 6 7 G J P J D D G X P L 6 6 P J J P 6 G D J P J G E J P D E J P 7 K E J P X 7 6 P E 7 P J J P X A P L G P J J P L L 6 J P 7 A J J P J A A J P J G A G P G 6 P G L J P E 7 E J P L K 6 L K P K G E P X J P 6 E L J P 6 G J L 7 P J 7 L P L J P 6 L D J P J X 7 6 P E D 6 P G L E 7 P E L 7 P 7 A P G L P A 6 J P X K P G J P D 6 A J P L 6 K J P 6 7 7 J P J X J 6 J P L 6 J P J A J P D J X J P 6 J 6 G A P D 6 G P X J P J E A J P J 6 K G 7 P 7 K P G J P J K G J P J L D 6 P 6 X 6 P J X E 7 P E G A P E 7 P X 7 P L D P 7 7 P 7 J P A J K J P 6 A J J P J X X J P J D G L P L 7 P G D 6 P G 7 7 J P J K D A G P 7 6 6 P G J P 7 D J J P 6 6 L L E P E 6 G P E J P J X K J P J 6 K 6 P J 7 J P 6 7 X D P X 7 K P D A P 7 J P 6 6 J P D J P G J P E K 7 J P 7 K A J P 6 D J J P J L L L P X J P 7 E J P L K E J P 6 J E 7 E P E J P L J P J L X J P J 6 L 7 7 P G 7 P J J P J K X J P J A 6 J P E G J P J 6 A 6 P E 7 6 P X 7 P E J P 6 L P K J P D J P G G K J P J L J J P J X D J P J G X D P E A P X K J P L A A J P 6 J A 7 E P A A P 7 J P J L D J P J J G 7 7 P G L P J J P J K D J P J J K J P K 6 J P 6 D A 6 P L 6 6 P A 7 P L J P 6 D P E G P 6 J P G 7 7 J P J D K J P J A K J P J 6 J A P J J P L X J P L L A J P 7 G D 7 X P J 6 L P X J P J L E J P J 7 D G 6 P K 6 G P 6 J P 6 7 E J P J A A 6 P 6 K J P D 6 A L P D 7 7 P A 7 P X J P K A P 7 6 P A J P G K E J P 6 K 6 J P J E 6 J P J G D A P X 7 P A 6 J J P D X L J P J G D G K P 7 6 P G J P J K K J P J 6 L G 7 P 7 J P L J P 6 J X J P J 6 6 6 P L G J P L L E 6 P K 6 G P 6 D P K 6 P K K P 6 6 P 6 J P A D J J P J 7 7 J P 7 6 7 J P J 7 7 L P 7 J P L 4567 Ͳ 8 6 J P A L 7 J P J D 7 6 G P 7 G P E J P J K E J P J 7 X L G P L A P G J P 7 J 6 J P J 6 L J P A J J P J G D D P K 6 J P L D P A G P 7 7 D P X 6 6 P K J P A K E J P J A 6 J P 7 X K J P J 7 D K P E 6 P A 7 J P L D G J P J L E L P D J P A J P J A E J P J 6 J A A P 6 J P A J P G J E J P J J 6 J P G G J P J 7 K L P 6 J P K G P J J P L L E P D 6 G P D J P E 6 K J P J G 6 J P L J 7 J P J G G E P E J P L G J P A G 7 J P J E X K P D J P A J P J 6 6 J P J 6 6 L X P D 6 P L J P 7 J J J P J 7 7 J P 7 7 J P J 7 E K P E 6 P 7 7 P 6 J P G 6 L P 6 K P 6 J P K D X J P 6 6 D J P A E G J P J K A L P E J P J L J P A X G J P 6 6 L 6 A P L D P L J P J 7 D J P J J J G X P X G P L J P 6 X D J P J L A J P L J J P 7 J D G P X A P 6 6 P K J P D 6 X P K 7 P 7 J P A X G J P J D L J P G 6 G J P J G E G P G J P A A J P L D 7 J P J D G E P 6 6 P J J P 7 G A J P J 7 K 7 L P E A P A J P 6 G D J P J 7 D J P E 6 J P G G L G P L 6 7 P 6 A P A 7 P G 6 X P J J P L J P G L X J P J E X J P 6 E G J P J L 7 7 P L J P E D J P D 7 X J P J 6 K 6 D P A G P 6 J P J A X J P J J X L J P L D P E J P 7 7 7 J P J 7 J J P L X J P J E D L P A 6 6 P A 7 P E J P A 6 K P K L P D J P A K D J P J E K J P 7 E A J P J G 6 G P L J P D E J P A G E J P 6 J K 6 7 P K J P 7 J P 6 J D J P J 6 G G X P L K P J J P 7 L K J P J G K J P A E J P J 7 A L P D J P J 7 P E J P A 6 7 P G 6 P L J P A 6 K J P J X X J P 7 7 K J P J L A G P 6 J P K K J P D 7 K J P 6 7 6 6 K P D G P K J P 6 J 6 J P J 7 7 L 6 P E 7 P 7 J P 7 E K J P J G G J P D D J P J A D K P 7 L P E G P J 6 P D 6 7 P 6 L P 6 J P A L L J P J 7 A J P 7 L 6 J P J 6 K G P L 6 P 6 X J P K L J J P J E G 7 7 P D D P K J P 6 7 X J P J 7 J L D P 6 L P 6 J P 7 X L J P J J X J P D 6 J P J L E K P X D P 7 A P 7 6 P J 7 6 P A A P L J P D 7 K J P J J L J P 7 X D J P J 6 G G P X J P G 6 J J P E 6 D J P 7 X 6 D P 6 6 P A J P J E 6 J P J 6 A L L P 6 7 P E J P 7 A D J P J 6 6 J P 7 G J P J G K L P D G P X D P L G P D 6 E P 6 D P L J P E 7 7 J P J E XͲ J P 7 K A 6 P 6 X L G P E J P E 254 Table 3. Stable isotope measurements for the LaClede stromatolites, averaged by layer. Stratigraphic height is a measure of the approximate distance of the layer from the bottommost layer of the stromatolite and serves to correlate layers from LC12-E and LC12-E2. Average values (avg) are the mean of individual drill holes from a layer; standard deviations ( ı ) are presented where multiple measurements were averaged. The value in grey were considered an outlier and excluded from subsequent analyses. ɷ 6 G 5 S" =Y0VTɷ 6 K Z S" = Y 0 V T 4*<-' M3'*3%.'*2 ;%& ;-%.; 3 $%&'( ) *+'% & ' -.%( / !"# ʍ !"# ʍ 4567 ͲG 6 6P6 /F* 7PLD JP76 Ͳ 6PD7 JPXL 7 7PJ JPXK JPGX Ͳ 7P6G JPGG G 7PE Ͳ JPG7 6PDG Ͳ LPAD 6P6X L LPL 6PGE 6PGJ Ͳ 7P76 6PGA A APE JPDG JP6A Ͳ LPLA 6PG7 D DP6 6PK6 JP6X Ͳ 6JPEG JPAK E EP6 6PXD JPGL Ͳ 6PG6 JPJE K EPD Ͳ JPLX JP7X Ͳ 6PK7 JPLK X KPL JP67 JP76 Ͳ JPEL JPJX 6J XPA GPE7 JPG7 ͲJPEE JPG7 4567 Ͳ8 6 6PJ GP67 JP7J Ͳ LPXJ JPX7 7 7P7 6 7PGD JP6J Ͳ DP66 JP6G G LPK G 7PLE JP6D Ͳ LPLG JP6X L LPX G 7P6A JPAA Ͳ GPEE JPX6 A AP7 7PAG JP6D Ͳ GPAX JPGG D APL L 7PLG JPAX Ͳ GPLE JPED E DPG L 6PLK JPAJ Ͳ APDJ 6PE6 K EPE A 7PAG JPEE Ͳ LPAK 6PGD X KPD A 7PAD JPLL ͲGPLL 6P6D 6J XPX A 7PDA JP6J Ͳ LPE7 JPGD 255 4567 Ͳ87 @6 6PL 6 GPDA JPJJ Ͳ LPGD JP6J @7 6PD 6 GPAD JPJ7 Ͳ LPEL JPJ7 @G 6PX 6 7PDL JPJJ Ͳ GPDG JPJA @L 7PJ 6Ͳ JPJG Ͳ 67PXJ @A 7PA 6 6PDK Ͳ DPA7 @D GPJ 7 GPXJ Ͳ 7PX7 @E GPA 7 GPDJ Ͳ GPGD @K GPX G JPAJ Ͳ EP67 @X LP7 G 6PEA JPJG Ͳ DP6A JP67 @6J LPD G 7PL7 Ͳ LP7L @66 LPK G JPK6 Ͳ DPG7 @67 LPX G 7PAA JPJ6 Ͳ GPJ6 JPJL @6G APA L 7P7K JPJ6 Ͳ 7PKA JP6J @6L DP6 L 7PLE JPJ6 Ͳ GP6G JPJ6 @6A DPG L 6PXE JPJL Ͳ APE6 JP66 @6D EPJ L 6PK6 Ͳ APGL @6E EPG L 6PLA JPJ6 Ͳ DPX7 JPJA @6K EPE A 7PKE JPJ7 Ͳ7PX6 JPJX @6X KP6 A 6PJ7 JPJG Ͳ APE6 JPJA @76 KPD A 7P6X Ͳ APG7 @77 XPJ A 7PGD Ͳ 6JPJK @7G XPG A JPKE Ͳ KPAK @7L XPD A 7PE6 Ͳ EPG7 256 Table 4. Assessment of basin closure using plots of į 13 C vs. į 18 O for different microfabric regions of the LaClede stromatolites. 4%/-*' +-13 )%3 3'-/9:%/- '-1 H :31 5:(1-9 + * 1 % / [ \ 4*<-'1 $%&'( ) *+'% & / 1:(2- , 4567 ͲG 6 ͲL +(33(> 6 G JPKJ JPKA ]-1 A Ͳ6J 3(2 6 EͲ JPJ7 JPJL R ( 4567 Ͳ8 &( >+ 6PJ Ͳ7PA 6 6 G JPGE JPAX R ( GPJ ͲLPX 7 ͲG 7 K JPAL JPKA ]-1 APL ͲEPG L 6 A JP7K JPKE ]-1 EPE ͲXPX A 6 K JP6E JPLK R ( \W1%/. ^*:+(3_1 ,`JPE &(C*'%*3%(/ &H3()) S^*:+( 3? 6XXJ TP 257 Table 5. Lake volume model results for LaClede stromatolite sequences that displayed closed basin covariation using the ion and isotope models. Isotope (1) results treated microfabric regions 2-3 and region 4 seperately. Isotope (2) results treated microfabric regions 1-4 as a continuous unit. The maximum change ǻ = maximum – minimum value for all measured layers of a given sequence. For LC12-E combined : ǻ MF2-3 is the change observed over microfabric regions 2-3, ǻ MF4 is the change over microfabric region 4, and ǻ All is the total change observed when micofabrics region 2-3 and 4 are treated separately for the Isotope (1) model, or the change observed over microfabric regions 1-4 when they are treated together as in the Isotope (2) model. Standard deviations are for calculations from multiple drill sites for a layer. !"#$% & '($" # $)*"%+ ,'-#.-'/)0 -' 1) 2%-.3) 41 3 5 6 ,'-#.-'/)0 -' 1) 0 ) 7/ 8 43 6 9%+ 9:%/%7) 4; 6 9:%/%7) 4< 6 9%+ 9:%/%7) 4; 6 9:%/%7) 4< 6 ='>)$ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ !"# ʍ =,;< Ͳ5 ; (%//%3 < 5 < ? ? ; ? 5 ; ? <@AB ;AC ;DA@ @AB < < B < C E ; B ; ? <;AF <A< ;DA5 @A< 5 < C 5 ; 5 F < ; 5 5 ; <;AF 5AE <@A; ;A@ F < B < < ? ? ; C ? < F <@A; BAB ;DAD @AE ȴ F ; ; D E B @ F B @AD ?AB <A; ;AB =,;< ͲG#%3(" +)0 ; ; ; D @ < < ; E < ; E ;EA@ @AC ;EAF @AB ;AF ; D ; < ;EA; @A; ;AB ; D D @ ;EA5 @A@ ;AE ; B D ; ;DAB @A@ <A< < @ D < < < ; ? 5 <@A@ @AC <@A< @A; <A? < < ? <@A? 5 < ; ? 5 ; ? ? ;DA@ ;DA; 5A? ; ? E ; B ; ;DA< ;DA5 5AE < 5 ? < 5 D <@AD <@AE FA< 5 < ; < 5 < ; ? 5 <@A; @A; <@A< @A; FAB ; C F ; C B ;DAD ;DAE FAD ; D ? 5 @ ; D C ; E ; D E < @ ;EA; ;A@ ;EA< @AB ;EA5 @AC FAE ; B B 5 < ; B ; ; 5 ; B 5 ; 5 ;DA? ;A< ;DA5 @A? ;DAF @A? 258 ?AF F ; ? 5 ? ; B 5 ; 5 ; B 5 ; < ;DA@ @A< ;DAF @A? ;DAF @AF ?A? ; ? 5 < ; ? 5 < ;DA@ @A; ;DA@ @A; BA; ; ? C @ ; ? C ;DA< @A@ ;DA< @A@ BA5 ; D F 5 B < @ < < E < @ < < E ;EA; ;A5 ;EAC @AE ;EAC @AE C ; E ? ; E ? ;EA? ;EA? CA5 < 5 F ; < 5 F ; <@AC @A@ <@AC @A@ ȴ H)I< Ͳ5 D < <AD ȴ H)IF D ; <AC ȴ J-- ? ? 5 C D ? ;AE ;AF <AE 259 REFERENCES Berelson, W.M., Corsetti, F.A., Pepe-Ranney, C., Hammond, D.E., Beaumont, W., and Spear, J.R., 2011, Hot spring siliceous stromatolites from Yellowstone National Park: assessing growth rate and laminae formation: Geobiology, v. 9, no. 5, p. 411–424, doi: 10.1111/j.1472-4669.2011.00288.x. Bradley, W.H., 1964, The geology of the Green River Formation and associated Eocene rocks in southwestern Wyoming and adjacent parts of Colorado and Utah: U.S. Geological Survey Professional Paper 496-A, 86 p. Bradley, W.H., 1929, The varves and climate of the Green River epoch: U.S. Geological Survey Professional Paper 158-E, 110 p. Carroll, A.R., and Bohacs, K.M., 1999, Stratigraphic classification of ancient lakes: Balancing tectonic and climatic controls: Geology, v. 27, no. 2, p. 99–102, doi: 10.1130/0091- 7613(1999)027<0099:SCOALB>2.3.CO;2. 