Quick viewing(Text Mode)

Open Kosei.Pdf

Open Kosei.Pdf

The Pennsylvania State University The Graduate School Department of Geosciences

GEOCHEMISTRY OF – BLACK : THE EARLY EVOLUTION OF THE ATMOSPHERE, OCEANS, AND BIOSPHERE

A Thesis in Geosciences by Kosei Yamaguchi

Copyright 2002 Kosei Yamaguchi

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2002 We approve the thesis of Kosei Yamaguchi Date of Signature

______Hiroshi Ohmoto Professor of Geochemistry Thesis Advisor Chair of Committee

______Michael A. Arthur Professor of Geosciences

______Lee R. Kump Professor of Geosciences

______Raymond G. Najjar Associate Professor of Meteorology

______Peter Deines Professor of Geochemistry Associate Head for Graduate Program and Research in Geosciences iii

ABSTRACT

When did the Earth's surface environment become oxic? The timing and mechanism

of the rise of atmospheric pO2 level in the early have been long debated but no consensus has been reached. The oxygenation of the atmosphere and oceans has significant impacts on the evolution of the biosphere and the geochemical cycles of -sensitive elements. In order to constrain the evolution of the atmosphere, oceans, biosphere, and geochemical cycles of elements, a systematic and multidisciplinary study of inorganic geochemistry and stable geochemistry was conducted using Archean– Paleoproterozoic black shales, graywackes, and red shales. The samples were collected from unweathered drillcores of the Swaziland Supergroup (3.25 Ga Sheba Formation of Fig Tree Group), the Supergroup (2.96 Ga Parktown Formation of the Group), the Ventersdorp Supergroup (2.71 Ga Rietgat Formation of the Platberg Group), and the Supergroup (2.64 Ga Black Reef Formation of the Wolkberg Group, 2.56 Ga Oak Tree Formation of the Chuniespoort Group, 2.22 Ga Timeball Hill Formation of the Group, and ~2.2 Ga Mapedi Formation of the Olifantshoek Group) in South and the Mt. Bruce Supergroup (2.72 Ga Pillingini Tuff Formation and 2.69 Ga Jeerinah and Lewin Formations of the Fortescue Group and >2.60 Ga Marra Mamba Iron Formation, 2.60 Ga Wittenoom and Carawine Dolomite Formations of the Hamersley Group) in . The above objective was pursued from an array of inter-related studies: (1) the systematics of organic C - S - ferric Fe - ferrous Fe - P contents and stable of organic C and pyrite S; (2) the Mo geochemistry; (3) the N isotope geochemistry; and (4) the U-Th geochemistry. For comparison, data from modern and Phanerozoic sedimentary rocks were complied from the literature. Based on these data, the geochemical iv cycles of C, S, N, Fe, P, Mo, and U in the Archean–Paleoproterozoic surface environments were compared with those of Phanerozoic through modern environments. From (1), I suggest that the contents and stable isotopic compositions of organic C and pyrite S in sediments have been primarily controlled by redox / biological processes in diverse sedimentary redox environments involving oxygenic photosynthesis, aerobic recycling of organic matter, bacterial sulfate reduction, and methanogenesis throughout geologic time. Iron has been oxidized during weathering and reduced during diagenesis, and the P-mediated redox stabilization of the atmosphere and oceans has been operating throughout geologic time. The geochemical cycles of C-S-Fe-P-O would have been essentially the same as today, and they were already in full operation at least 3.25 Ga ago. From (2), based on the kinetics of metal sulfide dissolution, I suggest that Mo- bearing were quantitatively oxidized during weathering and transported to the

-6 oceans if the pO2 was higher than 10 atm, i.e., >0.0005 % PAL (present atmospheric level). The dissolved Mo was fixed by organic matter and S in locally anoxic environments. The geochemical cycle of Mo during the Archean–Paleoproterozoic time was essentially the same as today. From (3), I suggest that the microbially mediated redox cycling of N involving biological N2-fixation, nitrification, denitrification, and ammonification was already in operation during the Archean, based on the characteristics of the N isotopic compositions of organic-bound N and -bound N observed in marine sediments throughout the geologic ages. High Fe content and a positive correlation between Mo and Corg contents of the shales support (although is not an evidence for) the operation of enzymes (FeMo cofactor)

responsible for microbial N2-fixation (nitrogenase) and denitrification (dinitrogenase and dinitrogenase-reductase) in the Archean oceans. From (4), I suggest that the continental weathering flux of U in the Archean was probably the same as today because of the enhanced weathering rate of U-bearing silicate v minerals () under a high pCO2 atmosphere. The weathering of is considered to be minor in the total continental weathering flux of U. Generally low U contents of the Archean-Paleoproterozoic shales (< 10 ppm) are explained by extensive submarine hydrothermal activity at mid-oceanic ridges as a major sink for the oceanic U and the lack of significant enrichment of organic matter in the shales. The sedimentary enrichment of U by organic matter and the secular increase of the Th/U ratios suggests an importance of the decoupling of U and Th and tectonic recycling of U throughout geologic time. The geochemical cycle of U in the Archean–Paleoproterozoic surface environments was essentially the same as it is today. The most important discovery of this study is the early development of the present day redox environments, microbial activity, and geochemical cycles of redox-sensitive

elements in the Archean. In the Archean, N2-fixers, photosynthesizers, nitrifyers, denitrifyers, sulfate-reducers, and methane-producing / consuming microorganisms had already formed complex ecosystems, like those of today, in the globally oxic world where the atmosphere and the oceans were oxic with local anoxic environments such as mid-depth

O2-minimum zone and anoxic / euxinic basins. This study has implications for the early evolution of an oxic atmosphere, oxic oceans, and complex microbial biosphere not only on the Earth but also on the other Earth-like planets distributed in the universe. Such astrobiological implications expand the possibilities for the discovery of extraterrestrial biosignatures in future space missions. vi

TABLE OF CONTENTS

LIST OF FIGURES...... xvi LIST OF TABLES ...... xxiii ACKNOWLEDGMENTS...... xxv

Chapter 1 General introduction: Evolution of the atmosphere and biosphere in the early Precambrian...... 1

1-1. Introduction ...... 1 1-2. Evolution of the atmosphere in the early Precambrian...... 2 1-2-1. Prebiotic atmosphere...... 2 1-2-2. Emergence of life...... 3 1-2-3. Rise of ...... 4

1-2-3-1. Source and sink of atmospheric O2...... 4 1-2-3-2. Controversy over the rise of atmospheric O2...... 5 1-2-3-3. Geological evidence bearing information on the atmospheric O2...... 6 Prevailing view: Low O2 level before 2.2 Ga...... 6 Emerging view: High O2 level since ~3.8 Ga...... 7 1-3. Evolution of the biosphere in the early Precambrian ...... 9 1-3-1. Chemofossils of possible photosynthesizers at 3.8 Ga...... 9 1-3-2. Microfossils and at 3.5 Ga...... 10 1-3-3. Sulfate-reducing bacteria at 3.5 ~ 3.4 Ga...... 10 1-3-4. Nitrogen-metabolizing bacteria at 3.5 G...... 11 1-3-5. Thermophilic microfossils at 3.2 Ga...... 11 1-3-6. Organic biomarkers for cyanobacteria, methanotrophic bacteria, and eukaryotes at 2.7 Ga ...... 11 1-3-7. Methanogens and methanotrophs at 2.7 Ga...... 12 1-3-8. Life on land at 2.6 Ga ...... 12 1-3-10. Eukaryotic megafossils at 2.1 Ga...... 12 vii 1-4. Evolution of the in the early Precambrian...... 12 1-5. Evolution of the geochemical cycles of redox-sensitive elements (C, S, N, Fe, P, Mo, and U) ...... 13 1-5-1. Geochemical cycles of redox-sensitive elements...... 14

1-5-2. Influence of pO2 on the geochemical cycles of redox-sensitive elements...... 17 1-6. Objectives of the thesis ...... 18 1-7. Organization of the thesis ...... 19 1-8. Significance and implications of the thesis...... 21 References ...... 22 Figures...... 37

Chapter 2 Geochemistry of Archean–Paleoproterozoic black shales: I. Diversity in the redox states of sedimentary environments inferred from C-S-Fe-P systematics and C-S stable isotopes ...... 41

Abstract...... 41 2-1. Introduction ...... 43 2-1-1. Evolution of the atmosphere, biosphere, and oceans...... 43 2-1-2. Previous approaches and associated problems / limitations...... 43 2-1-3. Geochemical cycles of C, S, Fe, P, and O: New approaches and focuses of this study ...... 44 2-2. Geological settings and samples ...... 45 2-2-1. Swaziland Supergroup...... 46 2-2-2. Witwatersrand Supergroup...... 47 2-2-3. Ventersdorp Supergroup...... 48 2-2-4. ...... 49 2-2-5. Mt. Bruce Supergroup...... 50 2-3. Analytical methods...... 52 2-3-1. Microscopy...... 52 2-3-2. Pulverization ...... 52 2-3-3. Major elements (Fe, Al, Ti, and P) analyses ...... 53 2-3-4. Elemental analyses of organic C, carbonate C, and sulfide S...... 53 2-3-5. Stable isotope analyses of organic C and pyrite S...... 54 viii 2-3-6. Degree of pyritization (DOP) analysis...... 55 2-4. Results...... 56 2-4-1. Total Fe, ferric Fe, and ferrous Fe contents...... 56 2-4-2. ∑Fe/Ti, Fe3+/Ti, Fe2+/Ti, and Fe3+/∑Fe ratios ...... 57 2-4-3. Pyrite-bound Fe, HCl-soluble Fe, and reactive Fe contents...... 58 2-4-4. Degree of pyritization (DOP) values...... 59 2-4-5. Aluminum and contents...... 60 2-4-6. Phosphorus contents...... 61 2-4-7. Organic C and carbonate C contents...... 62 2-4-8. Pyrite S contents...... 64 2-4-9. Pyrite S to organic C ratios...... 65 2-4-10. Carbon isotopic compositions of organic C...... 66 2-4-11. Sulfur isotopic compositions of pyrite S...... 67 2-5. Discussion...... 67 2-5-1. Geochemical cycle of Fe...... 67 2-5-1-1. Redox chemistry of Fe ...... 67 2-5-1-2. Geochemical cycle of Fe in anoxic vs. oxic environments ...... 69 2-5-1-3. Archean–Paleoproterozoic geochemical cycles of Fe...... 70 2-5-2. Geochemical cycle of P...... 71 2-5-2-1. Detrital, biological, and redox controls on heterogeneous distribution of P in sediments...... 71 2-5-2-2. Phosphorus in the Archean–Paleoproterozoic shales...... 72 2-5-3. Geochemical cycle of C ...... 73 2-5-3-1. Origin of organic matter in shales...... 73 2-5-3-2. Types of marine organisms indicated by C isotopic compositions of organic C...... 76 Effects of post-depositional alterations on C isotopic compositions of organic C...... 77 Antiquity of photosynthesis...... 78

Effects of CO2 concentration on biological isotopic effect...... 79 Isotopic fractionation by chemosynthesis ...... 81 2-5-3-3. Original contents of organic C in shales...... 82 2-5-3-4. Aerobic cycling and burial flux of organic matter...... 83 2-5-3-5. Redox structure of the oceans as indicated from the relationships between contents and isotopic compositions of organic C...... 85 ix 2-5-3-6. Redox structure of the oceans as indicated from the organic C to P ratios ...... 88 Changes in the source rocks compositions ?...... 89 Globally anoxic environment ?...... 89 ?...... 90 Diagenetic changes ?...... 90 Redox controls on the variability of the sedimentary organic C to P ratios...... 91 2-5-3-7. Redox cycles of C based on C isotopic compositions ...... 91 2-5-4. Geochemical cycle of S...... 92 2-5-4-1. Activity of sulfate-reducing bacteria indicated by S isotope...... 93 Previous study ...... 93 This study...... 94 Secular trend of S isotope compositions...... 95 2-5-4-2. Sedimentary distribution of S in shales ...... 96 2-5-4-3. Relationships of the organic C and S contents in various environments ...... 98 2-5-4-4. Paleoredox environments inferred from the pyrite S to organic C ratios and the degree of pyritization (DOP)...... 100 2-6. Significance and implications...... 103 2-7. Conclusions...... 103 References ...... 106 Figures...... 122 Tables ...... 152

Chapter 3 Geochemistry of Archean–Paleoproterozoic black shales: II. Geochemical cycle of molybdenum ...... 167

Abstract...... 167 3-1. Introduction ...... 169 3-1-1. Redox-sensitive metals in black shales...... 169 3-1-2. Previous studies about Mo as a proxy of bottom water anoxia ...... 170 3-1-3. Purpose of this study...... 172 3-2. Geological settings and samples ...... 173 x 3-3. Analytical methods...... 173 Mo and Zn analyses...... 174 3-4. Results...... 174 3-4-1. Relationship among Mo, organic C, and pyrite S contents...... 175 3-4-2. contents...... 176 3-5. Discussion...... 176 3-5-1. Molybdenum in geological materials ...... 176

3-5-2. The pO2 dependence of Mo weathering flux from continents...... 177 Rate of Mo weathering ...... 179 3-5-3. Source, sink, and sedimentary enrichment mechanisms of Mo ...... 183 3-5-3-1. Source and sink of Mo...... 183 3-5-3-2. Mechanisms of Mo enrichment...... 183 3-5-4. Burial flux of Mo in the Archean–Paleoproterozoic sediments...... 184 3-5-4-1. Molybdenum enrichment in shales: authigenic? hydrothermal? .....184 3-5-4-2. Molybdenum enrichment inferred from the Mo/Al ratios ...... 185 Archean–Paleoproterozoic ...... 185 Modern and Phanerozoic...... 186 Mechanisms to cause variations in the Mo/Al ratios ...... 187 Implications of variable Mo/Al ratios of shales...... 188 3-5-5. Redox controls on the relationships among Mo, organic C, and pyrite S contents...... 189 3-5-5-1. Archean–Paleoproterozoic...... 189 3-5-5-2. Modern and Phanerozoic...... 190 3-5-5-3. Schematic model...... 191 3-5-5-4. Significance of low Mo contents of the Archean– Paleoproterozoic black shales...... 193 3-6. Implications ...... 194 3-7. Conclusions...... 195 References ...... 197 Figures...... 210 Tables ...... 227 xi Chapter 4 Geochemistry of Archean–Paleoproterozoic black shales: III. Fixation and redox cycling of nitrogen by organisms...... 237

Abstract...... 237 4-1. Introduction ...... 239 4-1-1. Nitrogen isotope as an indicator of past biological activity...... 239 4-1-2. Nitrogen biogeochemistry in Precambrian...... 240 4-1-3. Objectives ...... 241 4-2. Background ...... 242 4-2-1. Nitrogen biogeochemical cycles...... 242 4-2-2. Nitrogen isotopic fractionations...... 243 4-2-3. Previous studies on N in the Archean ...... 244 4-3. Geological settings and samples ...... 246 4-4. Analytical methods...... 246 4-5. Results...... 247 4-5-1. Organic C contents ...... 247 4-5-2. Nitrogen contents...... 248 4-5-2-1. Bulk N contents...... 248 4-5-2-2. Nitrogen contents in organic matters and clays...... 249 4-5-3. Nitrogen to organic C ratios...... 250 4-5-4. Carbon isotopic compositions of organic C...... 251 4-5-5. Nitrogen isotopic compositions ...... 252 4-5-6. Molybdenum contents ...... 254 4-6. Discussion...... 255 4-6-1. Factors controlling the N and organic C contents of shales ...... 255 4-6-1-1. Type of organisms: Terrestrial vs. Marine ? ...... 255 4-6-1-2. Change in the N and organic C contents during diagenesis and metamorphism ...... 256 4-6-1-3. Total N, organic-bound N, and clay-bound N contents of the Archean shales...... 258 4-6-1-4. Relationship between the N and organic C contents ...... 259 4-6-2. Microbial activity inferred from the C isotopes of organic C ...... 260 4-6-3. Microbial activity inferred from the N isotopes...... 261 4-6-3-1. Effect of metamorphism and hydrothermal alteration ...... 261

4-6-3-2. N2-fixation ...... 262 xii 4-6-3-3. Redox cycling of N ...... 263 4-6-3-3. Non-redox cycling of ammonia...... 265

4-6-4. N2-fixation as a source of atmospheric O2 ?...... 266 4-6-5. Molybdenum availability to microorganisms ...... 267 4-6-6. Implications for the Archean atmosphere and oceans...... 269 4-6-6-1. Environmental prerequisites for the early evolution of

biological N2-fixation...... 269 4-6-6-2. Source of nitrate and the evolution of atmosphere ...... 269

4-6-6-3. Oceanic environments leading to biological N2-fixation...... 270 4-6-6-4. Diverse microbial biosphere in the Archean...... 271 4-6-7. Critical examination of Beaumont and Robert (1999)...... 272 4-7. Conclusions...... 274 References ...... 277 Figures...... 289 Tables ...... 298

Chapter 5 Geochemistry of Archean–Paleoproterozoic black shales: IV. Geochemical cycle of ...... 308

Abstract...... 308 5-1. Introduction ...... 310 5-2. Background ...... 312 5-2-1. Uranium and thorium in rocks and minerals...... 312 5-2-2. Decoupling of U and Th...... 313 5-2-3. Uranium as a proxy for sedimentary redox environments...... 313 5-2-3-1. Recent sediments...... 313 5-2-3-2. Phanerozoic black shales...... 314 5-2-3-3. Precambrian black shales...... 315 5-3. Geological settings and samples ...... 315 5-4. Analytical methods...... 316 5-5. Results...... 316 5-5-1. Contents of U, Th, Al, organic C, and carbonate C...... 316 5-5-2. Correction for U and Th contents with ages...... 317 5-5-3. Uranium to organic C ratios...... 318 xiii 5-5-4. Uranium to aluminum ratios ...... 319 5-5-5. Thorium to uranium ratios ...... 320 5-6. Discussion...... 321 5-6-1. Post-depositional effects on the fractionation of the U to organic C ratios, U/Al ratios, and Th/U ratios...... 321 5-6-1-1. Metamorphic effects...... 322 5-6-1-2. Source rock effects...... 322 5-6-1-3. Early diagenetic effects and/or changes in the redox depositional environments ...... 323 5-6-2. Geochemical cycle of U...... 324 5-6-2-1. Source of U: weathering of U-bearing minerals ...... 325 Uraninite dissolution rate...... 325 dissolution rate...... 326 5-6-2-2. Sink of U: sediments vs. mid-oceanic ridges...... 327 5-6-3. Sedimentary fixation of U by organic matter through geologic time...328 5-6-3-1. Archean-Paleoproterozoic shales...... 328 5-6-3-2. Modern sediments...... 339 5-6-4. Low U contents of the Archean–Paleoproterozoic black shales ...... 330 5-6-4-1. Continental weathering flux of U...... 331 5-6-4-2. Extensive submarine hydrothermal activity...... 332 5-6-4-3. Lack of significant enrichments of organic C ...... 334 5-6-5. Estimation of U concentration in the Archean oceans ...... 336 5-6-6. Tectonic recycling and decoupling of U and Th through geologic time...... 339 Secular change in the age-corrected Th/U ratios...... 340 5-7. Conclusions...... 342 References ...... 344 Figures...... 353 Tables ...... 370

Appendices ...... 382 xiv Appendix A Geological settings...... 383

A-1. Introduction ...... 383 A-2. ...... 383 A-2-1. Swaziland Supergroup...... 384 A-2-1-1. ...... 385 A-2-1-2. Fig Tree Group...... 385 A-2-1-3. ...... 386 A-2-2. Witwatersrand Supergroup...... 387 A-2-1. Ventersdorp Supergroup ...... 389 A-2-1. Transvaal Supergroup...... 389 A-2-4-1. Wolkberg Group...... 389 A-2-4-2. Chuniespoort Group...... 390 A-2-4-3. Pretoria Group...... 391 A-2-5. Metamorphism and deformation in the Witwatersrand Basin...... 392 A-3. Australia...... 393 A-3-1. Mt. Bruce Supergroup...... 393 A-3-1-1. Fortescue Group...... 394 A-3-1-2. Hamersley Group...... 395 A-3-1-3. Turee Creek Group...... 396 A-3-1-4. Lewin Shale and Carawine Dolomite...... 397 References ...... 399 Figures...... 410

Appendix B Samples...... 416

B-1. Introduction...... 416 B-2. Sampling method...... 416 B-3. Drillcore PU1308...... 418 B-4. Drillcore MRE10...... 419 B-5. Drillcore DRH13...... 420 B-6. Drillcore MSF6...... 421 B-7. Drillcore JPBR...... 422 B-8. Drillcore PTB3...... 422 B-9. Drillcore SA1677 ...... 423 xv B-10. Drillcore WRL1 ...... 424 B-11. Drillcore RHDH2A...... 425 References ...... 427 Figures...... 428

Appendix C Analytical methods...... 437

C-1. Sample treatment and pulverization...... 437 C-2. Preparation of thin sections and microscopic observation...... 438 C-3. Decomposition of the powdered samples ...... 438 C-3-1. Alkali fusion ...... 439 C-3-1. Acid digestion...... 441 C-4. Major and minor elements analysis ...... 443 C-5. Minor, trace, and rare earth elements analysis...... 444 C-6. Ferrous Fe titration...... 444 C-7. Loss on ignition...... 446 C-8. Mineralogical analysis by X-ray diffraction...... 447 C-9. Organic C, carbonate C, H, N, and S elemental analysis ...... 448 C-10. extraction ...... 448 C-11. Carbon and oxygen isotope analysis ...... 449 C-11-1. Off line isotopic analysis of organic C...... 449 C-11-2. On line isotopic analysis of organic C ...... 450 C-11-3. Carbon and oxygen isotopic analysis of carbonate ...... 450 C-11-4. Oxygen isotopic analysis of silicates ...... 451 C-12. Sulfur isotopic analysis ...... 451 C-13. Nitrogen isotopic analysis ...... 452 C-14. Degree of pyritization analysis ...... 452 References ...... 454 Figures...... 455 xvi

LIST OF FIGURES

Chapter 1

Figure 1-1. Two contrasting models for the evolution of atmospheric O2 level...... 39

Figure 1-2. Summary of controversial geological 'evidence' for pO2 in the early atmosphere ...... 40

Figure 1-3. Geochemical cycles of C, S, N, Fe, Mo, and U in the continents- oceans-sediments system...... 41

Figure 1-4. Schematic diagram showing the characteristic pE values for redox reactions of trace metals and major electron acceptors during decomposition of organic matter in sediments...... 42

Chapter 2

Figure 2-1. Simplified geological map of South Africa ...... 125

Figure 2-2. Lithostratigraphic column of the Transvaal, Ventersdorp, Witwatersrand, and Swaziland Supergroups in South Africa...... 126

Figure 2-3. Simplified geological map of the Pilbara-Hamersley district, ...... 127

Figure 2-4. Lithostratigraphic column of the lower Hamersley and upper Fortescue Groups of the Mt. Bruce Supergroup, Western Australia...... 128

Figure 2-5. Log-log plot of the Fe3+/Ti vs. Fe2+/Ti (wt.) ratios for the Archean– Paleoproterozoic samples of this study...... 129 xvii Figure 2-6. Plot of the Fe3+/Ti vs. Fe2+/Ti (wt.) ratios for the Archean– Paleoproterozoic samples of this study...... 130

Figure 2-7. Histogram of the Fe3+/∑Fe ratios for the Archean– Paleoproterozoic samples of this study...... 132

Figure 2-8. Plot of the Fepy vs. FeHCl contents for the Archean– Paleoproterozoic samples of this study...... 133

Figure 2-9. Histogram of the DOP (degree of pyritization) values for the Archean–Paleoproterozoic samples of this study...... 135

Figure 2-10. Histogram of the Al2O3 contents for the Archean– Paleoproterozoic samples of this study...... 136

Figure 2-11. Histogram of the P2O5 contents for the Archean– Paleoproterozoic samples of this study...... 137

Figure 2-12. Histogram of the organic C (Corg) contents for the Archean– Paleoproterozoic samples of this study...... 138

Figure 2-13. Histogram of the carbonate C (Ccarb) contents for the Archean–Paleoproterozoic samples of this study...... 139

Figure 2-14. Histogram of the pyrite S (Spy) contents for the Archean–Paleoproterozoic samples of this study...... 140

Figure 2-15. Histogram of the C isotopic compositions of organic C 13 (δ Corg) values for the Archean–Paleoproterozoic samples of this study ...... 141

Figure 2-16. Histogram of the S isotopic compositions of sulfide (mainly 34 pyrite) S (δ Spy) values for the Archean–Paleoproterozoic samples of this study ...... 142

Figure 2-17. The P/Al (wt.) ratios of the Archean–Paleoproterozoic samples of this study and of various samples from literature...... 143 xviii Figure 2-18. Selected photomicrographs of the representative Archean–Paleoproterozoic samples of this study...... 145

Figure 2-19. The Corg/Al (wt.) ratios for the Archean–Paleoproterozoic samples of this study and of various samples from literature...... 147

Figure 2-20. Overall negative correlation between the Corg contents and the δ13C values for the Archean–Paleoproterozoic samples of this study...... 148

Figure 2-21. The Corg/P (wt.) ratios for the Archean–Paleoproterozoic samples of this study ...... 149

Figure 2-22. Secular change in the δ13C values for kerogen and bulk organic matter throughout geologic time...... 150

Figure 2-23. The S/Al (wt.) ratios of the Archean–Paleoproterozoic samples of this study and of various samples from literature...... 151

Figure 2-24. Secular change in the δ34S values for sulfide and sulfate throughout geologic time ...... 152

Figure 2-25. Plot of the pyrite S (Spy) vs. organic C (Corg) contents for the Archean–Paleoproterozoic samples of this study...... 153

Chapter 3

Figure 3-1. Plots of Mo vs. Corg and Mo vs. S contents of the Archean–Paleoproterozoic samples of this study...... 219

Figure 3-2. Molybdenum distribution in sediments depending on the redox state of the environments...... 223

Figure 3-3. Eh-pH diagram for aqueous species and solids in the system

Mo-S-O2-H2O at 25 ˚C and 1 bar total pressure ...... 224

Figure 3-4. Relationship between rate of pyrite oxidation and pO2 values...... 225 xix

Figure 3-5. Plot of the Mo contents vs. Zn/Al2O3 ratios of the Archean–Paleoproterozoic samples of this study...... 226

Figure 3-6. Enrichment factors (EF) for the Mo contents relative to the Al contents of the Archean–Paleoproterozoic samples of this study...... 228

Figure 3-7. Enrichment factors (EF) for the Mo contents relative to the Corg contents of the Archean–Paleoproterozoic samples of this study...... 230

Figure 3-8. Plot of the Mo vs. Corg contents for modern marine carbonaceous sediments and Phanerozoic black shales...... 232

Figure 3-9. Plot of the Mo vs. S contents for modern marine carbonaceous sediments and Phanerozoic black shales...... 233

Figure 3-10. Schematic diagram showing relationship among Mo, Corg, and S contents of sediments / sedimentary rocks...... 234

Figure 3-11. Relationship between the Mo contents in modern sediments and

dissolved O2 contents in their overlying bottom waters...... 235

Chapter 4

Figure 4-1. Schematic diagram showing the N biogeochemical cycles...... 299

Figure 4-2. Histogram of the bulk N contents of the Archean– Paleoproterozoic samples of this study (a through h) and of the modern sediments and Neoarchean– Paleoproterozoic shales from literature ...... 300

Figure 4-3. Relationship between the Norg/∑N ratios and ages in Ga of the Archean–Paleoproterozoic samples of this study and some Phanerozoic shales from literature ...... 301

Figure 4-4. Cross plot showing relationship between the N and Corg contents of the Archean–Paleoproterozoic samples of this study...... 302 xx

Figure 4-5. Histogram of the δ15N values of the Archean–Paleoproterozoic samples of this study and the various samples from literature...... 303

Figure 4-6. Relationship between the N contents and the δ15N values for the Catalina metasedimentary rocks of various metamorphic grades ...... 305

Figure 4-7. Plot of the Mo and Corg contents of the Mesoarchean black shales, black shales, black shales, and modern carbonaceous sediments of the Black Sea...... 306

Figure 4-8. Schematic illustrations of the Archean oceanic environments proposed in this study...... 307

Chapter 5

Figure 5-1. Summary of the geological materials bearing on the O2 content of the Precambrian atmosphere...... 368

Figure 5-2. Plots of the U vs. Corg and U vs. Al2O3 contents for the Archean–Paleoproterozoic samples of this study...... 369

Figure 5-3. Enrichment factor (EF) for U relative to organic C (Corg) in the Archean–Paleoproterozoic samples of this study...... 373

Figure 5-4. Variations in the age-corrected U to organic C (ppm / wt.%) ratios for the Archean–Paleoproterozoic samples of this study and the black shales of Hayashi et al. (1997)...... 374

Figure 5-5. Enrichment factor (EF) for U relative to Al in the Archean– Paleoproterozoic samples of this study...... 375

Figure 5-6. Variations in the age-corrected U to Al (ppm / wt.%) ratios for the Archean–Paleoproterozoic samples of this study and the black shales of Hayashi et al. (1997)...... 376 xxi Figure 5-7. Histogram of the Th/U ratios for the Archean–Paleoproterozoic samples of this study ...... 377

Figure 5-8. Uranium contents of , , and granite, and of their rock-forming minerals ...... 378

Figure 5-9. Eh-pH diagram for aqueous species and solids in the system U-

O2-CO2-H2O at 25 ˚C and 1 bar total pressure ...... 379

Figure 5-10. A model for the secular change in the weathering flux of U from the continent and the efficiency of U removal from seawater into mid-oceanic ridges (MOR)...... 380

Figure 5-11. Schematic diagram of the geochemical history of U to cause good or poor relationship between organic C and U in black shales...... 381

Figure 5-12. A comparison of authigenic U mass accumulation rate (MAR) in a number of sedimentary environments plotted against the

bottom water dissolved O2 (DO) contents overlying the core...... 382

out out Figure 5-13. Relationship between the Fppt [Archean] / Fppt [Today] ratios and C1 for U ...... 383

Figure 5-14. Evolution of the Thcorr/Ucorr (wt.) ratios for various sedimentary rocks through geologic time...... 384

Appendix

Figure A-1. Geologic map of South Africa ...... 430

Figure A-2. Geologic map of the Barberton ...... 431

Figure A-3. Geologic map of the Witwatersrand Basin...... 432 xxii Figure A-4. Lithostratigraphic column of the Swaziland, Witwatersrand, Ventersdorp, and Transvaal Supergroups...... 433

Figure A-5. Geologic map of the Pilbara , Western Australia...... 434

Figure A-6. Lithostratigraphic column of the middle Mt. Bruce Supergroup...... 435

Figure B-1. Lithostratigraphy of the ~3.2 Ga Fig Tree Group, the Swaziland Supergroup, Barberton Greenstone Belt, South Africa...... 448

Figure B-2. Lithostratigraphy of the ~2.9 Ga Parktown Formation in the Hospital Hill Subgroup of the West Rand Group, Witwatersrand Supergroup, South Africa ...... 449

Figure B-3. Lithostratigraphic column of the Transvaal and Ventersdorp Supergroups, South Africa...... 450

Figure B-4. Lithostratigraphy of a Paleoproterozoic succession in Griqualand West, South Africa...... 451

Figure B-5. Lithostratigraphy at Paleoproterozoic succession in Griqualand West, South Africa...... 452

Figure B-6. Lithostratigraphy of the drillcore WRL1 covering the lower Hamersley and upper Fortescue Groups, Mt. Bruce Supergroup, Western Australia ...... 453

Figure B-7. Lithostratigraphic column for the drillcore RHDH2A covering the Carawine Dolomite and Lewin Shale Formations, eastern Hamersley, Western Australia...... 454

Figure C-1. Flow diagram of the sample handling toward pulverization...... 473

Figure C-2. Flow diagram of the sample treatments for various type of analyses ...... 474 xxiii

LIST OF TABLES

Chapter 2

Table 2-1. Geochemical data of the Archean–Paleoproterozoic samples of this study ...... 155

Table 2-2. Iron-related data of the Archean–Paleoproterozoic samples of this study ...... 160

Table 2-3. Stable isotopic compositions and various elemental ratios of the Archean–Paleoproterozoic samples of this study...... 165

Chapter 3

Table 3-1. Summary of the geological settings and samples of this study ...... 236

Table 3-2. Geochemical data of the Archean–Paleoproterozoic samples of this study ...... 238

Table 3-3. Molybdenum concentration of various rocks...... 243

Table 3-4. Comparison of the average Mo to organic C ratios and Mo/Al ratios among Archean–Paleoproterozoic samples of this study...... 244

Table 3-5. Source and sink fluxes for Mo ...... 245

Chapter 4

Table 4-1. Summary of the geological settings and samples of this study ...... 308 xxiv Table 4-2. Geochemical data of the Archean–Paleoproterozoic samples of this study ...... 310

Table 4-3. Comparison of the organic C and N contents, N/C ratios, and C isotopic compositions of organic matter among the Archean, Paleoproterozoic, and Phanerozoic shales...... 315

Table 4-4. Comparison of the total N, organic-bound N, and clay-bound N contents among sediments and shales...... 317

Chapter 5

Table 5-1. Typical concentrations of U and Th in various igneous, metamorphic, and sedimentary rocks, upper continental crust, and rock-forming minerals ...... 385

Table 5-2. Summary of the geological settings and samples of this study ...... 386

Table 5-3. Geochemical data of the Archean–Paleoproterozoic samples of this study ...... 388

Table 5-4. Comparison of the U to organic C ratios for the Archean– Paleoproterozoic sedimentary rocks...... 393

Table 5-5. Comparison of the U/Al ratios and EF (enrichment factor) for the Archean–Paleoproterozoic sedimentary rocks...... 394

Table 5-6. Comparison of the U and Th contents and the Th/U ratios of the Archean–Paleoproterozoic sedimentary rocks...... 395

Table 5-7. Source and sink fluxes for U...... 396

Appendix

Table B-1. Summary of drillcores used in this study...... 435 xxv

ACKNOWLEDGMENTS

It may not be an easy task to acknowledge all those to whom I am indebted for completing this dissertation. But I will try.

I would like to thank Dr. Hiroshi Ohmoto, my thesis advisor, for hundreds of hours of discussion and tireless encouragement during the course of this study (6.5 years). I am also grateful to the members of my doctoral committee, Drs. Mike Arthur, Lee Kump, and Ray Najjar for their discussion and constructive criticisms. This work would have been impossible without the technical help and support of the following people at Penn State. Mr. Denny Walizer has been a “superman”, and fixed every problem that I encountered during the lab work. I owe very much to him, and appreciate for his encouragement. Generous permissions from Drs. Kate Freeman, Mike Arthur, Sue Brantley, and Hu Barnes allowed me to do analytical work in their labs. I thank Mr. Henry Gong, Dr. Shaole Wu, Mr. Joe Bodkin, Mrs. Kay Bickle, and Mr. Mark Angello for help with various aspects of analytical inorganic geochemistry. I thank Dr. Michael Bau for sharing with me his expertise in trace element geochemistry and also for valuable discussion. Ms. Yumiko Watanabe and Dr. Shuhei Ono are especially thanked for their discussion, technical assistance, and encouragement. Mr. Frank Kachurak, Mr. Matt Hurtgen, and Ms. Kate Spangler are thanked for their professional proof reading of my papers. Dr. Peter Deines is thanked for reviewing my papers. Technical assistance for CHNS elemental and isotope analyses from Dr. Takeshi Kakegawa of Tohoku Univ., Dr. Hiroshi Naraoka of Tokyo Metropolitan Univ., and Dr. Simon Poulsen of Univ. of Nevada- Reno are appreciated. Dr. Jim Kasting is thanked for his stimulating discussion of every xxvi topic related to the rise of O2 controversy. Discussion with Dr. Ethan Grossman of Texas A&M Univ. has been beneficial. Dr. Don Canfield of Odense Univ. (Denmark) is thanked for sending me a S isotope data set. Dr. Clark Johnson of the Univ. - Madison is thanked for his encouragement and patience. Fruitful and enjoyable fieldwork in (1995), South Africa (1996 and 1998), Namibia (1998), and Australia (1999), where the samples of this study were collected, were made possible by unselfish support from Drs. Takeshi Kakegawa and Ken-ichiro Hayashi of Tohoku Univ., Dr. Hiroshi Naraoka of Tokyo Metropolitan Univ., and Drs. Munetomo and Yoko Nedachi of Kagoshima Univ. and Kagoshima Immaculate Heart College, respectively. Technical assistance is acknowledged during the fieldwork from Dr. Gerald Bennett of Geological Survey, Dr. W.E.L. Minter of Univ. of Cape Town, Drs. Nick Beukes and Jens Gutzmer of Rand Univ., Dr. K.-H. Hoffmann of Geological Survey of Namibia, Dr. Tony Prave of the Univ. of St. Andrews, and Drs. Mark Barley and Brian Krapez of the Univ. of Western Australia. The Sheba Mine, Agnes Mine, and Anglo American Prospecting, Ltd. in South Africa, and CRA Exploration Pty. Ltd. in Australia are appreciated for access to the drillcore samples of this study. Thanks also go to the landscapes and wildlife of Canada, South Africa, Namibia, and Australia for being beautiful, wild, and impressive. I would like to thank many friends and fellow graduate students / postdocs / visiting researchers (then) who have passed through the Deike building during my tenure. They include Mr. Nikolai Pedentchouk, Mr. Matt Hurtgen, Dr. Alex Pavlov, Dr. Mark Pagani (Yale Univ.), Dr. Rich Pancost (Univ. Bristol), Dr. Rick Wilkin (EPA), Dr. Melissa Nugent (SUNY Stony Brook), Dr. Liane Benning (Univ. Leeds), Dr. Tracy Frank (Univ. Queensland), Dr. Andy Kurtz, Ms. Sarah Das, Mr. Nate Kaleta, Mr. Johnson Olanrewaju, Dr. Bob Altamura, Mr. Gento Kamei (Japan Nuclear Cycle Development Institute), Ms. Katya Bazilevskaya, and Dr. Naoki Watanabe (Niigata Univ.). Dr. Yongsong Huang xxvii

(Brown Univ.) is especially acknowledged for his encouragement and discussion, frequently at midnight. Continued service to and the chairmanship of the Departmental Colloquium Committee have been memorable. Jogging on the beautiful, unpaved trails in State College allowed me to stay in shape, keep my sanity, participate in several marathon races including the Boston Marathon, and to win awards from the Steamtown Marathon (2nd place) and the 2000 US Orienteering Championship (1st place in the brown class). Special thanks go to Ms. Koya Ohmoto for her encouragement, great hospitality at numerous home parties, and her professional piano music. Encouragement from Dr. Ei-ichiro Ochiai and Ms. Katsuko Ochiai are appreciated.

Discussion and encouragement from my colleagues in Japan are also thanked. They include Dr. Naohiko Ohkouchi of Institute of Frontier Research on Earth Evolution (IFREE), Dr. Shoichi Kiyokawa of Kyushu Univ., Dr. Sanny Saito of Japan Marine Science and Technology Center (JAMSTEC), Dr. Asahiko Taira of Ocean Research Institute (ORI), the Univ. of Tokyo and JAMSTEC, Dr. Kiyoshi Suyehiro of JAMSTEC, Dr. Hidekazu Tokuyama of ORI, Dr. Terry Ishii of ORI, Dr. Hodaka Kawahata of Geol. Surv. Japan, and Dr. George Hashimoto of Center for Climate System Research, Univ. of Tokyo. Dr. Jin Akiyama of Tokai Univ. and Dr. Susumu Tonegawa of the Massachusetts Inst. of Technology are especially thanked for their introduction and encouragement toward me to earn a PhD in the . Dr. Jin Akiyama is also thanked for his guidance to the Graph Theory. A class at the Department of Mathematics at Penn State was enjoyable.

Financial supports for this work were provided by the NASA Astrobiology Institute (NCC2-1057), NASA Exobiology Program (NAG5-9089), National Science Foundation (EAR 9706279), Department of Geoscience and Astrobiology Research Center at Penn xxviii

State, Ohmoto Graduate Fellowship, Krynine Memorial Fund, and the Japanese Ministry of Education, Culture, Sports, Science, and Technology (#100411004). Ms. Linda Decker, Ms. Kathryn McClintock, Ms. Linda Altamura, and Ms. Lois Cock-Gibson are appreciated for their administrative supports. I would like to thank my parents Masaaki Yamaguchi and Sai Yamaguchi, my brother Yasunaka Yamaguchi, my grandmother Toshi Oda, my relatives especially Tomi Oda and the late Hatsu Oda, for their continued encouragement, understanding, and support in many aspects throughout my life. Without their generous support, I would not be where I am today. Ms. Yuko Tsurumi and her family are appreciated for their understanding and encouragement. Ms. Keiko Oka and her family are appreciated for their continued encouragement, support, and patience. Chapter 1

General introduction: Evolution of the atmosphere and biosphere in the early Precambrian

"... models or working hypotheses that have become widely accepted as organizing principles for the field ... may us to ask the wrong questions ... or disregard significant lines of evidence simply because they seem inconsistent with our model- dependent predictions." J.W. Schopf, in Earth's Earliest Biosphere

1-1. Introduction

The Earth was formed about 4.6 billion years (Ga) ago. Chemical evolution of the surface environments of the Earth, as well as its deeper part, is characterized by irreversible differentiation processes into and within the spheres: atmosphere, hydrosphere, lithosphere, and asthenosphere. All the spheres have influenced each other directly and indirectly, and co-evolved. An additional sphere, the biosphere, has been developed since the emergence of life, and its evolution over Earth's history has modified the surface environments. Oxygenic photosynthesis by microorganisms is believed to have been the major source of atmospheric oxygen that transformed a prebiotic, anoxic atmosphere into an oxic atmosphere. However, the timing, extent, and mechanism of the "rise of atmospheric 2 oxygen" has been contentious among scientists. Fundamental questions pursued in this thesis include how and when the biosphere facilitated atmospheric transformation and how we can extract information on the chemical evolution of the atmosphere and biosphere from existing and new geologic records. The chemical evolution of the atmosphere is significant by itself and deeply linked to the evolution of the biosphere. The present chapter provides a general introduction to the following chapters that discuss the evolution of the atmosphere, oceans, and biosphere. First, a summary of the current understanding of the chemical evolution of the atmosphere is provided. This summary briefly reviews the theories concerning the chemical composition of the prebiotic atmosphere and emergence of life. Then, two contrasting views concerning the rise of atmospheric oxygen and the controversial geologic 'evidence' for both of the views are introduced. Next, a brief summary of the current knowledge of the evolution of the biosphere during the Archean (3.8-2.5 Ga) -Paleoproterozoic (2.5-1.8 Ga) is provided. A short summary is then given for the chemical and volumetric evolution of the continental crust. Next, an introduction to the evolution of the geochemical cycles of C, S, N, Fe, P, Mo, and U (redox-sensitive elements) is provided, followed by a brief summary of a sequence of sedimentary redox reactions. Last, the objectives, organization, and significance of the thesis are presented.

1-2. Evolution of the atmosphere in the early Precambrian

1-2-1. Prebiotic atmosphere The Earth's earliest, prebiotic atmosphere was essentially devoid of molecular oxygen. After the main accretionary and core-forming events occurred during the first few tens of millions of years, oceans emerged by the cooling of the Earth which caused the 3

condensation of H2O. The residual atmosphere was probably dominated by CO2, N2 and

H2O, with lesser amounts of CO and H2 (e.g., Holland, 1984). In the early atmosphere, UV

radiation from the young Sun would have encouraged photodissociation of H2O vapor,

resulting in the loss of hydrogen (to space) and accumulation of O2 in the atmosphere (e.g., Canuto et al., 1983). The UV radiation in the early atmosphere must have been by far more

intense than that of today (e.g., Canuto et al., 1983). However, O2 would have been consumed during the atmospheric and surface oxidation of reduced chemical species and by interaction with the mantle through volcanism (and if plate tectonics operated at that time) (e.g., Holland, 1984; Kasting et al., 1993). The accumulation of more than trace

amounts of O2 would depend on such an O2-sink. If the removal of O2 by the reduced chemical species was rapid, as is likely due to active tectonics and volcanics in the early

Earth, the atmospheric O2 content would have been very low (Kasting and Walker, 1981; Kasting, 1993).

1-2-2. Emergence of life The origin of life, including timing and mechanisms, is not yet known. It could be exogenous (delivery of extra-terrestrial organic matter; e.g., Chyba et al., 1990; Chyba and Sagan, 1992; Chyba, 1993; Wallis and Wickramasinghe, 1995; Whittet, 1997) and/or endogenous (hydrothermal / lightning synthesis of organic matter on Earth; e.g., Miller, 1953; Miller and Urey, 1959; Farmer, 2000; Mancinelli and McKay, 1988; Navarro- González et al., 2001). The earliest emergence of liquid water on the Earth's surface is the crucial constraint on the timing of the emergence of life, because liquid water is necessary for life's sustainability, propagation and evolution. A very early existence of the continental crust and oceans, as old as 4.4 - 4.3 Ga ago, has been recently demonstrated (Wilde et al., 2001; Mojzsis et al., 2001). Therefore, life could have already existed by ~4.4 Ga ago. Researchers have speculated that life may have 4

emerged rapidly, almost instantaneously in geologic timescales, once the proper environment was provided on the early Earth (Overbeck and Fogleman, 1989). However, the very early forms of life may have been almost completely destroyed by the intense bombardments of planetary objects which continued until ~3.8 Ga (e.g., Maher and Stevenson, 1988). During that period, the early life could have repeatedly originated and then been destroyed. Although some could have survived in niches, the earliest organisms are not necessarily the common ancestor of modern organisms. The first form of life was probably not photosynthetic but chemotrophic. In a pre- photosynthetic world, early microorganisms (probably chemotrophs) utilized local redox gradients to obtain energy and nutrient elements such as Fe, P, Ni, and Mo (e.g., Nisbet, 1995; Farmer, 2000) for life, probably near marine / terrestrial hydrothermal systems.

1-2-3. Rise of oxygen

1-2-3-1. Source and sink of atmospheric O2

The most significant source of O2, photosynthesis, emerged on the Earth by at least the Neoarchean (~2.7 Ga: Buick, 1992; Beukes and Lowe, 1989; Brocks et al., 1999; Eigenbrode et al., 2001), and probably as old as 3.5 Ga (Schopf and Packer, 1987; Awramik et al., 1983, 1988; Schopf, 1993), possibly older than 3.8 Ga (Schidlowski, 1988; Mojzsis et al., 1996; Ohmoto, 1997; Rossing, 1999). Oxygenic photosynthesizers, such as cyanobacteria, utilize the light from the Sun to fuel growth and produce O2 as a by-product. The overall chemical reaction of oxygenic photosynthesis is as follows:

...... CO2 + H2O = CH2O + O2 (1-1) 5

where "CH2O" represents organic matter. As a result of this reaction, photosynthetic organisms started pumping O2 into the atmosphere and began making the way for the later

evolution of multicellular life. However, most of the O2 produced by oxygenic photosynthesizers was consumed by the backward reaction of Eq. 1-1. The O2 accumulation in the atmosphere becomes possible only when the backward reaction of Eq.

1-1 is prevented; i.e., the removal of CH2O from the system (the burial of organic matter in the marine sediments). The burial flux of organic matter is equal to the net O2 production

flux into the atmosphere. However, the atmospheric O2 budget reflects the balance between its net production by photosynthesis and its consumption by reduced volcanic gases and weathering (e.g., Holland, 1984; Berner and Canfield, 1989).

1-2-3-2. Controversy over the rise of atmospheric O2

The timing of the rise of O2 in the ancient atmosphere has been vigorously debated since 1950, and no consensus has been reached (Fig. 1-1. e.g., Berkner and Marshall, 1965; Cloud, 1968, 1972; Dimroth and Kimberley, 1976; Walker, 1977; Clemney and Badham, 1982; Holland, 1984, 1994, 1999; Kasting, 1987, 1993, 2001; Lambert and Donnelly, 1991; Kasting et al., 1992; DesMarais et al., 1992; Han and Runneger, 1992; Ohmoto et al., 1993, 2001; Canfield and Teske, 1996; Karhu and Holland, 1996; Ohmoto, 1996, 1997, 1999; DesMarais, 1997; Holland and Rye, 1997; Canfield, 1998; Rye and Holland, 1998; Beaumont and Robert, 1999; Rasmussen and Buick, 1999; Canfield et al., 2000; Farquhar et al., 2000; Kump et al. 2000; Catling et al., 2001; Phillps et al., 2001; Lasaga and

-13 -3 Ohmoto, 2002a, b). One school postulates a very low O2 level (10 to 10 PAL: present atmospheric level) before its dramatic rise to > 0.15 PAL between 2.2-1.9 Ga (e.g., Kasting, 1993; Holland, 1994). In contrast, another school postulates an essentially constant

atmospheric O2 level since at least 3.8 Ga (e.g., Dimroth and Kimberley, 1976; Ohmoto,

1997) (Fig. 1-1). We must base any inference of the history of the atmospheric O2 level on 6

indirect evidence because of the lack of a direct sample of the ancient atmosphere. Geological records may have great potential to provide useful and critical information concerning the redox state of the ancient atmosphere. However, because of its indirect nature, much of it is circumstantial and all of it is no better than semi-quantitative (Holland, 1994).

1-2-3-3. Geological records bearing information on the atmospheric O2 Figure 1-2 summarizes the lines of geological evidence to support the model of the

rise of atmospheric O2 level between 2.2-1.9 Ga. At face value, these observations appear to provide compelling evidence for a reducing atmosphere prior to 2.2 Ga. However, a detailed examination of each individual line of 'evidence' results in, without difficulty, the realization that it is ambiguous and maybe even misleading (Fig. 1-2). In many cases, the same 'evidence' may be used to support alternative interpretations. Previous investigators have drawn contrasting conclusions about the redox state of the ancient atmosphere based on research using similar sets of samples (see the examples below) and analytical methods. In this section, the lines of controversial geological indicators for the rise of the pO2 level between 2.2 and 1.9 Ga are briefly summarized and contrasted with alternative interpretations. For a more detailed discussion, see Holland (1994, 1999), Ohmoto (1997), Phillips et al. (2001), and Ohmoto et al. (in prep).

Prevailing view: Low O2 level before 2.2 Ga The following geological observations have been used by researchers to suggest that

the pO2 levels were low before 2.2 Ga: (1) loss or retention of Fe in paleosols (e.g., Button, 1979; Gay and Grandstaff, 1979; Grandstaff et al., 1986; Holland and Zbinden, 1988; Pinto and Holland, 1988; Zbinden et al., 1988; Holland et al., 1989; Feakes et al., 1989; Holland and Beukes, 1990; Sutton and Maynard, 1992; Macfarlane et al., 1994a, 1994b; Rye et al., 7

1995; Rye and Holland, 1998; Pan and Stauffer, 2000; Murakami et al., 2001); (2) mineralization mechanisms for the U (e.g., Davidson, 1953, 1957; Davidson and Cosgrave, 1955; Roscoe, 1973; Minter, 1976, 1999; Grandstaff, 1980, 1986; Robertson, 1981; Robinson and Spooner, 1982, 1984a, 1984b; Robb et al., 1990, 1992; Robb and

Meyer, 1995; Frimmel, 1997); (3) occurrence of O2-sensitive heavy minerals as detrital components in ~3 Ga (Rasmussen and Buick, 1999); (4) age-distribution of red beds (Cloud, 1968; Eriksson and Cheney, 1992); (5) low content of redox-sensitive trace metals (e.g., Mo and U) in black shales (Davy, 1983); (6) age distribution and formational mechanism of iron formations (e.g., Garrels et al., 1973; Beukes and Klein, 1992; Klein and Beukes, 1989; 1992, 1993); (7) discovery of eukaryotes (Han and Runneger, 1992); (8) S isotopic composition of sulfides and sulfates for the secular changes in the S cycle (e.g., Cameron, 1982; Hattori et al., 1983a, 1983b, Hattori et al., 1985; Cameron and Hattori, 1987b; Canfield, 1998; Canfield and Raisewell, 1999); (9) mass-independent S isotope fractionation (Farquhar et al., 2000); and (10) secular changes in the N cycle (Beaumont and Robert, 1999).

Emerging view: High O2 level since ~3.8 Ga In contrast, the following lines of 'evidence' have been used by researchers to suggest that pO2 levels were high in the early Precambrian: (1) discovery of laterites at the top of a ~2.3 Ga paleosol profile (Ohmoto et al., 1999; Beukes et al., 2001); (2) common occurrence of Fe loss in paleosol of all ages, including the Phanerozoic, caused by variable processes including alteration by hydrothermal fluids, organic acids produced by soil biota (Palmer et al., 1989; Ohmoto, 1996), and local factors such as climate / topography / groundwater filtration (Schau and Henderson, 1983; Maynard, 1992); (3) development of oxidized paleosols of ~2.7 Ga in age (e.g., Kimberly and Grandstaff, 1986), of ~2.5 Ga in age (Nedachi Y. et al., 1998) and of ~2.3 Ga in age (Panahi et al., 2000); (4) hydrothermal 8 mineralization of uraninite and pyrite in U ores (e.g., Phillips et al., 1987; Philips and Myers, 1989; Barnicoat et al., 1997; Phillips and Law, 1997, 2000; Nedachi M. et al., 1998; Yamaguchi et al., 1998); (5) survival of detrital uraninite and pyrite in Phanerozoic sediments (Maynard et al., 1991; Maynard, 1992; Ono, 2001); (6) post-depositional mineralization (rather than detrital transport) of siderite (Ohmoto, 1999); (7) discovery of 2.7 Ga old red beds (Shegelski, 1980); (8) occurrence of ferric oxide crust of pillow (Dimroth and Lichtblau, 1978); (9) occurrence of iron-formations in Neoproterozoic (e.g., Klein and Beukes, 1993) and Paleozoic (e.g., Peter, 2001); (10) negative Ce anomaly in iron-formations (Yamaguchi et al., 2000; Bau et al., in prep); (11) large variations in the S isotopic compositions of sulfides in sediments (Ohmoto et al., 1993; Kakegawa and Ohmoto, 1999; Kakegawa et al., 1999, 2000; Shen et al., 2001) and in volcanogenic massive sulfide deposits (Huston et al., 2001); (12) abundance of organic carbon in Archean shales suggesting an operation of aerobic recycling (Towe, 1990, 1991, 1994); (13) redox-sensitive metals in black shales (Naraoka et al., 2002; Yamaguchi and Ohmoto, 2002); and (14) discovery of biomarkers for cyanobacteria and eukaryotes in Archean black shales (Brocks et al., 1999; confirmed by H. Naraoka, pers. comm., 2000; Eigenbrode et al., 2001). The stability of an O2-rich atmosphere has been recently demonstrated by geochemical dynamic modelling of O-C-S-Fe system (Lasaga and Ohmoto, 2002a, 2002b).

The historical development of the hypothesis of the reducing Archean atmosphere has its root in the early-middle last century in the recognition of 'detrital' uraninite and pyrite (unstable in an oxygenated environment) in the Witwatersrand Basin, South Africa. Later, the age-distributions of sedimentary rocks such as BIFs (banded iron-formations) and red beds were used to support the hypothesis. However, as introduced above, the geological 'evidence' used to support the hypothesis is becoming ambiguous and inconclusive. Therefore, it is premature to accept the hypothesis of a reducing atmosphere in the Archean. 9

1-3. Evolution of the biosphere in the early Precambrian

The biosphere and atmosphere have been closely linked and have co-evolved. This dissertation deals not only with the evolution of the atmosphere in the Archean and Paleoproterozoic (discussed above), but also with the evolution of the biosphere (evolution of cyanobacteria, sulfate-reducing bacteria, methanogenic and methanotrophic bacteria, nitrogen-fixing bacteria, and nitrogen-reducing bacteria). It is therefore appropriate to briefly review the current understanding of the evolution of the microbial biosphere in the early Precambrian. In this section, a brief review is provided in chronological order (Fig. 1- 2). Only the oldest geological and geochemical evidence reported is mentioned. Calibration of the first appearance of various types of early organisms by molecular clock dating is, unfortunately, not covered in the following brief review due to a rather large uncertainty in the age estimate by that method and a discrepancy with the various lines of geologic records introduced below.

1-3-1. Chemofossils of possible photosynthesizers at 3.8 Ga The oldest geological evidence for biological activity on Earth has been inferred from the metasedimentary rocks in the , southwest Greenland. The C

isotopic compositions of Corg (organic carbon) in metasedimentary rocks may indicate a widespread microbial (probably photosynthetic) activity in the ~3.8 Ga oceans (e.g., Schidlowski, 1982, 1987, 1988, 2001; Schidlowski et al., 1983; Schidlowski and Aharon, 1992; Mojzsis et al., 1996; Rossing, 1999). However, the evidence is not conclusive because the Corg isotope signatures could have been created or significantly modified by a post- depositional abiotic process such as metamorphism (e.g., Naraoka et al., 1996). 10

1-3-2. Microfossils and stromatolites at 3.5 Ga Geologic records in western Australia (the ) and South Africa (the Onverwacht Group) have provided evidence of early life on Earth in the forms of microfossils and stromatolites (e.g., Awramik et al., 1983, 1988; Lowe, 1983; Walsh and Lowe, 1985; Schopf and Packer, 1987; Buick, 1988; Walsh, 1992; Schopf, 1993; Walter, 1994; Walsh and Lowe, 1999; Westall et al., 2001). Stromatolites are lithified layers of microbial materials within sediments. Morphological analysis of these revealed similarities to modern cyanobacteria, suggesting that they are ancestors of modern cyanobacteria - and hence oxygenic photoautotrophs (e.g., Schopf and Packer, 1987; Schopf, 1993). However, the validity of such microfossil 'evidence' has been controversial.

The Corg isotopic compositions of these microfossils have been individually analyzed using the SIMS (secondary ion mass spectrometry) method (Ueno et al., 2001), first developed by House et al. (2000) for much younger (0.8 ~ 2.1 Ga old) microfossils in stromatolitic .

1-3-3. Sulfate-reducing bacteria at 3.5 ~ 3.4 Ga Studies of the S isotopic composition of sulfate and sulfide minerals in 3.5-3.4 Ga old barite (Shen et al., 2001), chert (Ohmoto et al., 1993), and shales (Kakegawa and Ohmoto, 1999) in South Africa and Australia indicate the presence of sulfate-reducing bacteria in the Archean oceans. This further suggests that the aqueous environment, either globally or locally, may have contained sufficient amounts of sulfate to allow for bacterial sulfate reduction. This may imply that the atmosphere was oxygenated enough for the existence of sulfate produced by the oxidative weathering of sulfide minerals in the continent. However, sulfate in seawater could have been produced by hydrolysis of volcanic

SO2 to generate 1/10 of the present level of seawater sulfate (28 mM; e.g., Libes, 1992). 11

1-3-4. Nitrogen-metabolizing bacteria at 3.5 Ga Beaumont and Robert (1999) analyzed the N isotopic compositions of kerogen from the early Precambrian chemical sediments including 3.5 Ga and iron- formations, and proposed that ammonia-metabolizing bacteria were active in the 3.5 Ga oceans. Additionally, they proposed that nitrate-metabolizing bacteria were not yet evolved until ~2.5 Ga when the inferred rise of atmospheric O2 occurred. However, this result was subsequently questioned by Pinti and Hashizume (2001), who argued that the N isotopic signatures reported by Beaumont and Robert (1999) were those of chemosynthetic microorganisms living near submarine hydrothermal vents rather than those of "normal" marine organisms.

1-3-5. Thermophilic microfossils at 3.2 Ga Microfossils of a possible thermophiles were very recently found in the 3.2 Ga VMS deposit at Sulfur Spring, Western Australia (Rasmussen, 2000). Morphological and organic geochemical evidence for the existence of microorganisms which lived in hot environments has also been demonstrated from the contemporaneous 3.2 Ga hydrothermal vents in the Barberton Greenstone Belt, South Africa (de Wit et al., 1982; de Ronde and Ebbesen, 1996).

1-3-6. Organic biomarkers for cyanobacteria, methanotrophic bacteria, and eukaryotes at 2.7 Ga Biomarkers of cyanobacteria and eukaryotes were found in the 2.7 Ga black shales in Western Australia (Brocks et al., 1999). This finding was confirmed by Naraoka (pers. comm. 2000) and Eigenbrode et al. (2001) using different samples from the same drillcores 12

(WRL1) used by Brocks et al. (1999). Furthermore, Eigenbrode et al. (2001) reported the discovery of biomarkers indicative of aerobic methanotrophy.

1-3-7. Methanogens and methanotrophs at 2.7 Ga

The Corg isotopic compositions of 2.7 Ga sedimentary rocks suggest that methanogenic and methanotrophic bacteria were active in the Neoarchean oceans (Hayes et al., 1983; Hayes, 1994). Rye and Holland (2000) reported Corg isotopic evidence of methanotrophs in a possible terrestrial pond preserved in a 2.76 Ga paleosol in Australia.

1-3-8. Life on land at 2.6 Ga Martini (1994) identified a 2.6 Ga paleosol developed on mafic rocks in South Africa. From a geochemical investigation the paleosol, Watanabe et al. (2000) found the remnants of terrestrial microbial mats, possibly of cyanobacterial in origin.

1-3-9. Eukaryotic megafossils at 2.1 Ga Megafossils of eukaryotes (Grypania) were found in Paleoproterozoic BIFs in Michigan (Han and Runneger, 1992). Grypania appear to have been photosynthetic autotrophs (Han and Runneger, 1992).

1-4. Evolution of the continental crust in the early Precambrian

The Earth is characterized by the existence of oceans and continental crusts (granitic rocks), which are missing in the other planets in today's solar system. The origin of continental crust on Earth is likely related to the operation of plate tectonics. However, the timing of, area / volume of, and processes responsible for continental crust formation and its 13

evolution through geologic time have been debated for over 30 years. Among such numerous unknowns, the important questions related to the present thesis include the following: (1) if continents of mass and area similar to those of today existed during the Archean (and which, if any, would have been destroyed by erosion and subduction in the course of plate tectonic processes) and (2) if continental crust chemically evolved from mafic-dominated crust to -dominated crust. A variety of models addressing the evolution of continental crust has been forwarded, ranging from the early "big bang" theory (e.g., Armstrong, 1968, 1981, 1991; Reymer and Schubert, 1984), and the gradual growth theory (e.g., Hurley and Rand, 1969), to the episodic growth theory (e.g., Taylor and McLennan, 1985, 1995; Condie, 1989). Neodymium isotope studies and Nb-U systematics of some ancient mantle-derived rocks (from which the continental crust was extracted) suggest the signatures of "evolved mantle", and therefore suggest an early growth of the continental crust (e.g., Bowring and Housh, 1995; Hofmann, 1997; Sylvester et al., 1997; Collerson and Kamber, 1999; Kerrich et al., 1999). The recent discovery of detrital zircons in early Archean sediments in Australia (Jack Hills in the Narryer Complex) that yield SHRIMP (sensitive high resolution ion- microprobe) U-Pb dates of 4.3-4.4 Ga (Mojzsis et al., 2001; Wilde et al., 2001) suggest, although do not prove, the existence of still older granitic continental crust as a source of the zircons and also the existence of oceans.

1-5. Evolution of the geochemical cycles of redox-sensitive elements (C, S, N, Fe, P, Mo, and U)

The concentrations of elements in the atmosphere and oceans have been regulated by a variety of geochemical interactions with the continental and oceanic lithospheres. The 14 cycling of an element between the Earth's major reservoirs is referred to as the geochemical cycle of that element. The geochemical cycles of redox-sensitive elements in the Earth's surface environments have been mediated by complex redox reactions involving both biological and non-biological controls. The evolution of the geochemical cycles of redox- sensitive elements are therefore intimately linked to the evolution of the atmosphere, oceans, and biosphere. This study assesses the evolution of the geochemical cycles of some redox- sensitive elements (C, S, N, Fe, P, Mo, and U) through geologic time. In this section, the geochemical cycling of these redox-sensitive elements through the Earth's major reservoirs and the behaviors of these elements in typical modern marine sediments are briefly summarized. The geochemical cycles of the individual elements are treated in more detail in the following chapters.

1-5-1. Geochemical cycles of redox-sensitive elements We consider a simple mass-balance model for the geochemical cycles of C, S, N, Fe, P, Mo, and U using a system composed of four major reservoirs: continents, oceans, sediments, and mid-oceanic ridges (MOR) (Fig. 1-3). The atmosphere is implicitly included in the continents and oceans reservoirs. The mantle is another important reservoir; however, it is not explicitly considered here because we are primarily concerned with the geochemical cycles of those elements in the Earth's surface environments.

Through continental weathering (flux: Fweath), the physically / chemically weathered materials (fluxes: Fphysw and Fchemw) are transported through rivers to the oceans via two mechanisms: the detrital (particulate) transport (flux: Fdet) and the dissolved transport (flux:

Fdis). For an element i,

i i i i i ...... F weath = F physw + F chemw = F dis + F det (1-2) 15

Continental chemical weathering could be oxidative or reductive weathering depending on

the global / local redox conditions. The hydrothermal flux (Fhyd) from the submarine MOR system also enters into the oceans, although Fhyd is variable in space and time depending on the activity of the MOR system. Some elements are regenerated from the sediments into oceans (flux: Freg), depending on the redox conditions of bottom water. Therefore the input flux of an element to the oceans (dissolved component; Fin, ocean) is the sum of the Fdis,

Fhyd, and Freg.

i i i i ...... F in, ocean = F dis + F hyd + F reg (1-3)

Removal of an element from oceans (flux: Fout, ocean) occurs in two ways: burial in

sediments (flux: Fppt; ppt: precipitation) and fixation in MOR system through its seawater circulation (flux: FMOR).

i i i ...... F out, ocean = F ppt + F MOR (1-4)

i i At a steady state (F in, ocean = F out, ocean),

i i i i i ...... F dis + F hyd + F reg = F ppt + F MOR (1-5)

i i i i i i i i All of the fluxes F weath, F dis (F chemw), F det (F physw), F hyd, F reg, F ppt, and i F MOR are complex functions with numerous parameters, such as pO2, pCO2, topography, climate, ocean chemistry, heat flux, etc. We extract important differences in these fluxes mainly between globally oxic and globally anoxic environments. In a globally anoxic world, where the atmosphere and entire oceans are anoxic, continental weathering (Fweath) may be dominated by physical weathering (Fphysw) and by reductive chemical weathering. Organic 16

carbon, S, N, Fe, Mo, and U may be transported through rivers to the oceans mainly as detrital forms with their reduced valency (C0, S2-, N3-, Fe2+, Mo4+, and U4+; Fig. 1-3). Iron may also be transported as dissolved Fe2+ by the reductive dissolution of Fe3+-bearing minerals (e.g., magnetite) as well in congruent dissolution of the Fe2+-bearing minerals. Accordingly, the following relationships would be expected for the redox-sensitive elements.

For C, S, N, Mo, and U:

... Fphysw/Fchemw (anoxic atmosph.) > Fphysw/Fchemw (oxic atmosph.) (1-6)

For Fe:

... Fphysw/Fchemw (anoxic atmosph.) < Fphysw/Fchemw (oxic atmosph.) (1-7)

In contrast, in a globally oxic world where the entire atmosphere and oceans are oxic

with local anoxic environments (e.g., mid-depth O2 minimum zone and anoxic basins) as

today, continental weathering (Fweath) would be dominated by the oxidative chemical

weathering and the physical weathering (Fphysw). Organic carbon, S, N, Mo, and U would be transported through rivers to the oceans mainly as dissolved (oxidized) forms with their increased valency (C4+, S6+, N5+, Mo6+, and U6+; Fig. 1-3). Iron would be transported in oxidized (Fe3+), detrital forms. Global redox conditions would control the valence of elements and their total weathering flux.

For C, S, P, N, Mo, and U:

...... Fweath (anoxic atmosphere) < Fweath (oxic atmosphere) (1-8)

For Fe: 17

...... Fweath (anoxic atmosphere) > Fweath (oxic atmosphere) (1-9)

Phosphorus does not change its valency during chemical weathering; however, its behavior in the oceans does depend on the redox state of the water body. This is because P adsorbed on Fe (Mn)-oxyhydroxides formed in oxic oceans is released into seawater upon reductive dissolution of Fe (Mn)-oxyhydroxides during settling in local anoxic water columns and/or early diagenesis in sediments.

P P ...... F reg (anoxic bottom water) > F reg (oxic bottom water) (1-10)

A higher surface temperature and higher pCO2 in the Archean compared to today (e.g., Sclater et al., 1980; Kasting, 1993) would have enhanced continental weathering rates; however, this could have been compensated to some degree by the suppressed continental

(oxidative) chemical weathering rate if the Archean atmosphere had a low pO2 level. Additionally, the inferred higher heat flux in the Archean would have enhanced both the

Fhyd and FMOR values because of enhanced seawater hydrothermal circulation through MOR.

i i ...... F hyd (Archean) > F hyd (today) (1-11) i i ...... F MOR (Archean) > F MOR (today) (1-12)

1-5-2. Influence of pO2 on the geochemical cycles of redox-sensitive elements A sequence of redox reactions mediated by microorganisms in a typical profile of modern marine sediments is calculated using the thermodynamic data in literature and shown in Fig. 1-4. Organic matter is the principal source of energy to promote these

reactions. The general sequence of oxidants used in the decomposition of OM is O2 18

(5+) - (4+) (3+) (aerobic respiration) ---> N O3 (denitrification) ---> Mn O2 ---> Fe (OH)3 ---> (6+) 2- (6+) 2- (6+) 2- (4+) U O2(CO3)2 ---> Mo O4 ---> S O4 (sulfate reduction) ---> C O2 (methanogenesis). This sequence corresponds to decreases in the redox potential of the oxidants (electron donors) and thus decreases in the free energy available by respiration with the different reductants (electron acceptors) (Fig. 1-4). The diagram in Fig. 1-4 is also

-10 important in the evolution of the Fdis/Fdet ratios. At pO2 levels higher than 10 atm, all the redox-sensitive elements of interest (C, S, N, Fe, Mo, and U) are in their oxidized valence

states and Fdis is expected to be more important than Fdet, although the kinetics of chemical weathering reactions should be taken into account. The development of these redox sequences in marine sediments is controlled by a variety of microbiological organisms that depend on the development of anoxic conditions within sediments and/or in the overlying water column. Various types of authigenic minerals are formed in marine sediments during early diagenesis. Fe-sulfide is quantitatively the most important phase. Useful information regarding the geochemical cycling of elements, redox conditions and biological activity can be obtained by examining the enrichment patterns of redox-sensitive elements in sediments and sedimentary rocks. In particular, a coupled approach using the inorganic geochemistry of redox-sensitive elements and organic and stable isotope geochemistry has proven to be a powerful tool to extract (paleo)environmental information.

1-6. Objectives of the thesis

A suite of geochemical and stable isotopic analyses was performed on modern weathering-free Archean–Paleoproterozoic sedimentary rocks (mostly carbonaceous black shales, with minor graywackes and red shales) ranging from 3.25 to 2.2 Ga, in order to 19

constrain the redox evolution of the atmosphere and oceans, the evolution of the continental crust, and the evolution of the biosphere. The main objective of this study is to better understand the early Precambrian evolution of the atmosphere, oceans, crust, and biosphere.

This objective was pursued from an array of inter-related studies: (1) Corg-Spy-Fe-P systematics, (2) Mo geochemistry, (3) N isotope geochemistry, and (4) U-Th geochemistry. For comparison, data from modern sediments and Phanerozoic sedimentary rocks, similar to those of the Archean–Paleoproterozoic samples of this study (i.e., C, S, N, Fe, P, Mo, and U), were complied from literature. Based on these data, the geochemical cycles of C, S, N, Fe, P, Mo, and U in the Archean–Paleoproterozoic surface environments are compared with those of modern–Phanerozoic. Particular focus is placed on the changes in the Fdis/Fdet ratios. The subordinate objectives of each theme are described in its respective chapter.

1-7. Organization of the thesis

This PhD thesis is written based on the studies during the course of doctoral research at The Pennsylvania State University between 1995 and 2001 under the supervision of Prof. Hiroshi Ohmoto. Not all the studies performed are included in this thesis. The thesis is composed of the following five chapters and four appendices. Chapter 1 is intended as an overall introduction of the thesis. Brief reviews are

provided for the evolution of the atmosphere (the "Rise of O2" controversy), oceans, crust, biosphere, and geochemical cycles of redox-sensitive elements. It is followed by objectives, organization, and significance of this study.

Chapter 2 is about the systematics among organic C (Corg) - pyrite S (Spy) - Fe - P, titled "Geochemistry of Archean–Paleoproterozoic black shales, II: Diversity in the redox states of sedimentary environments inferred from C–S–Fe-P systematics and C–S stable 20

3+ 2+ isotopes". Based on the contents of Corg, Spy, Fe -Fe , and P and the stable isotopic compositions of Corg and Spy for the Archean-Paleoproterozoic black shales, we suggest that aerobic recycling of Corg, the present-day distribution of sedimentary Corg and Spy influenced by biological activity, redox control on the Corg/P ratios, oxidative weathering and post-depositional redox reactions for Fe, (oxygenic) photosynthetic activities, and bacterial sulfate reduction have operated throughout geologic time. Chapter 3 deals with the geochemistry of a redox-sensitive element, Mo, titled "Geochemistry of Archean–Paleoproterozoic black shales, II: Geochemical cycle of molybdenum". Continental weathering flux of Mo in the Archean–Paleoproterozoic ages is suggested to have been comparable to that of today based on the dissolution kinetics of Mo- bearing minerals. The diagenetic fixation of dissolved Mo by organic matter and biogenic sulfide is suggested to have occurred in the Archean-Paleoproterozoic black shales of this study. Chapter 4 addresses the evolution of N-metabolizing microorganisms, titled "Geochemistry of Archean–Paleoproterozoic black shales, IV: Fixation and redox cycling of nitrogen by organisms". Redox-cycling of nitrogen, including N2-fixation, nitrification, and denitrification by microorganisms in the Archean, is proposed based on the stable isotope signatures of N in organic matter and clay in the Archean–Paleoproterozoic sedimentary rocks and their similarities to those of modern and Phanerozoic sediments. Chapter 5 is concerned with the geochemical cycle of another important redox- sensitive element, U. It is titled "Geochemistry of Archean–Paleoproterozoic black shales, V: Geochemical cycle of uranium and the evolution of continental crust". We have found that the decoupling of U from Th in oxidizing environments and tectonic recycling of U have been important throughout geologic time. Crustal evolution in an oxidizing environment is implied. 21

Appendices complement the thesis with the detailed descriptions of the geologic settings, nature of samples, and analytical methods utilized in this study. Appendix A provides the detailed descriptions for the geologic setting (stratigraphy, tectonic setting, and metamorphism) of the Swaziland, Witwatersrand, Ventersdorp, and Transvaal Supergroups in South Africa and the Mt. Bruce Supergroup in Australia. In Appendix B, the samples used in this study are described in detail. Appendix C presents the detailed analytical procedures for the various types of chemical and isotopic analyses performed in this study. Appendix D presents a published paper on the mass-independent S isotope fractionation. A part of this paper has been based on a manuscript written by the author in Sep. 2000.

1-8. Significance and implications of the thesis

The most significant contribution from the series of studies presented in this thesis on the geochemistry of the Archean–Paleoproterozoic shales is the suggestion of the early development of the present day-style redox environments and microbial activity in the

Archean. In the Archean, N2-fixers, photosynthesizers, nitrifyers, denitrifyers, sulfate- reducers, and methane-producing / consuming microorganisms had already formed complex ecosystems, like those of today, in the globally oxic oceans with local anoxic / euxinic environments overlain by the globally oxic atmosphere. This study has implications for the early evolution of an oxic atmosphere, oxic oceans, and complex microbial biosphere not only on the Earth but also on the other Earth-like planets distributed in the universe. 22 References

Armstrong, R.L. (1968) A model for Sr and Pb isotope evolution in a dynamic Earth. Rev. Geophys. 6, 175-199.

Armstrong, R.L. (1981) Radiogenic isotopes: the case for crustal recycling on a near- steady-state no-continental-growth Earth. Phil. Trans. Royal Soc. Lond., A 301, 443-472.

Armstrong, R.L. (1991) The persistent myth of crustal growth. Austral J. Earth Sci. 38, 613-630.

Awramik, S.M., Schopf, J.W., and Walter, M.R. (1983) Filamentous bacteria from the Archean of Western Australia. Precam. Res. 20, 357-374.

Awramik, S.M., Schopf, J.W. and Walter, M.R. (1988) Carbonaceous filaments from North Pole, Western Australia: Are they fossil bacteria in Archaean stromatolites? A discussion. Precam. Res. 39, 303-309.

Barnicoat, A.C., Henderson, I.H.C., Knipe, R.J., Yardley, B.W., Napier, R.W., Fox, N.P.C., Kenyon, A.K., Muntigh, D.J., Strydom, D., Winkler, K.S., Lawrrence, S.R., and Conford, C. (1997) Hydrothermal mineralization in the Witwatersrand basin. Nature 386, 820-824.

Bau, M., Dulski, P., and Ohmoto, H. (in prep) Negative Ce anomaly in the 2.7 Ga Algoma- type Temagami iron-formations, Ontario, Canada.

Beaumont, V. and Robert, F. (1999) Nitrogen isotope ratios of in Precambrian cherts: a record of the evolution of atmosphere chemistry? Precam. Res. 96, 63-82.

Berkner, L.V. and Marshall, L.C. (1965) On the origin and rise of oxygen concentration in the Earth's atmosphere. J. Atmos. Sci. 22, 225-261.

Berner, R.A. and Canfield, D.E. (1989) A new model for atmospheric oxygen over Phanerozoic time. Am. J. Sci. 289, 333-361.

Beukes, N.J. and Lowe, D.R. (1989) Environmental control on diverse morphologies in the 3000 Myr Pongola Supergroup, South Africa. Sedimentology 36, 383-397.

Beukes, N.J. and Klein, C. (1992) Models for iron-formation deposition. In The Biosphere: A Multidisciplinary Study (eds. J.W. Schopf, and C. Klein), Cambridge University Press, Cambridge, , 147-152. 23

Beukes, N.J., Gutzmer, J., and Dorland, H. (2001) Tropical laterites, atmospheric O2 and CO2 levels, and life on land in the early Proterozoic. European Union of Geoscience Meeting, Abstract, Strasbourg, March 2001, p64.

Bowring, S.A. and Housh, T. (1995) The Earth's early evolution. Science 269, 1535-1540.

Brocks, J.J., Logan, G.A., Buick, R., and Summons, R.E. (1999) Archean Molecular Fossils and the Early Rise of Eukaryotes. Science 285, 1033-1036.

Buick, R. (1988) Carbonaceous filaments from North Pole, Western Australia: Are they fossil bacteria in Archaean stromatolites? A reply. Precam. Res. 39, 311-317.

Buick, R. (1992) The antiquity of oxygenic photosynthesis: Evidence from stromatolites in sulfate-deficient Archean lakes. Science 255, 74-77.

Button, A. (1979) Early Proterozoic weathering profile on the 2200 m.y. old Hekpoort Basalt, Pretoria Group, South Africa: Preliminary results. Econ. Geol. Res. Unit, Univ. of Witwatersrand, Info. Circ., 133.

Cameron, E.M. (1982) Sulphate and sulphate reduction in early Precambrian ocean. Nature 296, 145-148.

Cameron, E.M. and Hattori, K. (1987b) Archean sulphur cycle: Evidence from sulphate minerals and isotopically fractionated sulphides in Superior province, Canada. Chem. Geol. 65, 341-358.

Canfield, D.E. and Teske, A. (1996) Late Proterozoic rise in atmospheric oxygen concentration inferred from phylogenetic and sulfur-isotope studies. Nature 382, 127-132.

Canfield, D.E. (1998) A new model for Proterozoic ocean chemistry. Nature 396, 450-453.

Canfield, D.E. and Raisewell, R. (1999) The evolution of the sulfur cycle. Am. J. Sci. 299, 697-723.

Canfield, D.E., Habicht, K.S., and Thamdrup, B. (2000) The Archean sulfur cycle and the early history of atmospheric oxygen. Science 288, 658-661.

Canuto, V.M., Levine, J.S., Augustsson, T.T., Imhoff, C.L., and Giampapa, M.S. (1983) The young Sun and the atmosphere and photochemistry of the early Earth. Nature 305, 281-286.

Catling, D.C., Zahnle, K.J., McKay, C.P. (2001) Biogenic methane, hydrogen escape, and the irreversible oxidation of early Earth. Science 293, 839-843. 24 Chyba, C.F., Thomas, P.J., Brookshaw, L., and Sagan, C. (1990) Cometary delivery of organic molecules to the early Earth. Science 249, 366-373.

Chyba, C.F. and Sagan, C. (1992) Endogenous production, exogenous delivery, and impact- shock synthesis of organic molecules: An inventory for the origin of life. Nature 355, 125-131.

Chyba, C.F. (1993) The violent environment of the origin of life: Progress and uncertainties. Geochim. Cosmochim. Acta 57, 3351-3358.

Clemmey, H. and Badham, N. (1982) Oxygen in the Precambrian atmosphere: An evolution of the geological evidence. Geology 10, 141-146.

Cloud, P. (1968) Atmospheric and hydrospheric evolution on the primitive Earth. Science 160, 729-736.

Cloud, P. (1972) A working model of the primitive earth. Am. J. Sci. 272, 537-548.

Collerson, K.D. and Kamber, B.S. (1999) Evolution of the continents and the atmosphere inferred from Th-U-Nb systematics of the depleted mantle. Science 283, 1519-1522.

Condie, K.C. (1989) Plate tectonics and crustal evolution. 3rd ed., Pergamon Press, 476 pp.

Davidson, C.F. (1953) The gold-uranium ores of the Witwatersrand. Mining Mag. 88, 73- 85.

Davidson, C.F. (1957) On the occurrence of uranium in ancient conglomerates. Econ. Geol. 52, 668-693.

Davidson, C.F. and Cosgrave, M.E. (1955) On the importance of uraninite as a detrital minerals. Geol. Surv. Great Britain Bull. 10, 74-80.

Davy, R. (1983) A geochemical study of the Mount McRae shale and the upper part of the Mount Sylvia Formation in Core RD1, Rhodes Ridge, Western Australia. Geol. Surv. W. Aust. Record 1983/3. de Ronde, C.E. and Ebbesen, T.W. (1996) 3.2 b.y. of organic compound formation near sea-floor hot springs. Geology 24, 791-794. de Wit, M.J., Hart, R., Martin, A., and Abbot, P. (1982) Archean abiogenic and probable biogenic structures associated with mineralized hydrothermal vent systems and regional metasomatism, with implications for greenstone belt studies. Econ. Geol. 77, 1783-1802. 25 Des Marais, D.J., Strauss, H., Summons, R.E., and Hayes, J.M. (1992) Carbon isotope evidence for the stepwise oxidation of the Proterozoic environment. Nature 359, 605-609.

Des Marais, D.J. (1997) Long-term evolution of the biogeochemical carbon cycle. In Geomicrobiology: Interaction between microbes and minerals (eds. J.F. Banfield and K.H. Nealson), Reviews in Mineralogy 35, . Soc. Amer., Washington D.C.

Dimroth, E. and Kimberley, M.M. (1976) Precambrian atmospheric oxygen: evidence in the sedimentary distributions of carbon, sulfur, uranium, and iron. Can. J. Earth Sci. 13, 1161-1185.

Dimroth, E. and Lichtblau, A.P. (1978) Oxygen in the Archean ocean: Comparison of ferric oxide crusts on Archean and Cainozoic pillow . Neues Jajrbuch für Mineralogie, Abhandlungen, 113, 1-22.

Eigenbrode, J.L., Freeman, K.H., Brocks, J.J., Summons, R.E., and Logan, G.A. (2001) Late Archean biomarkers of carbonate and shale lithofacies from the Hamersley Basin, Pilbara Craton, Western Australia. 11th V.M. Goldschmidt Conf., Abstract 3461.

Eriksson, P.G. and Cheney, E.S. (1992) Evidence for the transition to an oxygen-rich atmosphere during the evolution of red beds in the Lower Proterozoic sequences of . Precam. Res. 54, 257-269.

Farmer, J.D. (2000) Hydrothermal systems: Doorways to early biosphere evolution. GSA Today, 10, 1-9.

Farquhar, J., Bao, H. and Thiemens, M. (2000) Atmospheric influence of earth's earliest sulfur cycle. Science 289, 756-758.

Frimmel, H.E. (1997) Detrital origin of hydrothermal Witwatersrand gold - A review. Terra Nova 9, 192-197.

Garrels, R.M., Perry, E.A.Jr., Mackenzie, F.T. (1973) Genesis of Precambrian iron- formations and the development of atmospheric oxygen. Econ. Geol. 68, 1173- 1179.

Gay, A.L. and Grandstaff, D.E. (1979) Chemistry and mineralogy of Precambrian paleosol at Elliot Lake, Ontario, Canada. Precam. Res. 12, 349-373.

Grandstaff, D.E. (1980) Origin of uraniferous conglomerates at Elliot Lake, Canada, and Witwatersrand, South Africa: Implications for oxygen in the Precambrian atmosphere. Precam. Res. 13, 1-26. 26 Grandstaff, D.E. (1986) Uraninite oxidation and the Precambrian atmosphere. US Geol. Surv. Prof. Paper 1161, C1 - C16.1161.

Grandstaff, D.E., Edelman, M. J., Foster, R. W., Zbinden, E. and Kimberley, M. M. (1986) Chemistry and mineralogy of Precambrian paleosols at the base of the Dominion and Pongola Groups (Transvaal, South Africa). Precam. Res. 32, 97-131.

Han, T.M. and Runneger, B. (1992) Megascopic eukaryotic algae from the 2.1 billion-year- old Negaunee Iron-Formation, Michigan. Science 257, 232-235.

Hattori, K., Campbell, F. A., and Krouse, H.R. (1983a) Sulphur isotope abundances in Aphebian clastic rocks: implications for the coeval atmosphere. Nature 302, 323- 326.

Hattori, K., Krouse, H.R., and Campbell, F.A. (1983b) The start of sulfur oxidation in continental environments: about 2.2 x 109 years ago. Science 221, 549-551.

Hattori, K., Campbell, F.A., and Krouse, H.R. (1985) Sulfur isotope abundances in sedimentary rocks, relevance to the evolution of the Precambrian atmosphere. Geochem. Int. 22, 97-114.

Hayes, J.M., Kaplan I.R., and Wedeking K.M. (1983) Precambrian organic geochemistry, Preservation of the record. In Earth's Earliest Biosphere: Its Origin and Evolution (ed. J.W. Schopf), pp. 93-135. Princeton Univ. Press, Princeton.

Hayes, J.M. (1994) Global methanotrophy at the Archean-Proterozoic transition. In Early Life on Earth. Nobel Symposium No. 84 (ed. S. Bengston). pp. 220-236. Columbia Univ. Press, New York.

Hofmann, A.W. (1997) Early evolution of continents. Science 275, 498-499.

Holland, H.D. (1984) Chemical evolution of the atmosphere and ocean. Princeton University Press, Princeton, 582p.

Holland, H.D. and Zbinden, E.A. (1988) Paleosols and the evolution of the atmosphere: Part 1. In Physical and Chemical Weathering in Geochemical Cycles (eds. A. Lerman. and M. Meybeck), Kluwer Academic Publishers, Dordrecht, 61-82.

Holland, H.D. and Beukes, N.J. (1990) A paleoweathering profile from Griqualand West, South Africa: Evidence for a dramatic rise in atmospheric oxygen between 2.2 and 1.9 b.y.b.p. Am. J. Sci. 290-A, 1-34. 27 Holland, H.D. (1994) Early Proterozoic atmospheric change. In Early Life on Earth. Nobel Symposium No. 84 (ed. S.Bengston) Columbia University Press, New York, 237- 244.

Holland, H. and Rye, R. (1997) Evidence in pre-2.2 Ga paleosols for the early evolution of atmospheric oxygen and terrestrial biota: Comment. Geology 25, 857-859.

Holland, H.D. (1999) When did the Earth's atmosphere become oxic? A Reply. Geochemical News 100, 20-22.

House, C.H., Schopf, J.W., McKeegan, K.D., Coath, C.D., Harrison, T.M., and Stetter, K.P. (2000) Carbon isotopic composition of individual Precambrian microfossils. Geology 28, 707-710.

Hurley, P.M. and Rand, J.R. (1969) Pre-Drift Continental Nuclei. Science 164, 1229-1242.

Huston, D.L., Brauhart, C.W., Drieberg, S.L., Davidson, G.J., and Groves, D.I. (2001) Metal leaching and inorganic sulfate reduction in volcanic-hosted massive sulfide mineral systems: Evidence from the paleo-Archean Panorama district, Western Australia. Geology 29, 687-690.

Kakegawa, T. and Ohmoto, H. (1999) Sulfur isotope evidence for the origin of 3.4 to 3.1 Ga pyrite at the Princeton gold mine, Barberton Greenstone Belt, South Africa. Precam. Res. 96, 209-224.

Kakegawa, T., Kawai, H., and Ohmoto, H. (1999) Origin of in ~2.5 Ga Mt. McRae shale, the Hamersley district, Western Australia. Geochim. Cosmochim. Acta 62, 3205-3220.

Kakegawa. T., Kasahara, Y., Hayashi, K., and Ohmoto, H. (2000) Sulfur and carbon isotope analyses of the 2.7 Ga Jeerinah Formation, Fortescue Group, Australia. Geochem. J. 34, 121-133.

Karhu, J.A. and Holland, H.D. (1996) Carbon isotopes and the rise of atmospheric oxygen. Geology 24, 867-870.

Kasting, J.F. (1987) Theoretical constraints on oxygen and carbon dioxide concentrations in the Precambrian atmosphere. Precam. Res. 34, 205-228.

Kasting, J.F. (1993) Earth's early atmosphere. Science 259, 920-926.

Kasting, J.F. (2001) The rise of atmospheric oxygen. Science 293, 819-820. 28 Kasting, J.F. and Walker, J.C.G. (1981) Limits on oxygen concentration in the prebiological atmosphere and the rate of abiotic fixation of nitrogen. J. Geophys. Res. 86, 1147-1158.

Kasting, J.F., Holland, H.D., and Kump, L.R. (1992) Atmospheric evolution: the rise of oxygen. In The Proterozoic Biosphere: A Multidisciplinary Study. (eds., J.W. Schopf and C. Klein ), Cambridge University Press, Cambridge, England, 159-164.

Kasting, J.F., Eggler, D.H., and Raeburn, S.P. (1993) Mantle redox evolution and the oxidation state of the Archean atmosphere. J. Geol. 101, 245-257.

Kerrich, R., Wyman, D., Hollings, P., and Polat, A. (1999) Variability of Nb/U and Th/La in 3.0 to 2.7 Ga Superior Province ocean plateau basalts: implications for the timing of continental growth and lithosphere recycling. Earth Planet. Sci. Lett. 168, 101-115.

Kimberley, M.M. and Grandstaff, D.E. (1986) Profiles of elemental concentrations in Precambrian paleosols on basaltic and granitic parent materials. Precam Res. 32, 133-154.

Klein, C. and Beukes, N.J. (1989) Geochemistry and sedimentology of a facies transition from to iron-formation deposition in the early Proterozoic Transvaal Supergroup, South Africa. Econ. Geol. 84, 1733-1774.

Klein, C. and Beukes, N.J. (1992) Time distribution, stratigraphy, and sedimentologic setting, and geochemistry of Precambrian iron-formations. In The Proterozoic Biosphere: A Multidisciplinary Study (eds. J.W. Schopf, and C. Klein), Cambridge University Press, Cambridge, England, 139-146.

Klein, C. and Beukes, N.J. (1993) Sedimentology and Geochemistry of the Glaciogenic Late Proterozoic Rapitan Iron-Formation in Canada. Econ. Geol. 88, 542-565.

Kump, L. R., Kasting, J. F.& Barley, M. E. (2000) Rise of atmospheric oxygen and the "upside-down" Archean mantle. Geochem. Geophys. Geosyst. 2, #2000GC000114.

Lambert, I.B. and Donnelly, T.H. (1991) Atmospheric oxygen levels in the Precambrian: a review of isotopic and geological evidence. Paleogeogr. Paleoclimatol. Paleoecol. (Global Planet. Change Sec.) 97, 83-91.

Lasaga, A.C. and Ohmoto, H. (2002a) The oxygen geochemical cycle: dynamics and stability. Geochim. Cosmochim. Acta 66, 361-381.

Lasaga, A.C. and Ohmoto, H. (2002b) Long term evolution of atmospheric oxygen and carbon dioxide. Am. J. Sci. (in review). 29 Libes, S.M. (1992) An introduction to marine biogeochemistry. John Wiley and Sons, Inc. New York. 734p.

Lowe, D.R. (1983) Restricted shallow-water of early Archean stromatolitic and evaporitic strata of the Strelley Pool chert, Pilbara Block, Western Australia. Precam. Res. 19, 239-283.

Macfarlane, A.W., Danielson A., Holland H.D., and Jacobsen S. B. (1994a) REE chemistry and Sm-Nd systematics of late Archean weathering profiles in the Fortescue Group, Western Australia. Geochim. Cosmochim. Acta 58, 1777-1794.

Macfarlane, A.W., Danielson, A., and Holland, H.D. (1994b) Geology and major and trace elements chemistry of late Archean weathering profiles in the Fortescue Group, Western Australia: implications for atmospheric pO2. Precam. Res. 65, 297-317.

Maher, K.A. and Stevenson, D.J. (1988) Impact frustration of the origin of life. Nature 331, 612-614.

Mancinelli, R.L. and McKay, C.P. (1988) The evolution of nitrogen cycling. Origins Life Evol. Biosph. 18,.311-325.

Maynard, J.B., Ritger, S.D., and Sutton, S.J. (1991) Chemistry of from the modern Indus River and the Archean Witwatersrand basin: Implications for the composition of the Archean atmosphere. Geology 19, 265-268.

Maynard, J.B. (1992) Chemistry of modern soils as a guide to interpreting Precambrian paleosols. J. Geol. 100, 279-289.

Miller, S.L. (1953) A production of amino acids under possible primitive Earth conditions. Science 117, 528-529.

Miller, S.L. and Urey, H.C. (1959) Organic compound synthesis on the primitive Earth. Science 130, 245-251.

Minter, W.E.L. (1976) Detrital gold, uranium, and pyrite concentrations related to sedimentology in the Precambrian Vaal Reef Placer, Witwatersrand, South Africa. Econ. Geol. Econ. Geol. 71, 157-176.

Minter, W.E.L. (1999) Irrefutable detrital origin of the Witwatersrand gold and evidence of eolian signatures. Econ. Geol. 94, 665-670.

Mojzsis, S.J., Arrhenius, G., McKeegan, K., Harrison, T.M., Nutman, A.P., and Friend, C.R.L. (1996) Evidence for life on Earth before 3,800 million years ago. Nature 384, 55-59. 30 Mojzsis, S.J., Harrison, T.M., and Pidgeon, R.T. (2001) Oxygen-isotope evidence from ancient zircons for liquid water at the Earth's surface 4,300 Myr ago. Nature 409, 178-181.

Murakami, T., Utsunomiya, S., Imazu, Y.& Prasad, N. (2001) Direct evidence of late Archean to early Proterozoic anoxic atmosphere from a product of 2.5 Ga old weathering. Earth Planet. Sci. Lett. 184, 523-528.

Naraoka, H., Ohtake, M., Maruyama, S., and Ohmoto, H. (1996) Non-biogenic graphite in 3.8 Ga metamorphic rocks from the Isua district, Greenland. Chem. Geol. 133, 251- 260.

Naraoka, H., Hayashi, K. and Ohmoto, H. (2001) Chemical and isotopic composition of Mount McRae Shale (~2.5 Ga) at Hamersley district, Western Australia: Its implication for the depositional environment and biological activity. Submitted to Geochim. Cosmochim. Acta.

Navarro-González, R., McKay, C.P., and Mvondo D.N. (2001) A possible nitrogen crisis for Archean life due to reduced nitrogen fixation by lightning. Nature 412, 61-64.

Nedachi, M., Yamanouchi, H., and Ohmoto, H. (1998) Detrital and hydrothermal origins for , pyrite, and uraninite in Au-U-rich conglomerates of the Witwatersrand gold field, South Africa. Mineral. Mag. 62A, 1060-1061.

Nedachi, Y., Nedachi, M., Bennett, G., and Ohmoto, H. (1998) Weathering of granite under an O2-and CO2-rich atmosphere 2.45 Ga at Pronto, Ontario, Canada. Mineral. Mag. 62A, 1062-1063.

Nisbet, E.G. (1995) Archaean ecology; a review of evidence for the early development of bacteria biomes, and speculations on the development of a global-scale biosphere. In Early Precambrian processes (eds., M.P. Coward and A.C. Ries), Geol. Soc. Spec. Pub. 95, Geol. Soc. London., UK, pp 27-51.

Ohmoto, H. (1996) Evidence in pre-2.2 Ga paleosols for the early evolution of atmospheric oxygen and terrestrial biota. Geology 24, 1135-1138.

Ohmoto, H. (1997) When did the Earth's atmosphere become oxic? The Geochemical News 93, 12-13 and 26-27.