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: Geological Society of America Bulletin, v. 100, no. 7, p. 1023–1039, doi: 10.1130/0016-7606(1988)100<1023:PAPSOL>2.3.CO;2. Doebbert, A.C., Carroll, A.R., Mulch, A., Chetel, L.M., and Chamberlain, C.P., 2010, Geomorphic controls on lacustrine isotopic compositions: Evidence from the Laney Member, Green River Formation, Wyoming: Geological Society of America Bulletin, v. 122, no. 1-2, p. 236–252, doi: 10.1130/b26522.1. Fischer, A.G., and Roberts, L.T., 1991, Cyclicity in the Green River Formation (Lacustrine Eocene) of Wyoming: Journal of Sedimentary Petrology, v. 61, no. 7, p. 1146–1154, doi: 10.1306/D4267852-2B26-11D7-8648000102C1865D. Kaufman, A.J., Jacobsen, S.B., and Knoll, A.H., 1993, The Vendian record of Sr and C isotopic variations in seawater: Implications for tectonics and paleoclimate: Earth and Planetary Science Letters, v. 120, no. 3-4, p. 409–430, doi: 10.1016/0012-821X(93)90254-7. Leggitt, V.L., and Cushman, R.A., 2001, Complex caddisfly-dominated bioherms from the Eocene Green River Formation: Sedimentary Geology, v. 145, no. 3-4, p. 377–396, doi: 10.1016/S0037-0738(01)00155-5. Leggitt, V.L., and Loewen, M.A., 2002, Eocene Green River Formation “Oocardium tufa” reinterpreted as complex arrays of calcified caddisfly (Insecta: Trichoptera) larval cases: Sedimentary Geology, v. 148, no. 1-2, p. 139–146, doi: 10.1016/S0037-0738(01)00214-7. 260 Morrill, C., and Koch, P.L., 2002, Elevation or alteration? Evaluation of isotopic constraints on paleoaltitudes surrounding the Eocene Green River Basin: Geology, v. 30, no. 2, p. 151, doi: 10.1130/0091-7613(2002)030<0151:EOAEOI>2.0.CO;2. Morrill, C., Small, E.E., and Sloan, L.C., 2001, Modeling orbital forcing of lake level change: Lake Gosiute (Eocene), North America: Global and Planetary Change, v. 29, no. 1-2, p. 57– 76, doi: 10.1016/S0921-8181(00)00084-9. Petryshyn, V.A., Corsetti, F.A., Berelson, W.M., Beaumont, W., and Lund, S.P., 2012, Stromatolite lamination frequency, Walker Lake, Nevada: Implications for stromatolites as biosignatures: Geology, v. 40, no. 6, p. 499–502, doi: 10.1130/G32675.1. Pietras, J.T., Carroll, A.R., and Rhodes, M.K., 2003, Lake basin response to tectonic drainage diversion: Eocene Green River Formation, Wyoming: Journal of Paleolimnology, v. 30, no. 2, p. 115–125, doi: 10.1023/A:1025518015341. Rhodes, M.K., Carroll, A.R., Pietras, J.T., Beard, B.L., and Johnson, C.M., 2002, Strontium isotope record of paleohydrology and continental weathering, Eocene Green River Formation, Wyoming: Geology, v. 30, no. 2, p. 167, doi: 10.1130/0091- 7613(2002)030<0167:SIROPA>2.0.CO;2. Roehler, H.W., 1993, Eocene climates, depositional environments, and geography, Greater Green River Basin, Wyoming, Utah, and Colorado: U.S. Geological Survey Professional Paper 1506-F: U.S. Geological Survey, 74 p. Roehler, H.W., 1973, Stratigraphic divisions and geologic history of the Laney Member of the Green River Formation in the Washakie Basin in Southern Wyoming; Geological Survey Bulletin 1372-E:, 28 p. Seard, C., Camoin, G., Rouchy, J.-M., and Virgone, A., 2013, Composition, structure and evolution of a lacustrine carbonate margin dominated by microbialites: Case study from the Green River formation (Eocene; Wyoming, USA): Palaeogeography, Palaeoclimatology, Palaeoecology, v. 381-382, p. 128–144, doi: 10.1016/j.palaeo.2013.04.023. Smith, M.E., Carroll, A.R., and Singer, B.S., 2008, Synoptic reconstruction of a major ancient lake system: Eocene Green River Formation, western United States: Geological Society of America Bulletin, v. 120, no. 1-2, p. 54–84, doi: 10.1130/B26073.1. Stanley, K., and Surdam, R.C., 1978, Sedimentation on the front of Eocene Gilbert-type deltas, Washakie Basin, Wyoming: Journal of Sedimentary Petrology, v. 48, no. 2, p. 557–573. 261 Surdam, R.C., and Stanley, K.O., 1980, Effects of changes in drainage-basin boundaries on sedimentation in Eocene Lakes Gosiute and Uinta of Wyoming, Utah, and Colorado: Geology, v. 8, no. 3, p. 135, doi: 10.1130/0091-7613(1980)8<135:EOCIDB>2.0.CO;2. Surdam, R.C., and Stanley, K.O., 1979, Lacustrine sedimentation during the culminating phase of Eocene Lake Gosiute, Wyoming (Green River Formation): Geological Society of America Bulletin, v. 90, p. 93–110. Surdam, R.C., and Wolfbauer, C.A., 1975, Green River Formation, Wyoming: A playa-lake complex: Geological Society of America Bulletin, v. 86, p. 335–345. Talbot, M.R., 1990, A review of the palaeohydrological interpretation of carbon and oxygen isotopic ratios in primary lacustrine carbonates: Chemical Geology: Isotope Geoscience section, v. 80, no. 4, p. 261–279. Wilf, P., 2000, Late Paleocene-early Eocene climate changes in southwestern Wyoming: Paleobotanical analysis: Geological Society of America Bulletin, v. 112, no. 2, p. 292–307, doi: 10.1130/0016-7606(2000)112<292:lpecci>2.0.co;2. 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. 279–83, doi: 10.1038/nature06588. Zachos, J.C., Pagani, M., Sloan, L., Thomas, E., and Billups, K., 2001, Trends, rhythms, and aberrations in global climate 65 Ma to present: Science, v. 292, no. 5517, p. 686–93, doi: 10.1126/science.1059412. 262 Appendix A: Supplemental stromatolite images 263 ! STROMATOLITE COLLECTION Boar’s Tusk stromatolites Figure A1. Stromatolite beds from which samples used in this study were obtained with the sampled areas outlined and numbered. A) BT12-CF-1 (1) and BT12-CF-2 (2) collected at 41.96609N, 109.25016W. B) BT12-CF-4 (4) collected at 41.96623N, 109.24997W. No image is available of the BT08 collection location. 2 4a 4b 1 A B 264 PREPARATION OF STROMATOLITE SLABS AND BILLETS Boar’s Tusk stromatolites Figure A2. BT12-CF-1 stromatolite. A-B) Whole BT12-CF-1 stromatolite. C) BT12-CF-1 stromatolite cut into four slabs. D-F) Cut slabs, BT12-CF-1A through BT12-CF-1D. G) BT12- CF-1B sectioned into four pieces for thin sections: 1Ba, 1Bb, 1Bc (small format) and 1Bd (large format). H) BT12-CF-1C slab polished and scanned. This face was used for microdrilling and mirrors the BT12-CF-1B face shown in panel G used to make thin sections. 265 ! Figure A2 1A 1B 1B 1C 1C 1D 1Ba 1Bb 1Bc 1Bd 1C A B C D G H E F 266 Figure A3. BT12-CF-2 stromatolite. A) Whole BT12-CF-2 stromatolite. B) BT12-CF-2 stromatolite cut into three slabs. The white line shows an additional cut that was made to further divide the stromatolite into four slabs: BT12-CF-2A through BT12-CF-2D. C) BT12-CF-2A cut for a large-format thin section. D) BT12-CF-2B polished and scanned. This face was used for microdrilling and mirrors the BT12-CF-2A face shown in panel C used to make thin sections. 267 ! Figure A3 2A 2B 2C 2D 2A 2B A B C D 268 Figure A4. BT12-CF-4 stromatolites. A) Whole BT12-CF-4 stromatolites. BT12-CF-4a was collected stratigraphically below BT12-CF-4b (see Figure A1). B) BT12-CF-4a stromatolite cut into three slabs: BT12-CF-4a-A through BT12-CF-4a-C. C) BT12-CF-4a-A cut for a large- format thin section. D) BT12-CF-4a-B polished and scanned. This face was used for microdrilling and mirrors the BT12-CF-4a-A face shown in panel C used to make thin sections. E) BT12-CF-4b stromatolite cut into two slabs: BT12-CF-4b-A and BT12-CF-4b-B. F) BT12- CF-4b-A cut for a large-format thin section. G) BT12-CF-4b-B polished and scanned. This face was used for microdrilling and mirrors the BT12-CF-4b-A face shown in panel F used to make thin sections. Although the slab broke up during cutting and polishing, the pieces were retained and labeled to indicate the original orientation. 269 ! Figure A4 4a 4b 4a-A 4a-B 4a-C 4b-A 4b-B 4b-A 4b-B 4a-A 4a-B A B E C F G D 270 LaClede Stromatolites Figure A5. LC12-3 stromatolite. A) Whole LC12-3 stromatolite. B) LC12-3 stromatolite cut into four slabs. C) LC12-3-A through LC12-3-D cut slabs. D) LC12-3-B cut into three pieces for large-format thin sections: LC12-3-Ba through LC12-3-Bc. E) LC12-3-C polished and scanned. This face was used for microdrilling and mirrors the LC12-3-B face shown in panel D used to make thin sections. 271 ! Figure A5 3-A 3-B 3-C 3-D 3-A 3-B 3-C 3-D 3-Ba 3-Bb 3-Bc 3-C A B C D E 272 Figure A6. LC12-JS-End (LC12-E) stromatolite. A) LC12-E stromatolite piece cut from one end of the whole LC12-JS-Loaf stromatolite. B) LC12-E stromatolite cut into three slabs. C) LC12- E-A through LC12-E-C cut slabs. D) LC12-E-A cut into six pieces for large-format thin sections: LC12-E-Aa through LC12-E-Af. E) LC12-E-B polished and scanned. This face was used for microdrilling and mirrors the LC12-E-A face shown in panel D used to make thin sections. 273 ! Figure A6 E-A E-B E-C E-A E-B E-C E-Aa E-Ab E-Ac E-Ad E-Ae E-Af E-B A B C D E 274 ! THIN SECTION PHOTOMOSAICS s e t i l o t amo r t s k s u T s ’ r a o B Figure A7. Photomicrograph mosaic of a thin section from BT08. 