Ohmoto, H. (1999) Redox state of the Archean atmosphere: Evidence from detrital heavy minerals in ca. 3250-2750 Ma sandstones from the Pilbara Craton, Australia: Comment. Geology 27, 1151-1152. 31 Ohmoto, H., Kakegawa, T., and Lowe, D.R. (1993) 3.4-billion-year-old biogenic pyrites from Barberton, South Africa: sulfur isotope evidence. Science 262, 555-557.

Ohmoto, H., Beukes, N.J., Gutzmer, J., and Nedachi, M. (1999) The formation of laterites approximately 2.3 billion years ago. Abst. with Programs, Geol. Soc. Amer. 31, 225-226.

Ohmoto, H., Yamaguchi, K.E, and Ono, S. (2001) Questions regarding Precambrian sulfur isotope fractionation. Science 292, 1959a.

Ohmoto, H. et al. (in prep) Precambrian atmospheric oxygen. A review. To be submitted to Ann. Rev. Earth Planet Sci.

Ono, S. (2001) Unpublished PhD thesis, Pennsylvania State University.

Overbeck, V.R. and Fogleman, G. (1989) On the possibility of life on early Mars. Abst. Lunar Planet. Sci. Conf. 20, 800-801.

Palmer, J.A., Phillips, G.N., and McCarthy, T.S. (1989) Paleosols and their relevance to Precambrian atmospheric composition. J. Geol. 97, 77-92.

Pan, Y. and Stauffer, M.R. (2000) Cerium anomaly and Th/U fractionation in the 1.85 Ga Flin Flon Paleosol: Clues from REE-and U-rich accessory minerals and implications for paleoatmospheric reconstruction. Am. Mineral. 85, 898-911.

Panahi, A., Young, G.M., and Rainbird, R.H. (2000) Behavior of major and trace elements (including REE) during Paleoproterozoic pedogenesis and diagenetic alteration of an Archean granite near Ville Marie, Québec, Canada. Geochim. Cosmochim. Acta 64, 2199-2220.

Peter, J.M. (2001) Ancient iron-rich metalliferous sediments (iron-formations): their genesis and use in the exploration for stratiform base metal sulphide deposits, with examples from the Bathurst Mining Camp. In Geochemistry of sediments and sedimentary rocks: secular evolutionary considerations to mineral deposit-forming environments (ed., D.R. Lentz), Geol. Assoc. Can., p1-38.

Phillips, G.N., Myers, R.E., and Palmer, J.A. (1987) Problems with the placer model for Witwatersrand gold. Geology 15, 1027-1030.

Phillips, G.N. and Myers, R.E. (1989) The Witwatersrand gold field, II: An origin for Witwatersrand gold during metamorphism and associated alteration.. Econ. Geol. Monogr. 6, 598-608. 32 Phillips, G.N. and Law, J.D.M. (1997) Hydrothermal origin for Witwatersrand gold. Soc. Econ. Geol. Newsletter 31.

Phillips, G.N. and Law, J.D.M. (2000) Witwatersrand gold field: geology, genesis and exploration. Rev. Econ. Geol. 13, 439-500.

Phillps, G.N., Law, J.D.M., and Myers, R.E. (2001) Is the redox state of the Archean atmosphere constrained? Soc. Econ. Geol. Newsletter 47.

Pinti, D.L. and Hashizume, K. (2001) 15N-depleted nitrogen in Early Archean kerogens: clues on ancient marine chemosynthetic-based ecosystem? A comment to Beaumont, V., Robert, F., 1999. Precam. Res. 105, 85-88.

Rasmussen, B. and Buick, R. (1999) Redox state of the Archean atmosphere: Evidence from detrital heavy minerals in ca. 3250-2750 Ma sandstones from the Pilbara Craton, Australia. Geology 27, 115-118.

Rasmussen, B. (2000) Filamentous microfossils in a 3,235-million-years-old volcanogenic massive sulphide deposit. Nature 405, 676-679.

Reymer, A. and Schubert, G. (1984) Phanerozoic addition rates to the continental crust and crustal growth. Tectonics 3, 63-77.

Robb, L.J., Davis, D.W., and Kamo, S.L. (1990) U-Pb ages on single detrital zircon grains from the Witwatersrand Basin, South Africa: Constraints on the age of sedimentation and on the evolution of granites adjacent to the basin. J. Geol. 98, 311-328.

Robb, L.J., Davis, D.W., Kamo, S.L., and Meyer, F.M. (1992) Ages of altered granites adjoining the Witwatersrand Basin with implications for the origin of gold and uranium. Nature 357, 667-680.

Robb, L.J. and Meyer, F.M. (1995) The Witwatersrand Basin, South Africa: Geological framework and mineralization processes. Geol. Rev. 10, 67-94.

Robertson, J.A. (1981) The Blind River uranium deposits: The ores and their setting. US Geol. Surv. Prof. Paper 1161-A-BB: U1-U23.

Robinson, A. and Spooner, T.C. (1982) Source of the detrital components of uraniferous conglomerates, Quirke ore zone, Elliot Lake Ontario, Canada. Nature 299, 622-624.

Robinson, A. and Spooner, T.C. (1984a) Postdepositional modification of uraninite-bearing quartz-pebble conglomerate from the Quirke ore zone, Elliot Lake, Ontario, Canada. Econ. Geol. 79, 297-321. 33 Robinson, A. and Spooner, T.C. (1984b) Can the Elliot Lake uranium-bearing quartz- pebble conglomerate be used to place limits on the oxygen content of the early Proterozoic atmosphere? J. Geol. 141, 221-228.

Roscoe, S.M. (1973) The Huronian Supergroup, a Paleoaphebian succession showing evidence of atmosphere evolution. Geol. Assoc. Can. Spec. Publ. 12, 31-38.

Rosing, M.T. (1999) 13C-depleted carbon microparticles in >3700-Ma sea-floor sedimentary rocks from West Greenland. Science 283, 674-676.

Rye, R., Kuo, P.H., and Holland, H.D. (1995) Atmospheric carbon dioxide concentrations before 2.2 billion years ago. Nature 378, 603-605.

Rye, R. and Holland, H.D. (1998) Paleosols and the evolution of atmospheric oxygen: a critical review. Am. J. Sci. 88, 621-672.

Rye, R and Holland, H.D. (2000) Life associated with a 2.76 Ga ephemeral pond?: Evidence from Mount Roe #2 paleosol. Geology 28, 483-486.

Schau, M. and Henderson, J.B. (1983) Archean chemical weathering at three localities on the . Precam. Res. 20, 189-224.

Schidlowski, M. (1982) Content and isotopic composition of reduced carbon in sediments. In Mineral Deposits and Evolution of the Biosphere (ed. by Holland, H.D. and Schidlowski, M.) Springer-Verlag, New York, 103-122.

Schidlowski, M., Hayes, J.M., and Kaplan, I.R. (1983) Isotopic inferences of ancient biochemistries: carbon, sulfur, hydrogen, and nitrogen. In Earth’s Earliest Biosphere: Its Origin and Evolution (ed., J.W. Schopf), Princeton Univ. Press, Princeton, 149-186.

Schidlowski, M. (1987) Application of stable carbon isotopes to early biochemical evolution on earth. Ann. Rev. Earth Planet. Sci. 15, 47-72.

Schidlowski, M. (1988). A 3,800-million-year isotopic record of life from carbon in sedimentary rocks. Nature 333, 313-318.

Schidlowski, M. and Aharon, P. (1992) Carbon Cycle and Carbon Isotope Record: Geochemical Impact of Life over 3.8 Ga of Earth History. In Early Organic Evolution: Implications for Mineral and Energy Resources (eds., M. Schidlowski, S. Golubic, M.M. Kimberley, D.M. McKirdy, and P.A. Trudinger), Springer- Verlag, New York, 147-175. 34 Schidlowski, M. (2001) Carbon isotopes as biogeochemical recorders of life over 3.8 Ga of Earth history: evolution of a concept. Precam. Res. 106, 117-134.

Schopf, J.W. and Packer, B.M. (1987) Early Archean (3.3 billion to 3.5 billion-years-old) microfossils from the Warrawoona Group, Western Australia. Science 237, 70-73.

Schopf, J.W. (1993) Microfossils of the early Archean Apex chert: new evidence of the antiquity of life. Science 260, 640-646.

Sclater, J.G., Jaupart, C., and Galson, D. (1980) The heat flow through oceanic and continental crust and the heat loss of the Earth. Rev. Geophys. Space Phys. 18, 269- 311.

Shegelski, R.J. (1980) Archean cratonization, emergence and red bed development, Lake Shebandowan area, Canada. Precam. Res. 12, 331-347.

Shen, Y., Buick, R. and Canfield, D. E. (2001) Isotopic evidence for microbial sulphate reduction in the early Archaean era. Nature 410, 77-81.

Sutton, S.J. and Maynard, J.B. (1992) Sediment- and basalt-hosted regoliths in the Huronian Supergroup: role of parent in middle Precambrian weathering profiles. Can. J. Earth Sci. 30, 60-76.

Sylvester, P.J., Campbell, I.H., and Bowyer, D.A. (1997) Niobium/uranium evidence for early formation of the continental crust. Science 275, 521-523.

Taylor, S.R. and McLennan, S.M. (1985) The Continental Crust: its Composition and Evolution. Blackwell Scientific Publications, 311p.

Taylor, S.R. and McLennan, S.M. (1995) The geochemical evolution of the continental crust. Rev. Geophys. 33, 241-265.

Towe, K.M. (1990) Aerobic respiration in the Archean? Nature 348, 54-56.

Towe, K.M. (1991) Aerobic carbon cycling and cerium oxidation: significance for Archean oxygen levels and deposition. Paleogeogr., Paleoclimatol., Paleoecol. (Global Planet. Change Sect.) 97, 113-123.

Towe, K.M. (1994) Earth’s early atmosphere: Constraints and opportunities for early evolution. In Early Life on Earth. Nobel Symposium No. 84. (ed., S. Bengston) Columbia University Press, New York, 36-47. 35 Ueno, Y., Isozaki, Y., Yurimoto, H., and Maruyama, S. (2001) Carbon isotopic signatures of individual Archean microfossils (?) from Western Australia. Int'l Geol. Rev. (in press).

Walker, J.C.G. (1977) Evolution of the Atmosphere. MacMillan, New York, 318 pp.

Wallis, M.K. and Wickramasinghe, N.C. (1995) Role of major terrestrial cratering events in dispersing life in the solar system. Earth Planet. Sci. Lett. 130, 69-73.

Walsh, M.M. (1992) Microfossils and possible microfossils from the early Archean Onverwacht Group, Barberton Mountain Land, South Africa. Precam. Res. 54, 271- 293.

Walsh, M.M. and Lowe, D.R. (1985) Filamentous microfossils from the 3500-Myr-old Onverwacht Group, Barberton Mountain Land, South Africa. Nature 314, 530-532.

Walsh, M.M. and Lowe, D.R. (1999) Modes of accumulation of carbonaceous matter in the Early Archean: A petrographic and geochemical study of the carbonaceous cherts of the Swaziland Supergroup. In Geologic Evolution of the Barberton Greenstone Belt, South Africa (D.R. Lowe and G.R, Byerly eds.). Geol. Soc. Amer. Spec. Paper 329. 115-132..

Walter, M.R. (1994) Stromatolites: The main geological source of information on the evolution of the early benthos. In Early Life on Earth, Nobel Symposium No. 84 (ed., S. Bengston), New York, Columbia, 270-286.

Watanabe, Y., Martini, J. E. J., and Ohmoto, H. (2000) Geochemical evidence for terrestrial ecosystem 2.6 billion years ago. Nature 408, 574-578.

Westall, F., de Wit, M.J., Dann, J., can der Gaast, S., de Ronde, C.E.J., and Gerneke, D. (2001) Early Archean fossil bacteria and biofilms in hydrothermally influenced, shallow water sediments, Barberton Greenstone Belt, South Africa. Precam. Res. 106, 91-112.

Whittet, D.C.B. (1997) Is extraterrestrial organic matter relevant to the origin of life on Earth? Origins Life Evol. Biosph. 27, 249-262.

Wilde, S.A., Valley, J.W., Peck, W.H., and Graham, C.M. (2001) Evidence from detrital zircons for the existence of continental crust and oceans on the Earth 4.4 Gyr ago. Nature 409, 175-178.

Yamaguchi, K.E., Ono, S., and Ohmoto, H. (1998) Diverse origins of pyrites in Paleoproterozoic uraniferous quartz-pebble conglomerate, Elliot Lake, Canada: 36 Evidence from laser-microprobe sulfur isotope analyses. Mineral. Mag. 62A, 1673- 1674.

Yamaguchi, K.E., Bau, M., and Ohmoto, H. (2000) Geochemistry of rare earth elements in Precambrian Banded Iron Formations: Are Ce anomalies real? In First Astrobiology Science Conference, NASA Ames Research Center, Moffett Field, CA. Apr. 2-5. Abstract, p 296.

Yamaguchi, K.E. and Ohmoto, H. (2002) Organic carbon, S, Mo, U, and V in Archean and Paleoproterozoic black shales. Astrobiology 3, 414. 37

Constant-O2 model (within ± 50% PAL) 100 %

[PAL] Atmospheric oxygen

10

1 Evolutionary model

Origin of life 0.1

~

10-13

4.6 4.0 2.2 1.9 Present Age [Ga]

Fig. 1-1. Two contrasting models for the evolution of atmospheric O2 level. The "Constant O2 model" proposes the emergence of cyanobacteria (oxygenic photosynthesizers) more than 3.8 Ga ago followed by a very rapid rise of pO2 to the present atmospheric level. Nearly constant level of pO2 has been maintained since then. In contrast, the "Evolutionary model" favors a dramatic rise of atmospheric pO2 level during 2.2-1.9 Ga (GOE: ; Holland, 1999). PAL stands for "present atmospheric level". 38

Geologic "evidence" for pO2 Evidence for high pO2 Evidence for low pO2

"Evidence" can not be used for pO2

Mt.Roe Hekpoort Paleosols

Shebandowan Red beds red bed

Witwatersrand Elliot Lake Uraniferous conglomerate

Hamersley Iron formations

S isotopes (δ34S)

S isotopes (∆33S)

N isotopes

Corg and S in black shales

Redox-sensitive metals in black shales

Sulfate evaporites

Biomarker Cyanobacteria in shales Stromatolite

Grypania Eukaryotes Biomarker

3.8 3.5 3.02.52.0 1.5 1.0 0.5 0 Age [Ga]

Fig. 1-2 Summary of controversial geological 'evidence' for pO2 in the early atmosphere. The vertical line at 2.2 and 1.85 Ga brackets the proposed periods of sudden O2 rise in the atmosphere (GOE: Great Oxidation event; Holland, 1984, 1994). See text for more information. Modified after Holland (1994) and Phillips et al. (2001). 39

Atmosphere

(4+) C O2 O2

Continents Oceans F (0) dis (4+) - OM (C H2O) HC O3 2- (6+) 2- Metal sulfide (S ) S O4 (3-) + N(5+)O - N H4 3 2+ Fe3+, Fe2+- oxide, Fe (6+) 2- silicate, sulfide, & Mo O4 carbonate: F Fhyd det U(6+)O 2+ (4+) 2 Mo S2 (4+) U O2 FMOR F ppt Freg Mid-oceanic uplift Sediments ridges

OM (C(0)H O) (4+) (4-) 2 C O2, C H4 2- Metal sulfide (S ) 2- Fburial Metal sulfide (S ) (3-) + N H4 (2+) (2-) Fe S2 subduction Fe3+, Fe2+- oxide, (4+) (2-) silicate, sulfide, & Mo S 2 carbonate: (4+) U O2 (4+) Mo S2 (4+) U O2

Mantle

Fig. 1-3. Geochemical cycles of C, S, N, Fe, Mo, and U in the continents- oceans-sediments system. The fluxes (F) between the resevoirs are also indicated. Fdet = detrital weathering flux, Fdis = dissolved weathering flux, Fppt = precipitation flux (biogenic, scavenging, etc.), Freg = regeneration flux, Fhyd = hydrothermal flux, and Fburial = burial flux. At a steady state, the input flux (Fin) to the oceans and the output flux (Fout) from the oceans are in balance: Fin = Fdis + Fhyd + Freg = Fout = Fppt + FMOR. 40

pO2 pE [atm] [mV] +15 1 + − 1 O2 (g) + H + e = H 2O 1 4 2 1 6 1 3 NO− + H+ + e− = N (g)+ H O 5 3 5 10 2 5 2 10-10

+10 1 + − 1 2+ MnO2 (s) + 2H + e = Mn (g) + H2O 10-20 2 2

Fe(OH) (s) + 3H+ + e− = Fe2+ + 3H O 10-30 3 2

+5 1 2- + − 1 - UO2 (CO3 )2 + H + e = UO2 (s) + HCO3 10-40 2 2 1 H VO- + 2H+ + e− = V O (s) + 2H O 10-50 2 4 2 2 4 2 0 1 3 1 3 UO (CO )4- + H+ + e− = UO (s) + HCO- 10-60 2 2 3 3 2 2 2 2 3 1 1 4 1 2 MoO2− + SO2− + H + + e− = MoS (s) + H O 18 4 9 4 3 18 2 3 2 -5 1 2- 9 + − 1 - 1 SO4 + H + e = HS + H2O 10-83.1 8 8 8 2 pH = 8 (pH2 = 1 [atm]) (seawater)

Fig. 1-4. Schematic diagram showing the characteristic pE values for redox reactions of trace metals and major electron acceptors during decomposition of organic matter in sediments. The corresponding pO2 levels are also shown. The pE - 2+ 2+ 2- values are for [NO3 ] = 50 µM, [Mn ] = 1 µM, [Fe ] = 5 µM, [SO4 ] = 28 mM, 2- 4- - [H2S] = 0.01 µM, [MoO4 ] = 106 nM, [UO2(CO3)3 ] = 13 nM, and [H2VO4 ] = 40 nM. The data used for calculation are from Crusius et al. (1996) and Piper (1994). Chapter 2

Geochemistry of Archean–Paleoproterozoic black shales: I. Diversity in the redox states of sedimentary environments inferred from C–S–Fe-P systematics and C–S stable isotopes

Abstract

3+ 2+ The relationships among organic C (Corg), S, Fe (both Fe and Fe ), and P and the stable

isotopic compositions of Corg and S in Archean–Paleoproterozoic carbonaceous shales and graywackes were examined to extract information on the paleoredox sedimentary environments, the geochemical cycles of C-S-Fe-P-O, and the evolution of the atmosphere, oceans, and biosphere. Geochemical and stable isotopic analyses were performed on 113 fresh drillcore samples from the 13 formations. The samples belong to the ~3.3 Ga Fig Tree, ~3.0 Ga West Rand, ~2.7 Ga Platberg, ~2.6 Ga Wolkberg, ~2.6 Ga Chuniespoort, and ~2.2 Ga Pretoria Groups in South Africa and the ~2.7 Ga Fortescue and ~2.6 Ga Hamersley

Groups in Australia. We found that (1) the Corg and S contents are variable within each formation, (2) the relationship between Corg and S is essentially similar to that of modern sediments and Phanerozoic sedimentary rocks (suggesting "normal marine" conditions), (3)

the Corg/P ratios show considerable variability of more than two orders of magnitude but fall within the two end members of the world average shale and the modern and ancient 42 anoxic sediments, (4) the P/Al ratios are essentially the same as the world average shales, (5) the Fe3+/Ti - Fe2+/Ti relationships of the normal shales (∑Fe/Ti < 20) indicate variable departures from those of their source rocks, (6) the degree of pyritization for Fe shows variability comparable to modern and Phanerozoic environments, (7) C isotopic compositions of organic matter are variable within each formation but their ranges are essentially the same as those of modern sediments, and (8) S isotopic compositions of sulfide are variable regardless of their ages.

These observations suggest that (1) sedimentary distribution of Corg and S are primarily controlled by redox / biological processes throughout geologic time, (2) aerobic recycling (oxidative degradation) of organic matter has been important throughout geologic time, (3) preferential regeneration of P in anoxic conditions has been important, (4) Fe was oxidized during Archean–Paleoproterozoic weathering and reduced during diagenesis, (5) sedimentary redox environments were already diverse at least since the Neoarchean, (6) oxygenic photosynthesis and methanogenesis have been important throughout geologic time, and (7) bacterial sulfate reduction has been important for sedimentary sulfide formation throughout geologic time. The above observations and suggestions further imply that the surface environment during the Archean–Paleoproterozoic was essentially the same as today. The atmosphere would have been oxic, the oceans would have been globally oxic but locally anoxic (e.g., mid-depth O2-minimum zones and anoxic basins), microbiological activities for photosynthesis and sulfate reduction would have been active, and P-mediated redox stabilization of the atmosphere and oceans would have been operating. The geochemical cycles of C-S-Fe-P-O would have been essentially the same as today, and they are already in full operation at least 3.25 Ga ago. This study has significant implications for the redox evolution of atmosphere and oceans and the evolution of biosphere in the early Earth. 43

2-1. Introduction

2-1-1. Evolution of the atmosphere, biosphere, and oceans

The timing and mechanisms for the rise of atmospheric O2 have remained poorly constrained and been intensely debated for decades (see Ohmoto, 1997; Holland, 1999; Phillips et al., 2001 for a summary of recent discussions). One school postulates a dramatic

-13 rise of atmospheric O2 from a very low concentration, possibly as low as 10 PAL (present atmospheric level), to 0.1 PAL at 2.2-2.0 Ga (GOE: Great Oxidation Event; see Holland, 1984, 1994, 1999; Kasting et al., 1992; Kasting, 1993), while another school postulates essentially constant O2 level at least since 3.8 Ga (e.g., Ohmoto, 1997). The rise

of atmospheric O2 is fundamentally important for the evolution of the biosphere, because it is related to the questions as to when the oxygenic photosynthesis began and became sufficiently important to transform the prebiotic reducing atmosphere to an oxidizing one

and when eukaryotes evolved. The rise of atmospheric O2 is also linked to the chemical evolution of the oceans through the direct interaction across an air-sea interface and the global/local ocean circulations involving sinking of atmosphere-influenced surface waters to deep oceans. The oxygenation of deep oceans almost certainly brought fundamental impacts on the marine microbial biosphere in the early Precambrian.

2-1-2. Previous approaches and associated problems / limitations To constrain the redox state of an ancient atmosphere, geological records such as paleosols, uranium ores, BIFs (banded iron-formations), shales, and red beds have been used to extract useful but indirect information on that subject (summary in Holland, 1994 and discussion in Ohmoto, 1997; Holland, 1999; Phillips et al., 2001; Ohmoto et al., in prep. See also Fig. 1-2). From those studies, however, contrasting conclusions are reached (see also Fig. 1-1 and 1-2) based on essentially the same sample sets and analytical data. 44

This is probably because of intrinsic ambiguity in such geological records due to post- depositional alteration and in geochemical data extracted from small number of samples which generally lacked in systematic and multidisciplinary approaches for sampling and analyses. Such examples evidenced in many previous studies are already given in Watanabe et al. (1997) and not repeated here. Systematic studies of the inorganic and organic geochemistry on Precambrian marine shales have great potential for elucidating the oceanic paleoredox environments and biological activity (e.g., Watanabe et al., 1997). However, in spite of their importance, the number of such studies is very limited and the collection of such data should be a high priority, as Canfield and Raiswell (1999) claim.

2-1-3. Geochemical cycles of C, S, Fe, P, and O: New approaches and focuses of this study In order to constrain the paleoredox conditions of the atmosphere and oceans and the evolution of biosphere in the early Precambrian, this study takes unparalleled advantages of (1) systematic collection of modern weathering-free, low metamorphic-grade drillcores samples of Archean–Paleoproterozoic siliceous and carbonaceous sedimentary rocks (93 black shales, 12 graywackes, and 8 red shales) that cover large space (South Africa and Australia) and time (13 formations ranging in age from ~3.3 to ~2.2 Ga), and of (2) systematic and multidisciplinary approaches for the analyses combining mineralogy, major, minor, trace, and rare earth element geochemistry, and C, S, and N elemental and stable isotope geochemistry. The focus of this study is placed on the understanding of Archean–Paleoproterozoic C-S-Fe-P-O geochemical cycles based on C-S-Fe-P geochemistry and C-S stable isotope geochemistry using a part of the data obtained. This is because the sedimentary distribution of organic C (Corg), pyrite-S (Spy), Fe, and P and the stable isotopic compositions of Corg and Spy in shales are influenced by non-biological redox processes, biological processes, 45

and biologically mediated redox processes (e.g., weathering of Fe, biological production of organic matter (OM), its aerobic/anaerobic decomposition, redox cycling of P mediated by Fe, sulfate reduction by bacteria, and formation of authigenic Fe-sulfide minerals utilizing bacteriogenic sulfide and reactive Fe). The geochemical cycles of those elements exert an significant influence on the overall redox evolution of the atmosphere-oceans-crust system (e.g., Garrels and Perry, 1974; Holland, 1978; Garrels and Lerman, 1984; Berner et al., 1983; Lasaga et al., 1985; Kump and Garrels, 1986; Lasaga and Ohmoto, 2002a, b). This study examines (1) the Fe redox chemistry during weathering of source rocks

and post-depositional processes from the Fe3+ and Fe2+ contents and the Fe3+/∑Fe ratios of the shales and their source rocks, (2) the extent of primary production, the OM burial

flux, and the net O2 production flux from the Corg contents of the shales, (3) the extent of

authigenic sulfide formation and the oceanic redox environments from the Spy contents of the shales, (4) the C-S-Fe systematics from the relationship between Corg and Spy contents, the S/Corg ratios, and the DOP (degree of pyritization) values in comparison with those of modern sediments and recent sedimentary rocks, (5) the P geochemical cycles from the

Corg/P ratios and P contents of the shales, (6) the diverse microbiological activity of different types of organisms recorded in the stable isotopic compositions of Corg and Spy, and (7) the redox structure of the oceans and the oceanic sulfate concentration. This study provides an extensive geochemical and isotopic data set produced by a variety of analytical methods from systematically collected samples of varying ages.

2-2. Geological settings and samples

We have analyzed 93 laminated / nonlaminated black shale samples, 12 graywacke samples, and 8 laminated red shale samples (in 13 formations) from modern weathering- 46

free, low metamorphic grade drillcores (113 samples in total). South African samples are from the ~3.3 Ga Fig Tree, the ~3.0 Ga West Rand, the ~2.7 Ga Platberg, the ~2.6 Ga Wolkberg, the ~2.6 Ga Chuniespoort, and the ~2.2 Ga Pretoria Groups. Australian samples are from the ~2.7 Ga Fortescue and the ~2.6 Ga Hamersley Groups. Information on stratigraphy, ages, and geological settings of the samples are presented in this section and in more detail in Appendices A and B.

2-2-1. Swaziland Supergroup The Swaziland Supergroup is exposed in the 3.4 to 3.1 Ga Barberton Greenstone Belt located in the eastern part of the (Fig. 2-1). The Swaziland Supergroup consists of three groups: from lower to upper, the 3.48 - 3.45 Ga Onverwacht, 3.33 - 3.23 Ga Fig Tree, and 3.22 - 3.10 Ga Moodies Groups (Armstrong et al., 1990; Kröner et al., 1991; Kamo and Davis, 1994). The Onverwacht Group is predominantly composed of mafic and ultramafic rocks with minor sedimentary and felsic volcanic units. The Fig Tree Group is composed of a succession of graywacke, shale, chert, dacitic flow, and fragmented . The lowermost Sheba Formation of the Fig Tree Group is composed mainly of coarse, immature turbiditic and thin interbedded units of siltstone and shale (Lowe and Byerly, 1999). The Moodies Group is composed of feldspathic and quartzose sandstone, chert- clast, and siltstone. Depositional settings of the sedimentary rocks of the Fig Tree and Moodies Groups are a foreland or foredeep basin (Jackson et al., 1987; de Wit et al., 1992; Heubeck and Lowe, 1994), or in evolving back-arc or passive continental margins from early to and shelf-rise environments (Windley, 1995). The black shale samples of this study are partly from the 3.25 Ga Sheba Formation of the Fig Tree Group (Fig. 2-2). Seventeen shale samples were collected from a ~90 m long drillcore (PU1308) at the Agnes Mine (Wagener, 1986) in the Barberton Greenstone 47

Belt (Fig. 2-2) and chemically analyzed for this study. Twelve graywacke samples were also collected from the 3.25 Sheba Formation, using drillcore MRE10 at the Sheba Mine (Wagener and Wiegand, 1986) in the same district, and chemically analyzed for this study. The sedimentary textures are well preserved except in extremely sheared rocks. These samples of this study were subjected to very minor shearing and only relatively low-grade regional metamorphism (greenschist facies) (Viljoen and Viljoen, 1969).

2-2-2. Witwatersrand Supergroup The Witwatersrand basin extends about 200 km SW from the Barberton area with an areal extent of ~80,000 km2 (Fig. 2-1). The Witwatersrand Supergroup, the basin infill, conformably overlies volcanic rocks of the Dominion Group and uncomformably overlies the Archean basement rock (greenstone-granitoid complexes) of the Kaapvaal Craton. The Witwatersrand Supergroup was deposited within the period 2,970 to 2,714 Ma (Robb et al., 1991). Recent reviews on the geology of the Witwatersrand Basin have been provided by Coward et al. (1995) and Robb and Meyer (1995). The Witwatersrand Supergroup is dominated by siliclastic deposits that largely divided into a mixed argillaceous-arenaceous lower phase (West Rand Group: 2,914±8 Ma; Armstrong et al., 1991) and a predominantly arenaceous upper phase (Central Rand Group). The arenaceous rocks include several horizons of predominantly quartz-pebble conglomerate units that host Au and U mineralization ('reefs'). Shallow marine or tidal marine with minor alluvial depositional environments have been suggested for the West Rand Group (Tankard et al., 1982). The black shale samples of this study are partly from the 2.96 Ga Parktown Formation of the Hospital Hill Subgroup of the West Rand Group (Fig. 2-2). Fifteen shale samples were collected from a ~120 m section of a >500 m long drillcore (DRH13) recovered in the NW margin of the Witwatersrand Basin (Fig. 2-1), and were chemically 48 analyzed for this study. These samples of this study were not subjected to deformation and only suffering from relatively low-grade regional metamorphism (greenschist facies: e.g., Pretorius, 1981; Phillps et al., 1989; Wronkiewicz and Condie, 1990).

2-2-3. Ventersdorp Supergroup The Ventersdorp Supergroup (e.g., Button, 1981a), dated at 2,709±4 and 2,714±8 Ma (Armstrong et al., 1991), conformably overlies the Witwatersrand Supergroup and the Dominion Group of the Kaapvaal Craton. The Ventersdorp Supergroup is mainly composed of continental flood basalt with minor sediments (Winter, 1976), and divided into, from lower to upper, the , Platberg, and Pniel Groups. The Klipriviersberg and Platberg Groups are largely composed of volcanics, and the Pniel Group is composed of clastic sediments and volcanics. An alluvial plain depositional setting has been suggested for the sediments of the Pniel Group (Tankard et al., 1982). The lithology and sedimentation of both the Witwatersrand and Ventersdorp Supergroups were strongly influenced by tectonic processes, such as subduction, formation of island arcs and rifting (Burke et al., 1985, 1986). The black shale samples of this study are partly from the ~2.6 Ga Rietgat Formation of the Platberg Group (Fig. 2-2). Seven shale samples of this formation were collected from a ~800 m section of the ~4.4 km long drillcore (MSF6) recovered in the NW margin of the Witwatersrand Basin (Fig. 2-1), and chemically analyzed for this study. The drillcore MSF6 covers from lower to upper, the Florida , Livingstone Conglomerate, Kimberley Conglomerate, Elsburg Quartzite, Kameelsdoorn, Goedgenoeg, and Rietgat Formations from the Ventersdorp Supergroup, and the Black Reef Quartzite, Oak Tree, Monte Christo, Lyttelton , Eccles, Rooihoget , Timeball Hill, and lower Hekpoort Andesite Formations from the Transvaal Supergroup (see section below). These samples of this study were not subjected to deformation and only suffering from relatively low-grade 49

regional metamorphism (greenschist facies: e.g., Pretorius, 1981; Phillps et al., 1989; Wronkiewicz and Condie, 1990).

2-2-4. Transvaal Supergroup The Transvaal Supergroup (e.g., Button, 1981b), mainly composed of sedimentary rocks, uncomformably overlies the Ventersdorp Supergroup. The Transvaal Group is divided into, from lower to upper, the Wolkberg, Chuniespoort, and Pretoria Groups (Fig. 2-2). The Wolkberg Group (2,630 - 2,560 Ma: Jahn et al., 1990; Armstrong et al., 1991) is composed of fluvial and shallow marine feldspathic quartzite, arkoses, siltstones, pelites, carbonate, and conglomerates, as well as subaerially to subaqueously extruded basalt, and probably deposited in a basin (Tyler, 1978; Söhnge, 1986; Button, 1986). The Chuniespoort Group (2,557 ± 49 Ma: Jahn et al., 1990) is mostly composed of stromatolitic carbonates and iron-formations, and probably deposited in shallow marine or intertidal environments from the presence of stromatolitic carbonates (Truswell and Eriksson, 1975; Eriksson, 1983; Button, 1986; Söhnge, 1986; Cledenin, 1988). The Pretoria Group (~2.3 - 2.0 Ga) is mainly composed of and shales with minor volcanics and carbonates, and probably deposited in shallow marine cratonic environments (Button, 1981b; Tankard et al., 1982). It has been suggested that the Kaapvaal Craton was tectonically stable during the depositions of the Chuniespoort Group (Söhnge, 1986) and the lower section of the Pretoria Group (Schreiber et al., 1992). The black shale samples of this study are partly from the 2.64 Ga Black Reef Formation of the Wolkberg Group, the 2.56 Ga Oak Tree Formation of the Chuniespoort Group, and the 2.22 Ga Timeball Hill Formation of the Pretoria Group (Fig. 2-2). Four samples of the Black Reef Formation were collected from a 110 m section of a >2500 m long drillcore (JPBR). Three samples of the Oak Tree Formation were collected from a 200 50 m section of the ~4.4 km long drillcore (MSF6). Four samples of the Timeball Hill Formation were collected from a 350 m section of the ~4.4 km long drillcore (MSF6). Both of the drillcores (JPBR and MSF6) were recovered in the NW margin of the Witwatersrand Basin (Fig. 2-1). The collected samples were chemically analyzed for this study. Another set of 12 samples of the 2.22 Ga Timeball Hill Formation were collected from a 100 m section of the ~400 m long drillcore (PTB3) in the Pilgrim's Rest -Sabie region of the eastern Transvaal (Fig. 2-1), and chemically analyzed for this study. These samples of this study were not subjected to deformation and only suffering from relatively low-grade regional metamorphism (greenschist facies: e.g., Pretorius, 1981; Phillps et al., 1989; Wronkiewicz and Condie, 1990).

2-2-5. Mt. Bruce Supergroup Archean and Paleoproterozoic strata are also exceptionally well-preserved and exposed on the Pilbara Craton in the Pilbara-Hamersley regions of NW Australia (Fig. 2- 3). In the southern part of the Pilbara Craton, the Neoarchean Mt. Bruce Supergroup uncomformably overlies the cratonic basement (e.g., Hickman, 1983). The Mt. Bruce Supergroup is the infill of the Hamersley Basin that extends over an area of 1,000,000 km2 (Fig. 2-3). The Mt. Bruce Supergroup is divided into, from lower to upper, the Fortescue, Hamersley, and Turee Creek Groups. The Fortescue Group (e.g., Thorne and Blake, 1990) is mainly composed of mafic volcanics (basalt and ) and volcaniclastic rocks, subordinate acid volcanics (), conglomerate, sandstone, shale, and carbonates (Fig. 2-4). The Hamersley Group (e.g., Thorne, 1990) comformably lies on the Fortescue Group. The Hamersley Group is well known to geoscientists because of its extensive iron-formations. Other than iron-formations, the Hamersley Group contains dolomite, carbonaceous shale, and rhyolite. The Hamersley Group is divided into, from lower to upper, the Marra Mamba Iron, Wittenoom Dolomite, Mt. Sylvia, Mt. McRae, 51

Brockman Iron, Weeli Wolli, Woongarra Volcanics, and Boolgeeda Iron Formations (Fig. 2-4). The Turee Creek Group comformably sits on top of the Hamersley Group, and is largely composed of graywacke, sandstone, siltstone, quartzite, and dolomite. The black shale samples used in this study are partly from the 2.72 Ga Pillingini Tuff Formation of the middle Fortescue Group, the 2.69 Ga Jeerinah Formation of the uppermost Fortescue Group and the >2.60 Ga Marra Mamba Iron and 2.60 Ga Wittenoom Dolomite Formations of the Hamersley Group (Fig. 2-4). Three samples from the Pillingini Tuff Formation, 6 samples from the Jeerinah Formation, 4 samples from the Marra Mamba Iron Formation, and 3 samples from the Wittenoom Dolomite Formation were collected from a 200 m section, a 50 m section, a 70 m section, and a 5 m section, respectively, of a >1500 m long drillcore (WRL1) recovered near the town of Wittenoom, Western Australia (Fig. 2-3). These samples were chemically analyzed for this study. The drillcore WRL1 has been previously used to successfully extract Archean cyanobacterial biomarkers from the Jeerinah Formation (Brocks et al., 1999), and for carbonate-carbon and oxygen isotope study by Lindsay and Braiser (2002). The Lewin Shale Formation and Carawine Dolomite Formation occur in the eastern region of the Hamersley Basin, which have been thought to be stratigraphically equivalent to the Jeerinah and Wittenoom Dolomite, respectively, in the central-western region of the Hamersley Basin (Fig. 2-3; Simonson et al., 1993, 1998; Simonson and Hassler, 1997; Woodhead et al., 1998). The black shale samples of this study are partly from these two Formations. Eight samples from the Lewin Shale and 6 samples from the Carawine Dolomite Formation were collected from a 500 m long drillcore (RHDH2A), recovered near the Ripon Hills in the eastern Hamersley region (Fig. 2-3). These samples were chemically analyzed for this study. The drillcore RHDH2A has been previously used for sulfur isotope study (Bottomley et al., 1992). The Australian samples of this study were not subjected to 52 deformation and only suffering from relatively low-grade regional metamorphism (greenschist facies).

2-3. Analytical methods

Detailed analytical methods are presented in the Appendix C. In this section, a brief summary of the analytical method used in this study is presented.

2-3-1. Microscopy Thin sections from selected drillcore samples were prepared by a standard technique. Both transmitted light and reflected light were used for observation under the petrographic microscopes (Nikon Optiphot-POL) at The Pennsylvania State University.

2-3-2. Pulverization Drillcore samples were rinsed by large volumes of water to remove contaminants such as encrusting dirt. Samples were crushed into coarse fragments (~5 - ~10 mm) using a jaw-crusher. The coarse fraction of the samples were rinsed with water for several times to remove attached dusts, and washed again with distilled deionized water (18 MΩ) in an ultrasonic cleaner for ~10 minutes (repeated when necessary). Then they were washed with acetone in an ultrasonic bath for 5 minutes to remove any organic contaminants from the samples. Cleaned chip samples were dried at ~60 ˚C until dryness. Parts of cleaned rock chips were pulverized until -150 mesh size (106 µm) using agate ballmill at the Pennsylvania State University or ACTLAB, Inc. Between each sample, cleaner was used to avoid cross-contamination between samples. To track any possible contamination 53 during the sample processing, control samples were processed together. No noticeable inorganic/organic contamination was identified.

2-3-3. Major elements (Fe, Al, Ti, and P) analyses A portion of powdered samples was digested by alkali fusion to measure the total Fe, Ti, and P contents by ICP-AES (Inductively Coupled Plasma - Atomic Emission Spectroscopy) at Penn State (Leeman Labs PS 3000UV) and ACTLAB, Inc. Good precision (± 0.5-1 %) was achieved with ICP-AES for simultaneous multi-elements analyses. XRF (X-Ray Fluorescence: Rigaku 3270) method was also used for some samples to measure total Fe, Ti, and P contents of the samples at the Ocean Research Institute (ORI), University of Tokyo. The glass bead samples were made by a standard technique. The calibration were made using in-house standard materials that had been calibrated against the international SRM (standard reference material). Procedure blank samples and procedure standard samples (SCo1, MAG1, BCR2, and W2: United States Geological Survey SRM; JB3 and JG1A: Geological Survey of Japan SRM) were also analyzed to track possible contamination during experimental procedures. Ferrous iron contents were determined by a standard wet chemical technique (titration) at Penn State and ACTLAB, Inc. The results at Penn State and ACTLAB, Inc. agreed very well for both of the samples of this study and the SRMs. The reproducibilities were mostly better than 5 %. Some samples with relatively low reproducibility (better than

10 %) were repeated for ≥4 times and the average values were adapted. Ferric iron contents were calculated as the difference between total Fe and Fe2+ contents.

2-3-4. Elemental analyses of organic C, carbonate C, and sulfide S A portion of powdered samples was treated with 2 N HCl overnight (or until bubbling ceased) at room temperature to remove carbonate minerals. Another portion of 54

powdered samples was treated first with organic solvent and then with HCl and HF in teflon bottles at elevated temperature (~80 ˚C) to extract kerogen (see detailed method in Watanabe et al. 1997). An elemental analyzer (EA) at Penn State (CE Instruments NA2500) was used

to measure C and S contents of bulk and decarbonated samples. Carbonate carbon (Ccarb)

contents were calculated as a difference between Ctotal and Corg contents. The detection limit was ~0.01 wt.%, and the reproducibilities were better than ± 0.2 wt.% for C and S analyses.

2-3-5. Stable isotope analyses of organic C and pyrite S

The CO2 and SO2 gases generated from samples during analysis with the EA were collected, cryogenically purified, and analyzed for C and S isotopic ratios by an off-line method using mass spectrometers (Finnigan MAT 252 for C and VG PRISM II for S) at

Penn State. For Corg isotopes, some samples were analyzed by Dr. Simon Poulson using an on-line method with an EA and a mass spectrometer at the University of Nevada-Reno. The isotopic compositions of Corg and S are expressed with standard δ (delta) notations in per mil (‰) relative to Vienna Pee Dee Belemnite (VPDB) for C and air Cañon Diablo Troilite (CDT) for S:

n 3 δ X = (Rsample/Rstandard -1) • 10 ...... (2-1)

where n = 13 or 34, X = C or S, and R = 13C/12C or 34S/32S, respectively. The

13 34 reproducibility of δ Corg and δ S values was better than ±0.1 ‰ and ±0.2 ‰, respectively. 55

2-3-6. Degree of pyritization (DOP) analyses Degree of pyritization (DOP) is a measure of the degree of pyritization of Fe, and defined as follows (e.g., Berner, 1970; Raiswell and Berner, 1985, 1986):

DOP = [Pyrite-bound Fe] / [Reactive Fe]...... (2-2) [Reactive Fe] = [Pyrite-bound Fe] + [HCl-extractable Fe] ...... (2-3)

In this study, pyrite-bound Fe was determined by calculating the stoichiometric value assuming that total S approximates pyritic S, since no sulfate minerals were identified in the samples of this study. Among the several methods to determine acid-soluble Fe contents (e.g., Berner, 1970; Raiswell et al., 1988, 1994; Leventhal and Taylor, 1990), the 24 h - 1 N HCl method was utilized in this study (Leventhal and Taylor, 1990). This is because laboratory handling of this method is easier and safer compared to that of the other methods. Weighed (10-20 mg) powdered samples were put in test tubes and 50 ml of 1 N HCl was added to each of the samples in the test tubes. After 24 hours, the samples were centrifuged and concentrations of dissolved Fe in the supernatant were measured by the atomic adsorption (flame emission spectrometry) method using Perkin Elmer 703 spectrophotometer at the Material Characterization Laboratory of Penn State. When necessary, the supernatant was further diluted with DDW to adjust Fe concentration suitable for AA analysis. Typically, Fe concentrations were diluted to reach ppm level. Measurements by AA were repeated, and the average values of Fe concentration were used to calculate leached Fe contents by HCl from the samples. 56

2-4. Results

2-4-1. Total Fe, ferric Fe, and ferrous Fe contents Total Fe, Fe3+, and Fe2+ contents considerably differ among the Archean– Paleoproterozoic samples of this study (Table 2-1). The 3.25 Ga Sheba Formation (shales)

has an average ∑Fe2O3 contents of 33.5 ± 12.8 (1σ) wt.%, while the coeval graywackes have a lower average ∑Fe2O3 contents of 8.4 ± 1.7 wt.%. The elevated Fe -contents in the (2+) 3.25 Ga shales are due to, as previously mentioned, abundance of siderite (Fe CO3). The samples of the 2.96 Ga Parktown Formation and the >2.60 Ga Marra Mamba Iron

Formation also have elevated ∑Fe2O3 contents (an average of 22.3 ± 9.7 wt.% and 32.6 ± 2.6 wt.%, respectively) compared to average shale (6.1 wt.%: PAAS; Taylor and McLennan, 1985). The former is attributed to abundance of Fe2+-bearing chlorite (15.0 ± 4.7 wt.% for FeO contents) and the latter is attributed to abundance of siderite, and magnetite, as confirmed by petrographic observation, XRD analysis, and magnet for magnetite. The 2.60 Ga Wittenoom Dolomite Formation and one samples from the contemporaneous Carawine Dolomite Formation have somewhat elevated FeO contents (9.4 to 12.5 wt.%), which is attributed to abundance of siderite as confirmed by the same method mentioned above. The red shales of the ~2.2 Ga Mapedi Formation are characterized by overwhelming abundance of Fe3+ (11.6 ± 4.5 wt.%; mostly due to hematite) with minor Fe2+ contents (1.5 ± 1.3 wt.%). The 2.64 Ga Black Reef Formation and the 2.56 Oak Tree Formation have depleted

∑Fe2O3 contents (3.1 ± 2.4 wt.% and 2.0 ± 0.6 wt.%, respectively) compared to that of average shale (6.1 wt.%).

Other samples show the ∑Fe2O3 contents more or less similar to that of the average shale (6.1 wt.%): the 2.72 Ga Pillingini Tuff Formation (10.9 ± 5.2 wt.%), the 2.71 Ga Rietgat Formation (7.2 ± 1.6 wt.%), the 2.69 Ga Lewin Shale (6.3 ± 2.4 wt.%) and Jeerinah (6.2 ± 3.6 wt.%) Formations, the 2.60 Ga Carawine Dolomite Formation (5.6 ± 1.6 wt.%, 57

except for one samples mentioned above), and the 2.22 Ga Timeball Hill Formation (8.3 ± 2.3 wt.% for Eastern Transvaal samples and 7.8 ± 1.3 wt.% for Central Transvaal samples).

2-4-2. ∑Fe/Ti, Fe3+/Ti, Fe2+/TI, and Fe3+/∑Fe ratios As indicators of the quantity of detrital materials, Al and/or Ti contents of sediments / shales have been commonly used in previous studies because Al and Ti are the least mobile major elements during weathering due to their extremely low solubilities in low temperature fluids in the Earth's surface conditions (e.g., Stumm and Morgan, 1996; Ayers and Watson, 1993; Ziemniak et al., 1993). Titanium contents are used in this study to normalize Fe contents to estimate the relative enrichment or depletion of Fe to clastic component (Ti) in shales (e.g., Ohmoto, 1996; Hayashi et al., 1997). The ∑Fe/Ti, Fe3+/Ti, Fe2+/TI, and Fe3+/∑Fe ratios are summarized in Table 2-2. The relationships between Fe3+/Ti and Fe2+/Ti ratios for the samples of this study are shown in Fig. 2-5 (all the samples) and Fig. 2-6-a through -h (each formation). As shown in Fig. 2-5 and Fig. 2-6, the Fe3+/Ti and Fe2+/Ti ratios are quite variable, ranging in ~3 orders of magnitude (~0.1 < Fe3+/Ti < ~100, ~1 < Fe2+/Ti < ~1000). The

∑Fe/Ti ratio of the average shale (PAAS) is 6.5 (Taylor and McLennan, 1985). We assume that ∑Fe/Ti = 20, approximately a factor of 3 larger than the ∑Fe/Ti for PAAS, represent the upper limit of normal shales composition. This upper limit of ∑Fe/Ti is shown as a curve in Fig. 2-5 and as a line in Fig. 2-6. The extreme enrichments of Fe relative to Ti

(∑Fe/Ti » 20) are observed mainly in the 3.25 Ga Sheba Formation (shales) (∑Fe/Ti = ~1000; Fig. 2-6-a), the 2.96 Ga Parktown Formation (∑Fe/Ti = ~300; Fig. 2-6-c), and the >2.60 Ga Marra Mamba Iron Formation (∑Fe/Ti = ~200; Fig. 2-6-f). These extreme values are, as previously mentioned, due to elevated abundance of siderite, magnetite, chlorite, and

hematite. About two thirds of the samples have ∑Fe/Ti ratios lower than 20 (Fig. 2-6-b, -d, - 58

e, -g, and -h). The samples of the 2.56 Ga Oak Tree Formation have very low ∑Fe/Ti ratios (< 5; Fig. 2-6-g).