10mm 275 ! Figure A8. Transmitted light photomicrograph mosaics of thin sections BT12-CF-1Ba-c from the BT12-CF-1 stromatolite. Thin sections the same stromatolite slab 1B is shown in Figure A2-A and Figure A2-G. 1Ba 1Bc 1Ba 1Bb 1Bc 5mm 1Ba 1Bb 1Bc 276 ! Figure A9. Photomicrograph mosaics of the thin sections BT12-CF-1Ba-c shown in Figure A8 from the BT12-CF-1 stromatolite under cross-polarized light. 5mm 1Ba 1Bb 1Bc 277 ! Figure A10. Transmitted light photomicrograph mosaic of thin section BT12-CF-1Bd from the BT12-CF-1 stromatolite. 5 mm 278 ! Figure A11. Photomicrograph mosaic of thin section BT12-CF-1Bd from the BT12-CF-1 stromatolite under cross-polarized light. 5 mm 279 ! LaClede Stromatolites Figure A12. Photomicrographs of the LC12-3 stromatolite. A Close-up view of a layer poor in circular holes. B Close-up view of a portion of the stromatolite composed almost entirely of circular holes. C Transmitted light photomicrograph mosaic of the LC12-3 stromatolite slab B, thin section Bc, showing the locations of A and B as insets. D. Cross-polarized light photomicrograph mosaic of the LC12-3 stromatolite slab B thin sections Ba-Bc. Bc Bb Ba 2 mm B C Bc A 2 mm 1 cm A B D 1 cm 280 ! Figure A13. A Transmitted light photomicrograph mosaic of the LC12-E stromatolite slab A thin sections Aa-Af. B Cross-polarized light photomicrograph mosaic of the LC12-E stromatolite slab A thin sections Aa-Af. 1 cm 1 cm A B Aa Ad Ab Ac Ae Af Aa Ad Ab Ac Ae Af 281 Appendix B: Supplemental ICP-MS results 282 FULL TABLE OF RESULTS Table B. ICP-MS results for all Green River Formation stromatolites measured in this study. Measurements in gray text were considered outliers and were excluded from analysis in this thesis. !"#$ !%&' !()* %%+$ ,,&- ,./0 .1#2 ."+3 ..4- 5567 8"59$ !"5: 6;7<=$;<*2;0 >$?07 6$=@*0 A /$B72C D D D D D D @@= @@= @@= D D @@= 9E15 8F1 " % =2C72;0 1F,, !"F( 1F!% " 5 1F!! "F( ( , !5 .1 1F(, 1F!, !( 9E15 8F1 " , =2C72;0 1F.! !5F1 1F"8 " ! 1F!" %F5 1F., 1F!8 !5 9E15 8F1 % % =2C72;0 1F,( !,F! 1F!" " ! 1F!! "F5 . % "% ." 1F55 1F!! !. 9E15 !F1 " 1 =2G0H 1F"% 88F8 1F88 " ( 1F8. !F5 ( , !, "% 1F(8 1F!. !. 9E15 !F1 " 8 =2G0H 1F"( 8%F8 1F8" " . 1F!1 !FI ( % !" %1 1F.5 1F!" !% 9E15 !F1 " ! =2G0H 1F%8 8.F% 1F!" " " 1F!( "F, ( " !( %8 1F., 1F!" !! 9E15 "F1 ! ( =2G0H 1F!5 5FI 1F8I " ( 1F". "F. 5 8 ", ,! 1F,, 1F8, 8% 9E15 "F1 ! 5 =2G0H 1F"" 88F, 1F1I " I 1F!I !F% ( 8 !1 "( 1F.. 1F!8 8. 9E15 "F1 ! I =2G0H 1F"" 88F! 1F85 " % 1F!1 "F! ( ! !5 %5 1F.8 1F!8 8( 9E15 %F1 ! ! J$- 1F!( ,F! 1F15 % 1 1F88 8F( ( " 8" %" 1F.! 1F8. 8! 9E15 %F1 ! % J$- 1F!8 "F. 1F15 % 1 1F8! 8F. I % "" "" 1F.1 1F8( 8" 9E15 %F1 ! , J$- 1F!1 "F" 1F1. " ( 1F88 !F5 1F,( 1F8, 88 9E15 %F1 ! . J$- 1F!" "F! 1F8% % 1 1F8! 8F% ( % 8, "% 1F.8 1F85 8! 9E15 ,F1 8 . =2C72;0 1F%I 8(F! 1F8I % " 1F85 %F( 5 ! ". %( 1F(( 1F!( !! 9E15 ,F1 8 ( =2C72;0 1F%1 8,F! 1F8, " ( 1F!" ,F" 1F.% 1F!8 85 9E15 ,F1 ! 1 =2C72;0 1F%( 8.F, 1F!. % 8 1F"8 ,F. I " %1 %% 1F.( 1F!! !1 9E15 .F1 8 " =2G0H 1F"! (F8 1F8, % ! 1F!1 !F. 1F(1 1F!! 8. 9E15 .F1 8 % =2G0H 1F", 5F( 1F8, % 8 1F8. !F5 I " "% ," 1F.I 1F8I 85 9E15 .F1 8 , =2G0H 1F!" ,F5 1F1, " 8 1F8! 8FI . " 8% "( 1F," 1F8, 8! 283 9E15 .F1 8 I =2G0H 1F!, .F% 1F8, % " 1F"1 ,F, 5 5 "I %. 1F." 1F8, 8% 9E15 (F1 8 J$- 1F8( 1F5 1F!1 % ! 1F!, 1F. 5 1 81 !! 1F," 1F8! 5 9E15 (F1 ! J$- 1F85 1F( 1F1. % " 1F8( 1F% 1F,I 1F8% I 9E15 (F1 " J$- 1F!1 1F5 1F15 % % 1F!1 1F% 5 8 8! !8 1F.1 1F8, 5 9E15 (F, % J$- 1F8I 8F5 1F88 % 8 1F8% 1F. ( . 8% !% 1F,. 1F8% 8! 9E15 (F, , J$- 1F8% 1F5 1F8! % 1 1F!1 1F, 1F,1 1F88 I 9E15 (F, . J$- 1F8% 8F1 1F15 % 8 1F8. 1F, ( ( ( 8( 1F,% 1F8% I 9E15 5F1 ( J$- 1F8, 8F. 1F1( % ! 1F8. 1F. ( I 88 !% 1F,I 1F8( 5 9E15 5F1 5 J$- 1F8" 8F% 1F15 % 1 1F8, 1F, 1F,% 1F8" I 9E15 5F1 I J$- 1F!" 8F! 1F15 % " 1F8" 1F. 5 1 81 ,8 1F.1 1F8( 88 9E15 5F1 ! 8 J$- 1F8, !F8 1F1, % 8 1F8, 8F1 1F.! 1F8( I 9E15 5F, 8 1 J$- 1F!" "F. 1F1, % 1 1F8I 1F5 8 11 8! "" 1F.. 1F!! 88 9E15 5F, 8 8 J$- 1F!! !F" 1F1I " I 1F85 1F. ( ( 8! !I 1F,5 1F8. I 9E15 5F, 8 ! J$- 1F8I 8F. 1F18 % 8 1F8, 1F, 1F." 1F8I 5 9E15 IF1 " . =2C72;0 1F(, !IF, 1F!1 ! . 1F!I IF! I ! .I (8 1F,5 1F85 !I 9E15 IF1 " ( =2C72;0 1F(. "1FI 1F!1 " " 1F"! .F. 5 " "( ,8 1F.( 1F!8 "1 9E15 IF1 " 5 =2C72;0 1F(( "8F, 1F85 ! I 1F!5 .FI 1F.8 1F8( !( 9E15 IF1 % ! =2C72;0 1FI% "(F% 1F8! " , 1F!5 ,F8 . . "I 81, 1F(, 1F!" !. 9E15 81F1 " I =2G0H 1F"5 81F! 1F8! " 5 1F8. !F1 1F,5 1F8! 8! 9E15 81F1 % 1 =2G0H 1F"5 81F( 1F8% " I 1F8. !F1 ( ! 8( "% 1F.1 1F8" 8! 9E15 81F1 % 8 =2G0H 1F%1 81FI 1F1I % 8 1F8( 8FI ( % 8I %1 1F." 1F8" 8! 9E8! Ͳ8Ͳ 8 8 J$- 1F8I !F% 1F8% % " 1F8. 1FI ( % "% 8, 1F.5 1F!, 8. 9E8! Ͳ8Ͳ 8 ! J$- 1F8, 8F! 1F1. % , 1F8% 1F. ( , 8" 81 1F(! 1F!I 8. 9E8! Ͳ8Ͳ 8 " J$- 1F85 !F" 1F!1 % , 1F". 1FI ( % 8! 85 1F.8 1F!8 8% 9E8! Ͳ8 1 % J$- 1F!. %F% 1F88 % ! 1F8. 8F% . 5 8, 8! 1F.5 1F!. 8% 284 9E8! Ͳ8 1 , J$- 1F!, %F" 1F8. " I 1F85 8F" . , 8I 8" 1F.! 1F!% 8, 9E8! Ͳ8 1 . J$- 1F!( "FI 1F85 % , 1F!1 8F" ( % 88 8I 1F.5 1F!. 8. 9E8! Ͳ8 8 ( J$- 1F!% 8F5 1F1( % I 1F85 1F5 5 ! ( 81 1F(5 1F"! 8% 9E8! Ͳ8 8 5 J$- 1F!8 1FI 1F1% % , 1F8. 1F, ( , 5 5 1F(% 1F"1 8" 9E8! Ͳ8 8 I J$- 1F!, 1F( 1F1! , 8 1F8I 1F, 5 . . I 1F5" 1F"% 8, 9E8! Ͳ8 ! 8 1 J$- 1F!" 8F, 1F81 , " 1F!. 1F5 I ! 5 85 1F(I 1F"! !8 9E8! Ͳ8 ! 8 8 J$- 1F8I 1F( 1F1! % I 1F8, 1F, 5 " , ( 1F(, 1F"8 !8 9E8! Ͳ8 ! 8 ! J$- 1F!! 8F! 1F8% , 8 1F!1 1F5 5 . 5 81 1F(I 1F"" !( 9E8! Ͳ8 " 8 " J$- 1F!1 8F( 1F!" % 5 1F!" 8F8 5 " 5 8. 1F(8 1F!5 !5 9E8! Ͳ8 " 8 % J$- 1F!, 8F( 1F8, , . 1F!8 1FI I , I 8( 1F5, 1F", !5 9E8! Ͳ8 " 8 , J$- 1F!1 1FI 1F1. % . 1F8( 1F. ( ( , ( 1F(8 1F!I !! 9E8! Ͳ8 % 8 . J$- 1F". .F! 1F1" , 8 1F85 8FI 5 5 ( 81 1F5( 1F", 85 9E8! Ͳ8 % 8 ( J$- 1F!I %F! 1F1" , 1 1F85 8F" 5 . . !. 1F5% 1F", 8( 9E8! Ͳ8 % 8 5 J$- 1F", (F1 1F85 % , 1F8I !F, 5 " 8% 85 1F(( 1F"! !8 9E8! Ͳ8 , 8 I J$- 1F!1 8FI 1F81 , 1 1F8. 8F1 5 . ( 8! 1F58 1F"! !8 9E8! Ͳ8 , ! 1 J$- 1F!1 !F! 1F8" , , 1F!% 8F1 I , 5 8, 1F5. 1F"% 8I 9E8! Ͳ8 , ! 8 J$- 1F!1 !F8 1F15 , 5 1F8I 1FI 8 11 5 8( 1FI8 1F". !! 9E8! Ͳ8 . ! ! J$- 1F"8 ,F" 1F1, % 5 1F85 8F. 5 8 ( 8. 1F(I 1F"8 8" 9E8! Ͳ8 . ! " J$- 1F", 5F, 1F15 % 1 1F8I !F8 ( 1 8I 8, 1F(! 1F!I 8! 9E8! Ͳ8 . ! % J$- 1F"! .F, 1F1" % % 1F8. 8F( ( % ( 81 1F(I 1F"" 8" 9E8! Ͳ8 ( ! , =2C72;0 1F"8 IF( 1F8( " % 1F!8 !F% . ! !" 8. 1F.1 1F!. 8! 9E8! Ͳ8 ( ! . =2C72;0 1F"1 (FI 1F"8 % 8 1F!! !F" ( " 8( !1 1F.. 1F!5 8, 9E8! Ͳ8 ( ! ( =2C72;0 1F!I (F, 1F"! " 5 1F!" !F8 . I 8( !" 1F.1 1F!% 8% 9E8! Ͳ8 5 ! 5 J$- 1F!% 8F1 1F1. % . 1F8" 1F( ( , ( I 1F(, 1F!5 8" 9E8! Ͳ8 5 ! I J$- 1F!! 8F1 1F1% % , 1F8! 1F. ( ! ( % 1F(" 1F!( 8" 285 9E8! Ͳ8 5 " 1 J$- 1F!5 8F( 1F1( , 8 1F8, 1F5 5 " ( 88 1F5" 1F"! 8! 9E8! Ͳ8 I " 8 J$- 1F%8 %F% 1F15 % , 1F8, 8F" ( " 81 8! 1F(( 1F"8 I 9E8! Ͳ8 I " ! J$- 1F%, "F5 1F1% , ! 1F8( 8F" 5 % I ( 1F5I 1F"( 81 9E8! Ͳ8 I " " J$- 1F"" %F8 1F1( % ( 1F8, 8F! ( % I 5 1F51 1F"" 81 9E8! Ͳ8 8 1 " % =2C72;0 1F", 88F8 1F!8 " 8 1F85 !F1 , ! 88 8( 1F,I 1F!, 8" 9E8! Ͳ8 8 1 " , =2C72;0 1F"I 8%F1 1F8, " " 1F8I !F, , , 8" 8, 1F., 1F!5 8% 9E8! Ͳ8 8 1 " . =2C72;0 1F", 88F% 1F8, " . 1F8( !F8 , 5 81 8% 1F(! 1F"! 8" 9E8! Ͳ! " 8 J$- 1F!I 8F% 1F8% % . 1F!1 1FI ( " , . 1F.I 1F!. !! 9E8! Ͳ! " ! J$- 8F1, "1F, 1F"I , 8 1F"I %F, 5 . 8I !! 8F18 1F%! !5 9E8! Ͳ! " " J$- 1F.. 85F5 1F". " 5 1F!. "F8 . . 8" 8( 1F(. 1F"" 8I 9E8! Ͳ! % % J$- 1F.5 85F, 1F!, % 8 1F!, "F1 ( 8 8" 8. 1F5! 1F". !1 9E8! Ͳ! % , J$- 1F!% 1F5 1F1I " . 1F8, 1F. , I " "( 1F,% 1F!1 8I 9E8! Ͳ! % . J$- 1F"8 8F% 1F8, % . 1F!! 1F5 ( , % I 1F.5 1F!, !1 9E8! Ͳ! . 5 J$- 1F%% (F8 1F1, % , 1F85 !F" ( ! ( I 1F(% 1F!5 !" 9E8! Ͳ! ( 8 1 =2C72;0 1F!% "F1 1F1( " " 1F88 1FI , % , 5 1F,% 1F!8 5 9E8! Ͳ! ( 8 8 =2C72;0 1F!5 %F5 1F1, " ( 1F8" 8F" . 8 , . 1F.8 1F!% I 9E8! Ͳ! ( 8 ! =2C72;0 1F!5 %F( 1F88 " ! 1F8! 8F" , ! . ( 1F,% 1F!! 5 9E8! Ͳ% 8 8 J$- 1F!" "F1 1F1% % ( 1F81 1F5 . ( 8% 5 1F." 1F!1 8" 9E8! Ͳ% 8 ! J$- 1F!8 8F5 1F1" % . 1F88 1F. . I 8% 5 1F.. 1F8I 8% 9E8! Ͳ% 8 " J$- 1F8I 1F. , ( 1F1I 1F% ( 5 % , 1F(1 1F!! 8! 9E8! Ͳ% ! % =2C72;0 1F%8 8!F" 1F1, " 5 1F8! !F" , " ., 81 1F." 1F!. 8% 9E8! Ͳ% ! , =2C72;0 1F"I 8!F1 1F1I ! % 1F88 !F! " . (5 8" 1F"I 1F8( 8! 9E8! Ͳ% ! . =2C72;0 1F"" 5F. 1F1! " ( 1F81 8F( , ! %1 5 1F,( 1F!" 88 9E8! Ͳ% " ( =2G0H 1F!" "F( 1F1( " 5 1F81 8F1 , , 88. . 1F,, 1F8I 8. 9E8! Ͳ% " 5 =2G0H 1F!( ,F% 1F81 " " 1F81 8F" % ( ,8 8! 1F%5 1F8( 81 286 9E8! Ͳ% " I =2G0H 1F!I .F8 1F81 " 5 1F8" 8F% , , 5" 8" 1F.! 1F!! 8! 9E8! Ͳ% % 8 1 =2G0H 1F8I 8F( 1F8, " . 1F8! 1F( , " "I 81 1F,% 1F85 8! 9E8! Ͳ% % 8 8 =2G0H 1F!" 1F. 1F1, % " 1F8! 1F, . " 85 5 1F., 1F!! 81 9E8! Ͳ% % 8 ! =2G0H 1F8. 8F8 1F1( " " 1F81 1F, % ( 81 , 1F,1 1F85 ( 9E8! Ͳ% , 8 " J$- 1F!I 1F5 1F1! % ( 1F8, 1F% ( 8 . I 1F(8 1F!" 5 9E8! Ͳ% , 8 % J$- 1F8I 1F( 1F1% % ! 1F8! 1F% . 1 . . 1F." 1F!! 5 9E8! Ͳ% , 8 , J$- 1F!! 1F. 1F1! % 1 1F88 1F% . 1 % % 1F.1 1F!8 I 9E8! Ͳ% . 8 . =2C72;0 1F,. !!F8 1F88 ! , 1F8I !FI " I . 8, 1F%5 1F85 8( 9E8! Ͳ% . 8 ( =2C72;0 1F.! !!F" 1F8. ! % 1F!8 "F" % 1 81 8( 1F%. 1F8I 8, 9E8! Ͳ% . 8 5 =2C72;0 1F%! 8"F. 1F!8 8 % 1F8" !F1 ! " ( 88 1F!. 1F81 88 9E8! Ͳ% ( 8 I =2G0H 1F!( "F. 1F88 % " 1F8, 8F1 . ( ( I 1F.. 1F!, 8" 9E8! Ͳ% ( ! 1 =2G0H 1F!( "F% 1F81 " ( 1F8( 1FI , ( ( I 1F,. 1F"8 81 9E8! Ͳ% ( ! 8 =2G0H 1F!I %FI 1F8" " 5 1F85 8F! , I ( 5 1F,5 1F!1 81 9E8! Ͳ% 5 ! ! < < 2 H K 1F"% (FI 1F%! " 1 1F8I !F! , " 8, 8I 1F%I 1F!! 81 9E8! Ͳ% 5 ! " < < 2 H K 1F8! 8F% 1F1I 8 , 1F1( 1F% ! " 5 , 1F!8 1F8" % 9E8! Ͳ% 5 ! % < < 2 H K 1F85 !F" 1F85 ! ( 1F8! 1F( % % . 8. 1F%8 1F"" 5 9E8! Ͳ% I ! , J$- 1F!! 1FI 1F15 % . 1F!! 1F, ( 8 % . 1F., 1F!, 8( 9E8! Ͳ% I ! . J$- 1F!, 1F. 1F1" % % 1F8% 1F% ( 1 % . 1F.% 1F!% 8" 9E8! Ͳ% I ! ( J$- 1F!1 1F. 1F1% % 8 1F8! 1F% . % % % 1F.1 1F!" 88 9E8! Ͳ% 8 1 ! 5 J$- 1F!8 8F8 1F88 % 8 1F!8 1F, . % , . 1F,( 1F!8 8. 9E8! Ͳ% 8 1 ! I J$- 1F!1 1FI 1F1I % . 1F!1 1F, ( 1 , . 1F." 1F!" 8, 9E8! Ͳ% 8 1 " 1 J$- 1F!% 1F5 1F1I % I 1F!8 1F, ( ( . ( 1F.I 1F!, 8, 9E8! Ͳ% 8 8 " 8 J$- 1F8, 1F5 1F1! % % 1F8" 1F% . I , 8% 1F.. 1F!% 8( 9E8! Ͳ% 8 8 " ! J$- 1F8( 8F. 1F1" % " 1F8" 1F. . ( " % 1F.( 1F!, 8, 9E8! Ͳ% 8 8 " " J$- 1F8. 1FI 1F18 % , 1F8" 1F% ( 1 % , 1F., 1F!" 8. 287 9E8! Ͳ% 8 ! " % J$- 1F!, 8F5 1F1, % 1 1F8! 1F. . % % ( 1F.8 1F!. 5 9E8! Ͳ% 8 ! " , J$- 1F!% 8F( 1F1% % " 1F8! 1F. . ( " . 1F.( 1F!( 5 9E8! Ͳ% 8 ! " . J$- 1F!, !F" 1F1, % % 1F8" 1F( . I % . 1F.I 1F!. 5 >+8! Ͳ" 8 8 *<' 1F%1 !IF1 1F88 ! . 1F81 1FI % % " , 1F"% 1F1. I >+8! Ͳ" 8 ! *<' 1F., %"F% 1F1. " . 1F8( 8F5 . " . ( 1F,. 1F8! 81 >+8! Ͳ" 8 " *<' 1F%, "1F8 1F1! ! , 1F88 8F! % % . , 1F"5 1F15 % >+8! Ͳ" ! % 1F", !%F8 1F1" ! ( 1F81 8F1 % ( ! ( 1F"" 1F1( " >+8! Ͳ" ! , 1F,% !5F( 1F88 " 1 1F81 1F( , " " !, 1F%" 1F1. . >+8! Ͳ" ! . 8F"" .5F5 1F8( ( ( 1F!! 8F, 8 51 . 8! 8F8" 1F8, ( >+8! Ͳ" " ( 1F5, %%F! 1F8( % I 1F8% 8F1 I 8 " I 1F(! 1F1I , >+8! Ͳ" " 5 1F!5 8%FI 1F1, % ! 1F8( 1F. ( % ! % 1F"% 1F1% " >+8! Ͳ" " I 1F"5 8!F( 1F8, ! ( 1F81 1F% % 5 8 % 1F!. 1F1" ! >+8! Ͳ" % 8 1 1F!1 (FI 1F1( 8 , 1F1, 1F! ! . 8 ! 1F8, 1F1! 8 >+8! Ͳ" % 8 8 1F5" ,,FI 1F8" % 5 1F8, !F! 5 " , 85 1F(% 1F8. I >+8! Ͳ" % 8 ! 8F8, 5!F. 1F"! . " 1F!% !F5 8 1I 81 8" 1FI. 1F85 !8 >+8! Ͳ" , 8 " 8F(8 88"F% 1F!% I I 1F"1 ,F8 8 (% 88 !1 8F," 1F". 8( >+8! Ͳ" , 8 % 1F.5 %,F, 1F1I " 5 1F8! !F1 I 5 , 5 1F.8 1F8% . >+8! Ͳ" , 8 , 1F,% !,F5 1F1. ! . 1F1, 1F% % 5 8 , 1F%1 1F1, " >+8! Ͳ" . 8 . !F,1 88"F5 1F"I I . 1F8( 8F5 8 .. 5 8I 8F(5 1F!% I >+8! Ͳ" . 8 ( 8F"5 .8F! 1F85 , ( 1F88 1FI 8 8( , 81 1FI5 1F8" , >+8! Ͳ" . 8 5 8F!. %,F! 1F"% " 5 1F1I 8F8 ( ( , 88 1F,5 1F8I , >+8! Ͳ" ( 8 I 8F!" .1F8 1F8, , ! 1F8! 8F, ! ,8 ( 81 1F(I 1F!( . >+8! Ͳ" ( ! 1 1F,. !(F, 1F1. ! % 1F1, 1F( . ( " , 1F"( 1F8! ! >+8! Ͳ" ( ! 8 1F%8 !5F1 1F1% ! 8 1F8! 1F( ". " % 1F"8 1F1( 88 >+8! Ͳ" 5 ! ! 1FI. ,5F, 1F81 , ( 1F!" 8F. 88! , 81 1F(! 1F8! 8% 288 >+8! Ͳ" 5 ! " 1F"5 !"FI 1F1, 8 I 1F15 1F( . 1 ! , 1F!( 1F1, , >+8! Ͳ" 5 ! % 1F," "8F! 1F1% ! , 1F1I 1FI % " ! I 1F"( 1F1( , >+8! Ͳ" I ! , 1F.5 %%F! 1F1, % 8 1F85 8F5 ( ! % ( 1F,( 1F88 5 >+8! Ͳ" I ! . 1F!8 8%F, 1F1! 8 ( 1F1( 1F. ! I ! % 1F!1 1F1% " >+8! Ͳ" I ! ( 1F%% !5F" 1F1( " ( 1F8" 8F! . " " , 1F"I 1F1( ( >+8! Ͳ" 8 1 ! 5 1F(" "IF1 1F15 " " 1F81 8F! ( ! 5 I 1F,5 1F!8 , >+8! Ͳ" 8 1 ! I 1F(1 ".F( 1F15 " ! 1F81 8F! , I , ( 1F,% 1F8I % >+8! Ͳ" 8 1 " 1 1F.. "5F5 1F81 " ! 1F8! 8FI 5 , 5 I 1F,. 1F!" % >+8! ͲL 8 8 1F%( 5FI 1F88 % 5 1F!8 1FI ( ( 81 !1 1F." 1F"! 81 >+8! ͲL 8 ! 1F,5 8,F5 1F1I " 5 1F85 1F, , . . %8 1F," 1F!( ( >+8! ͲL 8 " 1F,( 8%F( 1F1. % % 1F!8 1F, . ( % !1 1F.1 1F"1 I >+8! ͲL ! % 1F!. 8F! 1F1! " 5 1F8I 1F! , I ! 8% 1F,! 1F"! , >+8! ͲL ! , 1F%" %F! 1F1, , , 1F"8 1F" 5 % " "5 1F(% 1F%" 5 >+8! ͲL ! . 1F,1 ,F1 1F1. , , 1F"8 1F" 5 , " ,( 1F(1 1F"5 ( >+8! ͲL " ( 1F%% 5F% , 8 1F85 1F! 5 1 " !" 1FII 1F.. , >+8! ͲL " 5 1F,5 IF8 1F1! % 5 1F!" 1F! ( 5 ! I 1F(. 1F%I , >+8! ͲL " I 1F,( 5F" 1F11 % I 1F8I 1F! ( 5 ! 81 1F5. 1F,5 , >+8! ͲL % 8 1 1F.( 8IFI 1F1" % ! 1F!" 1F, . 5 ! !8 1F.% 1F"% % >+8! ͲL % 8 8 1F.8 (F( 1F18 . , 1F!1 1F" 8 1" ! 81 1FI, 1F,1 , >+8! ͲL % 8 ! 1F,8 81FI " ( 1F8. 1F" . 1 8 85 1F,, 1F!I " >+8! Ͳ L , 8 " 1F"I .F8 1F!! " 1 1F8. 8F8 , ( 5 8I 1F%" 1F!! " >+8! ͲL , 8 % 1F,1 5F1 1F!, 8 I 1F88 1F( " % . 8I 1F!( 1F8" ! >+8! ͲL , 8 , 1F,1 (F! 1F1I ! , 1F8% 1F% % 1 " ". 1F", 1F8( ! >+8! ͲL . 8 . 1F.8 8IF1 1F1. " , 1F!1 1F. , . " 8% 1F,! 1F!% " >+8! ͲL . 8 ( 1F." 8"F8 1F1, % 5 1F!% 1F% (( " 85 1F.( 1F"8 % 289 >+8! ͲL . 8 5 1F., 8(F, 1F1( " I 1F!! 1F, . 8 ! !% 1F,( 1F!5 % >+8! ͲL ( 8 I 1F%. 8!F( 1F81 " % 1F!! 1F. , , ! 88 1F%, 1F!1 " >+8! ͲL ( ! 1 1F"% !F1 1F8" " ! 1F8. 1F" , " " 8I 1F"" 1F8" " >+8! ͲL ( ! 8 1F.8 8"F1 1F8! % , 1F!( 1F. I 8 " 8" 1F,I 1F!. % >+8! ͲL 5 ! ! 1F," 8.F5 1F15 % % 1F!I 1F( ( 8 " 8! 1F,! 1F!! " >+8! ͲL 5 ! " 1F,I 8.F8 1F81 % 1 1F!% 1F. . " ! 5 1F," 1F!% " >+8! ͲL 5 ! % 1F(. !"F1 1F8! % 8 1F"1 1F( ( 1 , 8. 1F,( 1F!. , >+8! ͲL I ! , 1FI1 !5F. 1F8% % " 1F!5 1F. ( . . !( 1F." 1F"8 % >+8! ͲL I ! . 1F5( !%F1 1F88 % % 1F!I 1F. ( , , !1 1F.! 1F"1 % >+8! ͲL I ! ( 1F(. 8,F! 1F8% , 8 1F"1 1F. 5 . % 8( 1F." 1F!5 % >+8! ͲL 8 1 ! 5 1F%, (F% 1F15 % ! 1F!, 1F! 5 1 , 8" 1F." 1F". " >+8! ͲL 8 1 ! I 8F1" .F% 1F1, % . 1F!( 1F! 5 ( 88 !% 1F(, 1F%% % >+8! ͲL 8 1 " 1 1F.( %F, 1F15 , 1 1F!, 1F! 5 ( % 8% 1F(5 1F,1 % 290 SOURCES OF CONTAMINATION In order to determine the source of the intrinsic barium fresh samples were drilled from the fan and micrite fabrics of the BT08 stromatolite and from “pure” calcite (a calcite crystal) and aragonite (a clam shell) using different drill bits (to investigate possible contamination from the bits). Drilled samples were rinsed in 10mL ddH 2 O agitated on an orbital shaker at 200 rpm for 18 hours to wash out water-soluble salts, which were removed as supernatant to fresh tubes and analyzed after centrifuging rinsed samples for 10 min at 1350 g. The remaining washed carbonate material was then dried in a convection oven and then dissolved by reacting with 9 mL 0.22M nitric acid under agitation on an orbital shaker at 200 rpm for 1 hour. Insolubles were pelleted out via centrifugation for 10 minutes at 1350 g, and the dissolved carbonate supernatant was transferred to a fresh tube and analyzed. The results of this test (Figures B1-B2) suggested that while drill bit contamination may have been an issue for some elements—notably uranium—there was negligible barium contamination from the drill bit and the barium measured in the samples was sourced primarily from the carbonate (vs. the water-soluble fraction) in both fan and micrite microfabrics (Figures B3-B4). Barium results reported in text, tables, and figures for the BT08 stromatolites are measured values minus the amount of barium added to the samples for the CAS procedure. ! 291 ! Figure B1. Microdrill bit contamination test results showing measured values of key elements as a fraction of the values measured for the stromatolite fan control. Samples drilled to test bit contamination include a calcite crystal (red) using three different drill bits and a sample ground using a mortar and pestle, and an aragonite clam shell (green). The variation caused by drilling calcite with different bits (blue) is equivalent to the variation for the calcite powdered using the different methods. The only elements that appear to be significantly influenced (contaminated) by the microdrill bits are Calcium and Uranium. Fraction vs. stromatolite fan control 300% 320% 0% 20% 40% 60% 80% 100% 120% Na Mg Al Ca Mn Fe Sr Ba U Bit Variation Calcite Crystal Aragonite Clam Shell ! 292 ! Figure B2. Concentrations of select elements measured using different drill bits to powder different samples as a test of contamination by drill bits. Contamination values for all other elements were negligible (see Figure B1). Key to samples measured: 1- Stromatolite fan control drilled from Layer 8 using a fresh Type 106 Dremel ball bit. 2a- Calcite crystal ground using a clean mortar and pestle (no bit). 2b- Calcite crystal drilled using a fresh Type 106 Dremel ball bit. 2c- Calcite crystal drilled using an oxidized/burnt Type 106 Dremel ball bit. 2d- Calcite crystal drilled using a broken carbide bit. 3- Aragonite clam shell drilled using a fresh Type 106 Dremel ball bit. 0.0E+03 1.0E+03 2.0E+03 3.0E+03 4.0E+03 5.0E+03 6.0E+03 1 2a 2b 2c 2d 3 Ca 0.0E+00 1.0E+03 2.0E+03 3.0E+03 4.0E+03 5.0E+03 6.0E+03 7.0E+03 1 2a 2b 2c 2d 3 Fe 0.0 1.0 2.0 3.0 4.0 1 2a 2b 2c 2d 3 U Concentration in Sample (ppm) ! 293 ! Figure B3. Water-soluble (blue) vs. carbonate-bound (red) fraction of key elements in the BT08 stromatolite. Fan samples were measured from Layer 8, micrite samples were measured from layer 9, and blank samples show the negligible contamination from the extraction procedure. 0.0E+00 5.0E+05 1.0E+06 1.5E+06 2.0E+06 2.5E+06 Fan Micrite Blank Na 0.0E+00 2.0E+07 4.0E+07 6.0E+07 8.0E+07 Fan Micrite Blank Mg 0.0E+00 1.0E+05 2.0E+05 3.0E+05 4.0E+05 5.0E+05 Fan Micrite Blank Al 0.0E+00 1.0E+06 2.0E+06 3.0E+06 4.0E+06 Fan Micrite Blank Concentration in Sample (ppb) Ca 0.0E+00 4.0E+05 8.0E+05 1.2E+06 1.6E+06 2.0E+06 Fan Micrite Blank Mn 0.0E+00 4.0E+06 8.0E+06 1.2E+07 1.6E+07 2.0E+07 Fan Micrite Blank Fe 0.0E+00 4.0E+05 8.0E+05 1.2E+06 1.6E+06 Fan Micrite Blank Sr 0.0E+00 1.0E+05 2.0E+05 3.0E+05 4.0E+05 5.0E+05 6.0E+05 Fan Micrite Blank Ba 0.0E+00 2.0E+03 4.0E+03 6.0E+03 8.0E+03 1.0E+04 1.2E+04 Fan Micrite Blank U water soluble carbonate bound KEY: ! 