Figure 2-7 shows histograms of the Fe3+/∑Fe ratios for the samples of this study. Large variations are often observed. For example, the Fe3+/∑Fe ratios ranges from 0.09 to 0.56 for the 2.60 Carawine Dolomite Formation, from 0.21 to 0.51 for the >2.60 Ga Marra Mamba Iron Formation, from 0.21 to 0.83 for the 2.69 Jeerinah and Lewin Shale Formation, and from 0.05 to 0.53 for the 2.96 Ga Parktown Formation). The range of the

Fe3+/∑Fe ratios are different between the shales and graywackes of the 3.25 Ga Sheba Formation (0.01 to 0.27 and 0.04 to 0.11, respectively).

2-4-3. Pyrite-bound Fe, HCl-soluble Fe, and reactive Fe contents

Pyrite-bound Fe (Fepy), HCl-soluble Fe (FeHCl), and reactive Fe (FeR: sum of the pyrite-bound Fe and the HCl-soluble Fe) contents for the samples of this study are

summarized in Table 2-2. The relationships among the Fepy, FeHCl, and FeR are presented

in Fig. 2-8. The characteristics of the contents of the Fepy are essentially the same as those of the Spy contents described later, and not given here. Because of an apparent lack of sulfide minerals as indicated by Spy = ~0 wt.% for the red shales of the ~2.2 Ga Mapedi

Formation, the FeHCl and FeR were not measured.

The majority of the samples have the FeHCl contents below 1 wt.% with the large variabilities within each formation (Table 2-2). The samples of the 3.25 Ga Sheba

Formation (shale) is exceptional in that they have high FeHCl contents (2.25 ± 0.90 wt.%) due to the abundance of siderite. Siderite is not completely dissolved by the treatment used in this study (1N HCl for 24 hours), and it usually takes a longer time (days to weeks) to achieve complete dissolution of siderite with the concentrated HCl . A partial dissolution of

siderite contributed to the elevated abundance of FeHCl for the 3.25 Ga Sheba Formation

(shale). The coeval graywackes have the FeHCl contents of 0.33 ± 0.11 wt.%. The contents 59

of FeHCl obtained in this study are in contrast with the samples from the Transvaal Supergroup used by Strauss and Beukes (1996) in that only one-third of their samples have

FeHCl contents lower than 1 wt.%.

The FeR (= FeHCl + Fepy) contents of the 3.25 Ga Sheba Formation (shale) and the 2.69 Ga Jeerinah Formation (2.55 ± 1.09 wt.% and 2.69 ± 0.82 wt.%, respectively) are higher than those of the other samples of this study. This is because the former samples are

rich in FeHCl (siderite) as mentioned before and the latter samples are rich in Spy (thus

Fepy). The 2.69 Ga Lewin Shale Formation, the >2.60 Ga Marra Mamba Iron Formation, and the 2.60 Ga Wittenoom Dolomite Formation show somewhat elevated FeR contents (~1 wt.%), while the other samples of this study except for the 3.25 Ga Sheba Formation (shale)

and the 2.69 Ga Jeerinah Formation have FeR contents lower than 1 wt.%. The 2.22 Ga

Timeball Hill Formation have the lowest average FeR contents of 0.20 - 0.25 wt.%.

2-4-4. Degree of pyritization (DOP) values

The DOP values (Fepy/FeR or Fepy / [FeHCl + Fepy]) of the samples of the present study are summarized in Table 2-2 and also shown in the histograms in Fig. 2-9 with previously published DOP values of the Neoarchean–Paleoproterozoic sediments from the Transvaal Supergroup (Strauss and Beukes, 1996). Figure 2-9 also shows the DOP values

of this study with the FeHCl and Fepy contents. The DOP values of the samples of this study are variable within each formation (Fig. 2-9). The DOP values range, by definition, from 0 to 1, however, more than 0.5 differences are observed in the 3.25 Ga Sheba Formation (shale; 0.0 - 0.6) (Fig. 2-9-s), the 2.96 Ga Parktown Formation (0.0 - 0.67; Fig. 2-9-q), the 2.71 Ga Rietgat Formation (0.0 - 0.67; Fig. 2-9-o), the 2.69 Ga Lewin Shale Formation (0.0 - 0.67; Fig. 2-9-n), the >2.60 Ga Marra Mamba Iron Formation (0.0 - 0.67; Fig. 2-9-j), the 2.60 Ga Carawine Dolomite Formations (0.0 - 0.64; Fig. 2-9-i), the 2.60 Ga Wittenoom Dolomite Formation (0.0 - 60

0.67; Fig. 2-9-h), the 2.56 Ga Oak Tree Formation (0.0 - 0.67; Fig. 2-9-g), and the 2.22 Ga Timeball Hill Formation (0.0 - 0.67; Fig. 2-9-b). Such variability (> 0.5) is also observed in the previously published DOP values for Neoarchean–Paleoproterozoic samples (largely shales) from the Transvaal Supergroup (Fig. 2-9-a, d, k; Strauss and Beukes, 1996). The variable DOP values (0 ~ 1 in a depth profile) are quite commonly observed in modern sediments, Phanerozoic shales, and Proterozoic shales (e.g., Raiswell and Berner, 1985, 1986; Fisher and Hudson, 1987; Donnelly et al., 1988; Raiswell et al., 1988; Dean and Arthur, 1989; Raiswell and Al-Biatty, 1989; Jackson and Raiswell, 1991; Pearson et al., 1996; Murphy et al., 1999).

2-4-5. Aluminum and titanium contents

The Al2O3 and TiO2 contents of the studied samples are summarized in Table 2-1.

The Al2O3 contents are also shown in the histograms in Fig. 2-10. The Al contents are used as a measure of clastic component to normalize the Corg, P, and S contents (Fig. 2-18, 2-20, and 2-23, respectively).

Compared to the typical Al2O3 content of 17.8 wt.% (PAAS: Post-Archean Australian Average Shales; Taylor and McLennan, 1985), the shales of this study exhibit generally low and variable Al2O3 contents (the average Al2O3 contents range from ). Shales

of the 3.25 Ga Sheba Formation have the Al2O3 contents (17.5 ~ 20.8 wt.%, n = 4: Fig. 2- 10-n) comparable to PAAS; however, siderite-rich and silica-rich shales of the same

Formation have lower Al2O3 contents (1.5 ~ 10 wt.%; Fig. 2-10-n). The low Al2O3

contents are mainly because of the elevated Fe contents for siderite-rich shales (∑Fe2O3

contents 30 ~ 50 wt.%; Table 2-2) and the elevated SiO2 contents (~74 wt.%) for silica-rich shales (Fig. 2-10-n). Normal shales have ∑Fe2O3 contents of typically ~6 wt.% (e.g., PAAS). Abundant siderite in some shales of the Sheba Formation was confirmed by 61

mineralogical analyses using XRD and microscopic observation using petrographic

microscope. Their high Ccarb contents also support significant abundance of siderite. The 2.96 Ga Parktown Formation (Fig. 2-10-l), the >2.60 Ga Marra Mamba Iron Formation (Fig. 2-10-f), and the 2.60 Ga Wittenoom Dolomite Formations (Fig. 2-10-d)

also show depleted Al2O3 contents (in average, ~12 wt.%, ~11 wt.%, and ~7 wt.%, respectively: Table 2-2) because of high ∑Fe2O3 contents (in average, ~22 wt.%, ~33 wt.%, and ~14 wt.%, respectively: Table 2-2). Three samples of the 2.64 Ga Black Reef

Formation show very low Al2O3 contents (< 4 wt.%) due to the elevated Ccarb contents (11 ~ 12 wt.%) (Fig. 2-10-g and Table 2-2). The shales of the 2.22 Ga Timeball Hill Formation

displays somewhat elevated Al2O3 contents by as much as 10 wt.% with the lower SiO2 contents compared to PAAS (59 wt.%) by 5 ~ 10 wt.%. Iron-bearing minerals in shales are usually detrital / authigenic clays (biotite, chlorite), and authigenic sulfides. The excess Fe (over ~6 wt. %) in the shales described above are mainly due to siderite (Sheba Formation), chlorite (Parktown Formation), and oxide (Marra Mamba Iron, Carawine and Wittenoom Dolomite Formations).

The overall ranges of the average TiO2 contents of the Archean–Paleoproterozoic samples of this study are from 0.26 wt.% (3.25 Ga Sheba Formation (shales)) to 0.87 wt.%

(2.72 Ga Pillingini Tuff Formation). These values are slightly smaller than the TiO2 content of the average shales (PAAS; 0.9 wt.%). Red shale samples of the ~2.2 Ga Mapedi

Formation exhibit the elevated average TiO2 contents (1.45 wt.%). Characteristics of the Ti contents of the studied samples resemble their Al contents, because of their coherent behaviors during weathering as mentioned above.

2-4-6. Phosphorus contents Phosphorus contents for the samples of this study are summarized in Table 2-1 and are also shown in the histograms in Fig. 2-11 with the previously published average shales 62

(Taylor and McLennan, 1985; Condie, 1993) and the Devonian black shales (Ingall et al., 1993).

The average P2O5 contents of the samples are generally lower than the P2O5 contents of the average shale (PAAS, 0.15 wt.%: Taylor and McLennan, 1985). For

example, the mean values of P2O5 contents are 0.08 ± 0.05 wt.% for the 3.25 Ga Sheba Formation (shales), 0.10 ± 0.02 wt.% for the 3.25 Ga Sheba Formation (graywackes), 0.08 ± 0.05 wt.% for the 2.96 Ga Parktown Formation, 0.10 ± 0.03 wt.% for the 2.71 Ga Rietgat Formation, 0.08 ± 0.02 for the 2.69 Jeerinah Formation, 0.09 ± 0.03 for the 2.69 Ga Lewin Shale Formation, 0.032 ± 0.01 for the 2.64 Ga Black Reef Formation, 0.05 ± 0.01 for the >2.60 Ga Marra Mamba Iron Formation, 0.04 ± 0.02 for the 2.60 Ga Wittenoom Dolomite Formation, and 0.09 ± 0.13 for the 2.56 Ga Oak Tree Formation

(Table 2-1, Fig. 2-11-n, -m, -l,, -j, -i, -g, -f, -d, and -c). Some of these samples with low P2O5

contents are due to the dilution effect by the ∑Fe2O3 contents, reaching to over 40 wt.%, or by the carbonate contents (see Fe result section). The samples of the 2.22 Ga Timeball Hill

Formation show P2O5 contents (0.16 ± 0.10 and 0.13 ± 0.15 wt.%) comparable to that of the average shale (Fig. 2-11-b), although with considerable variability indicated by their standard deviations. Some red shale samples of the ~2.2 Ga Mapedi Formation have higher

P2O5 contents, reaching to 0.78 wt.%, with a mean P2O5 content of 0.24 ± 0.25 wt.% (Fig. 2-11-a).

2-4-7. Organic C and carbonate C contents

Distributions of the Corg and Ccarb contents for the samples of this study are shown in Fig. 2-12 and 2-13, respectively. The data are tabulated in Table 2-1 with average contents and 1σ for each formation.

The overall range of the Corg contents is from ~0.1 to ~12 wt.%. The Corg contents are high (reaching to 12 wt.%) in the 2.69 - 2.60 Ga samples (Jeerinah, Lewin Shale, Marra 63

Mamba Iron, and Carawine Dolomite Formations; Fig. 2-12-i, -h, -f, and –e, respectively).

On the other hand, the other samples have the Corg contents of less than 3 wt.% (the 3.25 Ga Sheba Formation (shales), the 2.71 Ga Rietgat Formation, the 2.60 Wittenoom Dolomite Formation, and the 2.56 Ga Oak Tree Formation; Fig. 2-12-n, -j, -d, and -c, respectively) or less than 1 wt.% (the 3.25 Ga Sheba Formation (graywackes), the 2.96 Ga Parktown Formation, the 2.72 Ga Pillingini Tuff Formation, the >2.60 Ga Marra Mamba Iron Formation, the 2.22 Ga Timeball Hill Formation, and the ~2.2 Ga Mapedi Formation (red shale); Fig. 2-12-m, -l, -k, -f, -b, and –a, respectively). The graywackes and shales from the

3.25 Ga Sheba Formation display distinct difference in the Corg contents (0.26 ± 0.13 wt.% for graywackes and 1.10 ± 0.72 wt.% for shales).

The Ccarb contents for the samples of this study are mostly lower than 3 wt.% (Fig.

2-13). High Ccarb samples are found in the 3.25 Ga Sheba Formation (shales: up to ~9 wt.%; Fig. 2-13-n), the 2.64 Ga Black Reef Formation (up to ~12 wt.%; Fig. 2-13-g), and the 2.60 Ga Carawine and Wittenoom Dolomite Formations (up to ~9 and ~11 wt.%, Fig.

2-13-e and –d, respectively). The high Ccarb samples of the 3.25 Ga Sheba Formation (shales) are due to the abundance of Fe carbonate (siderite), and those of the 2.64 Ga Black Reef Formation and the 2.60 Ga Carawine and Wittenoom Dolomite Formations are mainly due to the abundance of calcite and dolomite, as confirmed by optical observation under petrographic microscope, XRD analysis, and major element analysis including Fe, Ca, and Mg. Important findings are that (1) there is no significant geographical difference in the

Corg contents between the black shales from South Africa and those from Australia, (2) there is no significant temporal difference in the Corg contents among black shales, (3) there

is large differences in the Corg content between black shales and non-black shales. All of these characteristics are typically found in modern sediments and Phanerozoic sedimentary rocks. 64

2-4-8. Pyrite S contents The pyrite S contents of the samples are shown in Fig 3-14. The data are tabulated in Table 2-1 with an average contents and 1σ for each formation. Since sulfate minerals were neither found under a petrographic microscope nor detected by XRD analysis, the measured S contents of the samples represent sulfide (Spy: pyrite sulfur) abundance.

The Spy contents for the samples of this study are generally low (i.e., < 1 wt.%), but they are vary considerably. For example, the samples of the 2.69 Ga Lewin Shale and Jeerinah Formations (Fig. 2-14-h and -i) range in the S contents from 0.1 to 4.0 wt.% and 0.7 to 3.0 wt.%, respectively. The 2.96 Ga Parktown Formation, the 2.72 Ga Pillingini Tuff Formation, and the 2.71 Ga Rietgat Formation (Fig. 2-14-m, -l, and -k, respectively) shows less variable S contents (0.0 to 0.7 wt.%). Out of 31 samples of the 3.25 Ga Sheba Formation (both shales and graywackes; Fig. 2-14-n and -m), only 5 samples have S contents higher than 0.1 wt.%. Red shales of the ~2.2 Ga Mapedi Formation do not show detectable S abundance. Variable Spy contents of shales are commonly found in the modern marine sediments, Phanerozoic shales, and Proterozoic shales (0 to ~6 wt.%: Raiswell and Berner, 1985, 1986; Fisher and Hudson, 1987; Donnelly et al., 1988; Raiswell et al., 1988; Dean and Arthur, 1989; Raiswell and Al-Biatty, 1989; Jackson and Raiswell, 1991; Pearson et al., 1996; Murphy et al., 1999).

Generally the low Spy contents of the Archean–Paleoproterozoic samples of this study (< 1 wt.%, mostly < 0.5 wt.%; Fig. 2-14) are consistent with Watanabe et al. (1997) who reported the Spy contents of the 100 samples of the Archean–Paleoproterozoic shales from South Africa ranging from < 0.01 to 1.6 wt.% with a mean of 0.1 wt.%. The Spy contents of this study and Watanabe et al. (1997) also agree well with the S contents of average shales (~0.2 wt.%; Turekian and Wedepohl, 1961; Wedepohl, 1971, 1991). 65

2-4-9. Pyrite S to organic C ratios

The Spy/Corg wt. ratios for the samples of this study are summarized in Table 2-3.

Due to the generally low Spy contents, the majority of samples have low Spy/Corg ratios (<

1). The average Spy/Corg ratio for each formation is lower than 1 except for the 2.96 Ga Parktown Formation, the 2.64 Ga Black Reef Formation, and the >2.60 Ga Marra Mamba

Iron Formation. Large variation in Spy/Corg ratios are found in these three formations, and their average Spy/Corg ratios are higher than 1 (reaching to 11). Such variability in the 2.96 Ga Parktown Formation and the 2.64 Ga Black Reef Formation is not due to the enrichment of Spy but due to the low abundance of Corg (Table 2-1; see also section 2-4-7). The

variable Spy/Corg ratios for the >2.60 Ga Marra Mamba Iron Formation are due to one Spy- rich sample (3.2 wt.%) compared to others (< 0.4 wt.%). Despite the large variabilities, the average Spy/Corg ratios for each formation of this study are comparable to the Spy/Corg ratio of normal marine sediments and sedimentary rocks (0.36; Berner, 1982; Berner and Raiswell, 1983; 1984; Raiswell and Berner, 1985, 1986, Raiswell et al., 1988). The observation of the large variabilities and the overall low average values for the Spy/Corg ratios for the samples of this study is also consistent with those in other examples of the

Archean and Proterozoic shales. Strauss and Beukes (1996) reported the Spy/Corg ratios for shales of the Transvaal Supergroup, generally ranging between 0 and 2 with an average of 0.4 (n = ~50). There are six samples in the data set of Strauss and Beukes (1996) whose

Spy/Corg ratios are variable and higher than 2 (33, 22.5, 6.4, 6.0, 3.5, and 2.4). Jackson and

Raiswell (1991) reported the Spy/Corg ratios (between 0 and 5 with an average of 0.7, n = 22) for the shales of the Roper Group in the McArthur Basin of northern Australia. 66

2-4-10. Carbon isotopic compositions of organic C

13 The δ Corg values of the studied samples are shown in Fig. 2-15. The overall range 13 13 in the δ Corg values is from -48.9 to -22.6 ‰. The range in the δ Corg values for each formation is different from one another. Generally, the ~2.6 to ~2.7 Ga samples (Fig. 2-15-

13 d through -j) have depleted (less than -40 ‰) δ Corg values, while the ~2.2 Ga Mapedi Formation (Fig. 2-15-a), the 2.96 Ga Parktown Formation (Fig. 2-15-l), and the 3.25 Ga

13 Sheba Formation (shale: Fig. 2-15-n) show δ Corg values between -30 to -25 ‰. The 13 depleted δ Corg values for the ~2.6 to ~2.7 Ga samples are consistent with the previously published results (e.g., Hayes et al., 1983; Schidlowski et al., 1983; Schidlowski, 1988; Hayes, 1994). More than 2 ‰ variability within a formation is found in all the formations except for the red shale of the ~2.2 Ga Mapedi Formation (Fig. 2-15-a) and the 2.56 Ga Oak Tree Formation (Fig. 2-15-c). Especially, the 2.64 Black Reef Formation and the 2.71 Rietgat Formation display more than 10 ‰ variabilities within each formation.

13 There is a clear difference in the δ Corg values among shales / graywackes of the 13 Transvaal, Mt. Bruce, and Swaziland Supergroups. The δ Corg values of the shales for the Timeball Hill Formation obtained in this study (-28.5 ± 0.6 ‰) agree well with those previously obtained by Watanabe et al. (1997) for the same formation from a different

13 region (-30.3 ± 2.0 ‰). The δ Corg values of the shales for the Oak Tree Formation (- 33.5 ± 0.5 ‰) also agree well with those of the stratigraphically underlying Black Reef Formation (-32.8 ± 2.3 ‰) obtained by Watanabe et al. (1997). Shales that belong to the

13 Mt. Bruce Supergroup show the distinctively negative δ Corg values (reaching to near -50 ‰). The shales of the Sheba Formation and the Timeball Hill Formation have more or less

13 similar δ Corg value between about -25 and -30 ‰. Graywackes of the Sheba Formation 13 exhibit the heaviest δ Corg values (near -25 ‰) among samples analyzed in this study. 67

2-4-11. Sulfur isotopic compositions of pyrite S

Figure 2-16 shows the δ34S values of bulk rock sulfide in the Archean–Paleoproterozoic samples of this study. Ranges within each formation are noted. For example, the 3.25 Ga Sheba Formation (shale; Fig. 2-16-l) displays ~8 ‰ variations, the contemporaneous graywacke (Fig. 2-16-k) displays ~10 ‰ variations, the 2.69 Ga Jeerinah Formation displays ~13 ‰ variation with the highest value being 13.4 ‰, and the 2.60 Ga Carawine Dolomite Formation display ~8 ‰ variation with the highest value of 14.3 ‰ (the highest value of this study). The shales and graywackes of the 3.25 Ga Sheba Formation (Fig. 2-16-k and -l) show different distribution pattern (-3.1 to +4.0 ‰ and +0.6 to +9.3 ‰, respectively), with the latter toward more positive values.

2-5. Discussion

In the following discussion, the geochemical cycles of Fe, P, C, and S are treated separately. However, each geochemical cycle is inter-related to the others to form a coherent overall geochemical cycles of the elements. It is hoped to draw a consistent picture of the Archean–Paleoproterozoic surface redox environment.

2-5-1. Geochemical cycles of Fe

2-5-1-1. Redox chemistry of Fe Iron geochemistry (i.e., total, ferric, and ferrous contents) of soils, sediments, and sedimentary rocks has been important and widely used to constrain the past redox environments because of its redox-sensitive behavior during weathering, diagenesis, and metamorphism. Timing of the rise of pO2 has been debated based on the Fe geochemistry 68

of paleosols (e.g., Ohmoto, 1996; Holland and Rye, 1997; Rye and Holland, 1998), because the past atmospheric redox condition is expected to have had direct and significant impacts on the Fe behaviors in paleoweathering profiles. In most modern weathering profiles inevitably influenced by an oxic atmosphere, the total Fe content remains nearly constant and does not deviate greatly from those of parental rocks (e.g., Nesbitt and Wilson, 1992). This is because the solubility of Fe3+-oxides is extremely small at normal pH ranges of the Earth's surface environments (e.g., Morel and Hering, 1993; Stumn and Morgan, 1996). Iron in modern weathering profiles is fixed as nearly insoluble Fe3+-oxyhydroxides / hydroxides (e.g., limonite, goethite). Therefore, the retention of Fe in the uppermost part of paleoweathering profiles can be used as evidence supporting an oxic atmosphere during soil formation (weathering of source rocks) (e.g., Ohmoto, 1996). However, in paleoweathering profiles developed under an anoxic atmosphere, Fe from Fe2+-silicates would be removed as aqueous Fe2+ in reduced solutions. Such Fe removal may occur in bedrock covered by poorly ventilated swamps, even in present-day globally oxygenated environments. Weathered materials on the continents are transported through rivers as dissolved and particulate phases enter into the oceans, and form sediments. Shales formed by the

accumulation of weathered materials which increased the Fe3+/∑Fe ratios from their source rocks should also show increased Fe3+/∑Fe ratios. Although marine sediments may change their Fe chemistry during organic and inorganic diagenesis, the Fe3+/∑Fe ratios of marine sediments may be used to constrain the paleoweathering condition of their source rocks and thus paleoredox conditions of the atmosphere. Indeed, the recent hemipelagic marine sediments deposited near the Japan Trench and the Shikoku Basin in northeastern Pacific

Oceans show the elevated Fe3+/∑Fe ratios (~0.6, an average of hundreds of DSDP / ODP samples) compared to the Fe3+/∑Fe ratios of the average igneous rocks of the Japan (0.3) (Yamaguchi and Ohmoto, 1997). However, in spite of its potential importance, Fe 69

geochemistry of Precambrian shales has not generally been considered, except for Kump and Holland (1992) and Hayashi et al. (1997).

2-5-1-2. Geochemical cycle of Fe in anoxic vs. oxic environments Geochemical cycles of Fe in the Archean–Paleoproterozoic surface environments have been debated. The popular idea for the geochemical cycle of Fe before the inferred GOE at around 2.2-2.0 Ga is such that Fe is lost as Fe2+ from source rocks during weathering, transported from the continents to the oceans as Fe2+ both in dissolved and particulate forms, and precipitated on the ocean floors. Additional inputs of Fe2+ to the sediments would have been supplied from submarine hydrothermal activities that would have been more active in the early Precambrian than that of today (e.g., Kump and Holland, 1992). All of these processes lead to the Fe2+-rich (anoxic) oceans before the inferred GOE. On the other hand, the emerging idea for the geochemical cycle of Fe in the early Precambrian postulates essentially the same geochemical cycles of Fe as that of today: Fe2+ is converted to Fe3+ during oxidative weathering, very little Fe2+ is carried by rivers in either the dissolved or particulate form (Kump and Holland, 1992), and Fe deposition occurs on the ocean floors. During continental weathering occurring under an oxygenated atmosphere, the weathered materials show increases in the Fe3+/Ti ratios, decreases in the Fe2+/Ti ratios, and unchanged ∑Fe/Ti ratios compared to their parent rocks. No addition of Fe is expected except for sediments accumulated in environments influenced by submarine hydrothermal environments. Complicating factors in the Fe geochemistry of ancient sedimentary rocks are post- depositional changes during diagenesis, metamorphism, and hydrothermal alterations.

Diagenetic pyrite formation occurs in sediments through reactions between H2S produced by sulfate-reducing bacteria (SRB) and reactive Fe (e.g., Berner, 1984). The reactive Fe in 70

modern marine sediments is mostly Fe3+-(hydr)oxides, rather than Fe2+-silicates (e.g., Canfield, 1989a; Raiswell et al., 1994; Canfield et al., 1996; Raiswell and Canfield, 1998; Lyons et al., 2000). During early diagenesis of sediments, Fe3+ may be reduced by Fe- reducing bacteria. During metamorphism of sedimentary rocks, Fe3+ undergoes reduction by OM and/or sulfides at elevated temperature. Therefore, carbonaceous shales and highly

metamorphosed shales are expected to show low Fe3+/∑Fe ratios. During hydrothermal 3+ alteration and high grade metamorphism, Fe reduction may occur by H2, H2S, and / or 2+ CH4 (Ohmoto and Kerrick, 1978). Additional Fe , carried in hydrothermal solution, may be added to the shales or Fe2+ may be leached in solution from the shales (i.e., reductive

dissolution of Fe3+). All of the above processes lead to lowering of the Fe3+/∑Fe ratios compared to the Fe3+/∑Fe ratios of the original marine sediments and of the parental soils.

2-5-1-3. Archean–Paleoproterozoic geochemical cycle of Fe Figure 2-5, 2-6, and 2-7 demonstrate the overall variabilities in the Fe3+/Ti, Fe2+/Ti, and Fe3+/∑Fe ratios among the Archean–Paleoproterozoic shales of this study. Large variations in these rocks are unlikely to have been caused by rapid changes in the source types, because the range in the Fe3+/∑Fe ratios among typical igneous rocks (from basalt and granite) lies typically between 0.1 and 0.3 (Holland, 1984). Therefore, the large variations are more likely the results of diagenetic / metamorphic modifications of the Fe signatures that were originally acquired during varying degrees of oxidative weathering on the continents. The extreme Fe enrichments in shales suggested for the 3.25 Ga Sheba Formation (Fig. 2-5 and 2-6-a), the 2.96 Ga Parktown Formation (Fig. 2-5 and 2-6-c), and the >2.60 Marra Mamba Iron Formation (Fig. 2-5 and 2-6-f) were most likely due to the additions of Fe-bearing minerals precipitated from hydrothermally influenced seawater. Therefore, the Fe geochemistry of these rocks is useless for extracting information on the paleoredox 71

sedimentary environments. However, the "normal" shales that have not suffered from such an extreme Fe enrichment may provide important information on the paleoredox sedimentary environments. The 2.69 Ga Jeerinah Formation (Fig. 2-6-e) and coeval Lewin Shale Formation (Fig. 2-6-e) show overall increase in the Fe3+/Ti ratios and decrease in the

Fe2+/Ti ratios with nearly conserving the ∑Fe/Ti ratios, suggesting the oxidative weathering of source rocks on the continent and minor post-depositional effects. The 3.25 Ga Sheba Formation (graywackes; Fig. 2-6-b), the 2.72 Ga Pillingini Tuff Formation (Fig. 2-6-d), the 2.71 Ga Rietgat Formation (Fig. 2-6-d), and the 2.22 Ga Timeball Hill Formation (Fig. 2-6- h) exhibit the Fe3+/Ti, Fe2+/Ti, and ∑Fe/Ti ratios not much different from those of their source rocks. These observations may represent the oxidative weathering of the source rocks followed by post-depositional Fe3+ reduction during diagenesis and/or minor grade of metamorphism.

2-5-2. Geochemical cycle of P

2-5-2-1. Detrital, biological, and redox controls on the heterogeneous distribution of P in sediments Phosphorus is an essential nutrient for the marine ecosystem. The bio-availability of dissolved P has been thought to control biological productivity on geological time scales (e.g., Holland, 1978; Broecker and Peng, 1982). Phosphorus availability in the oceans is limited to the release of P during weathering of continental crusts, of which apatite is the primary source of P (e.g., Garrels and Mackenzie, 1971; Holland, 1984; Filippelli and

3- Delaney, 1994). Phosphorus is transported in groundwater and river water as PO4 . Non- reactive P, primarily consisting of detrital apatite, usually represents only a small fraction relative to the total P flux (e.g., Ruttenberg and Berner, 1993; Filippelli and Delaney, 1996). Delivery of reactive P from oceans to sediments have been mediated by precipitating OM 72

and Fe oxides, the latter of which has high sorption capacity for phosphate. Benthic phosphate regeneration in sediments is controlled by P recycling from Fe redox cycling and authigenic apatite formation (e.g., Froelich et al., 1982, 1988; Van Cappellen and Berner, 1988; Ruttenberg and Berner, 1993; Ingall et al., 1993; Ruttenberg, 1993). Aerobic conditions enhance the removal of oceanic P by sediments, whereas anaerobic conditions promote the return of dissolved P to the water column (Ingall et al., 1993; Ingall and Jahnke, 1994, 1997). The enhanced retention of P by sediments underlying oxygenated bottom waters is attributed to active bioaccumulation by aerobic benthic bacteria and phosphate sorption by Fe3+-(hydr)oxides (Ingall et al., 1993; Ingall and Jahnke, 1994). The sedimentary distribution of P has been governed by detrital, biological, and redox processes throughout geologic time.

2-5-2-2. Phosphorus in the Archean–Paleoproterozoic shales In a globally anoxic world where the atmosphere and the entire oceans are anoxic, sedimentary enrichment / depletion of P would have been governed by detrital and biological processes. Regeneration of P upon reductive dissolution of P-bound Fe- (oxyhydr)oxides would have been minor and the availability of reactive P to organisms would have been limited in the globally anoxic oceans. Bio-available P would have been provided to organisms by the apatite dissolution and the anaerobic decomposition of OM. However, it is not clear whether or not sedimentary enrichment / depletion of P differs greatly between the globally anoxic environments and the globally oxic environments as today. Since P contents of shales (Fig. 2-11) alone can not be used to infer the paleoredox environments, we examine the P/Al (wt.) ratios of the Archean–Paleoproterozoic shales and compare them with those of the modern sediments and Phanerozoic sedimentary rocks. The P/Al wt. ratios of the Archean–Paleoproterozoic shales of this study are summarized in Table 2-3 and shown in Fig. 2-17. Important observations are that the overall 73

average P/Al ratios of the Archean–Paleoproterozoic shales of this study are close to the P/Al ratios of the world average shales (0.007; PAAS: Taylor and McLennan, 1985), the average Canadian Aphebian (Paleoproterozoic) shales (0.007, n = 326) of Cameron and Garrels (1980), the average Canadian Archean shales (0.008, n = 406) of Cameron and Garrels (1980), and the South African Neoarchean–Paleoproterozoic shales (0.001 ~ 0.01, n = 240) of Wronkiewicz and Condie (1987, 1989, 1990). The observed similarity in the ranges of the P/Al ratios suggests that the present-day style geochemical cycle of P has been essentially in operation since at least Mesoarchean, and thus that the P contents of shales were governed by detrital, biological, and redox processes. In the later discussion (2-

5-3-6), we utilize the Corg/P ratios of the Archean–Paleoproterozoic shales of this study and those of modern sediments and Phanerozoic shales, in the hope that meaningful information on the redox conditions of paleoenvironments are extracted.

2-5-3. Geochemical cycle of C

2-5-3-1. Origin of organic matter in shales The modes of occurrence of OM in the black shales and graywackes are examined using their photomicrographs (Fig. 2-18-a through -d for South African samples and -e through -h for Australian samples). Figures 2-18-a (2.22 Ga Timeball Hill Formation), Fig. 2-18-d (3.25 Ga Fig Tree Group shale), Fig. 2-18-f (>2.60 Ga Marra Mamba Iron Formation), and Fig. 2-18-h (2.69 Ga Lewin Shale Formation) illustrate finely laminated features of carbonaceous part (darker parts) and clay-rich parts (lighter parts). A sample of the 2.56 Ga Oak Tree Formation (Fig. 2-18-b) displays relatively massive, poorly-laminated features, with dispersed Corg in clay . In the Fig. 2-18-c for a graywacke sample of

the 3.25 Ga Sheba Formation, concentration of Corg occurs at the boundary between fine- grained part (lower, darker part) and coarse-grained part (upper, lighter part). There appears 74 to be a layer rich in Corg in the upper coarse-grained part. Carbonate minerals (dolomite), as occurring relatively large crystals, are observed in the darker matrix rich in Corg in the Fig. 2-18-e for a carbonate-rich shale sample in the 2.60 Ga Wittenoom Dolomite Formation.

Figure 2-18-g shows concentration of Corg in the matrix sandwiched by carbonate minerals (dolomite). Sedimentary textures of the black shales in this study are more or less similar to each other and almost identical to black shales of younger ages, such as Cretaceous or modern carbonaceous sediments. The sedimentary features, such as (1) wavy or planer laminations of carbonaceous parts concordant to bedding planes and (2) soft sediment deformation involving carbonaceous laminations (e.g., Fig. 2-18-d), strongly suggest that the OM in the studied samples are in situ (autochthonous) or precipitated OM probably from surface oceans. Possible origins of OM in the black shales will be discussed later. The distribution of OM in recent marine sediments is determined by biologic, sedimentary, and diagenetic processes. Most OM is labile in the globally oxygenated oceans, and only ~0.2 % of the primary production escapes oxidation / decomposition during settling through seawater column and early diagenesis in sediments (e.g., Libes, 1992). This would be most likely true for the ancient sediments. Enrichment of OM in sediments has been explained by two fundamental mechanisms: (1) increased production of marine OM and/or enhanced input of terrestrial OM (e.g., Pedersen and Calvert, 1990; Calvert and Pedersen 1993; Calvert et al., 1996); (2) enhanced preservation of OM as a result of development of anoxia (e.g., Demaison and Moore, 1980). However, there are many factors controlling the preservation potential of sedimentary OM (e.g., Canfield, 1989b, 1994). Organic matter in recent marine sediments contains two components at the time of deposition. One is the autochthonous component, i.e., the remnant of organisms that lived in the overlying water body and in sediments during the sediment accumulation. The other is 75

the detrital component, i.e., remnant of kerogen from older sedimentary rocks and terrestrial plant debris that survived weathering. This is also true for the ancient carbonaceous sediments. However, an additional component, "migrated OM" during diagenesis and metamorphism, must be considered. The Archean–Paleoproterozoic shales of this study have, therefore, three possibilities for the origin of OM, namely: (A) detrital OM of old production that went through tectonic cycle(s); (B) precipitated OM of new production; and (C) introduced (migrated) OM by metamorphic/hydrothermal fluids during post- depositional processes. In each case, the source of the OM could be a variety of organisms.

Process (A) is such that Corg observed in the sediments are allochthonous (i.e., detrital material from continents that went through previous tectonic cycle(s) such as subduction and uplift). If (A) is the case, then we would expect that very fine C particles were produced during physical / chemical weathering and distributed throughout the sediments, sometimes forming a gradational distribution which depends on the grain sizes of the matrix. Since such features are not observed in any sample of this study, the possibility of (A) is considered unlikely. Process (C) is such that Corg is not syn- depositional (neither detrital (A) nor precipitate of primary production (B)) but post- depositionally introduced by metamorphic/hydrothermal fluids. If (C) is the case, then we

would expect that Corg is generally accumulated more in fractured parts than in bedding- parallel layers. However, such feature is not observed in any sample of this study.

It is reasonable to assume that the Corg observed and analyzed in this study was precipitated from biomass that flourished in the surface ocean or in the sediments at that time. This is in agreement with the characteristics of the facies-controlled distribution of OM in Precambrian sediments as outlined by Dimroth and Kimberley (1976). Biological, photoautotrophic production of OM found in carbonaceous Archean sedimentary rocks has been strongly suggested from numerous observations, including Corg isotopic compositions (Hayes et al., 1983; Schidlowski, 1988; Schidlowski and Aharon, 1992; de Ronde et al., 76

1991a; DesMarais et al., 1992; Strauss et al., 1992; Mojzsis et al., 1996; Watanabe et al., 1997; Rosing, 1999), associations with stromatolite (e.g., Walsh and Lowe, 1985, 1999; Schopf and Packer, 1987; Walsh, 1992; Schopf, 1993; Westall et al., 2001), facies-

controlled Corg distribution (Dimroth and Kimberley, 1976), and organic biomarker evidence (Brocks et al., 1999). However, even though the majority of OM in the samples of this study are the remnants of new organisms (the case B), the possibilities of (A) and (C) can not be completely discounted. Based on the above argument, in the following discussion we assume that all of the OM in the shales of this study are the remnants of organisms existed at the time of deposition.

2-5-3-2. Types of marine organisms indicated by C isotopic compositions of organic C

13 The δ Corg values of the early Precambrian sediments have been measured by many previous investigators to infer ancient biological activities (e.g., Hayes et al., 1983; Strauss, 1986, 1989; Schidlowski, 1988; de Ronde et al., 1991a; Des Marais et al., 1992; Schidlowski and Aharon, 1992; Strauss et al., 1992a, b; Mojzsis et al., 1996; Watanabe et

al., 1997; Rosing, 1999). This is because the δ13C values of marine organisms are primarily 13 - controlled by the δ C values of dissolved HCO3 (and CO2) in the seawater and the kinetic isotope effects associated with the various ways of assimilation of dissolved C sources for the organisms. Reviews of the mechanism of C isotopic fractionation by life in the generation of biomass have been given by O'Leary (1981), Farquhar et al. (1982), Roeske

13 and O'Leary (1984), and Schidlowski et al. (1983). A compilation of δ Corg values 13 showed that the δ Corg values (-25 ‰: a typical value for normal, oxygenic photosynthetic marine organisms) have changed little since 3.8 Ga ago, indicative of photosynthetic origin (e.g., Schidlowski et al., 1983; Schidlowski, 1987, 1988; 2001; Schidlowski and Aharon, 1992; Strauss et al., 1992a; Mojzsis et al., 1996; Rosing, 1999). Carbon isotopic

13 compositions of carbonate C (δ Ccarb) have also been relatively unchanged around 0 ‰ 77

since the Archean except for some excursion events in the geologic history (e.g., Lomagundi event at ~2.2-~2.0 Ga: Karhu and Holland, 1996; Melezhik et al., 1999). The

13 antiquity of photosynthesis and the nearly constant difference between δ Corg and 13 δ Ccarb values have profound implications for the C cycles and the evolution of biosphere. 13 However, there has been a controversy as to whether original δ Corg values at the time of deposition can be preserved during post-depositional processes. In order to

13 correctly interpret the δ Corg data obtained in this study, it is necessary to estimate the 13 degree of possible shift in the δ Corg values during overall post-depositional processes from their original values.

Effects of post-depositional alterations on C isotopic compositions of organic C It is suggested by Schidlowski et al. (1983) that diagenesis, catagenesis, and

13 metagenesis have little impact on modifying the original δ Corg values, except for the case

in which Corg undergoes isotopic exchange with coexisting carbonate minerals in high 13 grade metamorphism (δ Ccarb values are also affected). However, it is suggested by Hayes et al. (1983), Des Marais et al. (1992), Strauss et al. (1992a, b), and Hayes (1994) that diagenesis and metamorphism could result in a positive isotopic shift due to escape of carbon-bearing materials (hydrocarbons) such as methane that are enriched in the lighter isotope. This is especially the case for highly altered OM whose kerogen's H/C atomic ratios are less than 0.3. Des Marais et al. (1992) proposed an equation to estimate the

13 13 original δ Corg values from H/C atomic ratio of kerogen and measured δ Ckerogen. Watanabe et al. (1997) also suggested an alternative equation based on the theoretical model of carbon loss using the Rayleigh distillation process. The correction equation proposed by Des Marais et al. (1992) is a fitting curve made from the numerous data points and adjusted to go through the estimated original

13 δ Corg value of -32 ‰ and an H/C atomic ratio of 1.5. Therefore it dismisses the original 78

13 differences in δ Ckerogen values which most likely depend on environmental factors such

as pCO2 and CH4-related processes (discussed later). On the other hand, the correction 13 equation proposed by Watanabe et al. (1997) does not assume initial δ Corg values and is independent of the environmental factors. Furthermore, it agrees with the experimental data of the artificial alteration of OM performed by Peters et al. (1981), Simoneit et al. (1981), and Lewan (1983). Based on this information, the equation by Watanabe et al. (1997) is adopted for this study.

Antiquity of photosynthesis

13 The corrected δ Corg values of the Archean–Paleoproterozoic shales of this study based on the theoretical model by Watanabe et al. (1997) are consistent with previous results by many others suggesting that photosynthetic marine organisms (possibly cyanobacteria) utilizing the Calvin cycle were already very active in the Archean oceans (e.g., Reimer et al., 1979; de Ronde et al., 1991a; de Ronde and Ebbesen, 1996). This suggestion is also supported by the previous discovery of cyanobacteria-like microfossils in chert in both of the ~3.5 Ga Onverwacht Group underlying the Fig Tree Group and the coeval Warrawoona Groups in Australia (Schopf and Packer, 1987; Awramik et al., 1988; Schopf, 1992, 1993; Walsh, 1992), although such microfossils often invoke controversy as to whether they are true microfossils or not (e.g., Schopf et al., 2002; Braiser et al., 2002). It should be noted here that a cyanobacterial organic biomarker has been discovered from 2.7 Ga old black shales in Australia (Brocks et al., 1999). Although the Calvin cycle has been suggested to have operated in Archean–

13 Paleoproterozoic oceanic environments, some of the δ Corg values obtained in this study may not be readily explained solely by the operation of Calvin cycle. For example, the

13 corrected mean δ Corg values of -31 ‰ for the shales of the 3.25 Ga Sheba Formation are lighter by 6 ‰ and those of -40 ~ -50 ‰ for the ~2.7 to ~2.6 Ga shales (Rietgat, Lewin 79

Shale, Jeerinah, Marra Mamba, Wittenoom Dolomite, and Carawine Dolomite Formations;

13 Table 2-3) are much lighter than the typical δ Corg values for modern marine sediments (- 13 25 ‰: Schidlowski, 1988). Therefore, additional factors to lower the δ Corg values are required to explain the results of this study. Such factors can be sought from environmental and/or microbiological parameters. One possibility is an elevated concentration of

- atmospheric CO2 in the Archean and thus the elevated concentration of dissolved HCO3 in the coeval oceans. Another possibility is a contribution of methanogens and methanotrophs which result in an enrichment of 12C in the sedimentary OM.

Effects of CO2 concentration on biological isotopic effect The isotopic composition of photosynthesizers is controlled by the isotopic

13 composition of the CO2 source (atmosphere δ CCO2 = -7 ‰, or oceanic bicarbonate 13 δ CHCO3 = ± 0 ‰), the isotopic fractionation from CO2 assimilation and transformation 13 13 13 to organic matter (∆ C = δ Corg - δ CHCO3 = -20 ~ -30 ‰), and the quantity of 13 available CO2. Park and Epstein (1960) reported changes in plant matter δ C to more

negative values with increasing pCO2 concentration. A number of experiments at different pCO2 regimes have been carried out and confirm this result (e.g., Farquhar et al., 1982; Sharkey and Berry, 1985; Rau et al., 1989, 1991, 1992, 1997; Freeman and Hayes, 1992; Raven et al., 1993; Goericke and Fry, 1994; Hinga et al., 1994; Jasper et al., 1994; Heimann and Maier-Reimer, 1996; Maslin et al., 1996; Bidigare et al., 1997). Particularly at low cell densities and pCO2 level greater than 0.5 %, Corg becomes isotopically lighter. When pCO2 is high, the photosynthesizing cells readily fractionate the abundant 12C from 13C resulting in a large isotope effect. Conversely, when cell growth is limited by carbon dioxide availability at low pCO2, all available CO2 is used, regardless of its isotopic value. This 13 results in heavier (higher) δ Corg value. 80

The Earth’s early atmosphere is believed to have been enriched in CO2 in the Archean, by a factor of ~100 to ~1000 higher than the present--day (pre-industrial)

atmospheric CO2 level of ~350 ppm (e.g., Kasting, 1987, 1992, 1993). Thus it provides

ample available CO2 to early life. Higher past levels of pCO2 has been supported by the paleoclimate record of the Phanerozoic (Rau et al, 1989). It is therefore anticipated that

13 Archean δ Corg values would be isotopically lighter on average than typical post-Archean biomass (-27 ‰) (Des Marais, 1994), but the role of evolutionary changes in enzymatic mechanisms must also be considered.

13 The following equation relating ∆ C to pCO2 is used by Kump and Arthur (1999) and Lasaga and Ohmoto (2002b) based on the Bidigare et al. (1997) data set for haptophyte algae:

13 3- ...... ∆ C = {159.5 • [PO4 ] + 38.39} / {0.034 • pCO2} - 33 (2-5)

13 3- where ∆ C is in per mil (‰) and [PO4 ] is in µM and pCO2 is in ppm. This equation 13 3- suggests that the ∆ C is always heavier than -33 ‰ at any [PO4 ] and pCO2 condition 13 (see Fig. 2 of Kump and Arthur, 1999). From the observation that the δ Ccarb values in the Archean are essentially the same as those of today (~0 ‰; e.g., Schidlowski et al., 1983; Schidlowski, 1988) and the assumptions that the enzymes, cell size range, growth rate were

13 the same between Archean and today, we assume that the δ C value of atmospheric CO2 in the Archean was the same as that of today (-7 ‰; e.g., Faure, 1986). Certainly the inferred

13 higher pCO2 level in the Archean is an important factor to have controlled the δ Corg 13 values; however, it is not sufficient to cause the very negative δ Corg values of -40 ~ -50 ‰ for the ~2.6 to ~2.7 Ga samples of this study. 81

Isotopic fractionation by chemosynthesis Methanogenesis, driven by bacteria, is an ubiquitous process in anaerobic

- ecosystems where OM is decomposed without O2 and NO3 . Available C isotopic fractionation data have shown a range down to -38 ‰ (Schidlowski et al., 1983). The isotopic fractionation associated with methanotrophy would be small (Hayes, 1994);

13 however, the CH4 generated by methanogens from OM decay is strongly depleted in C, commonly with values as low as δ13C = -60 ‰, due to kinetic isotope effects associated with fermentation:

...... 2 CH2O ---> CH4 + CO2 (2-6)

Methylotrophs, CH4-consuming bacteria, carry out CH4 oxidation (reverse reaction

of Eq. 2-6). The O2 needed for CH4 metabolisms is provided by oxygenic photosynthesizers (frequently as a bacterial mat). Methylotrophs assimilate isotopically

light CH4 evolved from the degradation of OM in sediments. This recycling of OM produces isotopically lighter OM on average than biomass formed through photosynthesis. Such a community structure of microorganisms is found in contemporary microbial mats which are considered analogues for the most primitive ecological communities (e.g., Des Marais, 1995). Summons et al. (1994) have found that methylotrophic bacteria discriminate against 13C by an additional 15-30 ‰ relative to the initial fractionation of the isotopes by photosynthetic primary producers, hence resulting in very low lipid δ13C values of between –51.3 and –66.8 ‰ as measured in the field and laboratory. These organisms are primitive members of a branch (Archea) of the phylogenetic tree, and many of them are hyperthermophiles. They are likely to have participated in the C cycle of the early Archean along with the primary producers of the biosphere, photosynthetic bacteria. 82

13 The δ Corg values of shales of the 3.25 Ga Sheba Formation (-28.7 ‰) and the 2.22 Ga Timeball Hill Formation (-28.5 ‰) (Fig. 2-15; Table 4-1, 4-2) may be better explained by a high pCO2 hypothesis, and those of the 2.69 Ga Lewin Shale Formation (- 42.1 ‰), >2.60 Ga Marra Mamba Iron Formation (-40.4 ‰), 2.60 Ga Carawine Dolomite Formation (-45.0 ‰), and 2.60 Ga Wittenoom Dolomite Formation (-37.4 ‰) (Fig. 2-15; Table 4-1, 4-2) may be better explained by an involvement of methanogens and methanotrophs in the sedimentary C cycle (Hayes, 1994). It should be noted that

2- methanotrophs require O2 and/or possibly SO4 as oxidants of CH4 (e.g., Hoehler et al., 1994, 1996; Hinrichs et al., 1999; Boetius et al., 2000; Bian et al., 2001; Orphan et al., 2001; Schouten et al., 2001).