294 ! Figure B4. Fraction of total measured amount of element present in the water-soluble (salt) fraction. The rest of the element measured was bound in carbonate. Results are shown for a fan layer (layer 8, blue) and a micrite layer (layer 9, red). 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Na Mg Al Ca Mn Fe Sr Ba U Water soluble fraction Element Measured Fans Micrite 295 DILUTION CHECKS When some elements (Mg, Ca, and Fe) in the BT08 samples were measured with values outside the range for which the instrument was calibrated, we ran dilutions of samples to get accurate values. Results of the dilutions for the elements run showed that for most isotopes measured ( 23 Na, 24 Mg, 44 Ca, 55 Mn, 56 Fe, 88 Sr, 137 Ba, 138 Ba, and 238 U), undiluted values measured were approximately equal to the values measured for diluted samples even when the undiluted sample values fell well outside the calibrated range (Figure B5). Because of this, for future analyses, undiluted values for the “well-behaved” isotopes listed above were used even when they fell outside the instrument calibration range. The isotopes measured for which this was not the case were 27 Al, 43 Ca, and 86 Sr. 44 Ca and 88 Sr were used instead of 43 Ca and 86 Sr, respectively, in subsequent analyses so these deviations were not of concern in this study. For these and other isotopes that were not “well-behaved”, values measured outside the instrument calibration range were discarded and not used in subsequent data analysis. ! 296 ! Figure B5. Check on measured dilution values for ICP-MS measurements of different elements. Values determined by measuring undiluted samples are plotted on the x-axis and values determined by measuring 1:10 diluted samples corrected for the dilution value are plotted on the y-axis. The dashed line in each image is the 1:1 line that would be expected if the values read for dilutions, when adjusted for the dilution amount, give the same values as the initial samples. All values are in ppb. y = 0.92x R ! = 0.80 1.5E6 3.0E6 1.5E6 3.0E6 23Na y = 0.89x R ! = 1.00 5.E7 1.E8 5.E7 1.E8 24Mg y = 1.04x R ! = -3.07 6.0E5 1.2E6 6.0E5 1.2E6 27 Al y = 0.97x R ! = 0.92 0 2.E6 4.E6 2.0E6 4.0E6 Ca 43Ca 44Ca y = 0.87x R ! = 0.99 8.0E5 1.6E6 8.0E5 1.6E6 55Mn y = 0.93x R ! = 1.00 1.5E7 3.0E7 1.5E7 3.0E7 56Fe y = 0.90x R ! = 0.93 1.5E6 3.0E6 1.5E6 3.0E6 Sr 86Sr 87Sr 88Sr y = 0.87x R ! = 0.98 y = 0.89x R ! = 0.99 1.5E6 3.0E6 1.5E6 3.0E6 Value of Undiluted Sample 138Ba 137Ba 138Ba y = 0.86x R ! = 0.98 5.E3 1.E4 5.E3 1.E4 238U 0 0 0 0 0 0 0 0 Value of Diluted Sample 297 Appendix C: Supplemental standard stable isotope results 298 BOAR’S TUSK STROMATOLITES Table C1. All standard stable isotope measurements for the Boar’s Tusk stromatolites. Samples or measurements in gray text were considered outliers and were not included in analyses. 6;7<=$;<*2;0 >$?07 6$=@*0 A /$B72Cɷ 8 " + MN OPQ9Rɷ 8 5 S MN O P Q 9 R 9E15 8F1 " % =2C72;0 %F8.,Ͳ 8F.." 9E15 8F1 " , =2C72;0 %F!!.Ͳ 8F,55 9E15 8F1 % % =2C72;0 "F(!"Ͳ !F11I 9E15 !F1 " 1 =2G0H "F.!"Ͳ !F(," 9E15 !F1 " 8 =2G0H "F!!.Ͳ "F!5! 9E15 "F1 ! ( =2G0H !F1!(Ͳ %F.!% 9E15 "F1 ! 5 =2G0H 1F5""Ͳ .F.(I 9E15 "F1 ! I =2G0H 8F%.%Ͳ .F1%1 9E15 %F1 ! ! J$- !F1!1Ͳ %F!81 9E15 %F1 ! % J$- !F"5,Ͳ "F5., 9E15 %F1 ! , J$- 8F55!Ͳ %F%"8 9E15 ,F1 8 . =2C72;0 "F"..Ͳ !F(!1 9E15 ,F1 8 ( =2C72;0 !F%I,Ͳ ,F!,( 9E15 ,F1 ! 1 =2C72;0 !F,,.Ͳ %F!.. 9E15 .F1 8 " =2G0H !F5!1Ͳ "F%." 9E15 .F1 8 % =2G0H !FII1Ͳ "F8II 9E15 (F1 8 J$- Ͳ1F.,.Ͳ IF... 9E15 (F1 ! J$- 8F%.8Ͳ ,F.., 9E15 (F1 " J$- 8F5!IͲ %FI8, 9E15 (F, % J$- 1FI,8Ͳ .F"I( 9E15 (F, , J$- 8F!I(Ͳ .F!I( 9E15 (F, . J$- 8F",5Ͳ ,F.8% 9E15 5F1 ( J$- !F1!1Ͳ %F"I! 9E15 5F1 5 J$- 8F!,5Ͳ ,F5I. 9E15 5F1 I J$- 8F(,"Ͳ %F.!! 9E15 5F, 8 1 J$- 8F5I.Ͳ %F%(5 9E15 5F, 8 8 J$- 8F5I5Ͳ %F""8 9E15 5F, 8 ! J$- 8F"".Ͳ %FI.. 9E15 IF1 " . =2C72;0 %F8.5Ͳ 8F%1" 9E15 IF1 " ( =2C72;0 %F,",Ͳ 1F5"1 9E15 IF1 " 5 =2C72;0 %F%..Ͳ 1FI"1 9E15 81F1 " I =2G0H "F(1IͲ !FI%5 9E15 81F1 % 1 =2G0H "F(18Ͳ !F5I! 9E15 81F1 % 8 =2G0H "F(11Ͳ !FII. 9E8! Ͳ8Ͳ 8 8 J$- !F888Ͳ %F!81 9E8! Ͳ8Ͳ 8 ! J$- !F!""Ͳ "FI., 299 9E8! Ͳ8Ͳ 8 " J$- 8F"1,Ͳ .F"1" 9E8! Ͳ8 1 % J$- !F8,"Ͳ "FI(, 9E8! Ͳ8 1 , J$- 8FI5"Ͳ %F!81 9E8! Ͳ8 1 . J$- 8F(."Ͳ %F(.5 9E8! Ͳ8 8 ( J$- 8F(!,Ͳ %F!5I 9E8! Ͳ8 8 5 J$- 8F.%IͲ %F%5( 9E8! Ͳ8 8 I J$- 8F,".Ͳ %F,8" 9E8! Ͳ8 ! 8 1 J$- 8F"(.Ͳ ,F"%, 9E8! Ͳ8 ! 8 8 J$- 8F((.Ͳ %F8.! 9E8! Ͳ8 ! 8 ! J$- 8F.%IͲ %F.11 9E8! Ͳ8 " 8 " J$- 8F,1(Ͳ ,F8!I 9E8! Ͳ8 " 8 % J$- 8FI1.Ͳ %F1", 9E8! Ͳ8 " 8 , J$- 8F51!Ͳ %F"(% 9E8! Ͳ8 % 8 . J$- !F%,.Ͳ "F511 9E8! Ͳ8 % 8 ( J$- !F1..Ͳ %F!.1 9E8! Ͳ8 % 8 5 J$- !F%"%Ͳ %F1!I 9E8! Ͳ8 , 8 I J$- !F8%5Ͳ "F((( 9E8! Ͳ8 , ! 1 J$- 8F.!"Ͳ %FI,I 9E8! Ͳ8 , ! 8 J$- !F1.8Ͳ %F1"1 9E8! Ͳ8 . ! ! J$- !F."%Ͳ "F,!, 9E8! Ͳ8 . ! " J$- !F5("Ͳ "F!,( 9E8! Ͳ8 . ! % J$- !F(,5Ͳ "F".5 9E8! Ͳ8 ( ! , =2C72;0 !F.8!Ͳ "F1(% 9E8! Ͳ8 ( ! . =2C72;0 !F(%IͲ !FI!5 9E8! Ͳ8 ( ! ( =2C72;0 !FI!,Ͳ !F.%( 9E8! Ͳ8 5 ! 5 J$- !F%55Ͳ "F.8! 9E8! Ͳ8 5 ! I J$- !F%(!Ͳ "F5!% 9E8! Ͳ8 5 " 1 J$- !F"1%Ͳ "F5%. 9E8! Ͳ8 I " 8 J$- !F,58Ͳ "F,." 9E8! Ͳ8 I " ! J$- !F%.8Ͳ "F5%! 9E8! Ͳ8 I " " J$- !F""1Ͳ "F.(, 9E8! Ͳ8 8 1 " % =2C72;0 "F!"(Ͳ !F,.1 9E8! Ͳ8 8 1 " , =2C72;0 "F8,.Ͳ !F"!I 9E8! Ͳ8 8 1 " . =2C72;0 "F15%Ͳ !F%5. 9E8! Ͳ! " 8 J$- 8F%.(Ͳ ,F!"5 9E8! Ͳ! % , J$- !F,88Ͳ %F115 9E8! Ͳ! % . J$- !F,!(Ͳ "F5I! 9E8!Ͳ! . 5 J$- !F%8"Ͳ "FI!, 9E8! Ͳ! ( 8 1 =2C72;0 "F58,Ͳ 8F.1( 9E8! Ͳ! ( 8 8 =2C72;0 "F",.Ͳ !F1(! 9E8! Ͳ! ( 8 ! =2C72;0 "F%%1Ͳ !F1!( 300 9E8! Ͳ% $ 8 8 J$- !F11"Ͳ %F%(. 9E8! Ͳ% $ 8 ! J$- 8F",,Ͳ .F%.. 9E8! Ͳ% $ 8 " J$- 8F.!,Ͳ %FI88 9E8! Ͳ% $ ! % =2C72;0 "F8,"Ͳ !F,.8 9E8! Ͳ% $ ! , =2C72;0 "F8I8Ͳ !F.,I 9E8! Ͳ% $ ! . =2C72;0 "F8!(Ͳ !F(I, 9E8! Ͳ% $ " ( =2G0H !F18%Ͳ %F8"! 9E8! Ͳ% $ " 5 =2G0H !F!((Ͳ "F.,% 9E8! Ͳ% $ " I =2G0H 8F((IͲ ,F15! 9E8! Ͳ% $ % 8 1 =2G0H 8FI8.Ͳ %F%"1 9E8! Ͳ% $ % 8 8 =2G0H 8FI!5Ͳ %F"I% 9E8! Ͳ% $ % 8 ! =2G0H 8FI%!Ͳ %F"8. 9E8! Ͳ% $ , 8 " J$- 8FI%8Ͳ %F,%8 9E8! Ͳ% $ , 8 % J$- 8F(!5Ͳ %F"55 9E8! Ͳ% $ , 8 , J$- 8F5%1Ͳ %F%,! 9E8! Ͳ% $ . 8 . =2C72;0 %F,18Ͳ 8F"85 9E8! Ͳ% $ . 8 ( =2C72;0 %F8IIͲ 8F.I8 9E8! Ͳ% B ( 8 I =2G0H !F885Ͳ "F5(. 9E8! Ͳ% B ( ! 1 =2G0H !F%5.Ͳ "F,"" 9E8! Ͳ% B ( ! 8 =2G0H !F.(8Ͳ "F%8% 9E8! Ͳ% B 5 ! ! < < 2 H K !F"IIͲ "F88. 9E8! Ͳ% B 5 ! " < < 2 H K 8F,."Ͳ ,F.I" 9E8! Ͳ% B 5 ! % < < 2 H K !F.,.Ͳ "F(88 9E8! Ͳ% B I ! , J$- 1FII!Ͳ ,F5I% 9E8! Ͳ% B I ! . J$- 8F,88Ͳ %F(1" 9E8! Ͳ% B I ! ( J$- 8F%%!Ͳ %F5"! 9E8! Ͳ% B 8 1 ! 5 J$- 8F81%Ͳ ,F,I( 9E8! Ͳ% B 8 1 ! I J$- 8F1"IͲ ,FI!. 9E8! Ͳ% B 8 1 " 1 J$- 8F,1IͲ ,F1%! 9E8! Ͳ% B 8 8 " 8 J$- 8F.I(Ͳ %F!,! 9E8! Ͳ% B 8 8 " ! J$- 8F(1%Ͳ %F!%5 9E8! Ͳ% B 8 8 " " J$- 8F!,(Ͳ %F.I% 9E8! Ͳ% B 8 ! " % J$- 8F.!.Ͳ %F.58 9E8! Ͳ% B 8 ! " , J$- !FI!!Ͳ !F.!I 9E8! Ͳ% B 8 ! " . J$- !F1%8Ͳ %F81. 301 LACLEDE STROMATOLITES Table C2. All standard stable isotope measurements for the LaClede stromatolites. Measurements in gray text were considered outliers and were excluded from analysis in this thesis. 6;7<=$;<*2;0 >$?07 T0*F *$?07 MC = R 6$=@*0ɷ 8 " + MN OPQ9Rɷ 8 5 S MN O P Q 9 R >+8! Ͳ" 8 8F8 8 !F!%51Ͳ !F(8!1 >+8! Ͳ" 8 8F8 ! !F..(1Ͳ 8F88,1 >+8! Ͳ" 8 8F8 " !F%.,1Ͳ 8F1%(1 >+8! Ͳ" ! !F1 % 1F,%"1Ͳ !F,151 >+8! Ͳ" ! !F1 , 8F88"1Ͳ 8FI,"1 >+8! Ͳ" ! !F1 . 8F!5I1Ͳ 8FI"(1 >+8! Ͳ" " !F( ( 8F"!%1Ͳ "F!".1 >+8! Ͳ" " !F( 5Ͳ 1F",%1Ͳ ,F,%(1 >+8! Ͳ" " !F( IͲ 8FI%11Ͳ %FI1%1 >+8! Ͳ" % %F% 8 1 8F.!!1Ͳ 8F!((1 >+8! Ͳ" % %F% 8 8 !F,!"1Ͳ 8F,I11 >+8! Ͳ" % %F% 8 !Ͳ 1F1%.1Ͳ "F(,%1 >+8! Ͳ" , ,F( 8 " 1F51.1Ͳ %F",81 >+8! Ͳ" , ,F( 8 % 1F,",1Ͳ "F8I!1 >+8! Ͳ" , ,F( 8 , 1F,,%1Ͳ ,F5851 >+8! Ͳ" . .F8 8 . 8FI"!1Ͳ 81F5!%1 >+8! Ͳ" . .F8 8 ( 8F,I!1Ͳ 81F1I(1 >+8! Ͳ" . .F8 8 5 8FI1%1Ͳ 88F!,%1 >+8! Ͳ" ( (F8 8 I 8F,(!1Ͳ 8F"I(1 >+8! Ͳ" ( (F8 ! 1 !F8("1Ͳ 8F!(.1 >+8! Ͳ" ( (F8 ! 8 !F8"81Ͳ 8F!.(1 >+8! Ͳ" 5 (F. ! !Ͳ 1F58I1Ͳ !F1(I1 >+8! Ͳ" 5 (F. ! "Ͳ 1F!(.1Ͳ 8F!.%1 >+8! Ͳ" 5 (F. ! %Ͳ 1F"5(1Ͳ !F88!1 >+8! Ͳ" I 5F% ! ,Ͳ 1F8!"1Ͳ 1F.%!1 >+8! Ͳ" I 5F% ! . 1F!1!1Ͳ 1F(I(1 >+8! Ͳ" I 5F% ! ( 1F!511Ͳ 1F(5!1 >+8! Ͳ" 8 1 IF, ! 5 "F%(%1Ͳ 1F,551 >+8! Ͳ" 8 1 IF, ! I "F.151Ͳ 1F,551 >+8! Ͳ" 8 1 IF, " 1 %F1551Ͳ 8F8"(1 >+8! ͲL 8 8F8 8 !FI,.1Ͳ ,FI,"1 >+8! ͲL 8 8F8 ! "F1..1Ͳ %F!(!1 302 >+8! ͲL 8 8F8 " "F",!1Ͳ %F%.51 >+8! ͲL ! !F, % !F!.81Ͳ .F8511 >+8! ͲL ! !F, , !F%,!1Ͳ ,FI,I1 >+8! ͲL ! !F, . !F"(81Ͳ .F!111 >+8! ͲL " ,F1 ( !F,!11Ͳ %F"1(1 >+8! ͲL " ,F1 5 !F!I11Ͳ %F.%(1 >+8! ͲL " ,F1 I !F.1I1Ͳ %F"%%1 >+8! ͲL % ,F! 8 1 !F(581Ͳ !F("I1 >+8! ͲL % ,F! 8 8 8F((.1Ͳ %F%511 >+8! ͲL % ,F! 8 ! 8FI181Ͳ %F1(I1 >+8! ͲL , ,F% 8 " !F%I%1Ͳ "FI1,1 >+8! ͲL , ,F% 8 % !F(181Ͳ "F!,,1 >+8! ͲL , ,F% 8 , !F"I81Ͳ "F.811 >+8! ͲL . ,F, 8 . !FI1!1Ͳ !F5(51 >+8! ͲL . ,F, 8 ( 8F(.81Ͳ %F"",1 >+8! ͲL . ,F, 8 5 !F.8!1Ͳ "F!1.1 >+8! ͲL ( .F" 8 I !F1!I1Ͳ %F!5.1 >+8! ͲL ( .F" ! 1 8F1%I1Ͳ (F,"%1 >+8! ͲL ( .F" ! 8 8F",51Ͳ %FI(!1 >+8! ͲL 5 (F8 ! ! !FI..1Ͳ ,F5,51 >+8! ͲL 5 (F8 ! " 8F.%%1Ͳ %F("(1 >+8! ͲL 5 (F8 ! % !FI5"1Ͳ "F8%%1 >+8! ͲL I 5F! ! , !FI5(1Ͳ !F,8"1 >+8! ͲL I 5F! ! . !F,(.1Ͳ "F1,51 >+8! ͲL I 5F! ! ( !F8811Ͳ %F(%%1 >+8! ͲL 8 1 IF5 ! 