2-5-3-3. Original contents of organic C in shales

Contents of Corg in sediments generally decrease with increasing depth of burial during early diagenesis, primarily because of microbial consumption. During later stages of diagenesis (i.e., metagenesis, catagenesis) and subsequent stages such as metamorphism, C contents in sediments tend to continue to decrease. However, the degree of the decrease in

Corg contents in sediments is higher during early diagenesis than during the subsequent later metagenesis, catagenesis, and metamorphism.

The original Corg contents of Neoarchean and Paleoproterozoic shales were calculated by Watanabe et al. (1997) using an equation that takes into account the loss of volatile hydrocarbon from sedimentary OM during post-depositional processes. According

to that study, it is estimated that original contents of Corg at the time of deposition during

Neoarchean and Paleoproterozoic are approximately 5 times of the preserved Corg contents in the shales. If this scheme is applied to 3.25 Ga old black shales of this study, the original

Corg contents at the time of deposition are estimated to be 2 to 15 wt.%, corresponding to

the range of measured Corg contents of 0.4 to 3 wt.%. Such an amount is nearly comparable 83

to the Corg contents of modern carbonaceous sediments deposited in high productivity areas such as the Peruvian margin and Namibian shelf under vigorous coastal upwellings (de Vries and Pearch, 1982; Calvert and Price, 1983; Suess et al., 1987).

2-5-3-4. Aerobic cycling and burial flux of organic matter

Demonstrated similarities in the Corg contents between Archean–Paleoproterozoic black shales and Phanerozoic / modern sediments suggest an operation of aerobic recycling of Corg throughout the geologic time (e.g., Towe, 1990, 1991, 1994). Below we develop a

quantitative argument to support the operation of aerobic recycling of Corg in the Archean.

Organic matter (CH2O) is produced by photosynthesis in the following reaction:

CO2 + H2O ---> CH2O + O2 ...... (2-7)

This equation suggests that the burial of 1 mole of Corg generates 1 mole of O2. If all the 2+ 2+ O2 is consumed by Fe in the oceans in the following reactions, then 4 moles of Fe are converted to Fe3+.

2+ - 3+ 4 Fe + O2 + 4 H ---> 4 Fe + 2 H2O...... (2-8)

Without aerobic recycling (decomposition) of Corg in the oceans (i.e., all the carbon produced by photosynthesis is unrealistically preserved), Fe3+ contents in wt.% as a sink

for the produced O2 would be

4 [mol Fe/mol C] ÷ 12 [g/mol C] x 56 [g/mol Fe] x Corg [wt.%] ...... ≈ ~19 x Corg [wt.%] (2-9) 84

If the average Corg contents in Archean sedimentary rocks are 0.5 wt.%, then their average

Fe contents should be ~9 wt.% to consume all the O2 produced by photosynthesis. This amount is nearly a factor of two greater than the actual average Fe contents of 5 wt.%,

indicating that there was not enough Fe to consume all the O2 generated by a burial of Corg. Applied to the shales of the Sheba Formation, the oldest samples of this study, the average

3+ Corg contents of 2 wt.% would require 38 wt.% of Fe . This is more than 25 times greater than measured Fe3+ contents (an average Fe3+ content is 1.42 wt.%, n = 17) of the Sheba

Formation shales. Other than reduced Fe, reducing volcanic gases such as H2 and H2S would be an important sink for O2. Towe (1994) did not take into consideration such reducing volcanic gases and reducing sulfide minerals (on the continent and the seafloor by hydrothermal venting) as a sink for O2. However, even with the reducing capacity of these volcanic gases and sulfide minerals, the required Fe3+ amount is still way too high, when compared to the actual contents in the rocks. Furthermore, without the aerobic recycling of

Corg, the C reservoir will be depleted (Lasaga and Ohmoto, 2002a). Therefore, the excess

O2 must have been consumed by oxidation of Corg; the aerobic recycling of Corg is still 3+ necessary to balance the Corg and Fe budgets in Archean sedimentary rocks.

An operation of aerobic cycling of Corg in the Archean oceanic environments suggests that the marine geochemical cycle of C in the Archean was essentially similar to

that of today. This suggestion may be tested by examining the burial flux of Corg in

sediments through geologic time. The Corg/Al wt. ratios of the Archean–Paleoproterozoic shales of this study are shown in Fig. 2-19, together with those of the world average shales (Wedepohl, 1991), the carbonaceous sediments of the Black Sea (Hirst, 1974; Brumsack, 1989), the Cretaceous black shales (Dean and Arthur, 1986), the average Canadian Aphebian (Paleoproterozoic) shales (Cameron and Garrels, 1980), and the average Canadian Archean shales (Cameron and Garrels, 1980). Important observations in Fig. 2-19 is that 85

the Corg/Al ratios of the Archean–Paleoproterozoic samples of this study fall within a range between the Corg/Al ratios of the world average shale (Wedepohl, 1991) and the carbonaceous Black Sea sediments (Hirst, 1974; Brumsack, 1989). There appears no

significant difference in the Corg/Al ratios of sediments, i.e., a proxy of the burial flux of

Corg, throughout geologic ages. We suggest that the burial flux of Corg in sediments (i.e., the production rate of O2; Eq. 2-7) in the Archean–Paleoproterozoic environments was essentially the same as that of today.

2-5-3-5. Redox structure of the oceans as indicated from the relationships between the contents and isotopic compositions of organic C Mechanisms for the formation of black shales have been long debated. Increased primary production and/or enhanced preservation of OM seem to be essential for the black shale formation (e.g., Arthur and Sageman, 1994; Pedersen and Calvert, 1990; Demaison and Moore, 1980). Recently, an importance of clay minerals as agents for sorbing OM has been proposed for the preservation mechanism of OM and thus the formation of black shales (Kennedy et al., 2002). In an environment where OM is well preserved, the accumulated OM serves as energy sources for diverse benthic community (e.g., chemotrophs and heterotrophs). Because of such continued biological activity, the amount

of OM and its Corg isotopic compositions may be modified. For example, a participation of methanogens in the sedimentary C cycle generally drive the Corg isotopic compositions of accumulated OM to the significantly negative values (e.g., -60 ‰; Summons et al., 1994). Other than methanogenesis, a variety of reactions including sulfide-oxidation and ammonia- oxidation result in biomass depleted in 13C (Strauss and Beukes, 1996). The onset of such benthic microbial activity depends on the amount of available energy source, i.e., OM. Therefore it is expected that there is a first-order negative correlation between the amount of 86

Corg and its isotopic compositions. However, whether or not such relationships are observed in marine sediments depends on the development of redox gradients in sediments.

In an oxic world like today's, the operation of CH4-involving microbial processes is limited to environments where anoxic conditions are developed. To induce the local anoxic

- 2- conditions, primary oxidants such as dissolved O2, NO3 , and SO4 need to be consumed

by decay of OM before the onset of CH4-involving microbial processes. Therefore, at a first order approximation, anoxic water body and OM-rich sediments tend to reach the stage of

the CH4-involving microbial processes. Since the CH4-involving microbial processes tend 13 to drive the δ Corg values of OM to more negative values, the negative correlations between 13 the Corg contents and the δ Corg values may be expected. However, sedimentary OM with 13 δ Corg values lower than –30 ‰ is rare in present-day environments. This may be partly

because the microbially produced CH4 at depth in sediments is consumed by microbes upon upward diffusion of CH4 into the less reducing sediments where sulfate reducers are active (e.g., Hoehler et al., 1994, 1996; Hinrichs et al., 1999; Boetius et al., 2000; DeLong, 2000; Bian et al., 2001; Orphan et al., 2001; Schouten et al., 2001) and / or escapes into the overlying oxygenated bottom waters. In a globally anoxic world where the entire atmosphere and oceans are anoxic, the production of OM and the operation of the CH4-involving microbial processes need not be related with each other. This is because the CH4-involving microbial processes are expected to occur in all environment where metabolizable OM is available to such microorganisms. In

13 other words, the δ Corg values are expected to be the same regardless of the amount of

Corg deposited.

Figure 2-20 shows the overall negative correlations between the Corg contents and 13 13 the δ Corg values of the Archean–Paleoproterozoic shales of this study. The δ Corg

values range from ~-50 to ~-20 ‰ and the Corg contents range from < 0.1 to > 10 wt.%. Such negative correlation may be interpreted as representing a mixing of at least two end 87

13 members; one has the high Corg contents and the δ Corg values of ~-45 ‰, and the other 13 has the low Corg contents and the δ Corg values of –25 ‰. There appears to have existed at least two types of microorganisms, one accumulated in anoxic environments and produced

13 the negative δ Corg values (~-45‰) and the other accumulated in oxic environments and 13 produced the less negative δ Corg values (~-25 ‰). From the above discussion, this observation suggests that redox gradient in the water column played an important role in controlling the various biochemical pathways metabolizing OM. The Archean–Paleoproterozoic oceans would have been oxygenated enough to yield such redox gradient formed by dissolved O2 in the seawater and abundance of OM in sediments. From a study of the sedimentary rocks of the Transvaal Supergroup, Strauss and

Beukes (1996) have previously observed that the samples with low Corg contents have less 13 negative δ Corg values. Strauss and Beukes (1996) have attributed such observation to the result of post-depositional thermal alteration. The effects of thermal alteration on the

13 δ Corg values of the Precambrian sedimentary rocks should not be underestimated, 13 because it may significantly modify the original δ Corg characteristics. There is a 13 possibility that the observed negative correlations between the Corg contents and the δ Corg values of the Archean–Paleoproterozoic samples of this study are also due to the effects of post-depositional thermal overprints. However, Strauss and Beukes (1996) have noted that

13 such thermal effects are observed for the samples with their δ Corg values ranging between 13 -26.5 and -21.3 ‰, and that the samples with δ Corg values ranging between -43.4 and - 31.4 ‰ preserve the primary signatures as a result of biochemical processes. There are only several samples of this study falling in the 'altered' range of -26.5 to -21.3 ‰ suggested by Strauss and Beukes (1996), and the majority of the samples of this study fall in the region outside of such 'altered' range (Fig. 2-20). Furthermore, contrary to the suggestion by Strauss and Beukes (1996), Watanabe et al. (1997) and Des Marais (1997) have noted that

13 the thermal maturation does not change the δ Corg values by more than 3 ‰. Therefore, 88

13 these studies suggest that the observed δ Corg values of the Archean– Paleoproterozoic shales of this study represent those created by microorganisms at the time of deposition and minimally modified by post-depositional thermal maturation. The observed negative

13 correlations between the Corg contents and the δ Corg values of this study still stand and suggest the redox structure of the Archean–Paleoproterozoic oceans.

2-5-3-6. Redox structure of the oceans as indicated from the organic C to P ratios From mass balance calculations using a coupled model of the biogeochemical cycles of C, P, O, and Fe, Van Cappellen and Ingall (1996) have suggested the redox dependence of P burial in the oceans provides a powerful forcing mechanism for balancing production

and consumption of atmospheric O2 over geologic time. Phosphorus in sediments is therefore important as an agent for redox stabilization of the surface environment of the Earth. The redox dependence of P is owing to the regeneration of P upon reductive dissolution of P-bound Fe3+-oxyhydroxides from anoxic sediments where they are overlain by anoxic bottom water. This causes the same P to be used repeatedly by the organisms

living in the water column resulting in higher Corg/P ratios for sediments deposited under anoxic conditions compared to those under oxic conditions (Ingall et al., 1993; Ingall and Jahnke, 1994, 1997). Data from both modern and ancient marine sediments indicate that the

atomic Corg/P ratios of sediments accumulated under oxygenated bottom water are ~200 or less, while those under anoxic bottom water are at least on the order of 500 ~ 1000, or as high as 4000 (corresponding to ~0.01 wt.% P and ~15.5 wt.% Corg) (Van Cappellen and Ingall, 1996).

The Corg/P ratio of the Archean–Paleoproterozoic samples of this study is

summarized in Fig. 2-21. The variations in the Corg/P ratios within each formation are in

many cases more than two orders of magnitude. The observed variations in the Corg/P ratios 89

of the Archean–Paleoproterozoic samples of this study largely fall within the two end

members; one is the average shale (the Corg/P wt. ratio = 3.05: Wedepohl, 1991; Taylor and McLennan, 1985) and the other one is the modern and ancient anoxic sediments (200 ~ 400: Van Cappellen and Ingall, 1996). From these observations and the discussion provided below, we suggest that the Corg/P ratios of the Archean–Paleoproterozoic shales of this study were also controlled by the marine redox environments and represent the diversity of the redox environments in the Archean–Paleoproterozoic oceans.

Changes in the source rocks compositions ?

The observed variations in the Corg/P ratios are difficult to explain by changes in the source rock composition of the Archean–Paleoproterozoic shales of this study, because P contents do not differ significantly among various types of igneous rocks. According to

Condie (1993), the P2O5 contents of Archean–Paleoproterozoic igneous rocks are 0.16 wt.% for basalt, 0.22 wt.% for andesite, 0.08 for granite, and 0.12 wt.% for TTG (tonalite-

trondhjemite-granodiorite). It is hard to make the variations of the Corg/P ratios observed in this study simply by mixing the different types of source igneous rocks.

Globally anoxic environment ?

Moreover, it is also difficult to explain the variable Corg/P ratios of the Archean–Paleoproterozoic sediments if they deposited in a globally anoxic environment, i.e., anoxic atmosphere and globally anoxic oceans. In such an environment, there would be very

3+ little or no redox-recycling of Corg and P. Formation of Fe -(hydr)oxides would be very minor, and thus the release of P adsorbed onto Fe3+-(hydr)oxides upon their reductive dissolution would be also very minor. The release of P from OM upon anaerobic degradation would occur in a globally anoxic environment, and the Corg/P ratios of the

sediments would be increased to some degree. Therefore, the Corg/P ratios of sediments 90

deposited under in a globally anoxic environment would be equal or higher than the original

Corg/P ratios for organisms. The atomic Corg/P ratio of modern marine organisms is 106

(Redfield et al., 1963), and this value corresponds to the Corg/P wt. ratio of ~40 (see Fig. 2- 17). However, more than a half of the Archean–Paleoproterozoic shales of this study have

the Corg/P ratios of less 40, and the Corg/P ratios are much more variable than expected from a globally anoxic condition.

Metamorphism ?

Metamorphism can potentially modify the Corg/P ratios of the Archean–Paleoproterozoic shales of this study. Apatite has been believed to be more resistant to thermal stress than OM. Therefore, the preferential loss of Corg relative to P might have occurred to the samples of this study. If this is the case, then the observed

Corg/P ratios would represent the minimum values; i.e., the original Corg/P ratios before

metamorphism would have been larger than the observed Corg/P ratios. However, such

possibility of metamorphic increase for the Corg/P ratios of this study would have negligible effect on the argument developed above (only strengthen the importance of oxic

environment where the Corg/P ratios are expected to be low compared to those in anoxic environment).

Diagenetic changes ?

As discussed earlier, the original Corg contents of the Archean–Paleoproterozoic sediments at the time of deposition would be 5 times higher than the measured values, and their original P contents at the time of deposition can be approximated to be 2 times higher than the measured values assuming that the degree of diagenetic decrease of P in sediments is similar to N (see chapter 4 of this thesis). Therefore, the original Corg/P ratios of the Archean–Paleoproterozoic shales at the time of deposition can be approximated to be 2.5 91

times higher than the observed values. However, such increase in the Corg/P ratios would not significantly modify the overall characteristics of the observations (variabilities and distribution between the two end members mentioned above), nor modify the argument above. It should be noted that such corrections for the original values would counteract the

possible metamorphic modification of the Corg/P ratios.

Redox control on the variability of the sedimentary organic C to P ratios From the above discussion, we propose that the most straightforward interpretation

of the observed wide variations in the Corg/P ratios of the Archean–Paleoproterozoic shales of this study, falling within the above-mentioned two endmembers, is the redox control on the sedimentary P geochemistry. This suggestion implies that both oxic and anoxic bottom water / sediments conditions would have played an important role in controlling the P geochemistry since at least the 3.25 Ga ago. Together with the previous argument of the burial flux of P in the Archean–Paleoproterozoic sediments (see section 2-5-2-2), we suggest that the geochemical cycle of P in the Archean–Paleoproterozoic environments was essentially the same as that of today, i.e., controlled by redox processes involving biological activity.

2-5-3-7. Redox cycles of C based on C isotopic compositions

13 The δ Corg values obtained in this study are displayed in Fig. 2-18 with previously reported δ13C values between ~3.8 Ga and ~1.4 Ga in age (see references in Fig. 2-18). 13 The δ Corg values of shales and graywackes of the ~3.25 Ga Sheba Formation fill the gap 13 of the data that existed previously. The δ Corg values of the other samples of this study 13 generally overlap with the previously reported δ Corg values (Fig. 2-18). We emphasize 13 the previous suggestion that the nearly constant δ Corg values (~ -25 ‰), when combined 13 with the nearly constant δ Ccarb values (~ 0 ‰), indicate the present day-style C cycle was 92

established very early and maintained in full operation throughout the geologic ages (e.g., Schidlowski et al., 1983; Schidlowski, 1987, 1988, 2001; Schidlowski and Aharon, 1992). This further suggests that the autotrophic C fixation reflecting the isotope-discriminating properties of ribulose-1,5-biphosphate (RuBP) carboxylase as the principal enzyme of the Calvin cycle has been in full operation since the Archean.

2-5-4. Geochemical cycles of S In marine sediments, bacterial sulfate reduction and the subsequent pyrite formation are common and important biogeochemical processes occurring below the sediment-water interface and also within euxinic water bodies. Under anoxic conditions, seawater sulfate

2- 3+ (SO4 ) is reduced to H2S which reacts with detrital Fe -minerals (reactive Fe) to

ultimately form FeS2 via intermediate species (e.g., Goldhaber and Kaplan, 1974; Berner, 1984). In an euxinic environment, sulfate reduction occurs also in the water column. The overall reaction of pyrite formation by bacterial sulfate reduction is simplified as follows

(CH2O represent OM):

2- 15 CH2O + 8 SO4 + 4 Fe(OH)3 - - ---> 4 FeS2 + 15 HCO3 + OH + 13 H2O...... (2-10)

The amounts of sedimentary pyrite formed during bacterial sulfate reduction are often limited by the following major factors: (i) the concentrations of metabolizable OM; (ii) the concentration sulfate in the overlying water; and (iii) the availability of reactive Fe (e.g., Berner and Raiswell, 1983; Raiswell et al., 1988; Canfield, 1989a; Canfield and Raiswell,

1991; Canfield et al., 1992; Lyons et al., 1992). Therefore, the abundance of Spy and Corg in sediments are intimately linked and controlled by bacterial sulfate reduction, primary production of OM, and the redox structure of the oceans. 93

Bacterial sulfate reduction and the subsequent pyrite formation would have also been common and important processes in the Archean–Paleoproterozoic oceans (e.g., Ohmoto et al., 1993; Kakegawa and Ohmoto, 1999; Kakegawa et al., 1999). Activity of SRB has been long recognized in ancient sedimentary rocks from its distinct stable isotopic biosignatures of sulfide minerals, and in the gene sequencing of their 16S rRNA (e.g., Canfield and Raiswell, 1999). However, it has been debated intensely what the activity of SRB was like in the distant geologic past and when the oceans became rich in sulfate and bacterial sulfate reduction began. In order to extract information on the activity of SRB and the paleoredox

environments, the Spy isotope compositions, the Spy contents, the Corg-Spy relationships, the

Spy/Corg ratios, and the DOP values are examined in this section.

2-5-4-1. Activity of sulfate-reducing bacteria indicated by S isotope

Previous studies Primarily based on the bulk-rock S isotope analyses of sulfide minerals in Archean–Paleoproterozoic sedimentary rocks, it has been suggested that, before ~2.2 Ga, SRB was absent, oceanic sulfate concentration was low, and pyrite in sediments were produced by magmatic H2S (e.g., Cameron, 1982; Hattori et al., 1985; Cameron and Hattori, 1987; Schidlowski, 1989; Lambert and Donnelly, 1990, 1992). In such

environments, the bulk-rock δ34S values for pyrite and sulfate in Archean–Paleoproterozoic sedimentary rocks were less variable compared to those in younger sediments and sedimentary rocks and fall within a narrow range of 0 ± 2 ‰ (Ohmoto, 1992).

On the other hand, based on the micro-scale analyses of δ34S value for individual pyrite grains in Archean–Paleoproterozoic sedimentary rocks using laser-ablation microprobe technique, Ohmoto et al. (1993), Kakegawa and Ohmoto (1999), and Kakegawa 94

et al. (1999) found the large variations in their δ34S values and suggested that the pyrite grains in those rocks were formed by bacterial sulfate reduction. Further suggestions from these studies are, based on the magnitude of isotopic fractionation factors accompanying the bacterial sulfate reduction, that the sulfate contents in the Archean oceans were higher than ~ 10 mM, about a third of present value (28 mM). The Archean atmosphere and oceans would have been oxygenated enough to allow such a high content of oceanic sulfate.

This study The potential effect of minor grade of metamorphism (lower than the greenschist facies) that the samples of this study have suffered is considered here to be minor, and thus the observed values are expected to represent the original values. The overall range and

34 distribution of the δ Spy values of the samples of this study agree well with those of sedimentary rocks of the Transvaal Group previously reported by Strauss and Beukes (1996). The observed intra-formation variabilities (e.g., -3.1 to +4.0 ‰ for the 3.25 Ga Sheba Formation; +0.2 to +13.4 ‰ for the 2.69 Ga Jeerinah Formation) and the temporal variabilities (see Fig. 2-16) in the bulk-rock δ34S values are consistent with the view that pyrite crystals in the Archean–Paleoproterozoic sedimentary rocks were formed by bacterial

sulfate reduction. Some formations exhibit small ranges in the δ34S values compared to others (e.g., Fig. 2-16-b, -h, -j). Such small ranges (e.g., –0.3 to +0.8 ‰ for the 2.71 Ga Rietgat Formation; Fig. 2-16-h) may be attributed to the small number of analysis (2 to 5 data points) and the homogenization of the small pyrite minerals inevitably associated with bulk-rock analysis. However, it should be noted that the wide variations with different

distribution patterns in the δ34S values between shales (-3.1 to +4.0 ‰) and graywackes (+0.6 to +9.3 ‰) of the 3.25 Ga Sheba Formation (Fig. 2-16-k and -l) were observed even

with the bulk-rock analysis, and that such characteristics of the δ34S values may not be 34 readily explained by a magmatic H2S model. In the magmatic H2S model, the δ S values 95

of sulfide precipitates should show similar distribution, i.e., range and average, among different types (grain sizes) of sedimentary rocks.

Other factors affecting the degrees of the Spy isotope fractionations include the sulfate concentration level, the rate of bacterial sulfate reduction, and open vs. closed system conditions with respect to sulfate availability to sulfate-reducers (e.g., Ohmoto, 1992). The

small degree of the Spy isotope fractionations observed in this study could be due to either a combination of the low sulfate concentration, the rapid sulfate reduction in the Archean–Paleoproterozoic oceans probably warmer than modern oceans due to higher heat

flow and enhanced greenhouse effect of higher pCO2, or the closed system environments with respect to sulfate. However, it is premature to suggest which factor was responsible for the results of this study before more detail studies are performed (e.g., microscale S isotope analyses of individual grains of and/or within-single grain of pyrite by the laser ablation or the ion-microprobe methods).

Secular trend of S isotopic compositions Canfield and Raiswell (1999) compiled data of S isotopic compositions of biogenic sulfide and sulfate throughout geologic time, and present a figure showing secular variations in those δ34S values. We update the compilation of Canfield and Raiswell (1999) with the new data set obtained in this study, and overlaid our δ34S value on Fig. 2-24. This study expands the published δ34S record at ~3.25 Ga and ~2.7 Ga with the newly found more widely variable δ34S values (-3.1 ~ +9.3 ‰ and –0.3 ~ +13.4 ‰, respectively). These positive δ34S values could have been produced by a Rayleigh process in a closed system with respect to sulfate. It may be premature to suggest the δ34S value of oceanic sulfate could have reached +9.3 ‰ at 3.25 Ga and 13~14 ‰ at 2.7 ~ 2.6 Ga, before much more detailed studies of S isotopes, preferably by microscale analysis by laser ablation are performed. 96

2-5-4-2 Sedimentary distribution of S in shales Essentially all the pyrite in modern marine sediments are produced by bacterial sulfate reduction. In modern oceans and most likely in ancient oceans, sulfate is mainly supplied by (1) rivers carrying sulfate produced by oxidative weathering of sulfide minerals and by dissolution of sulfate minerals, (2) atmospheric precipitation (aerosols containing

sulfate including oxidation products of volcanic gases such as SO2 and H2S), and (3)

oxidation, by dissolved O2 of seawater, of dissolved sulfide emanated by hydrothermal fluids and sulfide minerals formed by hydrothermal activity.

The Spy contents of the Archean–Paleoproterozoic shales of this study are presented in Fig. 2-14. The potential effect of minor grade of metamorphism (lower than the greenschist facies) on the S contents of the samples of this study should not be underestimated. Watanabe et al. (1997) assumed that the preserved S contents in the Neoarchean and Paleoproterozoic shales from South Africa represent ~20 % of the original values, from the study of thermal alteration of the Tertiary shales. Loss of pyrite from sediments and sedimentary rocks during diagenesis and metamorphism mostly occurs by high temperature (> 100 ˚C) aqueous reactions (Ohmoto and Rye, 1979; Ohmoto and Goldhaber, 1997). As Watanabe et al. (1997) suggest, the thermal loss of pyrite probably

can explain why the Spy contents of ancient shales have generally lower Spy contents compared to those of modern marine sediments (~ 1 wt.%) (Ohmoto et al., 1990). However, it is not clear if such thermal loss of Spy occurred to the samples of this study, because of a lack of petrographic evidence strongly suggesting the massive (~80 % ?) acid leaching of

Spy. It is therefore reasonable to assume that the original Spy contents of the Archean–Paleoproterozoic shales of this study are between x 1 and x 5 of the measured values. 97

Then we consider the burial flux of S in the Archean–Paleoproterozoic sediments and compare it with that of younger sediments. Figure 2-23 shows the S/Al wt. ratios of the Archean–Paleoproterozoic samples of this study, together with those of the world average shale (Wedepohl, 1991), the Black Sea carbonaceous sediments (Hirst, 1974; Brumsack, 1989), the Cretaceous black shales (Dean and Arthur, 1986), the average Canadian Paleoproterozoic (Aphebian) shales (Cameron and Garrels, 1980), and the average Canadian Archean shales (Cameron and Garrels, 1980). An important observation is that the S/Al ratios of the Archean–Paleoproterozoic samples of this study fall within a range between those of the average shale and the carbonaceous sediments in the Black Sea. It should be noted that the range or the values of the S/Al ratios of the samples from literature shown in Fig. 2-23 are the average of hundreds of samples and therefore likely variable among individual samples (such variation is not shown in Fig. 2-23). The observed ranges in the S/Al ratios of the Archean–Paleoproterozoic samples of this study (Fig. 2-23) are not uncommon. It should be also noted that the characteristics of the S/Al ratios of the Archean–Paleoproterozoic samples of this study do not change even if we apply x 1 ~ x 5 correction for the original S contents in the studied samples. The characteristics of the shale S/Al ratios presented in Fig. 2-23 suggest that the burial flux of S (and the input flux of S as well) in the Archean–Paleoproterozoic sediments is essentially the same as today. The different distributions of S in the coeval shales and graywackes (i.e., relatively fine-grained clastic rocks with higher S contents and coarse-grained clastic rocks with lower S contents) of the 3.25 Ga Sheba Formation is similar to that of modern sediments and Phanerozoic sedimentary rocks. Such distribution is actually the opposite to what is expected if the detrital process was important to distribute heavy Fe sulfide minerals (the heavy minerals accumulate in proximal coarse-grained sediments rather than distal fine- grained sediments). The variations in the S contents within each formation of this study are also typically observed in recent sedimentary rocks. Again, if the detrital process was 98

important to distribute Fe sulfide minerals to sediments, such intra-formational large variations in S contents are not expected. The characteristics of sedimentary S distribution for the Archean–Paleoproterozoic shales of this study shown above are not consistent with the dominance of detrital processes but consistent with the diagenetic enrichment of S by bacterial sulfate reduction. Significance of the Archean–Paleoproterozoic shales with low or no S content should not be underestimated. In the globally sulfidic oceans in the Archean- Paleoproterozoic as suggested by Canfield and Raiswell (1999), shales with high Fe contents with little or no S would be very difficult to form because any excess Fe would

rapidly react with dissolved H2S and form Fe sulfides (the solubilities of FeS and FeS2 are very low). However, many of the samples of this study contain little or no S with Fe comparable to average shales. Iron contents are almost always in excess over S contents. These observation suggest that the diagenetic sulfide formation would have been limited by the availability of sulfide (formed by bacteria) and/or metabolizable OM, not by reactive Fe, as seen in modern deep sea sediments (Canfield and Raiswell, 1999).

2-5-4-3. Relationships of the organic C and S contents in various environments In modern normal marine sediments, i.e., those deposited under oxygenated bottom waters with normal salinity, there is a positive correlation between Corg and Spy (e.g., Berner and Raiswell, 1983, 1984; Morse and Berner, 1995). This relationship indicates that the extent of pyrite formation in normal marine sediments is limited mainly by the concentration of metabolizable OM (e.g., Berner and Raiswell, 1983; Berner, 1984). Linear regression of a plot of Corg-Spy gives a intercept on the S axis of 0 to 0.2-0.3 wt.% Spy. Most recent normal marine sediments fall in an envelope around the best fit regression line with a

Corg/Spy ratio of 2.8 ± 0.2 (or Spy/Corg ratio of ~0.36) (Berner, 1982; Berner and Raiswell, 99

1983). However, the Spy/Corg ratios increase due to OM decomposition / maturation during burial (Raiswell and Berner, 1986).

In the case of sediments under suboxic water column (containing very low O2 and

H2S), linear regression of the Corg-Spy plot gives an intercept at the origin. For sediments

deposited under anoxic and sulfidic water column, i.e., euxinic environment (H2S, no O2),

high concentrations of Spy can be expected in sediments due to precipitation of syngenetic

pyrite formed in the euxinic water column. The linear regression of the Corg-Spy plot shows a weak correlation and a positive intercept on the S axis that is greater than 1 wt.% (Leventhal, 1983; Berner, 1984; Raiswell and Berner, 1985; Lyons and Berner, 1992).

Diversity in the Corg-Spy relationships can be seen in the Fig. 2-20 in which the results of this study are summarized. Watanabe et al. (1997) also have shown the Corg-Spy relationship for Archean–Paleoproterozoic shales from South Africa. The major difference between this study and Watanabe et al. (1997) lies in the choice of samples. As seen in the

Corg contents, Watanabe et al. (1997) mainly used the non-carbonaceous samples. Almost

all the samples used by Watanabe et al. (1997) have Corg and Spy contents less than 2 wt.% and 1 wt.%, respectively, while this study mainly uses carbonaceous shales (black shales) with Corg and Spy contents reaching to 12 wt.% and 4 wt.%, respectively.

The distribution of the Corg and Spy contents displayed in Fig. 2-25 is, importantly, very similar to those of the Black Sea (see Fig. 8 of Watanabe et al., 1997). Some data

points plotted above the "normal marine" line (Berner and Raiswell, 1984) with Spy contents higher than 1 wt.%, as shown in Fig. 2-25-a, possibly represent the formation of sulfide in euxinic water column. Many data points in Fig. 2-25 appear to follow the "normal marine" line although with considerable scatters. Such scatter is commonly observed in the modern marine sediments of many localities (e.g., Morse and Berner, 1995). Many samples of this study appear to be plot below the "normal marine" line (Fig. 2-25). These data points may reflect sulfate-limiting conditions (bacterial sulfate reduction did not proceed despite of 100

available OM) and/or reoxidation of once-formed sulfides by the overlying oxic bottom water (e.g., Morse and Berner, 1995).

The Corg-Spy relationships for the Archean–Paleoproterozoic shales of this study shown in Fig. 2-25 can not be readily explained by the globally sulfidic deep oceans as suggested by Canfield (1998). This is because many of the Archean, Fe-rich samples of this study are poor in S (e.g., the 3.25 Ga Sheba Formation, the 2.96 Ga Parktown Formation, the 2.72 Ga Pillingini Tuff Formation, the >2.60 Ga Marra Mamba Iron Formation, and the 2.60 Wittenoom Dolomite Formation; see Fig. 2-13). In the inferred sulfidic Archean oceans, reactive Fe should have been the primary limiting factor for the Fe-sulfide formation, because dissolved sulfide and dissolved Fe2+ do not coexist from a thermodynamic reason and thus the solubility of Fe-sulfide in sulfidic water is extremely small. The results of this study suggest an importance of the biological and therefore redox

controls on the variable sedimentary distributions of the Corg and Spy in the Archean–

Paleoproterozoic oceans. The abundance of Corg and Spy in the Archean–Paleoproterozoic marine sediments was probably controlled by complex biological / redox processes similar to those in the modern oceans.

2-5-4-4. Paleoredox environments inferred from the pyrite S to organic C ratios and the degree of pyritization (DOP)

The usefulness of the Spy/Corg ratios to distinguish marine vs. freshwater sediments and of the DOP values as a paleoredox indicator have been demonstrated by numerous studies of modern sediments and Phanerozoic sedimentary rocks (e.g., Berner and Raiswell, 1983, 1984; Raiswell and Berner, 1985, 1986; Raiswell et al., 1988; Raiswell and Al-Biatty,

1989; Johns and Manning, 1994). It is therefore tempting to use the Spy/Corg ratios and DOP values in Archean-Paleoproterozoic sedimentary rocks to extract information on the 101

sulfate concentration in the distant past and to constrain the extent of bacterial sulfate reduction (e.g., Donnelly and Crick, 1988, 1992; Donnelly and Jackson, 1988; Carrigan and Cameron, 1991; Jackson and Raiswell, 1991; Raiswell and Al-Biatty, 1992; Strauss and Beukes, 1996; Watanabe et al., 1997; Lyons et al., 2000).

However, there are some limitations in the use of Spy/Corg ratios for the ancient sedimentary rocks to extract information on the S cycling, as pointed out by many previous investigators (e.g., Raiswell and Berner, 1986; Morse and Emeis, 1990; Calvert and Karlin, 1991; Lyons and Berner, 1992; Leventhal, 1995; Morse and Berner, 1995; Canfield and

Raiswell, 1999). It is claimed that the ancient sedimentary rocks with modern Spy/Corg ratios do not necessarily suggest the high rate of sulfate reduction comparable to modern

because the typical Spy/Corg ratios of modern marine sediments (~0.36) can be generated with much more reduced rates of sulfate reduction if there is no S loss (Canfield and Raiswell, 1999). Despite the above limitations, many researchers including Raiswell have worked on

the Paleozoic and Proterozoic shales for their Spy/Corg ratios (e.g., Raiswell and Al-Biatty, 1989, 1992; Hieshima and Pratt, 1991; Jackson and Raiswell, 1991; Donnelly and Crick, 1992; Meyer and Robb, 1996) and meaningful results have been reported. Furthermore, a recent addition of Lyons et al. (2000) and Kah et al. (2001) has expanded the Precambrian

record of the Corg-Spy systematics. This study further expands the pre-existing

Precambrian Corg-Spy record. It is noted (in section 3-4-7) that the distribution of Spy/Corg ratios of the Archean–Paleoproterozoic shales of this study are quite variable and the

Spy/Corg ratios for some samples are more or less close to 0.36, a mean Spy/Corg ratio representing the normal marine sediments overlain by oxic seawater. Such characteristics observed in the samples of this study are commonly found not only in many modern marine sediments and Phanerozoic shales but also in the other Proterozoic and Archean shales, which may suggest the similar depositional environments with respect to redox condition 102 and activity of SRB. However, careful approach is needed to interpret the Spy/Corg ratios bearing the above-mentioned possible limitations in mind. We turn to the discussion of DOP values in the following because DOP values may be much more useful over the

Spy/Corg ratios in recognizing different degrees of bottom water oxygenation (Raiswell et al., 1988). This is because DOP and bottom water oxygenation are linked to variations in

3+ the intensity of exposure of detrital reactive Fe -bearing minerals to bacteriogenic H2S, which in turn affect the efficiency with which these are converted to pyrite (Raiswell and Al- Biatty, 1989). It should be noted that Arthur and Sageman (1994) question the usefulness of DOP as an indicator of bottom water oxygenation. Generally, DOP values of less than ~0.1 indicate non-marine environments (lack of sulfate) or lack of the anoxic conditions that are necessary for the SRB to function (e.g.,

Raiswell et al., 1988); anoxic conditions are not generally present if the Corg contents are below ~0.2 %. DOP values of ~0.2 to ~0.5 are typical of normal marine conditions; that is, oxic water overlies anoxic sediments, and sulfate reduction is possible a few mm to a few cm below the sediment-water interface (e.g., Raiswell et al., 1988). Samples with DOP values of ~0.5 to ~0.7 were probably deposited below anoxic water column conditions (e.g.,

Raiswell et al., 1988). DOP values greater than about ~0.7 indicate a euxinic and H2S- containing (sulfidic) water column (e.g., Raiswell et al., 1988). Despite its potential usefulness in the recognition of the degree of bottom water oxygenation in the ancient oceans, DOP has been rarely applied to the Archean black shales. This paper provides the first extensive data set of DOP for the Archean- Paleoproterozoic shales. The DOP values of this study are variable within each formation (section 3-4-4), and overlap with every range of the DOP values (< ~0.1, ~0.2 to ~0.5, ~0.5 to ~0.7, and > ~0.7) for the various redox environments mentioned above (Fig. 2-9). No major difference exists in the characteristics of the distribution of the DOP values among modern sediments, Phanerozoic shales, Proterozoic shales, and Archean shales. This 103

similarity may suggest that the depositional environments with respect to redox conditions were already diverse and fluctuating since at least the Mesoarchean, 3.25 Ga ago (represented by the Sheba Formation of this study). This further suggests the early development of present-day style oceanic redox structure (i.e., globally oxic oceans with

local O2-deficient environments such as mid-depth O2-minimum zones and anoxic basins) and the activity of SRB supported by sufficient oceanic sulfate concentration in the Archean. Although this study nearly doubles the number of the pre-existing DOP data for Archean shales, much more Archean DOP data should be accumulated from sediments of various depositional settings and ages, and the use of DOP as an indicator of bottom water oxygenation should be further tested.

2-6. Significance and implications

From the Corg-Spy-Fe-P systematics and the Corg-Spy isotopic compositions of the Archean–Paleoproterozoic samples of this study, we suggest that the Archean–Paleoproterozoic environments were essentially the same as today (i.e., globally oxic atmosphere and oceans with local anoxic environments) and that the diversity and activity of microorganisms involving photosynthesizers, methanogens, and sulfate reducers were important. These suggestions imply an early redox evolution of atmosphere and oceans and the early evolution of complex biosphere at least 3.25 Ga ago. 104

2-7. Conclusion

From the studies of Corg-Spy-Fe-P systematics and C-S stable isotope geochemistry of the Archean–Paleoproterozoic shales of this study, we reached the following conclusions.

(1) The observed variabilities in the Fe3+/∑Fe ratios within each formation suggest the redox control on the Fe geochemistry. General increases in the Fe3+/∑Fe ratios with minor decreases when compared to the those of their source rocks suggest a complex history of Fe geochemistry involving the efficient oxidative weathering on the continent followed by diagenetic / metamorphic reduction of Fe3+ to Fe2+.

(2) The preserved OM in the Archean–Paleoproterozoic shales of this study represents the remnant of organisms, possibly (oxygenic?) photosynthesizers, that lived in the Archean– Paleoproterozoic oceans and sediments.

(3) The observed Corg contents of the Archean–Paleoproterozoic shales of this study are explained by aerobic recycling of OM during and after the sedimentation in the Archean–Paleoproterozoic oceans.

13 (4) The δ Corg values of the Archean–Paleoproterozoic shales of this study suggest the activity of photosynthesizers and chemosynthesizers (methanogens and methanotroph) in sediments and/or overlying water column of the Archean–Paleoproterozoic oceans. The

13 overall negative correlation between the Corg contents and the δ Corg values of the Archean–Paleoproterozoic shales of this study suggests an importance of the redox control

on the development of CH4-metabolizing microorganisms. At least two contrasting environments existed in the Archean–Paleoproterozoic oceans; one is with organisms 105

13 producing OM with the δ Corg values of -25‰, probably cyanobacteria in an oxygenated

water column resulting low Corg contents in sediments, and the other is with organisms 13 producing OM with the δ Corg values of -45 ‰, probably methanogens and methanotrophs in reducing water column / sediments resulting in enhanced preservation of

Corg in sediments.

(5) The variabilities in the Corg/P ratios of the Archean–Paleoproterozoic shales of this study which fall within the two end members (the average shale and the modern-ancient anoxic sediments) suggest the redox conditions played an important role in controlling the

Corg/P ratios in the Archean–Paleoproterozoic oceans.

(6) The variable Corg and Spy contents and their relationships in each formation of the Archean–Paleoproterozoic shales of this study were primarily controlled by microbiological activity involving sulfate reducing bacteria.

(7) The δ34S values of the Archean–Paleoproterozoic shales of this study suggest an activity of sulfate reducing bacteria in sediments and/or water columns of the Archean– Paleoproterozoic oceans. This further implies the availability to sulfate reducers of sufficient concentration of oceanic sulfate, metabolizable OM, and reactive Fe(3+) for the operation of bacterial sulfate reduction.

(8) The Archean–Paleoproterozoic surface environments were such that the atmosphere was oxic; the oceans were globally oxic with local anoxic environments. This suggestion of early development of an oxygenated environment in the Archean is supported by the recent modeling results of Lasaga and Ohmoto (2002a) for oxygen geochemical cycles. 106

References

Aplin, A.C. and Macquaker, H.S. (1993) C-S-Fe geochemistry of some modern and ancient anoxic marine and mudstones. Phil. Trans. R. Soc. Lond. A 344, 89-100.

Armstrong, R.A., Compston, W., de Wit, M.J., and Williams, J.S. (1990) The stratigraphy of the 3.5-3.2 Ga Barberton Greenstone Belt revisited: a single zircon microprobe study. Earth Planet Sci. Lett. 101, 90-106.

Arndt, N.T., Nelson, D.R., Compston, W., Trendall, A.F. and Thorne, A.M. (1991) The age of the Fortescue Group, Hamersley Basin, Western Australia, from iron microprobe zircon U-Pb results. Aust. J. Earth Sci. 38, 261-281.

Arthur, M.A. and Sageman, B.B. (1994) Marine black shales: Depositional mechanism and environments of ancient deposits. Ann. Rev. Earth Planet. Sci. 22, 499-551.

Awramik, S.M., Schopf, J.W. and Walter, M.R. (1988) Carbonaceous filaments from North Pole, Western Australia: Are they fossil bacteria in Archaean stromatolites? A discussion. Precam. Res. 39, 303-309.

Bein, A., Almogli-Labin, A., Sass, E. (1990) Sulfur sinks and organic carbon relationships in Cretaceous organic-rich carbonates: Implications for evaluation of oxygen-poor depositional environments. Am. J. Sci. 290, 882-91.

Berner, R.A. (1970) Sedimentary pyrite formation. Am. J. Sci. 268, 2-23.

Berner, R.A. (1982) Burial of organic carbon and pyrite sulfur in the modern ocean: Its geochemical and environmental significance. Am. J. Sci. 282, 451-473.

Berner, R.A. (1984) Sedimentary pyrite formation: An update. Geochim. Cosmochim. Acta 48, 605-615.

Berner, R.A. (1985) Sulfate reduction, organic matter decomposition and pyrite formation. Philos. Trans. Royal Soc. London Ser. A315, 25-38.

Berner, R.A. and Raiswell, R. (1983) Burial of organic carbon and pyrite sulfur in sediments over Phanerozoic time: a new theory. Geochim. Cosmochim. Acta 47, 855-862.

Berner, R.A. and Raiswell, R. (1984) C/S method for distinguishing freshwater from marine sedimentary rocks. Geology 12, 365-368. 107 Bidigare, R.R., Flugge, A., Freeman, K.H., Hanson, K.L., Hayes, J.M., Hollander, D., Jasper, J.P., King, L.L., Laws, E.A., Milder, J., Millero, F.J., Pancost, R.P., Popp, B.N., Steinberg, P.A., and Wakeham, S.G. (1997) Consistent fractionation of 13C in nature and in the laboratory: growth-rate effects in some haptophyte algae. Global Biogeochem. Cycles 11, 279-292.

Bottomley, D.J., Veizer, J., Nielsen, H., and Moczydlowska, M. (1992) Isotopic composition of disseminated sulfur in Precambrian sedimentary rocks. Geochim. Cosmochim. Acta 56, 3311-3322.

Braiser, M.D., Green, O.R., Jephcoat, A.P., Kleppe, A.K., Van Kranendonk, M.J., Lindsay, J.F., Steele, A., and Grassineau, N.V. (2002) Questioning the evidence for Earth’s oldest fossils. Nature 416, 76-81.

Brocks, J.J., Logan, G.A., Buick, R., and Summons, R.E. (1999) Archean Molecular Fossils and the Early Rise of Eukaryotes. Science 285, 1033-1036.

Broecker, W.S. and Peng, T.H. (1982) Traces in the Sea. Lamont-Doherty Geological Observatory, Palisades, 690 pp.

Burke, K., Kidd, W.S.F., and Kushy, T.M. (1985) Is the Ventersdorp rift system of southern Africa related to a between the Kaapvaal and at 2.64 Ga ago? Tectonophysics 115, 1-24.

Burke, K., Kidd, W.S.F., and Kusky, T.M. (1986) Archean tectonics of the Witwatersrand, South Africa. Tectonics 5, 436-456.

Button, A. (1981a) The Ventersdorp Supergroup. In Precambrian of the Southern Hemisphere (ed. D. R. Hunter), Elsevier, Amsterdam, Netherlands, 520-527.

Button, A. (1981b) The Transvaal Supergroup. In Precambrian of the Southern Hemisphere (ed. D. R. Hunter), Elsevier, Amsterdam, Netherlands, 527-536.

Button, A. (1986) The Transvaal sub-basin of the Bushveld floor in the eastern Transvaal. Trans. Geol. Soc. S. Afr. 79, 3-12.

Calvert, S.E. and Karlin, R.E. (1991) Relationships between sulphur, organic carbon, and iron in the modern sediments of the Black Sea. Geochim. Cosmochim. Acta 55, 2483-2490.

Calvert, S.E. and Pedersen, T.F. (1993) Geochemistry of recent oxic and anoxic marine sediments: Implications for the geological record. Mar. Geol. 113, 67-88. 108 Calvert, S.E., and Price, N.B. (1983) Geochemistry of Namibian shelf sediments. In Coastal upwelling: its sediment record, part A (eds., J. Thiede and E. Suess), Plenum press, New York, 337-375.

Calvert, S.E., Bustin, R.M., and Ingall, E.D. (1996) Influence of water column anoxia and sediment supply on the burial and preservation of organic carbon on marine shales. Geochim. Cosmochim. Acta 60, 1577-1593.

Cameron, E.M. (1982) Sulphate and sulphate reduction in early Precambrian ocean. Nature 296, 145-148.

Cameron, E.M. and Garrels, R.M. (1980) Geochemical compositions of some Precambrian shales from the Canadian Shield. Chem. Geol. 28, 181-197.

Cameron, E.M. and Hattori, K. (1987) Archean sulphur cycle: Evidence from sulphate minerals and isotopically fractionated sulphides in Superior province, Canada. Chem. Geol. 65, 341-358.

Canfield, D.E. (1989a) Reactive iron in marine sediments. Geochim Cosmochim Acta 53, 619-632.

Canfield, D.E. (1989b) Sulfate reduction and oxic respiration in marine sediments: Implications for organic carbon preservation in euxinic sediments. Deep-Sea Res. 36, 121-138.

Canfield, D.E. (1994) Factors influencing organic carbon preservation in marine sediments. Chem. Geol. 114, 315-329.

Canfield, D.E. (1998) A new model for Proterozoic ocean chemistry. Nature 396, 450-453.

Canfield, D.E. and Raiswell, R. (1999) The evolution of the sulfur cycle. Am. J. Sci. 299, 697-723.