5 !F,(I1Ͳ ,F11I1 >+8! ͲL 8 1 IF5 ! I !F.181Ͳ %F"!%1 >+8! ͲL 8 1 IF5 " 1 !F(..1Ͳ %F5!I1 >+8! ͲL! 8 8F, 8 "F.%(1Ͳ %F",5, >+8! ͲL! ! 8F( ! "F,,.1Ͳ %F(%", >+8! ͲL! " !F8 " !F."I1Ͳ "F."!, >+8! ͲL! % !F" %Ͳ 1F1!I1Ͳ 8!F5I51 >+8! ͲL! , !F( , 8F.(.1Ͳ .F,!11 >+8! ͲL! . "F" . "FI1"1Ͳ !FI8.1 >+8! ͲL! ( "F. ( "F.1%1Ͳ "F"."1 >+8! ͲL! 5 %F1 5 1F%I(1Ͳ (F8!%1 >+8! ͲL! I %F" I 8F(,"1Ͳ .F8%., >+8! ͲL! 8 1 %F( 8 1 !F%!!1Ͳ %F!%11 303 >+8! ͲL! 8 8 ,F1 8 8 1F58%1Ͳ .F"851 >+8! ͲL! 8 ! ,F! 8 ! !F,,!,Ͳ "F11.1 >+8! ͲL! 8 " ,F. 8 " !F!5!1Ͳ !F5%,, >+8! ͲL! 8 % ,F5 8 % !F%(%1Ͳ "F8"%1 >+8! ͲL! 8 , .F8 8 , 8FI(81Ͳ ,F(81, >+8! ͲL! 8 . .F. 8 . 8F5881Ͳ ,F""51 >+8! ͲL! 8 ( .FI 8 ( 8F%%51Ͳ .FI!11 >+8! ͲL! 8 5 (F8 8 5 !F5.(1Ͳ !FI8!1 >+8! ͲL! 8 I (F. 8 I 8F18,1Ͳ ,F(8!, >+8! ͲL! ! 8 5F, ! 8 !F8I11Ͳ ,F"!!1 >+8! ͲL! ! ! 5F5 ! ! !F"."1Ͳ 81F1(,1 >+8! ͲL! ! " IF8 ! " 1F5(11Ͳ 5F,(.1 >+8! ͲL! ! % IF% ! % !F(1.1Ͳ (F"!81 304 Appendix D: Detailed clumped isotope methods and results METHODS Measurement Approximately 8-12 mg samples of calcium carbonate (or 8 mg of a calcium carbonate standard) were reacted for 20 minutes on a 90°C online common phosphoric acid bath system described in detail elsewhere (Passey et al, 2010). Briefly, CO 2 gas was immediately frozen by liquid nitrogen, after passing through a dry ice/ethanol trap. Cryogenic purification of CO 2 was achieved using an automated online vacuum line. Additional automated sample cleanup steps included passing sample gas through a Porapak Q TM 120/80 mesh GC column at -20°C to remove potential organic contaminants and silver wool (Sigma-Aldrich) to remove sulfur compounds. G 13 C, G 18 2 , ' 47 and ' 48 in CO 2 derived from the phosphoric acid digestion of carbonates were determined using a Thermo Scientific MAT 253 gas-source mass spectrometer using published configuration and methods (Ghosh, et al 2006). Samples were run to ensure mass 44 voltages remained stable at 16 V throughout the course of each analysis. Typical precision was 0.005 to 0.009‰, equivalent to about 1-2 qC, consistent with other studies (Eiler, 2007; Huntington et al., 2009; Eagle et al., 2009; 2010; Tripati et al., 2010; Thiagarajan et al., 2011). Carbonate standards were run between every 3-4 samples and were prepared and analyzed in the same manner as samples. 1000 qC and 25 q C equilibrated gas standards were also run each day. Samples were screened for the presence of contaminating molecules (such as hydrocarbons and sulfur compounds) using mass 48 anomalies. ' 48 values are calculated in the same way as ' 47 values, by references to the ' 48 stochastic distribution as defined by the analysis 305 of heated gases. Samples with large measured deviations (>1‰) from the ' 48 heated gas line and large internal errors were considered potentially indicative of the presence of contaminants and were excluded from further analysis. In total we conducted 12 analyses of samples microdrilled from four locations, two each from two different BT08 stromatolite layers. Calculations to derive clumped isotope values and their errors As sample reactions were carried out at 90 qC, we applied a published (Passey 2010) empirically derived acid digestion fractionation correction of 0.08‰ for ' 47 values to facilitate comparison to the published calcite line calibration in which samples were reacted at 25 qC (Ghosh 2006). Data that have been collected thus far for aragonite, dolomite, and calcite indicate no discernible mineral-specific acid digestion effects. Errors in reported ' 47 values and calculated temperatures include the propagated uncertainty in heated gas determination and in sample measurement (Huntington et al., 2009). The majority of data for this study was collected before the proposition of an “absolute reference frame” for clumped isotope studies of CO 2 based on the analysis of water equilibrated CO 2 gases. We therefore present data in tables both relative to the stochastic value (i.e., the nomenclature used in most previous studies) and in the absolute reference frame, while results in figures and described in the main text are reported in the absolute reference frame. Our conclusions are insensitive to which reference frame the data are reported in. Data in the absolute reference frame are generated using an empirical transfer function developed using data for heated gases, an in-house Carrara Marble standard (which was run the most frequently), an NBS- 19 standard, and a vein calcite in-house standard (102-GC-AZ01) following the procedure described elsewhere (Dennis, 2011). Average measured values on for Carrara Marble (n = 12) 306 relative to the stochastic distribution are 0.353 ±0.006 while reported values based on a large (>60) number of measurements are 0.352‰. The average measured value for Carrara Marble on the absolute reference frame is 0.391‰, identical to the value of Dennis et al., 2011. The average measured G 18 O (V-PDB) for NBS-19 (n = 2) was -2.244, close the accepted value of -2.19‰ and the G 13 C (V-PDB) was 2.035‰, similar to the accepted value of 1.95‰. Sample carbonate G 18 O and G 13 C data are reported relative to the VPDB standard. 307 RESULTS Table D. Clumped isotope results. Isotope measurements presented represent the average (and corresponding standard deviation ı ) of 8 acquisitions on a single sample. Samples were microdrilled from the BT08 stromatolite. BTF1 and BTF2 samples were drilled from two different locations along calcite fan Layers 8-8.5. BTM1 and BTM2 were likewise drilled from two different locations along dolomicrite Layer 9. Samples in gray text were excluded from analyses due to suspicions of drilling into secondary cement phase (Sample 3) or high ȴ 48 ı values indicating contamination with organic material during measurement (Samples 1 and 6). Temperatures presented here were calculated by comparison to the stochastic value using the method described by Ghosh et al. (2006) as well as using the absolute reference frame described by Dennis et al. (2011). ɷ 8" + MPQ9R ɷ 85 S'$K M6&S UR ɷ 85 SC$* C2 ; 0 MPQ9R ɷ %( MSV R ȴ %( MSV R ɷ %5 MSV R ȴ %5 MSV R )C2H H2'0K; C<7 7 0C; 2 <- WX<KX E MY+R )BK<*3;0 70J F J7 $=0 C<7 7 0C; 2 <- Q0--2K E MY+R A 6$=@*0 ZQ !" # ʍ !" # !" # ʍ !" # ʍ !" # ʍ $%&''(' !" # !" # ʍ 8 9E/8 !F 11 ( 1F 11 ( "%F , . 5Ͳ %F ." " 1F 18 % 8%F 5 " 1F 1(Ͳ 1F 1I I 1F 1, 8 1F 18 5 !.F ( I " (F I1 " 8F (% ( 1F ." 8 !5F , 1F .( ( "!F ! ! 9E/8 %F ,% ! 1F 18 " "%F , , .Ͳ %F .% % 1F 18 . 8(F " % 1F 1,Ͳ 1F 1, ( 1F 1% % 1F 18 , !.F ( 8 5 (F 5% ! 1F 5! ( 1F .% . !,F 8 1F .I " !5F 5 " 9E/8 8F I. 8 1F 11 " "%F , 5 !Ͳ %F .! 1 1F 11 ( 8%F I " 1F 1"Ͳ 1F 18 1 1F 1! I 1F 18 1 ",F % % ! 8.F " 5 . 1F .! % 1F ,5 , "IF . 1F .% 8 %1F . % 9E/8 8F I! I 1F 11 " "%F , 5 (Ͳ %F .8 , 1F 11 % 8%F I , 1F 1% 1F 1" , 1F 1% 8 1F 18 , "%F 5 . 5 8,F 5 8 " 1F %I 1 1F ." 8 !5F , 1F .5 I !IF ( , 9E/! !F 1" I 1F 11 , "%F ( 1 %Ͳ %F ,1 " 1F 11 I 8,F 1 ! 1F 1,Ͳ 1F 15 " 1F 1" 5 1F 18 " !(F , ! 1 5F ", " 1F 5( , 1F .% . !,F 1 1F .I % !5F . . 9E/! %F ,, " 1F 11 . "%F , ( "Ͳ %F .! 5 1F 1" 8 8(F " I 1F 1.Ͳ 1F 1" 8 1F 1" ( 1F 18 " !(F % " 1 5F ,1 5 8F "5 . 1F .( % 85F I 1F (! " !!F . ( 9E/! !F 1" % 1F 11 ! "%F . 5 !Ͳ %F ,! " 1F 11 , 8,F 8 . 1F 1" 1F 1, 8 1F 1! % 1F 11 5 ",F , ! ( 8.F ! ( ! 1F ,1 , 1F .% ! !,F I 1F (1 8 !(F 8 5 9E&8 %F "! 1 1F 11 , "(F . " .Ͳ 8F .5 ! 1F 11 . !1F ! 8 1F 1.Ͳ 1F 1% I 1F 1, , 1F 18 I ""F ( ( I 5F (. . 1F 58 " 1F .8 I "8F " 1F .. , ",F 1 I 9E&8 %F "5 ! 1F 11 " "(F 5 . (Ͳ 8F %, I 1F 11 . !1F ( " 1F 1" 1F 8" I 1F 1! " 1F 11 5 %(F " % 1 !8F , . ! 1F %! ( 1F .1 I ""F ( 1F .. . "%F 5 81 9E&! %F !, 8 1F 11 " "(F I !!Ͳ 8F %1 ( 1F 11 " !1F . , 1F 1% 1F 8" % 1F 1" 5 1F 18 " %(F . . , !8F ( ( % 1F ,1 , 1F .1 , "%F ( 1F .. ! ",F 5 88 9E&! %F !. ! 1F 11 " "(F I ! 5Ͳ 8F %1 1 1F 11 . !1F . 5 1F 1" 1F 8" 5 1F 1" % 1F 18 ! %(F ! . 5 !8F " ( " 1F "5 5 1F .1 I ""F ( 1F .. . "%F 5 8! 9E&! %F !8 1 1F 11 ( "(F . ! (Ͳ 8F .I 1 1F 18 1 !1F 8 1 1F 1,Ͳ 1F 1% I 1F 1% % 1F 18 . "%F 8 ( 1 IF 8. , 1F (5 8 1F .! 1 "8F 8 1F .. . "%F 5 308 REFERENCES Dennis, K.J., Affek, H.P., Passey, B.H., Schrag, D.P., and Eiler, J.M., 2011, Defining an absolute reference frame for “clumped” isotope studies of CO2: Geochimica et Cosmochimica Acta, v. 75, no. 22, p. 7117–7131, doi: 10.1016/j.gca.2011.09.025. Ghosh, P., Adkins, J., Affek, H., Balta, B., Guo, W., Schauble, E.A., Schrag, D., and Eiler, J.M., 2006, 13C–18O bonds in carbonate minerals: A new kind of paleothermometer: Geochimica et Cosmochimica Acta, v. 70, no. 6, p. 1439–1456, doi: 10.1016/j.gca.2005.11.014. 