Canfield, D.E., Habicht, K.S., and Thamdrup, B. (2000) The Archean sulfur cycle and the early history of atmospheric oxygen. Science 288, 658-661.

Canfield, D.E., Raiswell, R., and Bottrell, S. (1992) The reactivity of sedimentary iron minerals toward sulfide. Am. J. Sci. 292, 659-683.

Cledenin, C.W., Charlesworth, E.G., and Maske, S. (1988) Tectonic style and mechanism of early Proterozoic successor basin development, southern Africa. Tectonophysics 156, 275-291. 109 Condie, K.C. (1993) Chemical compositions and evolution of the upper continental crusts: Contrasting results from surface samples and shales. Chem. Geol. 104, 1-37.

Coward, M.P., Spencer, R.M., and Spooner, C.E. (1995) Development of the Witwatersrand Basin, South Africa. In Early Precambrian Processes (eds., M.P. Coward and A.C. Reis), Geol. Soc. Spec. Pub. 95, 243-269. de Ronde, C.E. and Ebbesen, T.W. (1996) 3.2 b.y. of organic compound formation near sea-floor hot springs. Geology 24, 791-794. de Ronde, C.E.J., Spooner, E.T.E., and de Wit, M.J. (1991a) Geologic and carbon isotopic evidence for pelagic, photosynthetic marine biota at ~3.23 to 3.47 × 109 yr, Barberton greenstone belt, South Africa. In Frontiers of life: Proceedings (eds. J. Tran Thanh Van et al.), 3rd Recontres of Blois, Blois, France, October 14-19, 1991. Paris, Edintons Frontieres, 435-436. de Vries, T.J. and Pearch, W.G. (1982) Fish debris in sediments of the upwelling zone off central Peru: a late record. Deep Sea Res. 28, 87-109. de Wit, M.J., Roering, C., Hart, R.J., Armstrong, R.A., de Ronde, C.E.J., Green, R.W.E., Tredoux, M., Peberdy, E. and Hart, R.A. (1992) Formation of an Archean continent. Nature 357, 553-562.

Dean, W.E., Arthur, M.A. (1989) Iron-sulfur-carbon relationships in organic-carbon-rich sequences I: Cretaceous Western Interior Seaway. Am. J. Sci. 289, 708-743.

Demaison, G.J. and Moore, G.T. (1980) Anoxic environments and oil source bed genesis. Org. Geochem. 2, 9-31.

Des Marais, D.J. (1985) Carbon exchange between the mantle and crust and its effect upon the atmosphere, today compared to Archean time. In The Carbon Cycle and Atmospheric CO2: Natural Variations Archean to Present (eds. E.T. Sundquist and W.S. Broecker), Amer. Geophys. Union, Washington DC, pp 602-611.

Des Marais, D.J., Strauss, H., Summons, R.E., and Hayes, J.M. (1992) Carbon isotope evidence for the stepwise oxidation of the Proterozoic environment. Nature 359, 605-609.

Des Marais, D.J. (1994) Tectonic control of the crustal organic carbon reservoir during the Precambrian. Chem. Geol. 114, 303-314.

Dimroth, E. and Kimberley, M.M. (1976) Precambrian atmospheric oxygen: evidence in the sedimentary distributions of carbon, sulfur, uranium, and iron. Can. J. Earth Sci. 13, 1161-1185. 110 Farquhar, G.D., O'Leary, M.H., and Berry, J.A. (1982) On the relationship between carbon isotope discrimination and the intercellular CO2 concentration in leaves. Aust. J. Plant Physiol. 9, 121-137.

Filippelli, G.M. and Delaney, M.L. (1994) The oceanic phosphorus cycle and continental weathering during the Neogene. Paleoceanography 9, 643-652.

Filippelli, G.M. and Delaney, M.L. (1996) Phosphorus geochemistry of equatorial Pacific sediments. Geochim. Cosmochim. Acta 60, 1479-1495.

Froelich, P.N., Bender, M.L., Luedtke, N.A., Heath, G.R., and de Vries, T. (1982) The marine phosphorus cycle. Am. J. Sci. 282, 474-511.

Froelich, P.N. (1988) Kinetic control of dissolved phosphate in natural rivers and estuaries: A primer on the phosphate buffer mechanism. Limnol. Oceanogr. 33, 649-668.

Garrels, R.M. and Mackenzie, F.T. (1971) Evolution of Sedimentary Rocks. W. W. Norton, New York, 397 pp.

Garrels, R.M. and Perry, E.A. Jr. (1974) Cycling of carbon, sulfur, and oxygen through geologic time. In The Sea (ed., E.D. Goldberg), Wiley, New York, 303-336.

Geological Survey of Western Australia (1990) Geology and mineral resources of Western Australia. W. Aus. Geol. Surv., Memoir 3, 827p.

Goericke, R. and Fry, B. (1994) Variations of marine plankton δ13C with latitude, temperature, and dissolved CO2 in the world ocean. Global Biogeochem. Cycles 8, 85-90.

Goldhaber, M.B. and Kaplan, I.R. (1974) The Sulfur Cycle, in The Sea, vol. 5 (ed. by Goldberg, E.D.) A Wiley-International Publication, New York., 569-655.

Hattori, K., Campbell, F.A., and Krouse, H.R. (1985) Sulfur isotope abundances in sedimentary rocks, relevance to the evolution of the Precambrian atmosphere. Geochem. Int. 22, 97-114.

Hayes, J.M. (1994) Global methanotrophy at the Archean-Proterozoic transition. In Early Life on Earth. Nobel Symposium No. 84 (ed. S. Bengston).pp. 220-236. Columbia Univ. Press, New York.

Hayes, J.M., Kaplan I.R., and Wedeking K.M. (1983) Precambrian organic geochemistry, Preservation of the record. In Earth's Earliest Biosphere: Its Origin and Evolution (ed. J.W. Schopf), pp. 93-135. Princeton Univ. Press, Princeton. 111 Heimann, M. and Maier-Reimer, E. (1996) On the relationship between the oceanic uptake of CO2 and its carbon isotopes. Global Biogeochem. Cycles 10, 89-110.

Heubeck, C. and Lowe, D.R.(1994) Depositional and tectonic setting of the Archean Moodies Group, Barberton Greenstone Belt, South Africa. Precam. Res. 68, 257- 290.

Hickman, A.H. (1983) Geology of the Pilbara Block and Its Environs. Geol. Surv. W. Aus. Bull. 127. 268p.

Hieshima, G.B. and Pratt, L.M. (1991) Sulfur/carbon ratios and extractable organic matter of the Middle Proterozoic Nonesuch Formation, North American Midcontinent rift. Precam. Res. 54, 65-79.

Hirst, D. (1974) Geochemistry of sediments from Eleven Black Sea cores. In The Black Sea: Geology, Chemistry, and Biology (eds. E.T. Degens and D.A. Ross), Am. Assoc. Petrol. Geol. Memoir 20, 430-455.

Holland, H.D. (1978) The Chemistry of the Atmosphere and the Oceans. John Willey and Sons, New York.

Holland, H.D. (1984) Chemical evolution of the atmosphere and ocean. Princeton University Press, Princeton, 582p.

Holland, H.D. (1994) Early Proterozoic atmospheric change. In Early Life on Earth. Nobel Symposium No. 84 (ed. S.Bengston) Columbia University Press, New York, 237- 244.

Holland, H.D. (1999) When did the Earth's atmosphere become oxic? A Reply. Geochemical News 100, 20-22.

Ingall, E.D. and Jahnke, R. (1994) Evidence for enhanced phosphorus regeneration from marine sediments overlain by oxygen depleted waters. Geochim. Cosmochim. Acta 58, 2571-2575.

Ingall, E.D. and Jahnke, R. (1997) Influence of water column anoxia on the elemental fractionation of carbon and phosphorus during sediment diagenesis. Mar. Geol. 139, 219-229.

Ingall, E.D., Bustin, R.M., and Van Cappellen, P. (1993) Influence of water column anoxia on the burial and preservation of carbon and phosphorus in marine sediments. Geochim. Cosmochim. Acta 57, 303-316. 112 Jackson, M.J. and Raiswell, R. (1991) Sedimentology and carbon-sulphur geochemistry of the Velkerri Formation, a mid-Proterozoic potential oil source in northern Australia. Precam. Res. 54, 81-108.

Jackson, M.P.A., Eriksson, K.A., and Harris, C.W. (1987) Early Archean foredeep sedimentation related to crustal shortening: a reinterpretation of the Barberton Sequence, southern Africa. Tectonophysics 136, 360-366.

Jahn, B.M., Bertrand-Sarfati, J., and Macé, N.M.J. (1990) Direct dating of stromatolitic carbonates from the Schmidtsdrif Formation (Transvaal Dolomite), South Africa, with implications on the age of the Ventersdorp Supergroup. Geology 18, 1211- 1214.

Jasper, J.P., Hayes, J.M., Mix, A.C., and Prahl, F.G. (1994) Photosynthetic fractionation of 13 C and concentration of dissolved CO2 in the central equatorial Pacific during the last 255,000 years. Paleoceanography 9, 781-798.

Jones, B. and Manning, D.A.C. (1994) Comparison of geochemical indices used for the interpretation of paleoredox conditions in ancient mudstones. Chem. Geol. 111, 111-129.

Kakegawa, T. and Ohmoto, H. (1999) Sulfur isotope evidence for the origin of 3.4 to 3.1 Ga pyrite at the Princeton gold mine, Barberton Greenstone Belt, South Africa. Precam. Res. 96, 209-224.

Kakegawa, T., Kawai, H., and Ohmoto, H. (1999) Origin of pyrites in ~2.5 Ga Mt. McRae shale, the Hamersley district, Western Australia. Geochim. Cosmochim. Acta 62, 3205-3220.

Kamo, S.L. and Davies, D.W. (1994) Reassessment of Archean crustal development in the Barberton Mountain Land, South Africa, based on U-Pb dating. Tectonics 13, 167- 192.

Karhu, J.A. and Holland, H.D. (1996) Carbon isotopes and the rise of atmospheric oxygen. Geology 24, 867-870.

Kasting, J.F. and Walker, J.C.G. (1981) Limits on oxygen concentration in the prebiological atmosphere and the rate of abiotic fixation of nitrogen. J. Geophys. Res. 86, 1147-1158.

Kasting, J.F. (1987) Theoretical constraints on oxygen and carbon dioxide concentrations in the Precambrian atmosphere. Precam. Res. 34, 205-228. 113 Kasting, J.F. (1991) Box Models for the evolution of atmospheric oxygen: an update. Paleogeogr., Paleoclimatol., Paleoecol. (Global Planet. Change Sect.) 97, 125-131.

Kasting, J.F. (1992) Proterozoic climates: the effect of changing atmospheric carbon dioxide concentrations. In Proterozoic Biosphere: A Multidisciprinary Study (eds. J.W. Schopf and C. Klein), pp. 165-168. Cambridge Univ. Press, Cambridge.

Kasting, J.F. (1993) Earth's early atmosphere. Science 259, 920-926.

Kasting, J.F., Holland, H.D., and Kump, L.R. (1992) Atmospheric evolution: the rise of oxygen. In The Proterozoic Biosphere: A Multidisciplinary Study. (eds., J.W. Schopf and C. Klein ), Cambridge University Press, Cambridge, England, 159-164.

Kennedy, M.J., Pevear, D.R., and Hill, R.J. (2002) Mineral surface control of organic carbon in black shale. Science 295, 657-660.

Kröner, A., Byerly, G.R., and Lowe, D.R. (1991) Chronology of Early Archean granite- greenstone evolution in the Barberton Mountain Land, South Africa, based on precise dating by single zircon evaporation. Earth Planet. Sci. Lett. 103, 41-54.

Kump, L.R. and Holland, H.D. (1992) Iron in Precambrian rocks: Implications for the global oxygen budget of the ancient Earth. Geochim. Cosmochim. Acta 56, 3217- 3223.

Lambert, I.B. and Donnelly, T.H. (1990) The paleoenvironmental significance of trends in sulphur isotope compositions in the Precambrian. In Stable isotopes and fluid processes in mineralization(eds., H.K. Herbert and S.E. Ho), Univ. W. Australia Spec. Publ. 23, 260-268.

Lambert, I.B. and Donnelly, T.H. (1992) Global oxidation and a supercontinent in the Proterozoic: Evidence from stable isotopic trends. In Early Organic Evolution: Implications for Mineral and Energy Resources (eds., M. Schidlowski, S. Golubic, M.M. Kimberley, D.M. McKirdy, and P.A. Trudinger), Springer-Verlag, 408-414.

Lasaga, A.C. and Ohmoto, H. (2002a) The oxygen geochemical cycle: dynamics and stability. Geochim. Cosmochim. Acta 66, 361-381.

Lasaga, A.C. and Ohmoto, H. (2002b) Long term evolution of atmospheric oxygen and carbon dioxide. Am. J. Sci., in review.

Leventhal, J.S. (1983) An interpretation of carbon and sulfur relationships in Black Sea sediments as indicators of environments of deposition. Geochim. Cosmochim. Acta 47, 133-137. 114 Lewan, M.D. (1983) Effects of thermal maturation of stable organic carbon isotopes as determined by hydrous pyrolysis of Woodford Shale. Geochim. Cosmochim. Acta 47, 1471-1479.

Libes, S.M. (1992) An introduction to marine biogeochemistry. John Wiley and Sons, Inc. New York. 734p.

Lowe, D.R. and Byerly, G.R. (1999) Geologic Evolution of the Barberton Greenstone Belt, South Africa. Geol. Soc. Am. Spec. Paper 329. 319p.

Lyons, T.W., and Berner, R.A. (1992) Carbon-sulfur-iron systematics of the uppermost deep-water sediments of the Black Sea. Chem. Geol. 99, 1-27.

Lyons, T.W., Luepke, J.J., Schreiber, M.E., and Zieg, G.A. (2000) Sulfur geochemical constraints on Mesoproterozoic restricted marine deposition: lower Belt Supergroup, northwestern United States. Geochim. Cosmochim. Acta 64, 427-437.

Melezhik, V.A., Fallick, A.E., Medvedev, P.V., and Makarikhin, V.V. (1999) Extreme 13 Ccarb enrichment in ca. 2.0 Ga magnesite-stromatolite-dolomite-'red beds' association in a global context: a case for the world-wide signal enhanced by a local environment. Earth Sci. Rev. 48, 71-120.

Meyer, F.M. and Robb, L.J. (1996) The geochemistry of black shales from the Chuniespoort Group, Transvaal Sequence, Eastern Transvaal, South Africa. Econ. Geol. 91, 111-121.

Mojzsis, S.J., Arrhenius, G., McKeegan, K., Harrison, T.M., Nutman, A.P., and Friend, C.R.L. (1996) Evidence for life on Earth before 3,800 million years ago. Nature 384, 55-59.

Morse, J.W. and Berner, R.A. (1995) What determines sedimentary C/S ratios? Geochim. Cosmochim. Acta 59, 1073-1077.

O'Leary, M.H. (1981) Carbon isotope fractionation in plants. Phytochemistry 20, 553-567.

Ohmoto, H. (1992) Biogeochemistry of sulfur and mechanisms of sulfide-sulfate mineralization in Archean oceans. In Early Organic Evolution: Implications for Mineral and Energy Resources. (eds., M. Schidlowski, S. Golubic, M.M. Kimberley, D.M. McKirdy, and P.A. Trudinger), Springer-Verlag, New York, 378- 397.

Ohmoto, H. (1996) Evidence in pre-2.2 Ga paleosols for the early evolution of atmospheric oxygen and terrestrial biota. Geology 24, 1135-1138. 115 Ohmoto, H. (1997) When did the Earth's atmosphere become oxic? The Geochemical News 93, 12-13 and 26-27. Ohmoto, H. and Kerrick, D. (1977) Devolatilization equilibria in graphitic systems. Am. J. Sci. 277, 1013-1044. Ohmoto, H. and Rye, R.O. (1979) Isotopes of sulfur and carbon. In Geochemistry of Hydrothermal Ore Deposits (Barnes, H.L. ed.), 2nd Ed., New York, Wiley-Intersci., 509-567 Ohmoto, H. and Goldhaber, M. (1997) Applications of sulfur and carbon isotopes in ore deposit research. In Geochemistry of Hydrothermal Ore Deposits, Third ed. (ed. Barnes, H. L.), John Wiley and Sons, New York, 517-611.

Ohmoto, H., Kaiser, C.J., and Geer, K.A. (1990) Systematics of sulphur isotopes in recent marine sediments and ancient sediment-hosted base metal deposits. In Stable Isotopes and Fluid Processes in Mineralization, Herbert, H.K. and Ho, S.E. eds., Perth. Univ. West. Aust. Publication 23, 70-120.

Ohmoto, H., Kakegawa, T., and Lowe, D.R. (1993) 3.4-billion-year-old biogenic pyrites from Barberton, South Africa: sulfur isotope evidence. Science 262, 555-557.

Ohmoto, H. et al. (in prep) Precambrian atmospheric oxygen. A review. To be submitted to Ann. Rev. Earth Planet Sci.

Park, R. and Epstein, S. (1960) Carbon isotope fractionation during photosynthesis. Geochim. Cosmochim. Acta 21, 110-126.

Pavlov, A.A., Kasting, J.F., Eigenbrode, J.L., and Freeman, K.H. (2001) Organic haze in Earth's early atmosphere: source of low-13C Late Archean kerogens? Geology 29, 1003-1006.

Pedersen, T.F. and Calvert, S.E. (1990) Anoxic vs. Productivity: What controls the formation of organic-carbon-rich sediments and sedimentary rocks? Am. Assoc. Petrol. Bull. 74, 454-466.

Phillips, G.N., Myers, R.E., Law, J.D.M., Bailey, A.C., Cadle, A.B., Beneke, S.D., and Giusti, L. (1989) The Witwatersrand gold fields: Part I. Postdepositional history, synsedimentationary processes, and gold distribution. Econ. Geol. Monogr. 6, 585- 597.

Pretorius, D. A. (1981) The Witwatersrand Supergroup. In Precambrian of the Southern Hemisphere (ed. D.R. Hunter), Elsevier, Amsterdam, Netherlands, 511-520.

Raiswell, R. and Al-Biatty, H.J. (1989) Depositional and diagenetic C-S-Fe signatures in early Paleozoic normal marine shales. Geochim. Cosmochim. Acta 53, 1147-1152. 116 Raiswell, R. and Al-Biatty, H.J. (1992) Depositional and diagenetic C-S-Fe signatures and the potential of shales to generate metal-rich fluids. In Early Organic Evolution: Implications for Mineral and Energy Resources. (eds., M. Schidlowski, S. Golubic, M.M. Kimberley, D.M. McKirdy, and P.A. Trudinger), Springer-Verlag, New York, 415-425.

Raiswell, R. and Berner, R.A. (1985) Pyrite formation in euxinic and semi-euxinic sediments. Am. J. Sci. 285, 710-724.

Raiswell, R. and Berner, R.A. (1986) Pyrite and organic matter in Phanerozoic normal marine shales. Geochim. Cosmochim. Acta 50, 1967-1976.

Raiswell, R. and Canfield, D.E. (1998) Source of iron for pyrite formation in marine sediments. Am. J. Sci. 298, 219-245.

Raiswell, R., Buckley, F., Berner, R.A., Anderson, T.F. (1988) Degree of pyritization of iron as a paleoenvironmental indicator of bottom-water oxygenation. J. Sed. Petrol. 58, 812-819.

Raiswell, R., Canfield, D.E., and Berner, R.A. (1994) A comparison of iron extraction methods for the determination of degree of pyritisation and the recognition of iron- limited pyrite formation. Chem. Geol. 111, 101-110.

Rau, G.H., Takahashi, T., and Des Marais, D.J. (1989) Latitudinal variations in plankton 13 δ C: Implications for CO2 and productivity in past oceans. Nature 341, 516-518.

Rau, G.H., Froelich, P.N., Takahashi, T., and Des Marais, D.J. (1991) Does sedimentary 13 organic δ C record variations in Quaternary ocean [CO2aq]? Paleoceanography 6, 335-347.

Rau, G.H., Takahashi T., Des Marais, D.J., Repeta, D.J., and Martin, J.H. (1992) The 13 relationship between organic matter δ C and [CO2aq] in ocean surface waters: Data from a JGOFS site in the northeast and a model. Geochim. Cosmochim. Acta 56, 1413-1419.

Rau, G.H., Riebesell, U., and Wolf-Gladrow, D. (1997) CO2aq-dependent photosynthetic fractionation in the ocean: A model versus measurements. Global Biogeochem. Cycles 11, 267-278.

Raven, J.A., Johnston, A.M., and Turpin, D.H. (1993) Influence of changes in CO2 concentration and temperature on marine phytoplankton 13C/12C ratios: An analysis of possible mechanisms. Global Planet. Change 8, 1-12. 117 Redfield, A.C., Ketchum, B.H., and Richards, F.A. (1963) The influence of organisms on the composition of sea water. In The Sea (ed. M.N.Hill), Wiley, New York, 26-77.

Reimer, T.O. (1992) Carbonaceous high-alumina shale in the Transvaal Supergroup: Evidence of Early Proterozoic Karstic weathering in a marine environments. In Early Organic Evolution: Implications for Mineral and Energy Resources (eds., M. Schidlowski, S. Golubic, M. M. Kimberley, D. M. McKirdy, and P. A. Trudinger), Springer-Verlag, New York, 106-114.

Robb, L.J. and Meyer, F.M. (1995) The Witwatersrand Basin, South Africa: Geological framework and mineralization processes. Ore Geol. Rev. 10, 67-94.

Roeske, C.A. and O'Leary, M.H. (1984) Carbon isotope effect on the enzyme-catalyzed carboxylation of ribulose biphosphate. Biochemistry 1984, 6275-6284.

Rosing, M.T. (1999) 13C-depleted carbon microparticles in >3700-Ma sea-floor sedimentary rocks from West Greenland. Science 283, 674-676.

Ruttenberg, K.C. and Berner, R.A. (1993) Authigenic apatite formation and burial in sediments from non-upwelling continental margin environments. Geochim. Cosmochim. Acta 57, 991-1007.

Ruttenberg, K.C. (1993) Reassessment of the oceanic residence time of phosphorus. Chem. Geol. 107, 405-409.

Rye, R. and Holland, H.D. (1998) Paleosols and the evolution of atmospheric oxygen: a critical review. Am. J. Sci. 88, 621-672.

SACS (South African Committee for Stratigraphy) (1980) Stratigraphy of South Africa: Part I: Lithostratigraphy of the Republics of South Africa, South West Africa/Namibia and Republics of Bophuthatswana, , and Venda. Geol. Soc. S. Afr. Handbook 8, 690p.

Schidlowski, M. (1987). Application of stable carbon isotopes to early biochemical evolution on earth. Ann. Rev. Earth Planet. Sci. 15, 47-72.

Schidlowski, M. (1989). Evolution of the sulphur cycle in the Precambrian. In Evolution of the global biogeochemical sulphur cycle (eds., P. Brimblecombe and A. Yu). John Wiley, Chichester, U.K. 3-19.

Schidlowski, M. (1988). A 3,800-million-year isotopic record of life from carbon in sedimentary rocks. Nature 333, 313-318. 118 Schidlowski, M. (2001) Carbon isotopes as biogeochemical recorders of life over 3.8 Ga of Earth history: evolution of a concept. Precam. Res. 106, 117-134.

Schidlowski, M. and Aharon, P. (1992) Carbon Cycle and Carbon Isotope Record: Geochemical Impact of Life over 3.8 Ga of Earth History. In Early Organic Evolution: Implications for Mineral and Energy Resources (eds., M. Schidlowski, S. Golubic, M.M. Kimberley, D.M. McKirdy, and P.A. Trudinger), Springer- Verlag, New York, 147-175.

Schidlowski, M., Hayes, J.M., and Kaplan, I.R. (1983). Isotopic inferences of ancient biochemistries: carbon, sulfur, hydrogen, and nitrogen. In Earth’s Earliest Biosphere: Its Origin and Evolution (ed., J.W. Schopf), Princeton Univ. Press, Princeton, 149-186.

Schopf, J.W. (1993) Microfossils of the early Archean Apex chert: new evidence of the antiquity of life. Science 260, 640-646.

Schopf, J.W. and Packer, B.M. (1987). Early Archean (3.3 billion to 3.5 billion-years-old) microfossils from the Warrawoona Group, Western Australia. Science 237, 70-73.

Schopf, J.W., Kudryavstev, A.B., Agresti, D.G., Wdowiak, T.J., and Czaja, A.D. (2002) Laser-Raman imagenary of Earth’s earliest fossils. Nature 416, 73-76.

Sclater, J.G., Jaupart, C., and Galson, D. (1980) The heat flow through oceanic and continental crust and the heat loss of the Earth. Rev. Geophys. Space Phys. 18, 269- 311.

Sharkey, T.D. and Berry, J.A. (1985) Carbon isotope fractionation in algae as influenced by inducible CO2 concentrating mechanism. In Inorganic carbon uptake by aquatic photosynthetic organisms (eds., W.J. Lucas and J.A. Berry), Am. Soc. Plant Physiol., Rockville, MD, 389-401.

Simoneit, B.R.T., Brenner, S., Peters, K.E., and Kaplan, I.R. (1981) Thermal alteration of Cretaceous black shale by intrusions in the Eastern Atlantic: II. Effects on bitumen and kerogen. Geochim. Cosmochim. Acta 45, 1581-1602.

Simonson, B.M. and Hassler, S.W. (1997) Revised correlations in the Early Precambrian Hamersley Basin based on a horizon of resedimented impact spherules. Aus. J. Earth Sci. 44, 37-48.

Simonson, B.M., Davies, D., Wallace, M., Reeves, S., and Hassler, S.W. (1998) anomaly but no shocked quartz from Late Archean microkrystite layer: Oceanic impact ejecta? Geology 26, 195-198. 119 Simonson, B.M., Schubel, K.A. and Hassler, S.W. (1993) Carbonate sedimentology of the early Precambrian Hamersley Group of Western Australia. Precam. Res. 60, 287- 335.

Söhnge, A.P.G. (1986) Mineral province of southern Africa. In Mineral Deposits of Southern Africa (eds. C.R. Anhaeusser and S. Maske), Geol. Soc. S. Afr., , pp 1-23.

Strauss, H. (1986) Carbon and sulfur isotopes in Precambrian sediments from the Canadian Shield. Geochim. Cosmochim. Acta. 50, 2653-2662.

Strauss, H. (1989) Carbon and sulfur isotopic data for carbonaceous metasediments from the Kidd Creek massive sulfide deposit and vicinity, Timmins, Ontario. Econ. Geol. 84, 959-962.

Strauss, H. and Beukes, N.J. (1996) Carbon and sulfur isotopic compositions of organic carbon and pyrite in sediments from the Transvaal Supergroup, South Africa. Precam. Res. 79, 57-71.

Strauss, H. and Moore, T.B.(1992) Abundances and isotopic compositions of Carbon and Sulfur species in whole rock and Kerogen samples. In The Proterozoic Biosphere (ed. Schopf, J.W. and Klein, C.), Cambridge University Press, U.K, 709-798.

Strauss, H., DesMarais, D.J., Hayes, J.M. and Summons, R.E. (1992a) The Carbon- Isotopic Record. In The Proterozoic Biosphere: A multidisciplinary study (eds., J.W. Schopf and C. Klein), Cambridge University Press, Cambridge, UK, 117-127.

Strauss, H., DesMarais, D.J., Hayes, J.M., and Summons, R.E. (1992b) Proterozoic organic carbon - Its preservation and isotopic record. In Early Organic Evolution: Implications for Mineral and Energy Resources (eds., M. Schidlowski, S. Golubic, M.M. Kimberley, D.M. McKirdy, and P.A. Trudinger), Springer-Verlag, New York, 202-211.

Stumm, W. and Morgan, J.J. (1996) Aquatic Chemistry. 3rd. ed., John Wiley & Sons, 1022 pp.

Suess, E., Kulm, L.D., and Killingley, J.S. (1987) Coastal upwelling and a history of organic-rich mudstone deposition off Peru. In Marine Petroleum Source Rocks (J. Brooks and A.J. Fleet, eds.), Geol. Soc. Publ. 26, 181-198.

Summons, R.E., Jahnke, L.L., and Roksandic, Z. (1994) Carbon isotopic fractionation of lipids from methanotrophic bacteria: Relevance for interpretation of the geochemical record of biomarkers. Geochim. Cosmochim. Acta 58, 2853-2863. 120 Tankard, A.J., Jackson, M.P.A., Eriksson, K.A., Hobday, D.K., Hunter, D.R., and Minter, W.E.L. (1982) Crustal evolution of southern Africa, 3.8 billion years of the history. Springer, Berlin, 523 pp.

Taylor, S.R. and McLennan, S.M. (1985) The Continental Crust: its Composition and Evolution. Blackwell Scientific Publications, 311p.

Thorne, A.M. (1990) Hamersley Basin. In 3rd International Archean Symposium, Perth (1990) Excursion Guidebook (eds., S.E. Ho, J.E. Glover, J.S. Myers, and J.R. Muhling), Univ. of W. Aust, Dept. of Geol. and Univ. Extension Pub. No. 21, 13- 14.

Thorne, A.M. and Blake, T.S. (1990) Fortescue Group, in Ho, S.E., Glover, J.E., Myers, J.S., and Muhling, J.R., eds., Third International Archean Symposium, Perth (1990) Excursion Guidebook, University of Western Australia, Department of Geology and University Extension Publication No. 21, 14-18.

Towe, K.M. (1990) Aerobic respiration in the Archean? Nature 348, 54-56.

Towe, K.M. (1991) Aerobic carbon cycling and cerium oxidation: significance for Archean oxygen levels and banded iron formation deposition. Paleogeogr., Paleoclimatol., Paleoecol. (Global Planet. Change Sect.) 97, 113-123.

Towe, K.M. (1994) Earth’s early atmosphere: Constraints and opportunities for early evolution. In Early Life on Earth. Nobel Symposium No. 84. (ed., S. Bengston) Columbia University Press, New York, 36-47.

Turekian, K. and Wedepohl, K.H. (1961) Distribution of elements in some major units of the earth's crust. Geol. Soc. Am. Bull. 72, 175-192.

Tyler, N. (1978) A stratigraphic analysis of the pre-Chuniespoort Group strata around the Makoppa Dome, west-central Transvaal. M.Sc. Thesis, Univ., Witwatersrand, South Africa. 206 pp.

Van Cappellen, P. and Berner, R.A. (1988) A mathematical model for the early diagenesis of phosphorus and fluorine in marine sediments: apatite precipitation. Am. J. Sci. 288, 289-333.

Van Cappellen, P. and Ingall, E.D. (1996) Redox stabilization of the atmosphere and oceans by phosphorus-limited marine production. Science 271, 493-496.

Viljoen, M.J. and Viljoen, R.P. (1969) An introduction to the geology of the Barberton granite-greenstone terrain. Geol. Soc. South Afr. Spec. Pub. 2, 9-28. 121 Vine, J.D. and Tourtelot, E.B. (1970) Geochemistry of black shale deposits: A summary report. Econ. Geol. 65, 253-272.

Walsh, M.M. (1992) Microfossils and possible microfossils from the early Archean Onverwacht Group, Barberton Mountain Land, South Africa. Precam. Res. 54, 271- 293.

Walsh, M.M. and Lowe, D.R. (1985) Filamentous microfossils from the 3500-Myr-old Onverwacht Group, Barberton Mountain Land, South Africa. Nature 314, 530-532.

Walsh, M.M. and Lowe, D.R. (1999) Modes of accumulation of carbonaceous matter in the Early Archean: A petrographic and geochemical study of the carbonaceous cherts of the Swaziland Supergroup. In Geologic Evolution of the Barberton Greenstone Belt, South Africa (D.R. Lowe and G.R, Byerly eds.). Geol. Soc. Amer. Spec. Paper 329. 115-132..

Watanabe, Y., Naraoka, H., Wronkiewicz, D.J., Condie, K.C., and Ohmoto, H. (1997) Carbon, nitrogen, and sulfur geochemistry of Archean and Proterozoic shales from the Kaapvaal Craton, South Africa. Geochim. Cosmochim. Acta 61, 3441-3459.

Wedepohl, K.H. (1991) The composition of the upper Earth's crust and the natural cycles of selected metals. Metals in natural raw materials. Natural resources. In Metals and their compounds in the Environment (ed., E. Merian), VCH, Weinheim. 3-17.

Westall, F., de Wit, M.J., Dann, J., can der Gaast, S., de Ronde, C.E.J., and Gerneke, D. (2001) Early Archean fossil bacteria and biofilms in hydrothermally influenced, shallow water sediments, Barberton Greenstone Belt, South Africa. Precam. Res. 106, 91-112.

Windley, B.F. (1995) The Evolving Continents. 3rd ed. John Willey and Sons, 526 pp.

Winter, H. de la R. (1976) A lithostratigraphic characteristics of fluvial deposits. Soc. Econ. Paleont. Miner. Spec. Pub. 16, 84-97.

Woodhead, J.D., Hergt, J.M., and Simonson, B.M. (1998) Isotopic dating of an Archean bolide impact horizon, Hamersley Basin, Western Australia. Geology 26, 47-50.

Wronkiewicz, D.J. and Condie, K.C. (1990) Geochemistry and mineralogy of sediments from Ventersdorp and Transvaal Supergroup, South Africa: Cratonic evolution during the early Proterozoic. Geochim. Cosmochim. Acta 54, 343-354. 122

20˚E 30˚E N MOÇANBIQUE Sabie - Pilgrim's Rest BOTSWANA region

Barberton 25˚S Greenstone Belt Pretoria NAMIBIA

Witwatersrand Basin SWAZILAND LESOTHO 30˚S Atlantic Ocean

SOUTH AFRICA Transvaal Supergroup

Ventersdorp Supergroup

Witwatersrand Supergroup Cape Town Barberton Greenstone Belt 400 km

Fig. 2-1. Simplified geological map of South Africa showing the distribution of the Transvaal, Ventersdorp, and the Witwatersrand Supergroups, the Sabie- Pilgrim's Rest region, the Witwatersrand Basin, and the Barberton Greenstone Belt. Modified after SACS (1980). Modified after thickness isnottoscale.Samplesweretakenfrom shadedGroups. Witwatersrand, andSwazilandSupergroupsinSouth Africa.Vertical Fig. 2-2.LithostratigraphiccolumnoftheTransvaal, Ventersdorp, Supergroup Dominion

Legend Swaziland Witwatersrand Ventersdorp Transvaal Supergroup Supergroup Supergroup Supergroup Onverwacht Fig Tree Moodies Platberg Pniel Pretoria Rand West Rand Central Wolkberg oort Chuniesp berg Kliprivers Group quartzite conglomerate Lowe andByerly(1999) v v v + v v + v v v v v + v v v v v v + + v dolomite shale/siltstone graywacke (U-Pb zirconbySHRIMP) (Syferfontein FelsicLava) 3,074 Ma (U-Pb zirconbySHRIMP) (Crown ) 2,914 Ma (U-Pb zirconbySHRIMP) (Klipriversberg Lavas) 2,714 Ma (Pb-Pb singlezircon) (Malmani Gp,) 2,557 Ma (Rb-Sr wholerRock) (Hekpoort Lava) 2,224 Ma (U-Pb zirconbySHRIMP) 3 (U-Pb zirconbySHRIMP) 3 ,453 Ma ,230 Ma . ++ v v granite/granitoid mafic lava 123 124

20 ˚S N 117 ˚E 120 ˚E Port Hedland

Karratha

Ripon Hill

Wittenoom

23 ˚S

Turee Creek Group Mt. Bruce Hamersley Group Supergroup Fortescue Group Perth Granite-granitoid Archean Mafic basement Basement

Fig. 2-3. Simplified geological map of the Pilbara-Hamersley district, Western Australia, showing the distribution of the Turee Creek, Hamersley, and Fortescue Groups of the Mt. Bruce Supergroup. Drillcore samples of this study were taken from near Wittenoom in the central Hamersley Basin and near Ripon Hill in the eastern Hamersley region. Modified after Geological Survey of Western Australia (1990). 125

Central-western Hamersley

Formation Ages (number of samples) Brockman 2,470 Ma Iron

Mt.McRae / Mt.Sylvia Shale Eastern Hamersley Witteonoom n = 3 Dolomite 2,603 Ma Formation Age Hamersley Group Marra Carawine Mamba Dolomite 2,548 Ma Iron n = 4 n = 6 2,687 Ma Jeerinah Lewin v v Shale n = 8 Mt. Bruce Supergroup v Maddina v v Basalt v v Legend 500 m Fortescue Group iron-formation v v v shale/siltstone v v mafic lava dolomite

Fig. 2-4. Lithostratigraphic column of the lower Hamersley and upper Fortescue Groups of the Mt. Bruce Supergroup, Western Australia. Scale bar represents 500m in vertical thickness in an idealized section. Samples were taken from shaded Formations (Wittenoom Dolomite, Marra Mamba Iron, Carawine Dolomite, and Lewin Shale Formations). Correlation of Carawine Dolomite and underlying Lewin Shale in the eastern region to the Hamersley / Fortescue Groups in the central region has not been firmly established; it is suggested that the Carawine Dolomite and Lewin Formations correspond to the Wittenoom Dolomite and Jeerinah Formations, respectively (e.g., Simonson and Hassler, 1997). n: number of samples. Modified after Geological Survey of Western Australia (1990) and Simonson and Hassler (1997). 126

1000

100

∑Fe/Ti = 20

10 ∑Fe/Ti = 5 Fe3+/Ti [wt.ratio]

1 = 10 2+ /Fe 3+ Fe ∑

0.1 = 1 2+ = 0.1 /Fe 2+ 3+ /Fe Fe 3+ ∑ Fe ∑ 0.01 0.01 0.1 1 10 100 1000 Fe2+/Ti [wt.ratio]

Legend 3.25 Ga Sheba Fm, shale 2.69 Ga Lewin Shale Fm 3.25 Ga Sheba Fm, siderite >2.60 Ga Marra Mamba Iron Fm 3.25 Ga Sheba Fm, chert 2.60 Ga Wittenoom Fm 3.25 Ga Sheba Fm, graywacke 2.60 Ga Carawine Dolomite 2.96 Ga Parktown Fm 2.64 Ga Black Reef Fm 2.72 Ga Pillingini Tuff Fm 2.56 Ga Oak Tree Fm 2.71 Ga Rietgat Fm 2.22 Ga Timeball Hill Fm 2.69 Ga Jeerinah Fm ~2.2 Ga Mapedi Fm

Fig. 2-5. Log-log plot of the Fe3+/Ti vs. Fe2+/Ti (wt.) ratios for the Archean– Paleoproterozoic samples of this study. The curves represent constant ∑Fe/Ti ratios (20 and 5). The line represent the constant Fe3+/Fe2+ ratios (0.1, 1, and 10). 127

(a)100 (b) 20 shale 3.25 Ga Sheba Fm siderite-rich 3.25 Ga Sheba Fm cherty Shales Graywackes 80 15

60 Fe3+/Ti [wt.ratio] 10 40

5 20

0 0 0 200 400 600 800 1000 051015 20

Fe2+/Ti Fe2+/Ti [wt.ratio] [wt.ratio]

(c)100 (d) 20 2.96 Ga Parktown Fm 2.71 Ga Rietgat Fm 2.72 Ga Pillingini Tuff Fm 80 15

60 Fe3+/Ti [wt.ratio] 10 40

5 20

0 0 020406080100 051015 20

Fe2+/Ti Fe2+/Ti [wt.ratio] [wt.ratio]

Fig. 2-6. Plot of the Fe3+/Ti vs. Fe2+/Ti (wt.) ratios for the Archean– Paleoproterozoic samples of this study. The solid and dashed lines represent ∑Fe/Ti = 20 and ∑Fe/Ti = 10, respectively. 128

(e)20 (f) 200 2.69 Ga Jeerinah Fm 2.6 Ga Carawine Dolomite Fm 2.69 Ga Lewin Shale Fm 2.60 Ga Wittenoom Dolomite Fm >2.60 Ga Marra Mamba Iron Fm 15 150

Fe3+/Ti 10 100 [wt.ratio]

5 50

0 0 051015 20 050100 150 200

Fe2+/Ti Fe2+/Ti [wt.ratio] [wt.ratio]

(g)20 (h) 20 2.56 Ga Oak Tree Fm 2.22 Ga Timeball Hill Fm (Eastern) 2.22 Ga Timeball Hill Fm 15 15 (Central) ~2.2 Ga Mapedi Fm

Fe3+/Ti 10 [wt.ratio] 10

5 5

0 0 051015 20 051015 20

Fe2+/Ti Fe2+/Ti [wt.ratio] [wt.ratio]

Fig. 2-6. Plot of the Fe3+/Ti vs. Fe2+/Ti (wt.) ratios for the Archean– Paleoproterozoic samples of this study. The solid and dashed lines represent ∑Fe/Ti = 20 and ∑Fe/Ti = 10, respectively. 129

3 (a) ~2.2 Ga Mapedi Fm 0 5 (b) central Transvaal 2.22 Ga Timeball Hill Fm eastern Transvaal 0 (c) 2 2.56 Ga Oak Tree Fm 0 2 (d) 2.60 Ga Wittenoom Dolomite Fm 0 2 (e) 2.60 Ga Carawine Dolomite Fm 0 (f) 1 0 >2.60 Ga Marra Mamba Iron Fm 2 (g) 2.69 Ga Lewin Fm 0 2 (h) 2.69 Ga Jeerinah Fm 0 5 Count (i) 2.71 Ga Rietgat Fm 0 2 (j) 2.72 Ga Pillingini Tuff Fm 0 5 (k) 2.96 Ga Parktown Fm

0

6 5 3.25 Ga Sheba Fm (l) (graywackes) 0 8 siderite- (m) 5 rich 3.25 Ga Sheba Fm

shale 0 0 0.2 0.4 0.6 0.8 1

average mafic rocks average Archean igneous rocks average intermediate rocks average felsic rocks

0 0.2 0.4 0.6 0.8 1 Fe3+/∑Fe

Fig. 2-7. Histogram of the Fe3+/∑Fe ratios for the Archean–Paleoproterozoic samples of this study. Also shown are the Fe3+/∑Fe ratios for typical igneous rocks (data from Holland, 1984: p145. original references therein). 130

(a) 4

4

DOP = 0.75 3 DOP = 0.5 3

Fepy Spy [wt.%] 2 [wt.%] Fe 2 R = 4 Fe R = 3 Fe 1 DOP = 0.25 R = 2 1

0 0 0 1 2 3 4 (b) 1

1

DOP = 0.75 0.75 DOP = 0.5 0.75

Spy Fepy 0.5 Fe [wt.%] [wt.%] R = 1 0.5

0.25 Fe DOP = 0.25 0.25 R = 0.5

0 0 0 0.25 0.5 0.75 1

FeHCl [wt.%] 131

Legend 3.25 Ga Sheba Fm, shale 2.69 Ga Lewin Shale Fm 3.25 Ga Sheba Fm, siderite >2.60 Ga Marra Mamba Iron Fm 3.25 Ga Sheba Fm, chert 2.60 Ga Wittenoom Fm 3.25 Ga Sheba Fm, graywacke 2.60 Ga Carawine Dolomite 2.96 Ga Parktown Fm 2.64 Ga Black Reef Fm 2.72 Ga Pillingini Tuff Fm 2.56 Ga Oak Tree Fm 2.71 Ga Rietgat Fm 2.22 Ga Timeball Hill Fm 2.69 Ga Jeerinah Fm ~2.2 Ga Mapedi Fm

Fig. 2-8. Plot of Fepy versus FeHCl contents for the Archean–Paleoproterozoic shales of this study. The FeR (= Fepy+ FeHCl) contents are indicated by the lines with negative slopes and the DOP values (Fepy / FeR ) are indicated by the lines with positive slopes in both (a) all the samples and (b) blow-up of (a). In (a), Spy contents are also indicated on the right y axis. 132

(a) 2 ~2.2 Ga Mapedi Shale Fm 0 (Strauss and Beukes, 1992) 13

4 eastern Transvaal (b) central Transvaal 2.22 Ga Timeball Hill Fm 0 3 2.53 Ga Reivillo Fm (c) 0 (Strauss and Beukes, 1992) 4 2.54 Ga Monteville Fm (d) (Strauss and Beukes, 1992) 0 2.55 Ga Lokamonna Fm 2 (e) 0 (Strauss and Beukes, 1992) (f) 2 2.56 Ga Boomplas Fm 0 (Strauss and Beukes, 1992) 1 (g) 0 2.56 Ga Oak Tree Fm (h) 1 0 2.60 Ga Wittenoom Dolomite Fm 2 (i) 2.60 Ga Carawine Dolomite Fm 0 2 (j) >2.60 Ga Marra Mamba Iron Fm 0 4 2.64 Ga Black Reef Fm (k) Count 0 (Strauss and Beukes, 1992) 4 (l) 2.64 Ga Vryburg Fm 0 (Strauss and Beukes, 1992) 4 (m) 2.69 Ga Jeerinah Fm 0 3 (n) 2.69 Ga Lewin Fm 0 2 (o) 2.71 Ga Rietgat Fm 0 1 2.72 Ga Pillngini Tuff Fm (p) 0 3 (q) 2.96 Ga Parktown Fm 0 11 (r) 3.25 Ga Sheba Fm 1 (graywackes) 0 14 4 silica-rich siderite- (s) rich shale 3.25 Ga Sheba Fm 0 0 0.2 0.4 0.6 0.8 1 DOP Fig. 2-9. Histogram of the DOP (degree of pyritization) values for the Archean– Paleoproterozoic samples of this study. Published data from Strauss and Beukes (1996) are also shown. DOP is defined as a ratio of pyrite-bound Fe to reactive Fe (pyrite-bound Fe plus HCl-leachable Fe). See text for more detailed explanation. (Post-Archean Australian AverageShale: samples ofthisstudy.The verticallineatAl Fig. 2-10.Histogramof theAl Count (m) (h) (g) (e) (d) (b) (a) (n) (k) (c) (f) (l) (j) (i) 0 0 4 3 0 2 0 2 0 4 0 5 0 2 0 2 0 2 1 1 0 3 0 3 0 5 0102 silica-rich central Transvaal eastern Transvaal siderite-rich Al 2 O PAAS 3 [wt.%] 2 O 3 contentsfortheArchean–Paleoproterozoic 030 shale 9 Taylor andMcLennan,1985 2 O 3 =17.8wt.%representPAAS 2.64 GaBlackReefFm >2.60 GaMarraMambaIronFm 2.60 GaCarawineDolomiteFm 2.60 GaWittenoomDolomiteFm 2.56 GaOakTreeFm 2.22 GaTimeballHillFm 3.25 GaShebaFm (graywacke) 3.25 GaShebaFm 2.96 GaParktownFm 2.72 GaPillinginiTuffFm 2.71 GaRietgatFm 2.69 GaJeerinahFm 2.69 GaLewinFm ~2.2 GaMapediFm ). 133 134

2 (a) ~2.2 Ga Mapedi Fm 0 3 0.7 0.8 (b) 2.22 Ga Timeball Hill Fm 0 2 (c) 2.56 Ga Oak Tree Fm 0 2 (d) 2.60 Ga Wittenoom Dolomite Fm 0 3 (e) 2.60 Ga Carawine Dolomite Fm 0 2 (f) >2.6 Ga Marra Mamba Iron Fm 0 3 (g) 2.64 Ga Black Reef Fm 0 3 (h) 2.69 Ga Lewin Fm 0 Count 3 (i) 2.69 Ga Jeerinah Fm 0 2 (j) 2.71 Ga Rietgat Fm 0 2 (k) 2.72 Ga Pillingini Tuff Fm 0 4 (l) 2.96 Ga Parktown Fm 0 4 (m) 3.25 Ga Sheba Fm (graywackes) 0 4 (n) 3.25 Ga Sheba Fm 0 0 0.1 0.2 0.3 0.4 0.5

Av. shale (PAAS) by Taylor and McLennan (1985) Av. Archean cratonic shale by Condie (1993) Devonian laminated shales (n = 56) by Ingall et al. (1993) Devonian bioturbated shales (n = 33) by Ingall et al. (1993)

0 0.1 0.2 0.3 0.4 0.5

P2O5 [wt.%]

Fig. 2-11. Histogram of the P2O5 contents for the Archean–Paleoproterozoic samples of this study. The average shales and the Devonian shales are also shown for comparison. 135

7 (a) 1 ~2.2 Ga Mapedi Fm 0 6 5 eastern Transvaal (b) 2.22 Ga Timeball Hill Fm central Transvaal 0 2 (c) 0 2.56 Ga Oak Tree Fm 1 2.60 Ga Wittenoom Dolomite Fm (d) 0 2.60 Ga 1 (e) 0 Carawine Dolomite Fm 45 1 (f) 0 >2.6 Ga Marra Mamba Iron Fm 4 (g) 2.64 Ga Black Reef Fm 0 2 (h) 2.69 Ga Lewin Fm 0 6.5 7 1 Count 2.69 Ga Jeerinah Fm (i) 0 0 3691215 1 (j) 0 2.71 Ga Rietgat Fm 2 (k) 2.72 Ga Pillingini Tuff Fm 0 5 (l) 2.96 Ga Parktown Fm 0 6 5 (m) 3.25 Ga Sheba Fm (graywackes) 0 4 silica-rich (n) siderite-rich shale 3.25 Ga Sheba Fm 0 0 0.5 1 1.5 2 2.5 3

Average shale by Corg [wt.%] Wedepohl (1991)

Av. of 100 Archean–Paleoproterozoic shales from South Africa by Watanabe et al. (1997). Range: 0.06 - 2.8 [wt.%]

Fig. 2-12. Histogram of the organic C (Corg) contents for the Archean– Paleoproterozoic samples of this study. Note that (i) has a different scale (0 ~ 15). 136

8 (a) ~2.2 Ga Mapedi Fm 0 16 central Transvaal (b) 12 eastern Transvaal 2.22 Ga Timeball Hill Fm 1 0 3 (c) 2.56 Ga Oak Tree Fm 0 2 (d) 2.60 Ga Wittenoom Dolomite Fm 0 2 (e) 2.60 Ga Carawine Dolomite Fm 0

2 (f) 0 >2.60 Ga Marra Mamba Iron Fm 2 (g) 2.64 Ga Black Reef Fm 0

8 (h) 2.69 Ga Lewin Fm Count 0 5 (i) 2.69 Ga Jeerinah Fm 0 7 (j) 2.71 Ga Rietgat Fm 0 3 (k) 2.72 Ga Pillingini Tuff Fm 0 14 (l) 2.96 Ga Parktown Fm 0 7 3 3.25 Ga Sheba Fm (m) (graywackes) 0

5 silica- (n) rich siderite-rich shale 3.25 Ga Sheba Fm 0 051015

Ccarb [wt.%]

Fig. 2-13. Histogram of the carbonate C (Ccarb) contents for the Archean–Paleoproterozoic samples of this study. 137

8 (a) ~2.2 Ga Mapedi Fm 0 16 2.22 Ga Timeball Hill Fm (b) 4 central Transvaal eastern Transvaal 0 (c) 1 2.56 Ga Oak Tree Fm 0 1 (d) 0 2.60 Ga Wittenoom Dolomite Fm 1 (e) 0 2.60 Ga Carawine Dolomite Fm (f) 1 >2.60 Ga Marra Mamba Iron Fm 0 8 3.5 (g) 1 0 2.64 Ga Black Reef Fm 2 (h) 2.69 Ga Lewin Fm 0 4 4.5 Count (i) 1 0 2.69 Ga Jeerinah Fm 3.5 2 (j) 2.71 Ga Rietgat Fm 0

(k) 2 2.72 Ga Pillingini Tuff Fm 0 7 (l) 2 2.96 Ga Parktown Fm 0 13 3.25 Ga Sheba Fm (m) 1 (graywackes) 0 12 4 3.25 Ga Sheba Fm (n) silica-rich siderite-rich shale 0 01 23

Average shale Spy [wt.%]

Av. of 100 Archean–Paleoproterozoic shales from South Africa by Watanabe et al. (1997). Range: < 0.01 - 1.6 [wt.%]

Fig. 2-14. Histogram of the pyrite S (Spy) contents for the Archean–Paleoproterozoic samples of this study. 138

5 (a) ~2.2 Ga Mapedi Fm 0 7 4 eastern Transvaal (b) 2.22 Ga Timeball Hill Fm central Transvaal 0 2 (c) 2.56 Ga Oak Tree Fm 0 (d) 1 0 2.60 Ga Wittenoom Dolomite Fm 2 (e) 2.60 Ga Carawine Dolomite Fm 0

(f) 1 0 >2.60 Ga Marra Mamba Iron Fm 2 (g) 2.64 Ga Black Reef Fm 0 2 (h) 2.69 Ga Jeerinah Fm

Count 0 3 (i) 2.69 Ga Lewin Fm 0 2 (j) 2.71 Ga Rietgat Fm 0 (k) 1 0 2.72 Ga Pillingini Tuff Fm 7 3 (l) 2.96 Ga Parktown Fm 0 5 3.25 Ga Sheba Fm (m) (graywackes) 0 5 shale siderite- 3.25 Ga Sheba Fm (n) rich 0 -50 -45 -40 -35 -30 -25 -20 δ13C [‰]

13 Fig. 2-15. Histogram of the C isotopic compositions of organic C (δ Corg) values for the Archean–Paleoproterozoic samples of this study. 139

eastern Transvaal 1 2.22 Ga Timeball Hill Fm (a) 0

1 2.56 Ga Oak Tree Fm (b) 0

1 (c) 0 2.60 Ga Wittenoom Dolomite Fm

2 (d) 2.60 Ga Carawine Dolomite Fm 0

(e) 1 >2.60 Ga Marra Mamba Iron Fm 0

2 (f) 2.69 Ga Jeerinah Fm 0

2 (g) 2.69 Ga Lewin Fm 0 Count 3 (h) 2.71 Ga Rietgat Fm 0

2 (i) 2.69 Ga Pillingini Tuff Fm 0

(j) 1 2.96 Ga Parktown Fm 0

5 (k) 3.25 Ga Sheba Fm (graywackes) 0

5 shale siderite- (l) 3.25 Ga Sheba Fm rich 0 -10 -5 0 5 10 15 34 δ Spy [‰]

Fig. 2-16. Histogram of the S isotopic compositions of sulfide (mainly 34 pyrite) S (δ Spy) values for the Archean–Paleoproterozoic samples of this study. The shaded area represents the typical δ34S values for sulfides in igneous rocks (0 ± 2 ‰). 140

(a) ~2.2 Ga Mapedi Fm

central Transvaal eastern Transvaal (b) 2.22 Ga Timeball Hill Fm

(c) 2.56 Ga Oak Tree Fm

(d) 2.60 Ga Wittenoom Dolomite Fm

(e) 2.60 Ga Carawine Dolomite Fm

(f) >2.6 Ga Marra Mamba Iron Fm

(g) 2.64 Ga Black Reef Fm

(h) 2.69 Ga Lewin Fm

(i) 2.69 Ga Jeerinah Fm

(j) 2.71 Ga Rietgat Fm

(k) 2.72 Ga Pillingini Tuff Fm

(l) 2.96 Ga Parktown Fm

(m) 3.25 Ga Sheba Fm (graywackes)

shale cherty siderite-rich (n) 3.25 Ga Sheba Fm

10-4 0.001 0.01 0.1

(o) 1 Average shale (PAAS) 2 Black Sea sediments Cretaceous black shales 3 4 Devonian black shales 5 Average Paleoproterozoic shale 6 Average Archean shale 7 Transvaal Supergroup 8 Ventersdorp Supergroup 9 Witwatersrand Supergroup 10 Pongola Supergroup

10-4 0.001 0.01 0.1

P/Al [wt. ratio] 141

Fig. 2-17. The P/Al (wt.) ratios of the Archean–Paleoproterozoic samples of this study (a through n) and of various samples from literature (o). The vertical line at P/Al = 0.007 represents average shale (PAAS: Taylor and McLennan, 1985). 1. Average shale (PAAS: Taylor and McLennan, 1985); 2. Black Sea sediments, n = 253: Hirst (1974), Calvert and Batchelor (1978), Brumsack (1989); 3. Cretaceous black shales, n = 16, Dean and Arthur (1986); 4. Devonian black shales, n = 41, Caplan and Bustin (1996, 1998); 5. Average Paleoproterozoic (Aphebian) shales, n = 406, Cameron and Garrels (1980); 6. Average Archaen shales, n = 406, Cameron and Garrels (1980); 7. Transvaal Supergroup shales, n = 91, Wronkiewicz and Condie (1990); 8. Ventersdorp Supergroup shales, n = 11, Wronkiewicz and Condie (1990); 9. Witwatersrand Supergroup shales, n = 75, Wronkiewicz and Condie (1987); 10. Pongola Supergroup shales, n = 63, Wronkiewicz and Condie (1989). n = samples of samples. 142

(a) (c)

(b) (d)

500 µm

Fig. 2-18. Selected photomicrographs of the representative South African rock samples of this study. Under plain-polarized light. Scale bar is 500 µm. (a) a black shale of the 2.22 Ga Timeball Hill Formation; (b) a black shale of the 2.56 Ga Oak Tree Formation; (c) a graywacke and (d) a black shale of the 3.25 Ga Sheba Formation. 143

(e) (g)

(f) (h)

500 µm

Fig. 2-18. Selected photomicrographs of the representative Australian rock samples of this study. Under plain-polarized light. Scale bar is 500 µm. (e) a carbonate-rich black shale of the 2.60 Ga Wittenoom Dolomite Formation; (f) a black shale of the >2.60 Ga Marra Mamba Iron Formation; (g) a carbonate-rich black shale of the 2.60 Ga Carawine Dolomite Formation; and (h) a black shale of the 2.69 Ga Lewin Shale Formation. 144

(a) ~2.2 Ga Mapedi Fm

(b) 2.22 Ga Timeball Hill Fm

(c) 2.56 Ga Oak Tree Fm

(d) 2.60 Ga Wittenoom Dolomite Fm

(e) 2.60 Ga Carawine Dolomite Fm (f) >2.6 Ga Marra Mamba Iron Fm (g) 2.64 Ga Black Reef Fm

(h) 2.69 Ga Lewin Fm

(i) 2.69 Ga Jeerinah Fm

(j) 2.71 Ga Rietgat Fm

(l) 2.72 Ga Pillingini Tuff Fm

(k) 2.96 Ga Parktown Fm (m) 3.25 Ga Sheba Fm (graywackes)

siderite-rich silica-rich (n) shale 3.25 Ga Sheba Fm

0.01 0.1 1 5

1 Black Sea sediments (o) Average 2 Cretaceous black shales shale 3 Average Paleoproterozoic shale 4 Average Archean shale

0.01 0.1 1 5

Corg/Al [wt. ratio]

Fig. 2-19. The Corg/Al (wt.) ratios for the Archean–Paleoproterozoic samples of this study (a through n) and of various samples from literature (o). The vertical line at Corg/Al = 0.02 represents average shale (Wedepohl, 1991). 1. Black Sea sediments, n = 194: Hirst (1974), Brumsack (1989); 2. Cretaceous black shales, n = 16, Dean and Arthur (1986); 3. Average Paleoproterozoic (Aphebian) shales, n = 326, Cameron and Garrels (1980); 4. Average Archaen shales, n = 406, Cameron and Garrels (1980). n = number of samples. 145

100

10

Corg 1 [wt.%]

0.1

0.01 -50 -45 -40 -35 -30 -25 -20 δ13C [‰]

Legend 3.25 Ga Sheba Fm, shale 2.69 Ga Lewin Shale Fm 3.25 Ga Sheba Fm, siderite >2.60 Ga Marra Mamba Iron Fm 3.25 Ga Sheba Fm, chert 2.60 Ga Wittenoom Fm 3.25 Ga Sheba Fm, graywacke 2.60 Ga Carawine Dolomite 2.96 Ga Parktown Fm 2.64 Ga Black Reef Fm 2.72 Ga Pillingini Tuff Fm 2.56 Ga Oak Tree Fm 2.71 Ga Rietgat Fm 2.22 Ga Timeball Hill Fm 2.69 Ga Jeerinah Fm ~2.2 Ga Mapedi Fm

13 Fig. 2-20. Overall negative correlation between the Corg contents and the δ C values for the Archean–Paleoproterozoic samples of this study. 146

(a) ~2.2 Ga Mapedi Fm

eastern Transvaal central Transvaal (b) 2.22 Ga Timeball Hill Fm

(c) 2.56 Ga Oak Tree Fm

(d) 2.60 Ga Wittenoom Dolomite Fm

(e) 2.60 Ga Carawine Dolomite Fm

(f) >2.60 Ga Marra Mamba Iron Fm

(g) 2.64 Ga Black Reef Fm

(h) 2.69 Ga Jeerinah Fm

(i) 2.69 Ga Lewin Fm

(j) 2.71 Ga Rietgat Fm

(k) 2.72 Ga Pillingini Tuff Fm

(l) 2.96 Ga Parktown Fm

3.25 Ga Sheba Fm (m) (graywacke)

siderite-rich cherty shale (n) 3.25 Ga Sheba Fm

0.5110100 500 Av.Sh. Redfield Modern and Phanerozoic anoxic sediments Corg/P [wt. ratio]

Fig. 2-21. The Corg/P (wt.) ratios for the Archean–Paleoproterozoic samples of this study. The vertical lines at Corg/P = 3.05 and 41 represent average shale (Wedepohl, 1991 and Taylor and McLennan, 1985) and Redfield ratio (e.g., Redfield et al., 1963), respectively. The shaded band of Corg/P = 200 ~ 400 represents modern and Phanerozoic anoxic marine sediments deposited under completely anoxic bottom water (Ingall et al., 1993; Ingall and Jahnke, 1994). 147

0

Kerogen Total OM This study

-10

-20

δ13C -30 [‰]

-40

-50

-60

3.8 3.5 3.0 2.5 2.0 1.5

Age [Ga]

Fig. 2-22. Secular change in the δ13C values for kerogen and bulk organic matter throughout geologic time. Literature data are from compilation of Schidlowski et al. (1983), Schidlowki (1988), and Pavlov et al. (2001). 148

(a) ~2.2 Ga Mapedi Fm

(b) eastern Transvaal 2.22 Ga Timeball Hill Fm

(c) 2.56 Ga Oak Tree Fm

(d) 2.60 Ga Wittenoom Dolomite Fm

(e) 2.60 Ga Carawine Dolomite Fm

(f) >2.60 Ga Marra Mamba Iron Fm

(g) 2.64 Ga Black Reef Fm

(h) 2.69 Ga Lewin Fm

(i) 2.69 Ga Jeerinah Fm

(j) 2.71 Ga Rietgat Fm

(k) 2.72 Ga Pillingini Tuff Fm

(l) 2.96 Ga Parktown Fm

(m) 3.25 Ga Sheba Fm (graywacke) siderite-rich shale (n) 3.25 Ga Sheba Fm

0 0.01 0.1 1 10

1 Black Sea sediments (o) Average shale 2 Cretaceous black shales 3 Average Paleoproterozoic shale 4 Average Archean shale

0 0.01 0.1 1 10 S/Al [wt.ratio]

Fig. 2-23. The S/Al (wt.) ratios of the Archean–Paleoproterozoic samples of this study (a through n) and of various samples from literature (o). The vertical line at S/Al = 0.02 represents average shale (Wedepohl, 1991). 1. Black Sea sediments, n = 194: Hirst (1974), Brumsack (1989); 2. Cretaceous black shales, n = 16, Dean and Arthur (1986); 3. Average Paleoproterozoic (Aphebian) shales, n = 326, Cameron and Garrels (1980); 4. Average Archaen shales, n = 406, Cameron and Garrels (1980). n = number of samples. 149

60 Sulfate: Data from literature Sulfide: Data from literature Sulfide: This study

40

20

δ34S 0 [‰]

-20

-40

-60 4 3.5 3 2.5 2 1.5 1 0.5 0

Age [Ga]

Fig. 2-24. Secular change in the δ34S values for sulfide and sulfate throughout geologic time. Literature data are from compliation of Canfield and Raiswell (1999). Original refereneces therein. 150

(a) 4

Normal marine

3 Spy [wt.%] 2

1

0 024681012 14

0.8 (b)

0.6 Normal marine Spy [wt.%] Watanabe 0.4

et al.

0.2 Av. Sh. (1997)

0 0 0.5 1 1.5 2 2.5 3

Corg [wt.%]

Legend 3.25 Ga Sheba Fm, shale 2.69 Ga Lewin Shale Fm 3.25 Ga Sheba Fm, siderite >2.60 Ga Marra Mamba Iron Fm 3.25 Ga Sheba Fm, chert 2.60 Ga Wittenoom Fm 3.25 Ga Sheba Fm, graywacke 2.60 Ga Carawine Dolomite 2.96 Ga Parktown Fm 2.64 Ga Black Reef Fm 2.72 Ga Pillingini Tuff Fm 2.56 Ga Oak Tree Fm 2.71 Ga Rietgat Fm 2.22 Ga Timeball Hill Fm 2.69 Ga Jeerinah Fm ~2.2 Ga Mapedi Fm 151

Fig. 2-25. Plot of pyrite sulfur (Spy) versus organic carbon (Corg) contents for the Archean–Paleoproterozoic shales of this study. The average shale composition is also shown. The lines with a slope of 0.36 (= S/Corg) representing the trend of "normal marine" sediments (Berner and Raisewell, 1984) are shown in both (a) all the samples and (b) blow-up of (a). The shaded area represents 100 Archean– Paleoproterozoic shales from South Africa by Watanabe et al. (1997). 152 Table 2-1. Geochemical data of the Archean–Paleoproterozoic samples of this study.

Samples ∑Fe2O3 Fe2O3 FeO Al2O3 TiO2 P2O5 Corg Ccarb Stotal [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Carbonate-rich black shales, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) PU1308-01 43.50 3.17 36.30 3.87 0.13 0.09 0.68 8.07 0.15 PU1308-02 42.94 0.43 38.25 5.54 0.26 0.10 0.94 8.20 0.00 PU1308-03 36.50 0.99 31.95 1.91 0.05 0.06 0.43 6.91 0.16 PU1308-04 37.92 3.92 30.60 5.06 0.16 0.07 0.94 7.27 0.67 PU1308-05 31.62 1.29 27.29 8.15 0.30 0.02 1.43 6.25 0.00 PU1308-06 43.46 2.39 36.95 3.21 0.10 0.05 1.02 8.38 0.00 PU1308-07 13.25 1.45 10.62 17.46 0.65 0.09 2.97 2.19 0.00 PU1308-08 40.21 3.43 33.10 5.84 0.16 0.08 1.29 7.83 0.00 PU1308-09 11.73 1.68 9.04 0.98 0.03 0.01 0.46 3.24 0.00 PU1308-10 10.59 0.13 9.41 19.16 0.37 0.14 1.87 1.84 0.00 PU1308-11 17.55 3.25 12.87 20.77 0.38 0.20 0.22 2.78 0.00 PU1308-12 38.36 1.97 32.74 6.47 0.21 0.08 1.31 6.71 2.15 PU1308-13 25.32 6.76 16.70 18.19 0.64 0.17 2.28 3.88 2.70 PU1308-14 49.31 1.36 43.14 3.48 0.18 0.01 1.08 9.22 0.00 PU1308-15 42.63 0.59 37.83 1.46 0.05 0.02 0.49 8.14 0.00 PU1308-16 43.11 1.47 37.47 5.83 0.22 0.06 0.93 8.07 0.00 PU1308-17 41.14 0.24 36.80 1.77 0.04 0.05 0.42 7.80 0.00

Average 33.48 2.03 28.30 9.64 0.26 0.08 1.10 6.28 0.34 S.D. 12.76 1.68 11.67 8.27 0.20 0.05 0.72 2.46 0.81

Graywackes, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) MRE10-01 6.08 0.58 4.95 9.48 0.43 0.08 0.16 1.77 0.00 MRE10-02 9.89 0.62 8.34 13.42 0.64 0.10 0.52 1.52 0.00 MRE10-03 10.84 0.38 9.41 18.13 0.82 0.14 0.43 0.91 0.00 MRE10-04 9.13 0.46 7.80 11.35 0.58 0.10 0.21 1.21 0.00 MRE10-05 9.11 0.80 7.48 14.76 0.72 0.12 0.39 0.21 0.00 MRE10-06 9.09 0.58 7.66 15.29 0.66 0.11 0.12 2.17 0.00 MRE10-07 6.87 0.60 5.64 9.43 0.46 0.09 0.17 1.15 0.00 MRE10-08 7.17 0.73 5.80 10.41 0.47 0.08 0.16 1.51 0.00 MRE10-09 7.43 0.60 6.15 10.02 0.49 0.09 0.17 1.21 0.00 MRE10-10 6.15 0.62 4.98 8.10 0.36 0.07 0.17 2.83 0.00 MRE10-11 10.56 1.13 8.48 13.39 0.65 0.08 0.28 1.63 0.00 MRE10-12 9.01 0.70 7.48 10.88 0.56 0.09 0.29 2.91 0.21

Average 8.44 0.65 7.01 11.54 0.55 0.10 0.26 1.59 0.02 S.D. 1.65 0.19 1.47 2.62 0.12 0.02 0.13 0.77 0.06 153

Table 2-1. (continued)

Samples ∑Fe2O3 Fe2O3 FeO Al2O3 TiO2 P2O5 Corg Ccarb Stotal [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Iron-rich shales, Parktown Fm, West Rand G, Witwatersrand SG (2.96 Ga) DRH13-01 24.68 3.98 18.62 8.94 0.33 0.06 0.23 0.04 0.08 DRH13-02 25.55 4.87 18.60 6.95 0.26 0.05 0.08 0.15 0.08 DRH13-03 28.96 8.91 18.04 6.16 0.23 0.09 0.11 0.28 0.00 DRH13-04 25.47 3.53 19.74 10.51 0.39 0.13 0.06 0.02 0.00 DRH13-05 24.59 4.15 18.40 9.27 0.34 0.12 0.17 0.07 0.00 DRH13-06 23.92 3.39 18.48 10.41 0.39 0.15 0.07 0.02 0.69 DRH13-07 27.47 8.15 17.38 8.00 0.27 0.08 0.04 0.02 0.46 DRH13-08 41.72 20.22 19.35 4.64 0.17 0.16 0.14 0.08 0.00 DRH13-09 36.14 19.23 15.22 2.77 0.08 0.03 0.25 0.09 0.12 DRH13-10 17.17 2.23 13.45 13.70 0.48 0.05 0.51 0.00 0.30 DRH13-11 15.83 1.50 12.89 13.90 0.50 0.02 0.07 0.02 0.00 DRH13-12 14.88 0.83 12.64 12.93 1.62 0.19 0.70 0.42 0.00 DRH13-13 12.03 0.58 10.30 13.32 0.46 0.06 0.18 0.05 0.41 DRH13-14 8.00 1.28 6.04 17.72 0.67 0.07 0.27 0.01 0.35 DRH13-15 7.85 1.42 5.78 17.88 0.69 0.01 0.22 0.00 0.18

Average 22.28 5.62 14.99 12.11 0.58 0.08 0.21 0.08 0.18 S.D. 9.70 6.23 4.67 5.55 0.47 0.05 0.18 0.12 0.22

Shales, Pillingini Tuff Fm, Fortescue G, Mt. Bruce SG (2.72 Ga) WRL1-01 16.96 1.75 13.69 15.79 1.08 0.13 0.21 0.01 0.09 WRL1-02 8.34 1.84 5.85 14.07 1.14 0.13 0.26 0.12 0.07 WRL1-03 7.48 0.66 6.14 11.01 0.78 0.11 0.14 0.42 0.13

Average 10.93 1.42 8.56 11.61 0.87 0.12 0.20 0.18 0.10 S.D. 5.24 0.66 4.45 4.49 0.31 0.01 0.06 0.21 0.03

Black shales, Rietgat Fm, Platberg G, Ventersdorp SG (2.71 Ga) MSF6-01 4.61 0.17 4.00 8.92 0.49 0.09 0.32 0.24 0.13 MSF6-02 6.48 0.32 5.54 9.74 0.48 0.08 0.27 0.30 0.00 MSF6-03 6.46 0.58 5.29 11.96 0.68 0.07 0.59 0.50 0.00 MSF6-04 6.78 0.53 5.63 13.92 0.65 0.11 0.95 0.00 0.64 MSF6-05 7.94 0.79 6.43 17.90 0.77 0.06 1.30 0.03 0.38 MSF6-06 8.18 0.72 6.71 19.94 0.77 0.13 1.59 0.00 0.67 MSF6-07 9.61 0.72 8.00 16.64 0.87 0.14 1.10 0.04 0.53

Average 7.15 0.55 5.94 14.14 0.67 0.10 0.87 0.16 0.34 S.D. 1.59 0.23 1.26 4.19 0.15 0.03 0.50 0.19 0.29 154 Table 2-1. (continued)

Samples ∑Fe2O3 Fe2O3 FeO Al2O3 TiO2 P2O5 Corg Ccarb Stotal [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Black shales, Lewin Shale Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) RHDH2A-01 4.25 1.94 2.08 13.26 0.77 0.10 2.67 0.34 0.21 RHDH2A-02 4.12 2.35 1.59 15.37 0.82 0.13 6.83 0.32 1.44 RHDH2A-03 4.76 2.32 2.20 16.16 0.75 0.05 1.56 0.19 0.07 RHDH2A-04 5.74 0.88 4.37 15.84 0.75 0.07 2.66 0.03 0.07 RHDH2A-05 7.52 1.91 5.05 15.69 0.96 0.11 1.83 0.04 0.59 RHDH2A-06 8.63 2.06 5.91 14.49 0.64 0.06 2.50 0.44 0.58 RHDH2A-07 10.60 0.18 9.38 14.21 0.85 0.10 2.05 0.10 4.04 RHDH2A-08 4.77 1.36 3.07 16.29 0.75 0.07 2.18 0.85 0.28

Average 6.30 1.63 4.21 15.16 0.79 0.09 2.78 0.29 0.91 S.D. 2.37 0.76 2.59 1.07 0.09 0.03 1.68 0.27 1.34

Black shales, Jeerinah Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) WRL1-04 4.40 2.63 1.59 15.55 0.71 0.07 5.52 0.14 1.55 WRL1-05 3.45 2.87 0.52 13.79 0.53 0.06 7.39 2.32 1.84 WRL1-06 4.71 3.72 0.89 12.7 0.65 0.11 12.04 0.38 2.98 WRL1-07 5.79 3.72 1.86 12.92 0.59 0.08 8.19 0.34 3.03 WRL1-08 13.34 3.05 9.26 13.87 0.84 0.10 2.73 0.31 1.95 WRL1-09 5.40 1.16 3.82 11.23 0.53 0.08 3.87 0.10 0.74

Average 6.18 2.86 2.99 13.34 0.64 0.08 6.62 0.60 2.02 S.D. 3.60 0.94 3.28 1.44 0.12 0.02 3.36 0.85 0.88

Carbonate-rich black shales, Black Reef Fm, Wolkberg G, Transvaal SG (2.64 Ga) JPBR-01 0.76 0.65 0.10 0.77 0.03 0.01 0.09 12.02 0.00 JPBR-02 1.18 0.09 0.98 0.86 0.03 0.04 0.09 11.71 0.64 JPBR-03 5.12 0.18 4.45 3.26 0.11 0.04 0.33 11.83 0.04 JPBR-04 5.16 1.37 3.41 26.57 0.83 0.03 1.11 0.07 0.00

Average 3.06 0.57 2.24 16.41 0.58 0.03 0.41 8.91 0.17 S.D. 2.41 0.59 2.03 7.58 0.27 0.01 0.48 5.90 0.31 155 Table 2-1. (continued)

Samples ∑Fe2O3 Fe2O3 FeO Al2O3 TiO2 P2O5 Corg Ccarb Stotal [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Black shales, Marra Mamba Iron Fm, Hamersley G, Mt. Bruce SG (>2.60 Ga) WRL1-10 28.95 6.02 20.64 7.14 0.15 0.04 0.19 0.01 0.07 WRL1-11 34.33 14.20 18.12 2.47 0.12 0.03 0.37 1.51 0.37 WRL1-12 32.38 16.59 14.21 4.46 0.31 0.05 0.93 1.40 3.16 WRL1-13 34.70 12.87 19.65 4.00 0.29 0.06 0.74 3.80 0.14

Average 32.59 12.42 18.16 10.93 0.52 0.05 0.56 1.68 0.94 S.D. 2.63 4.54 2.83 6.48 0.30 0.01 0.34 1.57 1.49

Black shales, Wittenoom Dolomite Fm, Hamersley G, Mt. Bruce SG (2.60 Ga) WRL1-14 10.62 0.00 9.56 1.57 0.06 0.02 0.74 10.59 0.08 WRL1-15 13.01 2.61 9.36 14.28 0.56 0.05 2.78 0.13 1.64 WRL1-16 17.08 3.47 12.25 12.24 0.57 0.06 2.67 0.04 0.14

Average 13.57 2.03 10.39 7.06 0.32 0.04 2.06 3.59 0.62 S.D. 3.27 1.81 1.61 4.49 0.19 0.02 1.15 6.06 0.88

Carbonate-rich black shales, Carawine Dolomite Fm, Fortescue G, Mt. Bruce SG (2.60 Ga) RHDH2A-09 15.19 1.31 12.49 5.87 0.32 0.07 1.41 7.21 0.44 RHDH2A-10 4.07 1.67 2.16 4.99 0.29 0.03 1.02 9.22 0.08 RHDH2A-11 7.73 4.32 3.07 7.32 0.42 0.06 2.91 6.10 0.73 RHDH2A-12 5.44 2.57 2.58 9.5 0.70 0.07 1.91 6.68 0.14 RHDH2A-13 6.74 3.37 3.03 10.76 0.67 0.08 4.24 2.15 0.36 RHDH2A-14 4.20 1.98 2.00 15.77 1.00 0.06 4.45 7.29 1.23

Average 7.23 2.54 4.22 9.04 0.57 0.06 2.66 6.44 0.50 S.D. 4.15 1.14 4.07 3.95 0.27 0.02 1.45 2.35 0.43

Black shales, Oak Tree Fm, Chuniespoort G, Transvaal SG (2.56 Ga) MSF6-08 1.99 0.10 1.70 17.05 0.70 0.25 2.30 0.15 0.27 MSF6-09 2.60 0.08 2.27 13.74 0.56 0.02 1.64 0.05 0.94 MSF6-10 1.49 0.09 1.26 15.53 0.73 0.02 1.66 0.00 0.00

Average 2.03 0.09 1.74 12.57 0.56 0.09 1.87 0.07 0.40 S.D. 0.56 0.01 0.51 5.90 0.21 0.13 0.37 0.07 0.49 156 Table 2-1. (continued)

Samples ∑Fe2O3 Fe2O3 FeO Al2O3 TiO2 P2O5 Corg Ccarb Stotal [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) PTB3-01 7.44 1.26 5.56 24.69 0.72 0.04 0.41 0.00 0.00 PTB3-02 4.85 0.63 3.80 21.62 0.67 0.06 0.35 0.88 0.32 PTB3-03 8.85 0.14 7.84 24.88 0.70 0.20 0.46 0.00 0.00 PTB3-04 8.74 0.36 7.54 25.95 0.76 0.17 0.39 0.00 0.00 PTB3-05 10.18 0.57 8.65 24.73 0.72 0.40 0.39 0.00 0.00 PTB3-06 9.42 1.34 7.27 25.09 0.73 0.13 0.39 0.40 0.06 PTB3-07 9.96 1.39 7.71 24.91 0.71 0.10 0.36 0.28 0.00 PTB3-08 2.87 0.04 2.55 29.44 0.88 0.11 0.50 0.38 0.42 PTB3-09 9.04 0.71 7.49 26.36 0.77 0.17 0.43 0.01 0.00 PTB3-10 10.63 1.53 8.19 25.30 0.74 0.26 0.47 0.45 0.00 PTB3-11 8.61 1.09 6.77 25.54 0.72 0.14 0.43 0.00 0.00 PTB3-12 9.15 1.54 6.85 24.36 0.71 0.10 0.30 0.45 0.00

Average 8.31 0.88 6.68 25.72 0.75 0.16 0.41 0.24 0.07 S.D. 2.28 0.54 1.83 1.62 0.06 0.10 0.06 0.28 0.14

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) MSF6-11 9.63 0.58 8.14 18.43 0.72 0.07 0.28 0.00 0.00 MSF6-12 7.68 0.22 6.70 20.60 0.69 0.05 0.71 0.00 0.00 MSF6-13 6.96 1.27 5.12 18.85 0.63 0.36 0.55 0.00 0.00 MSF6-14 6.74 0.63 5.50 20.07 0.73 0.06 0.71 0.00 0.00

Average 7.75 0.67 6.37 19.48 0.69 0.13 0.56 0.00 0.00 S.D. 1.31 0.43 1.36 1.02 0.04 0.15 0.20 0.00 0.00

Red shales, Mapedi Fm, Pretoria G, Transvaal SG (~2.2 Ga) SA1677-01 10.91 8.04 2.58 16.98 1.07 0.08 0.10 0.00 0.00 SA1677-02 10.76 5.97 4.31 16.98 1.00 0.06 0.08 0.09 0.00 SA1677-03 9.52 7.98 1.39 19.47 1.07 0.09 0.09 0.00 0.00 SA1677-04 10.24 9.25 0.89 19.78 0.94 0.10 0.07 0.00 0.00 SA1677-05 13.26 12.42 0.76 18.79 1.02 0.09 0.07 0.00 0.00 SA1677-06 18.54 17.69 0.76 12.37 1.13 0.29 0.06 0.20 0.00 SA1677-07 14.75 14.19 0.50 14.50 2.32 0.78 0.06 0.72 0.00 SA1677-08 18.43 17.49 0.84 14.97 3.09 0.44 0.06 0.00 0.00

Average 13.30 11.63 1.50 16.73 1.45 0.24 0.07 0.13 0.00 S.D. 3.62 4.51 1.31 2.63 0.80 0.25 0.02 0.25 0.00

Fm: Formation, G; Group, SG: Supergroup, S.D. Standard deviation. 157 Table 2-2. Iron-related data for the Archean–Paleoproterozoic samples of this study.

Samples Fe3+/Ti Fe2+/Ti Fe3+/∑Fe HCl-Fe Py-Fe Reactive Fe DOP atomic ratio [wt.%] [wt.%] [wt.%]

Carbonate-rich black shales, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) PU1308-01 24.40 310.50 0.07 3.03 0.13 3.16 0.04 PU1308-02 1.66 163.95 0.01 3.06 0.00 3.06 0.00 PU1308-03 21.46 766.54 0.03 2.66 0.14 2.80 0.05 PU1308-04 24.51 212.67 0.10 2.69 0.59 3.28 0.18 PU1308-05 4.26 100.26 0.04 2.36 0.00 2.36 0.00 PU1308-06 23.47 402.64 0.06 2.90 0.00 2.90 0.00 PU1308-07 2.24 18.17 0.11 1.00 0.00 1.00 0.00 PU1308-08 21.45 230.05 0.09 2.77 0.00 2.77 0.00 PU1308-09 57.32 342.31 0.14 0.79 0.00 0.79 0.00 PU1308-10 0.36 28.50 0.01 0.74 0.00 0.74 0.00 PU1308-11 8.57 37.68 0.19 0.99 0.00 0.99 0.00 PU1308-12 9.40 173.69 0.05 2.53 1.87 4.40 0.43 PU1308-13 10.57 29.02 0.27 1.50 2.35 3.85 0.61 PU1308-14 7.63 268.28 0.03 3.72 0.00 3.72 0.00 PU1308-15 11.71 833.30 0.01 2.52 0.00 2.52 0.00 PU1308-16 6.71 189.74 0.03 2.45 0.00 2.45 0.00 PU1308-17 5.55 946.80 0.01 2.51 0.00 2.51 0.00

Average 14.19 297.30 0.07 2.25 0.30 2.55 0.08 S.D. 13.92 288.11 0.07 0.90 0.70 1.09 0.17

Graywackes, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) MRE10-01 1.35 12.80 0.10 0.30 0.00 0.30 0.00 MRE10-02 0.97 14.49 0.06 0.59 0.00 0.59 0.00 MRE10-03 0.46 12.76 0.04 0.39 0.00 0.39 0.00 MRE10-04 0.79 14.95 0.05 0.27 0.00 0.27 0.00 MRE10-05 1.11 11.55 0.09 0.20 0.00 0.20 0.00 MRE10-06 0.88 12.91 0.06 0.26 0.00 0.26 0.00 MRE10-07 1.31 13.63 0.09 0.23 0.00 0.23 0.00 MRE10-08 1.55 13.72 0.10 0.25 0.00 0.25 0.00 MRE10-09 1.23 13.96 0.08 0.34 0.00 0.34 0.00 MRE10-10 1.72 15.38 0.10 0.34 0.00 0.34 0.00 MRE10-11 1.74 14.48 0.11 0.29 0.00 0.29 0.00 MRE10-12 1.24 14.81 0.08 0.48 0.18 0.66 0.28

Average 1.20 13.79 0.08 0.33 0.02 0.34 0.02 S.D. 0.38 1.12 0.02 0.11 0.05 0.14 0.08 158 Table 2-2. continued

Samples Fe3+/Ti Fe2+/Ti Fe3+/∑Fe HCl-Fe Py-Fe Reactive Fe DOP atomic ratio [wt.%] [wt.%] [wt.%]

Iron-rich shales, Parktown Fm, West Rand G, Witwatersrand SG (2.96 Ga) DRH13-01 11.97 62.21 0.16 0.79 0.07 0.86 0.08 DRH13-02 18.96 80.45 0.19 - 0.07 - - DRH13-03 38.22 86.03 0.31 0.66 0.00 0.66 0.00 DRH13-04 9.06 56.27 0.14 - 0.00 - - DRH13-05 12.04 59.39 0.17 1.26 0.00 1.26 0.00 DRH13-06 8.78 53.27 0.14 1.09 0.60 1.69 0.35 DRH13-07 30.03 71.12 0.30 0.42 0.40 0.82 0.49 DRH13-08 118.34 125.85 0.48 0.49 0.00 0.49 0.00 DRH13-09 239.75 210.81 0.53 - 0.11 - - DRH13-10 4.61 30.91 0.13 0.49 0.26 0.75 0.35 DRH13-11 3.03 28.89 0.09 - 0.00 - - DRH13-12 0.51 8.69 0.06 0.29 0.00 0.29 0.00 DRH13-13 1.25 24.79 0.05 0.18 0.35 0.53 0.67 DRH13-14 1.91 9.97 0.16 - 0.30 - - DRH13-15 2.07 9.33 0.18 0.17 0.15 0.32 0.48

Average 33.37 61.20 0.21 0.59 0.15 0.77 0.24 S.D. 64.36 52.85 0.14 0.37 0.19 0.43 0.25

Shales, Pillingini Tuff Fm, Fortescue G, Mt. Bruce SG (2.72 Ga) WRL1-01 1.62 14.08 0.10 0.13 0.08 0.21 0.37 WRL1-02 1.61 5.69 0.22 0.29 0.06 0.35 0.18 WRL1-03 0.85 8.78 0.09 0.37 0.11 0.48 0.23

Average 1.36 9.52 0.14 0.26 0.08 0.35 0.26 S.D. 0.44 4.24 0.07 0.12 0.03 0.14 0.10

Black shales, Rietgat Fm, Platberg G, Ventersdorp SG (2.71 Ga) MSF6-01 0.34 9.12 0.04 0.11 0.11 0.22 0.50 MSF6-02 0.68 12.87 0.05 0.14 0.00 0.14 0.00 MSF6-03 0.85 8.68 0.09 0.15 0.00 0.15 0.00 MSF6-04 0.80 9.58 0.08 0.16 0.56 0.72 0.77 MSF6-05 1.03 9.32 0.10 0.34 0.33 0.66 0.50 MSF6-06 0.93 9.67 0.09 0.14 0.59 0.73 0.80 MSF6-07 0.83 10.20 0.08 0.33 0.46 0.79 0.59

Average 0.78 9.92 0.07 0.20 0.29 0.49 0.45 S.D. 0.22 1.39 0.02 0.09 0.26 0.30 0.33 159 Table 2-2. continued

Samples Fe3+/Ti Fe2+/Ti Fe3+/∑Fe HCl-Fe Py-Fe Reactive Fe DOP atomic ratio [wt.%] [wt.%] [wt.%]

Black shales, Lewin Shale Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) RHDH2A-01 2.53 3.01 0.46 0.07 0.18 0.25 0.72 RHDH2A-02 2.87 2.16 0.57 0.05 1.26 1.31 0.96 RHDH2A-03 3.10 3.27 0.49 0.09 0.06 0.15 0.40 RHDH2A-04 1.17 6.45 0.15 0.15 0.06 0.21 0.29 RHDH2A-05 1.99 5.86 0.25 0.15 0.51 0.66 0.77 RHDH2A-06 3.21 10.24 0.24 0.17 0.50 0.67 0.75 RHDH2A-07 0.21 12.23 0.02 0.13 3.51 3.64 0.96 RHDH2A-08 1.83 4.58 0.29 0.10 0.24 0.34 0.71

Average 2.11 5.97 0.31 0.11 0.79 0.91 0.70 S.D. 1.04 3.59 0.19 0.04 1.17 1.17 0.24

Black shales, Jeerinah Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) WRL1-04 3.72 2.50 0.60 0.97 1.35 2.31 0.58 WRL1-05 5.44 1.10 0.83 0.98 1.60 2.58 0.62 WRL1-06 5.76 1.53 0.79 0.99 2.60 3.59 0.72 WRL1-07 6.31 3.51 0.64 0.98 2.64 3.62 0.73 WRL1-08 3.65 12.32 0.23 0.85 1.70 2.55 0.67 WRL1-09 2.21 8.09 0.21 0.82 0.64 1.47 0.44

Average 4.52 4.84 0.55 0.93 1.75 2.69 0.63 S.D. 1.57 4.44 0.27 0.07 0.76 0.82 0.11

Carbonate-rich black shales, Black Reef Fm, Wolkberg G, Transvaal SG (2.64 Ga) JPBR-01 20.98 3.59 0.85 - 0.00 - - JPBR-02 3.34 40.36 0.08 - 0.56 - - JPBR-03 1.65 45.40 0.04 - 0.04 - - JPBR-04 1.65 4.56 0.27 - 0.00 - -

Average 6.90 23.48 0.31 - 0.15 - - S.D. 9.42 22.50 0.38 - 0.27 - - 160 Table 2-2. continued

Samples Fe3+/Ti Fe2+/Ti Fe3+/∑Fe HCl-Fe Py-Fe Reactive Fe DOP atomic ratio [wt.%] [wt.%] [wt.%]

Black shales, Marra Mamba Iron Fm, Hamersley G, Mt. Bruce SG (>2.60 Ga) WRL1-10 40.16 153.01 0.21 0.08 0.06 0.14 0.42 WRL1-11 115.52 163.82 0.41 0.27 0.32 0.60 0.54 WRL1-12 54.43 51.81 0.51 0.83 2.75 3.59 0.77 WRL1-13 43.80 74.32 0.37 0.09 0.12 0.21 0.57

Average 63.48 110.74 0.38 0.32 0.81 1.14 0.57 S.D. 35.22 55.99 0.13 0.35 1.30 1.65 0.14

Black shales, Wittenoom Dolomite Fm, Hamersley G, Mt. Bruce SG (2.60 Ga) WRL1-14 0.00 166.11 0.00 0.07 0.07 0.14 0.51 WRL1-15 4.70 18.72 0.20 1.42 1.43 2.85 0.50 WRL1-16 6.05 23.73 0.20 0.12 0.12 0.25 0.50

Average 3.58 69.52 0.13 0.54 0.54 1.08 0.50 S.D. 3.17 83.68 0.12 0.77 0.77 1.54 0.00

Carbonate-rich black shales, Carawine Dolomite Fm, Fortescue G, Mt. Bruce SG (2.60 Ga) RHDH2A-09 4.16 44.09 0.09 0.78 0.39 1.17 0.33 RHDH2A-10 5.76 8.28 0.41 0.16 0.07 0.23 0.30 RHDH2A-11 10.22 8.07 0.56 0.19 0.63 0.82 0.77 RHDH2A-12 3.67 4.10 0.47 0.21 0.12 0.33 0.37 RHDH2A-13 5.06 5.05 0.50 0.22 0.31 0.53 0.59 RHDH2A-14 1.98 2.22 0.47 0.12 1.07 1.19 0.90

Average 5.14 11.97 0.42 0.28 0.43 0.71 0.54 S.D. 2.80 15.91 0.17 0.25 0.37 0.41 0.25

Black shales, Oak Tree Fm, Chuniespoort G, Transvaal SG (2.56 Ga) MSF6-08 0.14 2.71 0.05 0.05 0.23 0.28 0.84 MSF6-09 0.15 4.54 0.03 0.06 0.82 0.88 0.93 MSF6-10 0.13 1.92 0.06 0.35 0.00 0.35 0.00

Average 0.14 3.06 0.05 0.15 0.35 0.50 0.59 S.D. 0.01 1.34 0.02 0.17 0.42 0.33 0.51 161 Table 2-2. continued

Samples Fe3+/Ti Fe2+/Ti Fe3+/∑Fe HCl-Fe Py-Fe Reactive Fe DOP atomic ratio [wt.%] [wt.%] [wt.%]

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) PTB3-01 1.74 8.54 0.17 0.06 0.00 0.06 0.00 PTB3-02 0.94 6.29 0.13 0.28 0.28 0.56 0.50 PTB3-03 0.20 12.38 0.02 0.08 0.00 0.08 0.00 PTB3-04 0.47 11.09 0.04 0.10 0.00 0.10 0.00 PTB3-05 0.80 13.37 0.06 0.14 0.00 0.14 0.00 PTB3-06 1.83 11.05 0.14 0.16 0.05 0.21 0.25 PTB3-07 1.96 12.03 0.14 0.16 0.00 0.16 0.00 PTB3-08 0.05 3.21 0.01 0.13 0.36 0.49 0.74 PTB3-09 0.92 10.76 0.08 0.10 0.00 0.10 0.00 PTB3-10 2.06 12.23 0.14 0.22 0.00 0.22 0.00 PTB3-11 1.52 10.48 0.13 0.06 0.00 0.06 0.00 PTB3-12 2.16 10.71 0.17 0.26 0.00 0.26 0.00

Average 1.22 10.18 0.10 0.15 0.06 0.20 0.12 S.D. 0.75 2.88 0.06 0.07 0.12 0.16 0.25

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) MSF6-11 0.81 12.66 0.06 0.39 0.00 0.39 0.00 MSF6-12 0.33 10.83 0.03 0.28 0.00 0.28 0.00 MSF6-13 2.00 8.99 0.18 0.26 0.00 0.26 0.00 MSF6-14 0.87 8.42 0.09 0.08 0.00 0.08 0.00

Average 1.00 10.23 0.09 0.25 0.00 0.25 0.00 S.D. 0.71 1.92 0.07 0.13 0.00 0.13 0.00

Red shales, Mapedi Fm, Pretoria G, Transvaal SG (~2.2 Ga) SA1677-01 7.55 2.69 0.74 - 0.00 - 0.00 SA1677-02 6.01 4.81 0.56 - 0.00 - 0.00 SA1677-03 7.44 1.44 0.84 - 0.00 - 0.00 SA1677-04 9.85 1.05 0.90 - 0.00 - 0.00 SA1677-05 12.14 0.83 0.94 - 0.00 - 0.00 SA1677-06 15.65 0.75 0.95 - 0.00 - 0.00 SA1677-07 6.13 0.24 0.96 - 0.00 - 0.00 SA1677-08 5.66 0.30 0.95 - 0.00 - 0.00

Average 8.80 1.51 0.85 - 0.00 - 0.00 S.D. 3.53 1.54 0.14 - 0.00 - 0.00

Fm: Formation, G; Group, SG: Supergroup, S.D.: Standard deviation. 162

Table 2-3. Stable isotopic compositions and various elemental ratios for the Archean– Paleoproterozoic samples of this study.