309 Appendix E: Matlab code CODE DESCRIPTIONS Code included in this appendix is heavily commented to make the function of code elements understandable. General descriptions of the purpose of the code functions in this appendix follow here. importSpreadsheet Code for this function imports the Excel spreadsheet containing the compiled ICP-MS and isotope data (along with outlier-deleted data) and packages and pre-processes the key the data for plotting and additional analysis in Matlab. ionModel Code for this function calculates lake volumes using inputted conserved ion concentrations averaged by stromatolite layer and an assumed minimum lake volume and lake depth. The equations used in this model are explained in the model methods section of this thesis. The code also plots lake volume and depth changes corresponding to each layer. This function references the function herrorbar, which is plots data with horizontal (x- axis) error bars and was downloaded from the Matlab Central File Exchange (van der Geest, 2006). carb2water Code for this function converts measured carbonate ɷ 18O (VPDB) values to the water isotope values from which the carbonates precipitated assuming equilibrium precipitation. This calculation takes temperature and mineralogy (calcite vs. dolomite) into account using the 310 calculations discussed in the methods section on calculating water isotope values from carbonate isotope values. scaleT Code for this function calculates a temperature from an XMg value assuming that XMg scales linearly with temperature. This scale is based on measured XMg and clumped isotope temperature values for BT08 stromatolite Layers 8-8.5 (fan) and Layer 9 (micrite). RayleighModel Code for this function calculates lake volume and depth from a set of water ɷ 18 O values. Also needed are a list of layer numbers, water temperatures corresponding to each į 18 O value, the assumed į 18 O of input waters, and the assumed minimum lake depth and volume. This function references the function herrorbar, which is plots data with horizontal (x-axis) error bars and was downloaded from the Matlab Central File Exchange (van der Geest, J., 2003, File ID#3963: http://www.mathworks.com/matlabcentral/fileexchange/3963-herrorbar, accessed November 2012). BTModelSensitivityPlot Code for this function runs the oxygen isotope model (using the function RayleighModel) for seven different temperature scenarios and for the full range of freshwater input į i values in order to assess the impact of these parameters on the maximum lake depth (and therefore lake depth change) calculated. It was used to produce the plots shown in Chapter 2, Figures 29-32. findClosedSequences Code for this function finds layer sequences that display potential closed system behavior, i.e. those layer sequences for which a best-fit line through a crossplot of ɷ 18 O vs. ɷ 13 C gives an r value ш 0.7. It returns a table of the sequences and their best-fit line statistics. 311 CODE importSpreadsheet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` SUHIL[ ^ DOO RG ` 1DPHVRIWKHFROXPQKHDGHUVWRNHHSPD\DOVRQHHGWRDGGPRGL I\FRGH EHORZLIWKLVLVFKDQJHG PHWDKHDGV ^ 6WURP &RGH /D\HU 6DPSOH1XPEHU )DEULF&RGH ` 'DWDKHDGV GDWDKHDGV ^ 1D 0J $O &D 0Q )H 1L &X =Q 6U %D 8 70 ;0J G&93'% G293'% ` FKHPFROV ^ 1D 0J $O &D 0Q )H 1L &X =Q 6U %D 8 70 ;0 J G& G2 ` 6WURPDWROLWHQDPHVFRUUHVSRQGLQJWRWKHVWURPDWROLWHQXPEHUVL Q 6WURP VWURPQDPHV ^ %7 %7 %7 %7 /& /&( ` VWURPQXPV >@ &2'('(6&5,37,21 6HOHFW UHDGLQWKHVSUHDGVKHHW >ILOHQDPHILOHSDWK@ XLJHWILOH^ [OV[ [OV ` )RUORRSLPSRUWVHDFKVKHHWDVDGLIIHUHQWYDULDEOH IRU V VL]HVKHHWQDPHV 5HDGLQWKHVKHHWDVVXPHVWKDWWKHILUVWFROXPQLVDGDWD FROXQQRWKHUZLVHZLOOQHHGWRDGGEODQNFROXPQVWRWKHIUR QWRIWKH QXPEHUVPDWUL[WRPDNHWKHGDWDDQGWKHKHDGHUVPDWFKXS >QXPEHUVVWULQJV@ [OVUHDGVWUFDWILOHSDWKILOHQDPHVKHH WQDPHV^V` 312 7ULPWKHGDWD E QXPEHUV ILQGGDWDVWDUWDQGHQG >KHDGURZVDPSOHFRO@ LQGVXEVL]HVWULQJVVWUPDWFK 6DPSOH 1XPEHU VWULQJV GHOHWHURZVWKDWDUHQ WVDPSOHV ELVQDQEVDPSOHFRO >@ *UDEWKHGDWDRILQWHUHVW FRQYHUWSUREOHPDWLFWR >aSUREFRO@ LQGVXEVL]HVWULQJVVWUPDWFK 3UREOHPDWLF" VWULQJV >SUREURZVa@ LQGVXEVL]HVWULQJVVWUPDWFK VWULQJVSUREFRO ESUREURZVSUREFRO EESUREFROa SUREFRO JUDEFROXPQVZLWKGHVLUHGGDWD ORRSWKURXJKDQGJUDEWKHPHWDGDWDRILQWHUHVWIRUHDFK VDPSOH PHWD 1D1VL]HEVL]HPHWDKHDGV IRU P VL]HPHWDKHDGV JUDEFRO LQGVXEVL]HVWULQJVVWUPDWFKPHWDKHDGV^P`VWULQJVKHD GURZ H[DFW PHWDP EJUDEFRO HQG ORRSWKURXJKDQGJUDEHDFKHOHPHQW FKHPGDWD 1D1VL]HEVL]HGDWDKHDGV IRU P VL]HGDWDKHDGV JUDEFRO VWUPDWFKGDWDKHDGV^P`VWULQJVKHDGURZ H[DFW FKHPGDWDP EJUDEFRO HQG 6DYHWKHGDWD IRU Q VL]HVWURPQXPV URZV PHWD VWURPQXPVQ HYDO> GDWD VWURPQDPHV^Q` SUHIL[^V` FKHPGDWD FKHPGDWDURZV @ HYDO> GDWD VWURPQDPHV^Q` SUHIL[^V` PHWD PHWDURZV @ HQG HYDO> GDWD SUHIL[^V` UDZGDWD QXPEHUV @ HYDO> GDWD SUHIL[^V` UDZWH[W VWULQJV @ HQG GDWDFKHPFROV FKHPFROV ionModel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carb2water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ayleighModel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`2B^ZDWHU`SHUPLO9602: 317 \ODEHO /D\HU 3ORWFDOFXODWHGODNHYROXPHV VXESORW KHUURUEDUDYJVDYJVDYJV [ODEHO &DOFXODWHGODNHYROXPHNPA \ODEHO /D\HU 3ORWFDOFXODWHGODNHGHSWKV VXESORW KHUURUEDUDYJVDYJVDYJV [ODEHO &DOFXODWHGODNHGHSWKP \ODEHO /D\HU BTModelSensitivityPlot IXQFWLRQ %70RGHO6HQVLWLYLW\3ORW *HQHUDWHDSORWRIPD[ODNHGHSWKFKDQJHYVGLDVVXPHGIRUDO O7PRGHOV 70RGHOV 7V7F 0RGHOIL[HG7PLQ&& 0RGHOIL[HG7PD[&& 0RGHOIL[HG7PLG&& 0RGHOVFDOHG71D&&VFDOHGZLWK1D 0RGHOVFDOHG7;0J&&VFDOHGZLWK;0J 0RGHOVHDVRQDOVHH7F&IDQV&PL[HG& PLFULWH 0RGHOWKHUPRFOLQH&&IDQV&PL[HG& PLFULWH 7PRGHO1DPHV >^ IL[HG7PLQ `^ IL[HG7PD[ `^ IL[HG7PLG `^ VFDOHG 71D `^ VFDOHG7;0J `^ VHDVRQDO `^ WKHUPRFOLQH `^ ,RQ0RGHO `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` 5D\OHLJK0RGHO%7VWDFNHGG2ZP7VP GLQ )LQGPD[GHSWKFDOFXODWHG FRUUHVSRQGLQJVWGHY PD[YDOVQ P PD[DYJV^QP` 319 PD[YDOVQ P DYJV^QP`DYJV^QP`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` QXPVWUGLQOHJHQGWH[W^Q ` OHJHQGWH[W ^Q ` VXESORW K KHUURUEDUDYJV^Q`DYJV^Q`DYJV^Q` VHWK &RORU FRORUVQ VXESORW K KHUURUEDUDYJV^Q`DYJV^Q`DYJV^Q` VHWK &RORU FRORUVQ HQG OHJHQGOHJHQGWH[W 0DNHDOD\HUSORWRIFDOFXODWHGODNHYROXPHDQGGHSWKIRUWK HGLIIHUHQW 320 WHPSHUDWXUHPRGHOVXVLQJGL FRORUV MHW ILJXUH VXESORW [ODEHO &DOFXODWHGODNHYROXPHNPA \ODEHO 6WDFNHGOD\HU [OLP>@ KROG RQ VXESORW [ODEHO &DOFXODWHGODNHGHSWKP \ODEHO 6WDFNHGOD\HU [OLP>@ KROG RQ OHJHQGWH[W FHOO IRU Q OHJHQGWH[W^Q ` 7PRGHO1DPHV^Q`OHJHQGWH[W^Q ` OHJHQGWH[W ^Q ` VXESORW K KHUURUEDUDYJV^Q`DYJV^Q`DYJV^Q` VHWK &RORU FRORUVQ VXESORW K KHUURUEDUDYJV^Q`DYJV^Q`DYJV^Q` VHWK &RORU FRORUVQ HQG OHJHQGOHJHQGWH[W 0DNHDER[SORWRIFDOFXODWHGYROXPHVIRUHDFKIDEULF &DOFXODWHLRQPRGHOUHVXOWV LRQ5HV LRQ0RGHO%7VWDFNHG%7VWDFNHG &RPELQHPRGHOUHVXOWVLQWRDVLQJOHPDWUL[ PRG5HV >LRQ5HVLRQ5HVDYJV^`DYJV^` @ )LQGIDEULFV PRG5HV >1D1VL]HPRG5HVPRG5HV@ IRU Q VL]HPRG5HV D %7VWDFNHGPRG5HVQ %7VWDFNHG PRG5HVQ D HQG IDQGDWD$OO PRG5HVPRG5HV PL[GDWD$OO PRG5HVPRG5HV PLFGDWD$OO PRG5HVPRG5HV IDQGDWD%7 IDQGDWD$OOIDQGDWD$OO PL[GDWD%7 PL[GDWD$OOPL[GDWD$OO PLFGDWD%7 PLFGDWD$OOPLFGDWD$OO 3UHSPDWUL[ZLWKFROXPQV I[PIRULRQPRGHO%7 IRULVR 7PLQPRGHOIRUWKHUPRFOLQHPRGHOVDPHXVLQJDOOVW URPVVWDFNHG PRGHOWLWOH >^ ,RQ0RGHO `^ ,VRWRSH0RGHO7B^PLQ` `^ ,VRWRSH0RGHO 7KHUPRFOLQH `@ IRU Q 3ORW%7 VXESORWQ YDOV 1D1VL]HPRG5HV YDOVVL]HIDQGDWD%7 IDQGDWD%7 YDOVVL]HPL[GDWD%7 PL[GDWD%7 YDOVVL]HPLFGDWD%7 PLFGDWD%7 ER[SORWYDOV>^ IDQV `^ PL[HG `^ PLFULWH `@ \OLP>@ 321 [ODEHO IDEULF \ODEHO &DOFXODWHGGHSWKP WLWOHPRGHOWLWOHQ 3ORW$OO VXESORWQ YDOV 1D1VL]HPRG5HV YDOVVL]HIDQGDWD$OO IDQGDWD$OO YDOVVL]HPL[GDWD$OO PL[GDWD$OO YDOVVL]HPLFGDWD$OO PLFGDWD$OO ER[SORWYDOV>^ IDQV `^ PL[HG `^ PLFULWH `@ \OLP>@ [ODEHO IDEULF \ODEHO &DOFXODWHGGHSWKP WLWOHPRGHOWLWOHQ HQG findClosedSequences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
Abstract (if available)
Abstract
Stromatolites, defined descriptively by Semikhatov et al. (1979) as ""attached, laminated, lifhified, sedimentary growth structures, accretionary away from a point or limited surface of initiation"", are considered important structures for understanding the evolution of life on earth and potentially elsewhere. ❧ In general, Archean and Proterozoic stromatolites are fine-grained whereas most modern marine examples are comparatively coarse-grained. Given that the modern marine forms are commonly studied as analogues to the ancient forms, it is important to understand the processes responsible for the textural differences. Cyanobacteria are typically considered the dominant stromatolite builders through time, but it is well known that many modern marine stromatolites also contain a eukaryotic component (various algae, diatoms, etc.). Thus, we conducted experiments to test the grain trapping and binding capabilities of filamentous cyanobacterial mats (dominated by the trichome-forming Coleofasciculus chtonoplastes) versus algal mats (Chaetomorpha) in order to better understand the grain-size trends in stromatolites through time. The mats were cut into coupons, inclined at angles from 0-75° in saltwater tanks to approximate the angle of lamina observed in typical stromatolites, and grains of various sizes (fine sand, coarse sand, and fine pebbles) were delivered to their surface. We measured both trapping and binding as a function of mat properties such as filament length and mesh density. Experiments were done under very low flow and moderate flow conditions. ❧ The cyanobacterial mats were able to trap fine grains consistently at all angles. At angles beyond the angle of repose, medium and coarse grains were not trapped as efficiently, but some (although very few) coarse grains were trapped even at high angles depending on the maturity of the mat and filament bundle length. Dense algal mats trapped medium and coarse grains efficiently at all angles, but were poor at trapping fine grains. The cyanobacteria bound the grains over time, regardless of angle, over time by physically wrapping them. The algae, in contrast, were unable to bind grains and tended to shed trapped grains over time. When flow was added to the experiment, trapping was significantly reduced in the cyanobacterial experiments, but not the algal experiments. ❧ Our experiments suggest that the presence of grains beyond the abiotic angle of repose can be considered a biosignature in ancient stromatolites where biogenicity is in question. Although we cannot conclude that all fine-grained stromatolites were formed by cyanobacteria, our results suggest that stromatolites where coarse grains are present at high angles at much the same frequency as at low angles (e.g., most modern marine stromatolites) may require a filamentous eukaryotic component in order to efficiently trap coarse grains beyond the angle of repose, and give insight into the evolution of stromatolite microfabrics through time. ❧ In addition to their utility as potential records of the communities that formed biogenic stromatolites, stromatolites are also potentially useful for reconstructing paleoenvironmental conditions. The Eocene Green River Formation (Sweetwater County, Wyoming) represents a system of lakes that covered parts of what is now Wyoming, Colorado, and Utah, and captures the Early Eocene Climatic Optimum (EECO, 52-50 Ma), a period of very high global temperatures representing the warmest period of the Cenozoic Era that is associated with very high levels of atmospheric CO₂. Lakes, especially closed basin lakes, can respond dramatically to environmental change because of their sensitivity to precipitation and evaporation. In this study, I 1) use stromatolites from the Rife Bed of the Green River Formation as fine-scale records of terrestrial paleoenvironmental change during a global hothouse climate, and 2) investigate how the environmental dynamics within the lake system affect the growth of stromatolites. ❧ The stromatolites we studied are composed of branching microdigitate columns laminated on the 10-100 µm scale. Laminae are grouped in cm-scale layers that alternate between micritic, calcite fan, and mixed microstructures. The micrite layers contain evidence for a biogenic origin and are enriched in ¹⁸O, Na, and Mg/Ca. The fan layers, in contrast, appear to have formed abiogenically and are relatively depleted in ¹⁸O, Na, and Mg/Ca compared to the micrite layers. The δ¹³C and δ¹⁸O strongly co-vary, suggesting the stromatolites formed in a closed basin lake (e.g., Talbot, 1990). In addition, clumped isotope analyses provide the first quantitative values for water temperatures in lake water from the Green River Formation. The different microfabrics are associated with significantly different lake water temperatures: ~35°C for micrite and ~28°C for fan layers. ❧ Given that the stromatolites grew in a closed basin lake, we can link changes in the δ¹⁸O (via a Rayleigh distillation model) and sodium ion concentration (assuming it behaves conservatively) to periods of evaporation or recharge, and thus model changes in lake volume/level during stromatolite growth. The two models converge upon similar results that suggest that dramatic lake volume changes occurred many times during the accretion of the ~30 cm thick stromatolite, with lake level rising and falling as much as 8 meters. Because of broad, flat bathymetry of the lake, such lake volume changes would have been accompanied by shoreline shifts on the order of tens of kilometers while the stromatolites were actively growing, throwing the traditional interpretation of stromatolites as shoreline facies into question. As a reality check, we note that the modern Great Salt Lake, a similarly broad and flat lake, experienced similar shoreline shifts over several years in the 1980s. ❧ The calculated lake volume—and consequently lake depth—changes also provide insight into the formation of the stromatolites studied. The micrite microfabric formed when the lake was shallow and warm, whereas the fan microfabric formed in cooler waters when the lake was deeper, possibly below a thermocline. I hypothesize that the alternation between biogenic and abiogenic microfabrics present in the Rife Bed stromatolites are the result of dramatic changes in lake level influencing the microbiology and chemistry of the waters in which the stromatolites were forming, potentially due to the at least intermittent existence of thermal and chemical stratification in the lake. ❧ The isotope and conservative ion lake volume models used with the Rife Bed stromatolites were also applied to stromatolites from the Lower LaClede Bed of the Laney Member of the Green River Formation, which captured major changes in basin hydrology, including transient periods of basin closure during a time when the basin is generally considered to have been balanced-filled. Our high-resolution sampling reveals significant changes in ¹⁸O and elemental composition on the cm-scale, providing a finer scale resolution to the filling and evaporation of Lake Gosuite during the waning Eocene Climatic Optimum, and is complementary to broader scale studies on the hydrology of the basin (e.g., Doebbert et al, 2010). Our results potentially indicate that several filling and evaporation stages representing depth changes of ~3 m occurred over the course of a few cm of stromatolite accretion, potentially indicating the magnitude of short-term climate change during the Eocene Climatic Optimum, the period with the highest temperatures and atmospheric CO₂ levels in the Cenozoic. In addition, periods of basin filling are often marked by sudden changes in stromatolite microfabric. ❧ In summary, this thesis highlights the need for continued experimental study to understand the specific mechanisms responsible for forming the wide range of stromatolite fabrics observed in the fossil record. It also demonstrates the potential usefulness of stromatolites as fine-scale records of the environment in which they formed, as well as of the types of biological communities that contributed (or did not contribute) to their growth.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Green River formation stromatolites as a paleoclimate indicator: an investigation of the early Eocene climatic optimum through mass spectrometry, micro-X-ray fluorescence spectroscopy, and petrography
PDF
Stromatolites in the ancient and modern: new methods for solving old problems
PDF
Antarctic climate variability from greenhouse to icehouse world
PDF
Great Salt Lake ooids: insights into rate of formation, potential as paleoenvironmental archives, and biogenicity
PDF
Germanium and silicon isotope geochemistry in terrestrial and marine low-temperature environments
Asset Metadata
Creator
Frantz, Carie Marie
(author)
Core Title
Stromatolites as biosignatures and paleoenvironmental records: experiments with modern mats and examples from the Eocene Green River Formation
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Geological Sciences
Publication Date
09/13/2013
Defense Date
08/16/2013
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
angle of repose,cyanobacteria,Early Eocene Climatic Optimum,Eocene,evaporation,facies,Green River Formation,ichnology,Lake Gosiute,microbial mats,OAI-PMH Harvest,paleoenvironment,paleolimnology,paleontology,Proterozoic,stromatolites
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Corsetti, Frank A. (
committee chair
), El-Naggar, Mohamed Y. (
committee member
), Nealson, Kenneth H. (
committee member
)
Creator Email
cariefrantz@gmail.com,cfrantz@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-329275
Unique identifier
UC11295007
Identifier
etd-FrantzCari-2038.pdf (filename),usctheses-c3-329275 (legacy record id)
Legacy Identifier
etd-FrantzCari-2038.pdf
Dmrecord
329275
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Frantz, Carie Marie
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
angle of repose
cyanobacteria
Early Eocene Climatic Optimum
Eocene
evaporation
facies
Green River Formation
ichnology
Lake Gosiute
microbial mats
paleoenvironment
paleolimnology
paleontology
Proterozoic
stromatolites