13 34 Samples δ C δ S Spy/Corg Corg/P Corg/Al P/Al S/Al [‰] [‰] wt. ratio wt. ratio wt. ratio wt. ratio wt. ratio

Carbonate-rich black shales, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) PU1308-01 -29.27 4.02 0.22 17.3 0.33 0.07 0.019 PU1308-02 -29.86 2.27 0.00 21.8 0.32 0.00 0.015 PU1308-03 -27.55 2.64 0.38 15.8 0.43 0.16 0.027 PU1308-04 -28.56 2.05 0.71 30.8 0.35 0.25 0.011 PU1308-05 -27.72 - 0.00 149.0 0.33 0.00 0.002 PU1308-06 -29.16 1.98 0.00 43.9 0.60 0.00 0.014 PU1308-07 -28.96 0.59 0.00 77.3 0.32 0.00 0.004 PU1308-08 -29.92 1.44 0.00 36.9 0.42 0.00 0.011 PU1308-09 -29.17 - 0.00 104.9 0.89 0.00 0.008 PU1308-10 -27.55 -2.72 0.00 31.3 0.18 0.00 0.006 PU1308-11 -27.38 -1.20 0.00 2.5 0.02 0.00 0.008 PU1308-12 -31.23 2.04 1.64 38.3 0.38 0.63 0.010 PU1308-13 -29.30 2.08 1.18 30.7 0.24 0.28 0.008 PU1308-14 -28.63 -3.11 0.00 173.1 0.58 0.00 0.003 PU1308-15 -28.07 -1.82 0.00 49.9 0.64 0.00 0.013 PU1308-16 -28.55 0.31 0.00 38.2 0.30 0.00 0.008 PU1308-17 -27.46 0.04 0.00 19.3 0.45 0.00 0.024

Average -28.73 0.71 0.24 51.8 0.40 0.08 0.011 S.D. 1.05 2.10 0.49 47.7 0.20 0.17 0.007

Graywackes, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) MRE10-01 -24.97 3.32 0.00 4.6 0.03 0.00 0.007 MRE10-02 -22.60 - 0.00 12.0 0.07 0.00 0.006 MRE10-03 -24.44 9.29 0.00 7.1 0.05 0.00 0.006 MRE10-04 -25.67 - 0.00 4.8 0.03 0.00 0.007 MRE10-05 -24.47 1.82 0.00 7.3 0.05 0.00 0.007 MRE10-06 -25.54 - 0.00 2.5 0.01 0.00 0.006 MRE10-07 -26.22 - 0.00 4.2 0.03 0.00 0.008 MRE10-08 -26.24 - 0.00 4.7 0.03 0.00 0.006 MRE10-09 -26.04 0.99 0.00 4.2 0.03 0.00 0.007 MRE10-10 -25.67 - 0.00 5.7 0.04 0.00 0.007 MRE10-11 -24.82 0.55 0.00 7.6 0.04 0.00 0.005 MRE10-12 -24.38 3.06 0.73 7.2 0.05 0.04 0.007

Average -25.09 3.17 0.06 6.0 0.04 0.00 0.007 S.D. 1.04 3.19 0.21 2.5 0.01 0.01 0.001 163

Table 2-3. (continued)

13 34 Samples δ C δ S Spy/Corg Corg/P Corg/Al P/Al S/Al [‰] [‰] wt. ratio wt. ratio wt. ratio wt. ratio wt. ratio

Iron-rich shales, Parktown Fm, West Rand G, Witwatersrand SG (2.96 Ga) DRH13-01 -28.59 3.08 0.32 8.3 0.05 0.02 0.006 DRH13-02 -25.94 3.40 1.00 3.5 0.02 0.02 0.006 DRH13-03 -26.70 - 0.00 2.8 0.03 0.00 0.012 DRH13-04 - - 0.00 1.1 0.01 0.00 0.010 DRH13-05 -26.85 - 0.00 3.2 0.03 0.00 0.010 DRH13-06 -26.60 2.77 9.25 1.2 0.01 0.12 0.012 DRH13-07 -26.63 3.72 11.08 1.1 0.01 0.11 0.009 DRH13-08 -26.57 - 0.00 1.9 0.06 0.00 0.029 DRH13-09 -26.97 - 0.49 21.8 0.17 0.08 0.008 DRH13-10 -26.90 - 0.59 21.4 0.07 0.04 0.003 DRH13-11 -27.64 - 0.00 7.0 0.01 0.00 0.001 DRH13-12 -26.82 - 0.00 8.6 0.10 0.00 0.012 DRH13-13 -28.94 1.58 2.32 7.3 0.02 0.06 0.003 DRH13-14 -27.40 - 1.29 8.7 0.03 0.04 0.003 DRH13-15 -27.77 1.75 0.79 51.2 0.02 0.02 0.000

Average -27.17 2.72 1.81 9.9 0.04 0.03 0.008 S.D. 0.82 0.87 3.47 13.2 0.04 0.04 0.007

Shales, Pillingini Tuff Fm, Fortescue G, Mt. Bruce SG (2.72 Ga) WRL1-01 -33.69 2.04 0.44 3.6 0.02 0.01 0.007 WRL1-02 -34.05 -0.08 0.27 4.5 0.03 0.01 0.008 WRL1-03 -32.54 2.86 0.95 2.9 0.02 0.02 0.008

Average -33.43 1.61 0.55 3.7 0.03 0.01 0.008 S.D. 0.79 1.52 0.35 0.8 0.01 0.01 0.001

Black shales, Rietgat Fm, Platberg G, Ventersdorp SG (2.71 Ga) MSF6-01 -34.25 0.75 0.40 7.9 MSF6-02 -34.50 - 0.00 7.6 0.05 0.00 0.007 MSF6-03 -36.64 - 0.00 18.0 0.09 0.00 0.005 MSF6-04 -40.60 0.40 0.68 19.7 0.13 0.09 0.007 MSF6-05 -44.28 -0.17 0.29 50.4 0.14 0.04 0.003 MSF6-06 -45.14 0.39 0.42 27.0 0.15 0.06 0.006 MSF6-07 -44.99 -0.30 0.48 17.4 0.12 0.06 0.007

Average -40.06 0.21 0.33 21.1 0.11 0.04 0.006 S.D. 4.91 0.44 0.25 14.6 0.04 0.04 0.002 164 Table 2-3. (continued)

13 34 Samples δ C δ S Spy/Corg Corg/P Corg/Al P/Al S/Al [‰] [‰] wt. ratio wt. ratio wt. ratio wt. ratio wt. ratio

Black shales, Lewin Shale Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) RHDH2A-01 -39.42 7.83 0.08 61.2 0.38 0.03 0.006 RHDH2A-02 -42.82 5.97 0.21 120.3 0.84 0.18 0.007 RHDH2A-03 -39.08 5.56 0.05 71.2 0.18 0.01 0.003 RHDH2A-04 -42.23 3.40 0.03 87.1 0.32 0.01 0.004 RHDH2A-05 -41.96 2.05 0.32 38.2 0.22 0.07 0.006 RHDH2A-06 -45.37 2.61 0.23 95.4 0.33 0.08 0.003 RHDH2A-07 -43.73 4.15 1.97 47.0 0.27 0.54 0.006 RHDH2A-08 -42.37 4.01 0.13 71.3 0.25 0.03 0.004

Average -42.12 4.45 0.38 74.0 0.35 0.12 0.005 S.D. 2.08 1.91 0.65 26.6 0.21 0.18 0.002

Black shales, Jeerinah Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) WRL1-04 -41.06 13.39 0.28 180.5 0.67 0.19 0.004 WRL1-05 -40.94 7.86 0.25 282.1 1.01 0.25 0.004 WRL1-06 -42.85 6.44 0.25 250.7 1.79 0.44 0.007 WRL1-07 -39.77 3.51 0.37 234.6 1.20 0.44 0.005 WRL1-08 -36.28 7.30 0.72 62.4 0.37 0.27 0.006 WRL1-09 -42.41 0.18 0.19 110.9 0.65 0.12 0.006

Average -40.55 6.45 0.34 186.9 0.95 0.29 0.005 S.D. 2.37 4.45 0.19 85.7 0.51 0.13 0.001

Carbonate-rich black shales, Black Reef Fm, Wolkberg G, Transvaal SG (2.64 Ga) JPBR-01 - - 0.00 21.7 0.23 0.00 0.011 JPBR-02 - 6.39 7.14 5.2 0.20 1.40 0.038 JPBR-03 -22.80 6.64 0.12 19.1 0.19 0.02 0.010 JPBR-04 -27.63 3.11 0.00 84.9 0.08 0.00 0.001

Average -25.22 5.38 1.82 32.72 0.18 0.36 0.015 S.D. 3.42 1.97 3.55 35.56 0.07 0.70 0.016 165 Table 2-3. (continued)

13 34 Samples δ C δ S Spy/Corg Corg/P Corg/Al P/Al S/Al [‰] [‰] wt. ratio wt. ratio wt. ratio wt. ratio wt. ratio

Black shales, Marra Mamba Iron Fm, Hamersley G, Mt. Bruce SG (>2.60 Ga) WRL1-10 - 11.41 0.00 10.9 0.05 0.02 0.005 WRL1-11 -37.66 13.73 1.01 27.9 0.28 0.28 0.010 WRL1-12 -40.35 5.27 3.41 42.5 0.39 1.34 0.009 WRL1-13 -43.31 8.41 0.19 28.4 0.35 0.07 0.012

Average -40.44 9.71 1.15 27.4 0.27 0.43 0.009 S.D. 2.83 3.67 1.57 12.9 0.15 0.62 0.003

Black shales, Wittenoom Dolomite Fm, Hamersley G, Mt. Bruce SG (2.60 Ga) WRL1-14 -40.41 4.77 0.11 84.6 0.89 0.10 0.011 WRL1-15 -39.05 0.07 0.59 127.4 0.37 0.22 0.003 WRL1-16 -32.04 3.76 0.05 101.9 0.41 0.02 0.004

Average -37.17 2.87 0.25 104.6 0.56 0.11 0.006 S.D. 4.49 2.47 0.30 21.5 0.29 0.10 0.004

Carbonate-rich black shales, Carawine Dolomite Fm, Fortescue G, Mt. Bruce SG (2.60 Ga) RHDH2A-09 -41.21 10.64 0.31 46.2 0.45 0.14 0.010 RHDH2A-10 -45.26 9.09 0.08 78.2 0.39 0.03 0.005 RHDH2A-11 -48.85 7.21 0.25 111.0 0.75 0.19 0.007 RHDH2A-12 -43.72 7.86 0.07 62.5 0.38 0.03 0.006 RHDH2A-13 -47.64 14.30 0.09 121.5 0.75 0.06 0.006 RHDH2A-14 -43.10 9.97 0.28 169.9 0.53 0.15 0.003

Average -44.96 9.85 0.18 98.2 0.54 0.10 0.006 S.D. 2.88 2.53 0.11 45.2 0.17 0.07 0.002

Black shales, Oak Tree Fm, Chuniespoort G, Transvaal SG (2.56 Ga) MSF6-08 -32.83 - 0.12 21.2 0.25 0.03 0.012 MSF6-09 -33.39 1.61 0.57 191.1 0.23 0.13 0.001 MSF6-10 -32.19 - 0.00 252.4 0.20 0.00 0.001

Average -32.80 1.61 0.23 154.9 0.23 0.05 0.005 S.D. 0.60 - 0.30 119.7 0.03 0.07 0.006 166 Table 2-3. (continued)

13 34 Samples δ C δ S Spy/Corg Corg/P Corg/Al P/Al S/Al [‰] [‰] wt. ratio wt. ratio wt. ratio wt. ratio wt. ratio

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) PTB3-01 -29.28 - 0.00 22.1 0.03 0.00 0.001 PTB3-02 -27.49 0.72 0.90 12.8 0.03 0.03 0.002 PTB3-03 -29.32 - 0.00 5.3 0.03 0.00 0.007 PTB3-04 -28.79 - 0.00 5.2 0.03 0.00 0.005 PTB3-05 -28.44 6.27 0.00 2.2 0.03 0.00 0.013 PTB3-06 -28.49 5.22 0.15 6.9 0.03 0.00 0.004 PTB3-07 -28.29 - 0.00 8.2 0.03 0.00 0.003 PTB3-08 -28.63 2.70 0.83 10.6 0.03 0.03 0.003 PTB3-09 -28.35 - 0.00 5.8 0.03 0.00 0.005 PTB3-10 -28.05 - 0.00 4.2 0.04 0.00 0.008 PTB3-11 -29.38 - 0.00 7.0 0.03 0.00 0.004 PTB3-12 -27.76 - 0.00 7.0 0.02 0.00 0.003

Average -28.52 3.73 0.16 8.1 0.03 0.00 0.005 S.D. 0.60 2.50 0.33 5.2 0.00 0.01 0.003

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) MSF6-11 -29.94 - 0.00 9.7 0.03 0.00 0.003 MSF6-12 -32.50 2.60 0.00 33.9 0.07 0.00 0.002 MSF6-13 -32.31 - 0.00 3.5 0.05 0.00 0.016 MSF6-14 -31.81 - 0.00 26.1 0.07 0.00 0.003

Average -31.64 2.60 0.00 18.3 0.05 0.00 0.006 S.D. 1.17 - 0.00 14.1 0.02 0.00 0.007

Red shales, Mapedi Fm, Pretoria G, Transvaal SG (~2.2 Ga) SA1677-01 -27.53 - 0.00 3.1 0.01 0.00 0.004 SA1677-02 -27.02 - 0.00 2.8 0.01 0.00 0.003 SA1677-03 -26.89 - 0.00 2.1 0.01 0.00 0.004 SA1677-04 -26.45 - 0.00 1.7 0.01 0.00 0.004 SA1677-05 -26.86 - 0.00 1.7 0.01 0.00 0.004 SA1677-06 -27.10 - 0.00 0.5 0.01 0.00 0.019 SA1677-07 -27.32 - 0.00 0.2 0.01 0.00 0.044 SA1677-08 -26.89 - 0.00 0.3 0.01 0.00 0.024

Average -27.01 - 0.00 1.6 0.01 0.00 0.013 S.D. 0.33 - 0.00 1.1 0.00 0.00 0.015

Fm: Formation, G; Group, SG: Supergroup, S.D.: Standard deviation. Chapter 3

Geochemistry of Archean–Paleoproterozoic black shales: III. Geochemical cycle of molybdenum

Abstract

Redox-sensitive metals such as Mo in sediments and sedimentary rocks have been suggested to have a great potential as indicators of sedimentary redox conditions, because dissolved Mo can be fixed by organic matter (OM) and sulfide. In this study, relationships among Mo, organic carbon (Corg), and S in the Archean–Paleoproterozoic black shales were examined to extract information on the paleoredox sedimentary environments. From geochemical analyses of 113 drillcore samples in the Fig Tree, West Rand, Platberg, Wolkberg, Chuniespoort, and Pretoria Groups in South Africa and the Fortescue and Hamersley Groups in Australia, we found that (1) Mo was fixed by sedimentary OM and sulfide in the Archean–Paleoproterozoic sediments and that (2) the Mo/Al and Mo/Corg ratios have large variations within each formation; the variations are similar to those in modern and Phanerozoic sediments. From the negative correlations between Mo and

Zn/Al2O3 ratios, we suggest that the relative Mo enrichment in shales did not result from precipitation of minerals from hydrothermal fluids. Molybdenum in common rocks is hosted mostly in pyrite and molybdenite. Based on the available data on the kinetics of pyrite oxidation, we estimate that pyrite crystals of 168 less than 1 mm3 in volume would be completely oxidized in soils or during transport within

4 5 -6 10 -10 years (i.e., a typical soil retention time) if the pO2 were higher than 10 atm, i.e., >0.0005 % PAL (present atmospheric level). The oxidation rates of molybdenite are probably similar to those of pyrite. Therefore, Mo was quantitatively transferred from continents to the oceans. The data obtained in this study lead us to suggest that the geochemical cycle of Mo during Archean–Paleoproterozoic time was essentially the same as today: Mo-bearing minerals were quantitatively oxidized during weathering, dissolved Mo was transported to the oceans, and the dissolved Mo was fixed by OM and S in locally anoxic marine environments. These suggestions imply that the Phanerozoic-style redox structure of the oceans (i.e. generally oxic ocean with localized anoxic/euxinic environments where bacterial sulfate reduction is active) may have already developed in the Archean. 169

3-1. Introduction

When did the Earth’s atmosphere become oxic? The rise of atmospheric O2 has far- reaching implications for the chemical evolution of atmosphere and biosphere. However, the timing and mechanisms of the rise of atmospheric O2 (e.g. 2.2-2.0 Ga GOE: Great Oxidation Event; see Holland, 1984, 1994) have been intensely debated for decades (see chapter 1), but no consensus has been reached. Because of a lack of a direct sample of a Precambrian atmosphere, we must use indirect approaches to constrain the oxygen contents of an ancient atmosphere. Geological indicators such as paleosols, iron-formations, uranium ores, and red beds have provided useful but indirect information on that subject (for recent summary of discussion, see Holland 1994, Ohmoto, 1997; Holland, 1999; Phillips et al., 2001). This study attempts to use the Mo geochemistry of Archean–Paleoproterozoic black shales to constrain the atmospheric pO2 level.

3-1-1. Redox-sensitive metals in black shales Black shales, i.e., carbonaceous shales, are commonly enriched in redox-sensitive metals (RSM) relative to average shales (Vine and Toutelot, 1970; Arthur and Sageman, 1994; Leventhal, 1993; Wignall, 1994), because organic matter (OM) and biologically produced sulfide are the major sink of RSM. There are many RSM that may take different valence states in the Earth's surface environments (Fig. 1-4). Such RSM typically show

different solubilities depending on the O2 content of water (e.g., Morford and Emerson,

1999). It has been suggested that concentration relationships between organic carbon (Corg) and RSM such as Mo, U, and V in marine shales are useful indicators of the atmospheric

O2 level (Holland, 1984, 1994). According to Holland's theory, an oxic atmosphere promotes the dissolution of RSM from rocks during weathering. This increases the concentrations of RSM in the 170

ocean. During the early diagenesis of carbonaceous sediments that are overlain by oxic bottom water, a redox gradient forms near or below the sediment-water interface (SWI)

because of consumption of dissolved O2 during degradation of sedimentary OM. Depending on the intensity and steepness of developed redox gradients, some RSM change their valence state and therefore decrease or increase their solubilities in pore waters (Fig. 3-

1). Therefore, characteristics of the RSM enrichments in ancient Corg-rich sedimentary rocks could potentially provide valuable information on paleoredox states of depositional environments (Leventhal, 1993). On the other hand, under an anoxic atmosphere such sedimentary fixation would not be effective because of significantly reduced oceanic concentration of RSM. Therefore, RSM geochemistry of Precambrian shales has potential to constrain the redox state of the Precambrian atmosphere.

3-1-2. Previous studies about Mo as a proxy of bottom water anoxia Among RSM, Mo has received particular attention because of its great potential as a proxy of paleoredox conditions of sedimentary environments. Extensive studies on Mo have been done for the recent marine sediments in the Black Sea (Hirst, 1974; Pilpchuk and Volkov, 1974; Volkov and Fomina, 1974; Calvert and Batchelor, 1978; Emelyanov et al., 1978; Holland, 1979; Brumsack, 1989; Calvert, 1990; Ravizza et al., 1991; Crusius et al., 1996), Cariaco Trench (Jacobs et al., 1987; Calvert and Pedersen, 1993; Dean et al., 1999), Saanich Inlet (Francois, 1988; Calvert and Pedersen, 1993; Crusius et al., 1996; Calvert et al., 2001; Morford et al., 2001; Russell and Morford, 2001), Baltic Sea (Salonen et al., 1995; Sternbeck, 2000; Sohlenius et al., 2001), Framvaren (Skei, 1981), Mediterranean Sea (Thomson et al., 1995; Nijienhuis et al., 1998; Warning and Brumsack, 2000), Gulf of California (Brumsack, 1986a; Brumsack, 1989; Brumsack and Gieskes, 1983), offshore California (Shimmield and Price, 1986; Shaw et al., 1990; Dean et al., 1997; Zheng et al., 2000), Gulf of Mexico (Huerta-Diaz and Morse, 1992), other Pacific, Atlantic, and Indian 171

oceans (Bertine and Turekian, 1973; Koide et al., 1986; Pedersen et al., 1988; Emerson and Husted, 1991; Thomson et al., 1993, 1996; Piper and Isaacs, 1995; Crusius et al., 1996; Morford and Emerson, 1999; Crusius and Thomson, 2000; Schaller et al., 2000; Thomson et al., 2001; Pailler et al., 2002), Japan Sea (Crusius et al., 1996), Antarctica (Bishop et al., 2001), and lakes (Brown et al., 2000) and estuaries (Malcom, 1985; Adelson et al., 2001). Although relatively fewer in number compared to modern sediments, many studies on Mo have been done for Phanerozoic black shales of Miocene (Piper and Isaacs, 1994, 2001; Hoppie and Garrison, 2001), Eocene (Holland, 1979), Cretaceous (Lange et al., 1977; Migdisov et al., 1980; Brumsack, 1980, 1986a, b; Arthur et al., 1990; Alberdi-Genolet and Tocco, 1999), (Holland, 1979; Coveney and Martin, 1983; Coveney et al., 1987, 1991; Coveney and Glascock, 1989; Desborough et al., 1991; Hatch and Leventhal, 1992; Caplan and Bustin, 1996, 1998), Devonian (Leventhal and Hosterman, 1982; Leventhal et al., 1983; Robl and Barron, 1988; Caplan and Bustin, 1996, 1998; Murphy et al., 2000; Joachimski et al., 2001), and (Armands, 1972; Leventhal, 1990) age. In contrast to the increasing number of published papers on Mo in modern and Phanerozoic sediments, however, very few studies on Mo have been performed for Precambrian (especially Archean) black shales to elucidate paleoenvironmental conditions.

According to Holland (1994), only three Archean Corg-rich shales have been studied for their Mo enrichment. One example is the 2.7 Ga low metamorphic-grade Jeerinah Formation in Western Australia (Davy and Hickman, 1988). The second example is the 2.5 Ga low metamorphic-grade Mt. McRae shale, again in Western Australia (Davy, 1983).

Very recently a significant positive correlation between Mo and Corg was found from the Mt. McRae Shale (Naraoka et al., submitted). The third example is the ~2.0 Ga highly metamorphosed rocks (mineralized ore) in Outokumpu, Finland (Loukola-Ruskeeniemi, 1991, 1999; Loukola-Ruskeeniemi et al., 1991; Loukola-Ruskeeniemi and Heino, 1996). Meyer and Robb (1996) studied the 2.5-2.6 Ga old black shales of the Chuniespoort 172

Group, South Africa. Geochemistry of Mo in Precambrian shales is crucially important not only for the potential to constrain the paleoredox environments of ancient atmosphere and hydrosphere. It is also important in understanding environmental conditions during the

evolution of biosphere, because Mo is a bio-essential nutrient (e.g., Mo for the N2-fixing enzyme nitrogenase).

3-1-3. Purpose of this study In order to examine the geochemical cycle of Mo in the Archean–Paleoproterozoic surface environments, we performed a geochemical investigation of the

Archean–Paleoproterozoic Corg-rich shales. The samples of this study are unweathered drillcore samples from the Swaziland, Witwatersrand, Ventersdorp, and Transvaal Supergroups in South Africa and Mt. Bruce Supergroup in Australia. They range in age

from 3.3 to 2.2 Ga. We provide a new set of geochemical data for their Mo-Corg-S contents from 113 samples from 12 Formations, mainly black shales (95) and some graywacke (10) and red shale (8) samples.

By examining their Mo-Corg-S relationships, we aim to evaluate the causes and environmental factors of Mo enrichment in the Archean–Paleoproterozoic black shales. From literature data, we also estimate the continental weathering rates of Mo as a function of

pO2. The ultimate goal of the present study is to constrain the redox state of the ancient atmosphere / hydrosphere. We also give suggestions about the evolution of the biosphere in terms of nutrient availability. This work is one of the studies designed to constrain paleoredox environments from the geochemistry of Archean–Paleoproterozoic black shales (see chapters 2, 4, and 5 of this thesis). 173

3-2. Geological settings and samples

The 113 samples used in this study belong to the Swaziland Supergroup, the Witwatersrand Supergroup, the Ventersdorp Supergroup, and the Transvaal Supergroup in South Africa and the Mt. Bruce Supergroup in Australia. The South African samples include the ~3.3 Ga Fig Tree Group (the 3.25 Ga Sheba Formation: 17 shales and 10 graywackes), the ~3.0 Ga West Rand Group (the 2.96 Ga Parktown Formation: 15 shales), the ~2.7 Ga Platberg Group (the 2.71 Ga Rietgat Formation: 7 shales), the ~2.6 Ga Wolkberg Group the (2.64 Ga Black Reef Formation: 4 shales), the ~2.6 Ga Chuniespoort Group (the 2.56 Ga Oak Tree Formation: 3 shales), the ~2.2 Ga Pretoria Group (the 2.22 Ga Timeball Hill Formation: 4 shales from Central Transvaal and 12 shales from Eastern Transvaal), and ~2.2 Ga Olifantshoek Group (the ~2.2 Ga Mapedi Formation: 8 shales). The Australian samples include the ~2.7 Ga Fortescue Group (the 2.72 Ga Pillingini Tuff Formation, the 2.69 Ga Jeerinah Formation, and the 2.69 Ga Lewin Shale Formation: 3, 6, and 8 shales, respectively) and the ~2.6 Ga Hamersley Group (the >2.60 Ga Marra Mamba Iron Formation: 4 shales, the 2.60 Ga Wittenoom Dolomite Formation: 3 shales, and the 2.60 Ga Carawine Dolomite Formation: 6 shales). More information on stratigraphy, locality, ages, geological / tectonic settings, and the number of the samples are summarized in Table 3-1.

3-3. Analytical methods

The methods for pulverization and elemental analyses of Al, Fe, Corg, Ccarb, and S are described in chapter 2 (see section 2-3-2 for pulverization, section 2-3-3 for Al and Fe analyses, and section 2-3-4 for Corg, Ccarb, and S elemental analyses). 174

Mo and Zn analyses A portion of powdered samples was completely digested either by mixed acid

(HNO3-HF-HClO4) method or alkali fusion method. To achieve complete-digestion of the samples containing refractory minerals with the acid mixture method, a "teflon-bomb" method (digestion in sealed teflon vessels with increased pressure) was used. Aluminum analysis was performed by ICP-AES (Inductively Coupled Plasma - Atomic Emission Spectrometry) at the Pennsylvania State University or by XRF (X-ray Fluorescence) at Ocean Research Institute, University of Tokyo. Mo analysis was performed by ICP-MS (Mass Spectrometry) at Penn State (Finnigan ELEMENT) and ACTLAB, Inc. (ELAN 6000). Procedure blank samples and several international standard samples (SCo1, MAG1, BCR2, and W2: US Geological Survey standard reference materials) were also analyzed to ensure analytical accuracy and to track possible contamination during experimental procedures. The detection limits for Mo and Zn are 0.1 ppm. The reproducibilities for Mo and Zn were better than ± 5 %.

3-4. Results

The results of chemical analyses (Mo, Corg, Ccarb, S, Al2O3, ∑Fe2O3, and Zn contents) of the Archean–Paleoproterozoic shales and graywackes of this study are summarized in Table 3-2. Average contents and one standard deviation (1σ) of elemental concentrations are calculated and shown for each formation. See the sections 2-4-5, 2-4-1,

2-4-7, and 2-4-8 for the results of Al, Fe, Corg, Ccarb, and S contents of the studied samples. 175

3-4-1. Relationship among Mo, organic C, and pyrite S contents

Figure 3-1 shows the relationships among Mo, Corg, and S contents of the samples of this study. See sections 2-4-7 and 2-4-8 for the results of Corg and S contents, respectively.

Positive correlations between Mo and Corg are seen in the 3.25 Ga Sheba Formation (Fig. 3-1-a and -c), the 2.96 Ga Parktown Formation (Fig. 3-1-e), the 2.72 Ga Pillingini Tuff Formation and the 2.71 Ga Rietgat Formation (Fig. 3-1-g), the 2.69 Ga Jeerinah and Lewin Shale Formation (Fig. 3-1-i), the >2.60 Ga Marra Mamba Iron Formation and the 2.60 Ga Wittenoom and Carawine Dolomite Formations (Fig. 3-1-k), the 2.64 Black Reef Formation (Fig. 3-1-m), and the 2.22 Ga Timeball Hill Formation (Fig. 3-1-o). However,

the slopes of regression lines for each Mo-Corg plot are different; the majority of the

samples exhibit lower Mo/Corg ratios compared to the Mo/Corg ratios of average shale and average black shale (Vine and Tourtelot, 1970; Wedepohl, 1991; Warning and Brumsack, 2000). The 3.25 Ga Sheba Formation (graywackes), the 2.96 Ga Parktown Formation, and the 2.72 Ga Pillingini Tuff Formation show Mo/Corg ratios comparable to average shale

(Fig. 3-1-c, -e, -g), all of which have low Corg contents (< ~0.5 wt.%). No sample in this study exceed 10 ppm for Mo contents. Positive correlations between Mo and S, if any, are less obvious than those between

Mo and Corg. The majority of the black shales in this study lacks enrichment of S. For example, all the samples except for three (i.e., 24 samples) in the 3.25 Ga Sheba Formation have S contents less than 0.2 wt.%, and the other samples are also typically low in S (< 1 wt.%) except for the 2.69 - 2.60 Ga samples (Jeerinah, Lewin Shale, Marra Mamba Iron, Carawine and Wittenoom Dolomite Formations). 176

3-4-2. Zinc contents The Zn content of shales is used in a later discussion as an indicator of hydrothermal influence, since Zn mineralization typically occurs in sediments / sedimentary rocks influenced by precipitation of minerals from hydrothermal fluids. Compared to the Zn content of PAAS (85 ppm; Taylor and McLennan, 1985), the 3.25 Ga Sheba Formation (shales) show higher Zn contents with wide variations (213 ± 246 ppm; Table 3-2). The nearly contemporaneous graywackes have lower Zn contents of 115 ± 39 ppm. The 2.69 Ga Jeerinah and Lewin Shale Formation have variable Zn contents with a mean of 133 ± 175 ppm (range in 0 - 816 ppm) for the Jeerinah Formation and 309 ± 333 ppm (range in 0 - 527 ppm) for the Lewin Shale Formation. The other samples have average Zn contents lower than 100 ppm and less variable compared to the 2.69 Ga Formations. Since many of the samples of this study contain significant amounts of chemical precipitates such as carbonate and Fe-oxide, a direct comparison of Zn contents between them and PAAS may not be adequate. Later, Zn contents are discussed after normalization

against the clastic component using Al2O3 to examine relative Zn enrichments in the samples.

3-5. Discussions

3-5-1. Molybdenum in geologic materials Molybdenum contents of felsic igneous rocks are slightly higher than in mafic rocks (1.1 ~ 1.8 ppm vs. 0.3 ~ 1.2 ppm; Table 3-3). In granitic rocks, Mo contents higher

than the average crustal value of ~1 ppm may be largely attributable to MoS2, whereas excess Mo contents in mafic rocks is chiefly associated with Fe-containing minerals. Metamorphic rocks also have similarly low Mo contents (< 1 ppm; Table 3-3). Sedimentary 177

rocks have variable Mo contents. Carbonaceous sediments / shales and ferromanganese nodules have significantly elevated Mo contents (Table 3-3); however, other sediments / sedimentary rocks have low (< 2 ppm) Mo contents (Table 3-3). Most of the Mo in sedimentary, metamorphic and igneous rocks occurs as a trace

constituent of pyrite, minor molybdenite (MoS2), and kerogen, and therefore they are considered here to be quantitatively important for the continental weathering flux of Mo. Ideally, from experimentally-determined oxidation rates of Mo-bearing pyrite, molybdenite,

and kerogen under various pO2 conditions, it is possible to quantitatively estimate the

weathering flux of Mo from the continents to the oceans at a given pO2 level. Unfortunately, however, because of a dearth of such data sets, we have to rely on approximation.

3-5-2. The pO2 dependence of Mo weathering flux from continents Molybdenum is introduced into the oceans mainly through runoff from the continental weathering and through hydrothermal emanation from the seafloor, and removed from the oceans mainly by burial into the sediments (Fig. 3-3). Both of the influx of Mo to and the outflux from the ocean depend on the atmospheric and oceanic redox environments.

The principal reaction governing the transport and deposition of molybdenite (MoS2) is:

2- 2- + ...... MoS2 + 4.5 O2 + 3 H2O <---> MoO4 + 2 SO4 + 6 H (3-1)

For comparison, that of pyrite (FeS2) is:

2- +...... FeS2 + 4.5 O2 + 2 H2O <---> Fe(OH)3 + 2 SO4 + 4 H (3-2)

An Eh-pH diagram for the Mo-S-O2-H2O system constructed using thermodynamic data in Brookins (1988) is shown in Fig. 3-3. Molybdenum oxanions 178

2- - (MoO4 and HMoO4 ) are stable forms of Mo under “normal” oxic and pH conditions. -60 Molybdenite is stable only under extremely reducing conditions (i.e., pO2 < 10 ). Under globally reducing atmosphere and oceans, one may expect that the weathering flux of Mo into the ocean would be dominated by detrital flux (Mo in reduced form) rather than by dissolved flux (Mo in oxidized form), and that the magnitude of the Mo weathering flux itself would be small compared to that in oxidizing environments (Fig. 3-3, 3-4). This is because the oxidation of Mo-bearing minerals (e.g., molybdenite) may be ineffective under low pO2. In low pO2 environments, Mo contents of marine sediments are expected to be (1) higher in coarse-grained sediments such as near-shore sands in proximal setting and lower in fine-grained sediments such as pelagic in distal settings, because the sedimentary distribution of Mo would have been governed by physical transport processes

of Mo-bearing, typically heavy minerals such as molybdenite (MoS2), and (2) not variable among fine-grained sediments because physical transport such as ocean currents and suspension flow followed by redeposition is the only effective process for the distribution of Mo-bearing minerals into the oceans. On the other hand, under globally oxic atmosphere and oceans with localized anoxic basins (like today's environments), the weathering flux of Mo into the oceans is dominated

2- by dissolved flux (molybdate MoO4 ) rather than unoxidized forms (Fig. 3-3, 3-4). Mo contents of marine sediments are expected to be highly variable depending on the development of local anoxic conditions, whether it is in the water column or in sediments, which enhance Mo fixation into the sediments (detailed mechanisms will be discussed later). No study has experimentally and quantitatively evaluated whether the weathering flux of Mo is indeed enhanced in oxidizing environments compared to reducing environments and to what extent it can be influenced by different degrees of oxygenation in the weathering environments. Therefore, it is necessary to examine the dependency of 179

weathering flux of Mo on the pO2 levels of the atmosphere before we proceed to the discussion of Mo geochemistry in Precambrian black shales.

Rate of Mo weathering Molybdenite is a very minor constituent of rocks compared to other Mo-bearing phases, pyrite and kerogen. The dissolution rates of Mo between pyrite and molybdenite may be expected to be similar because of their geochemical similarities, including crystal structures. Therefore, as a first-order approximation, the weathering flux of Mo during soil formation can be quantitatively evaluated by experimental data on the kinetics of pyrite

oxidation under different pO2 conditions (e.g., Mathews and Robins, 1974; Bailey and Peters, 1976; McKibben and Barnes, 1988; Nicholson et al., 1988, 1990; Williamson and Rimstidt, 1994; Holmes and Crundwell, 2000; Kamei and Ohmoto, 2000) and by a theoretical model of oxidation of kerogen in soil (Lasaga and Ohmoto, 2002). Here we attempt to estimate the continental weathering flux of Mo from pyrite oxidation. Continental weathering flux of Mo from kerogen is not considered here. Williamson and Rimstidt (1994) compiled experimental data from the literature for the oxidation of pyrite under different pO2 conditions, and reported the rate equation of pyrite oxidation indicating its kinetics as a function of dissolved O2 and pH:

-8.19 (±0.10) 0.5 (±0.04) 0.11 (±0.01) ...... r = 10 x [mDO ] / [mH+ ] (3-3)

2 where r is the rate of pyrite oxidation in units of [mol/m /s], mDO is the dissolved O2

concentration in [mol/kg], and mH+ is the proton concentration in [mol/kg]. Recently, Holmes and Crundwell (2000) reported a very similar rate equation for pyrite oxidation. It is not the purpose of the present study to detail the mechanisms of pyrite oxidation nor to examine various types of rate equations published by others (e.g., Mathews and Robins, 180

1974; Bailey and Peters, 1976; McKibben and Barnes, 1988; Nicholson et al., 1988, 1990; Williamson and Rimstidt, 1994; Kamei and Ohmoto, 2000). For this study we adopt the rate equation 3-3 for pyrite oxidation proposed by Williamson and Rimstidt (1994). We calculated the rate of pyrite oxidation under the conditions which seems to have existed in the Archean weathering environments: T = 40 ˚C and pH = 5. Average surface temperature in the Archean might have been higher than 40 ˚C due to significant greenhouse effects by denser CO2 atmosphere (e.g., x ~100 ~ 1000 PAL; e.g., Kasting, 1987). Previous studies (Knauth and Epstein, 1976; Knauth and Lowe, 1978; Karhu and Epstein, 1986) estimated the Archean seawater temperature to be as high as 70-80 ˚C from oxygen isotope studies of Archean cherts and phosphates. With the increase of surface temperature, solubility of oxygen in the seawater decreases (Henry's law constant KH = 0.00120

[mol/l/am] at T = 40 ˚C and KH = 0.00107 [mol/l/atm] at T = 80 ˚C; Stumm and Morgan, 1996). The pH in surface environments also might have been lower, i.e., more acidic, due to higher pCO2 (e.g., Kasting, 1987). Today's rainwater pH is ~5.7, and the Archean rainwater pH was very likely to have been lower than the present value if Archean pCO2 was higher than today (e.g., Ohmoto, 1999). According to the rate equation 3-3, a decrease in pH and an increase in surface temperature result in a decrease in the rate of pyrite oxidation.

Using the Henry's law constant KH = 0.00120 [mol/l/atm] at T = 40 ˚C (Stumm and Morgan, 1996), pH = 5, and the molar volume of pyrite (23.94 [cm3/mol]: Seward and Barnes, 1997), equation 3-3 becomes as follows:

-8.19 0.5 0.11 2 r = 10 x [(pO2 x KH) ] / [mH+ ] [mol/m /s] -9.10 0.5 2 = 10 x [pO2] [mol/m /s] 0.5 3 4 ...... = 0.60 x [pO2] [cm /10 yr] (3-4) 181

Figure 3-4 shows rate of pyrite oxidation under varying pO2 levels for a given 4 5 duration for reaction (10 and 10 years). For example, under the present-day pO2 level, pure pyrite crystals of less than 1 mm3 in volume will be completely oxidized within 104 years, which is a conservative estimate of time for soil retention in a weathering profile. A pyrite crystal size of 1 mm3 in volume is used here because disseminated pyrite and molybdenite crystals in carbonaceous shales are quantitatively the most important constituent of Mo with respect to continental weathering flux and their sizes are typically much less than 1 mm3. A decrease in pH by 1 unit will decrease the rate by a factor of ~0.8, and if the surface (reaction) temperature of pyrite oxidation is 80 ˚C then the rate will decrease by a factor of 0.03. Our calculation is rather conservative, and the actual rate of pyrite oxidation will be greatly enhanced by orders of magnitude. This significant underestimate is most likely due to the crude assumptions that the pyrite crystals remain free of crystal defects, fractures, and surface roughness during the oxidation process, and that the congruent reaction is purely an inorganic process. In the real world, pyrite crystals are almost certainly extensively fractured as weathering proceeds, and thus the surface area for oxidation will be increased by several orders of magnitude. It has also been demonstrated by many previous workers that microbially-mediated reactions will significantly accelerate the weathering rate of pyrite (e.g., Baldi et al., 1992; Mustin et al., 1993; Karavaiko et al., 1994; Nordstrom and Southam, 1997; Edwards et al., 1998; Yu et al., 2001). These fracturing and microbial processes will very likely offset the above-mentioned effect of pH and temperature, and greatly increase the overall rate of pyrite decomposition. Figure 3-4 also includes the case ('black large dot' in Fig. 3-4) where the surface area (roughness) of pyrite weathering is increased by a factor of 103, suggesting that pure pyrite crystals of less than 1 mm3 in volume will be completely oxidized within 104 years if

-6 pO2 is higher than 10 [atm]. Molybdenum contained in pyrite and molybdenite of less 182

than ~1 mm3 in volume would therefore be completely leached from a soil horizon into solution within a reasonable amount of time (e.g., an order of 104~105 years as a typical soil retention time; Lasaga and Ohmoto, 2002) and eventually be transported to the oceans,

-6 provided that pO2 is higher than 10 atm.

Rye and Holland (1998) proposed that pO2 value before 2.2 Ga was equal or less than 0.0002 atm (= 0.1 % PAL), which is larger by a factor of 200 than the value used above. If pO2 value was much higher than 0.0002 atm in the Archean as some researchers propose (e.g., Dimroth and Kimberley, 1976; Clemney and Badham, 1982; Towe, 1990, 1991, 1994; Ohmoto, 1996, 1997), then it would greatly increases the weathering flux of Mo contained in pyrite. Molybdenum resides in not only pyrite (molybdenite) but also in or associated with organic matter (kerogen: quantitatively significant for total Mo flux) and in silicate minerals (quantitatively very minor). Weathering of these Mo-bearing phases increases the overall weathering flux of Mo from continent to the oceans. Importantly, it has been recently proposed that pyrite crystals may be decomposed

in the absence of O2 by a reaction involving H2O2 (Borda et al., 2001). Furthermore, pyrite crystals may be decomposed by UV radiation (Borda et al., 2001). The UV flux to the Earth from the young Sun would have been much higher in the Archean than now. These observations suggest that pyrite may be qualitatively decomposed independently of the atmospheric pO2. Considering all of the factors discussed above, it appears very likely that Mo was quantitatively transported from continent to the ocean by oxidative weathering in the

Archean, to the same degree as modern Mo weathering, even though the pO2 value was as low as 10-6 PAL (Holland, 1994). 183

3-5-3. Source, sink, and sedimentary enrichment mechanisms of Mo

3-5-3-1. Source and sink of Mo The sources and sinks of Mo and their fluxes in present-day environments are summarized in Table. 3-5. The river flux (i.e., continental weathering flux) dominates the input flux of Mo to the oceans. The input of Mo from submarine hydrothermal activity is considered to be negligible (Table 3-5). The major sink of Mo from the oceans is both oxic and anoxic sediments. The uptake of Mo in MOR through hydrothermal circulation is less than ~10 % of the total sink (Table 3-5). Therefore, in the discussion below we consider the continental weathering flux of Mo and the flux of Mo to sediments with respect to the geochemical cycle of Mo.

3-5-3-2. Mechanisms of Mo enrichment Several hypotheses have been proposed as a mechanism for Mo enrichment in marine sediments. Dissolved Mo in the oceans enter in the sediments via: (1) coprecipitation

2- with Mn (and Fe) oxyhydroxides followed by release of molybdate MoO4 (Bertine and Turekian, 1973; Brumsack and Gieskes, 1983; Shimmield and Price, 1986; Shaw et al., 1990; Emerson and Huested, 1991; Crusius et al., 1996), (2) coprecipitation with Fe sulfide minerals (Bertine, 1972), (3) precipitation with settling OM containing Mo (Bertine, 1972; Calvert and Morris, 1977; Brumsack and Gieskes, 1983), and (4) diffusion from bottom water across the sediment-water interface into anoxic sediments (Francois, 1988; Shaw et al., 1990; Emerson and Huested, 1991; Crusius et al., 1996; Zheng et al., 2000). In order for Mo to be enriched in the sediments, it needs to be “fixed” in mineral form. Fixation occurs within sediments typically by sulfide (Huerta-Diaz and Morse, 1992; Helz et al., 1996; Adelson et al., 2000). 184

3-5-4. Burial flux of Mo in the Archean–Paleoproterozoic sediments

3-5-4-1. Molybdenum enrichment in shales: authigenic? hydrothermal? Although Mo is enriched in carbonaceous sediments through diagenetic fixation by OM and bacteriogenic sulfide, syn- and/or post-depositional hydrothermal fluids may precipitate Mo and obscure the geochemical signature of Mo. To examine the possible effects of hydrothermal overprints, we examine the relationships between Mo and Zn in the shales. Zinc is used as a representative element for a hydrothermal component (e.g., James and Elderfield, 1996; Nijenhuis et al., 1998; Naraoka et al., submitted) because significant Zn enrichments are typically found in hydrothermally influenced sediments, such as those near the East Pacific Rise and in the Red Sea.

Figure 3-5 shows the relationship between Mo contents and Zn/Al2O3 ratios. Zinc

contents are normalized against the clastic component, Al2O3, in order to distinguish the hydrothermal vs. detrital Zn. Detrital fractions have an appreciable amount of Zn (85 ppm

Zn and Zn/Al2O3 = 4.8 for PAAS; Taylor and McLennan, 1985). Positive correlations

between Mo contents and Zn/Al2O3 ratios are not observed. Instead, the data exhibit slight negative correlations with considerable scatter. For example, trends of increasing Mo

contents without a significant increase in Zn/Al2O3 ratios are observed in samples of the 3.25 Ga Sheba Formation (Fig. 3-5-a, -b), the 2.72 Pilingini Tuff Formation and the 2.71 Ga Rietgat Formation (Fig. 3-5-d), the 2.64 Ga Black Reef Formation and the 2.56 Ga Oak Tree Formation (Fig. 3-5-g), and the 2.22 Ga Timeball Hill Formation (Fig. 3-5-h). More importantly, overall negative correlations between Mo contents and Zn/Al2O3 ratios are observed for the 3.25 Ga Sheba Formation (Fig. 3-5-a, -b), the 2.96 Ga Parktown Formation (Fig. 3-5-c), the 2.69 Ga Jeerinah and Lewin Shale Formations (Fig. 3-5-e), and the 2.60 Ga Wittenoom Dolomite Formation (Fig. 3-5-f). 185

Carbonaceous sediments are good scavengers of metals dissolved in the oceans, and

not only Mo but also Zn can be fixed in the Corg-rich sediments. However, the negative

correlations between Mo and Zn/Al2O3 (and, therefore, negative correlation between Corg

and Zn/Al2O3: because of the positive correlations between Mo and Corg) suggest that the enrichments of Mo in the Archean–Paleoproterozoic shales of this study were not significantly affected by precipitation of minerals from hydrothermal fluids but were the results of diagenetic, authigenic enrichment by sedimentary OM and biogenic sulfides.

3-5-4-2. Molybdenum enrichment inferred from the Mo/Al ratios When we evaluate the amount of Mo fixed from seawater into sediments and sedimentary rocks based on the total Mo contents of the rocks, a problem may arise if the contribution of Mo in detrital materials (e.g., Mo-rich silicates) is high. Furthermore, a 'dilution' by clastic materials and other chemical components (e.g., carbonate and Fe oxides) may greatly affect total Mo contents in samples. Therefore, with an assumption that all the Al in sediments is detrital, we normalized the measured Mo contents against Al (i.e., aluminosilicate fraction) and estimated the degree of Mo enrichment in the shales. The Mo/Al (ppm/wt.%) ratios of the samples of this study are further normalized against the

Mo/Al ratio of average shale ([Mo/Al]Av.Sh. = 0.15; Turekian and Wedepohl, 1961; Taylor and McLennan, 1985; Wedepohl, 1991; Warning and Brumsack, 2000), and presented in histograms for each formation in Fig. 3-6. The data are summarized in Table 3-4, together with published data for the 1.9 Ga Labrador shales (Hayashi et al., 1997).

Archean–Paleoproterozoic Figure 3-6 clearly displays variable degrees of Mo enrichment within each formation of this study, typically up to a factor of 10, and consistent enrichment of Mo compared to the average shale (i.e., 1 < [Mo/Al]sample / [Mo/Al]Av. Sh. < 10 ). Graywackes 186

also show similar degree of variability in their [Mo/Al]sample / [Mo/Al]Av. Sh. ratios (Fig. 3- 6-m) when compared to the other shales samples (Fig. 3-6). Greater variability is found in the 3.25 Ga Sheba Formation (shale), 0.2 ~ 21.3, and the 2.22 Ga Timeball Hill Formation, 0.5 ~ 5.5 (Table 3-4; Fig. 3-6-n and -b, respectively). One sample in the 3.25 Ga Sheba

Formation (shales) show a depletion of Mo (i.e., [Mo/Al]sample / [Mo/Al]Av. Sh. < 1). The ~2.2 Ga red shales of the Mapedi Formation exhibit generally depleted Mo contents relative to Al (Fig. 3-6-a). The black shales of the 1.9 Ga Ramah Group in Labrador (Hayashi et al.,

1997) also display a similar distribution (average [Mo/Al]sample / [Mo/Al]Av. Sh. ratio = 1.5; Table 3-4) to the samples of this study.

Modern and Phanerozoic

The [Mo/Al]sample / [Mo/Al]Av. Sh. ratios for modern (carbonaceous) sediments and Phanerozoic sedimentary rocks (mainly black shales) are also included in Fig. 3-6 for comparison. Those sediments and sedimentary rocks display large variability in the

[Mo/Al]sample / [Mo/Al]Av. Sh. ratios (as large as 4 log unit) regardless of their ages, suggesting the importance of the redox controls on the enrichment / depletion of Mo in the sediments. Modern and Phanerozoic non-carbonaceous sediments are generally depleted in Mo, while carbonaceous ones are often, but not always, enriched in Mo. Extreme enrichment of Mo (i.e., [Mo/Al]sample / [Mo/Al]Av. Sh. > ~50) in modern and Phanerozoic sediments has been frequently observed in euxinic environments (e.g., Black Sea, Saanich Inlet, Cariaco Trench, Framvaren Fjord, etc.). In euxinic environments, the increased preservation of OM and formation of sulfide create favorable conditions for sedimentary Mo enrichments. The overlying oxic water body acts as source of Mo to the underlying euxinic water body; the former provides continuous supply of dissolved sulfate and Mo to the latter. 187

Mechanisms to cause variations in the Mo/Al ratios

There are striking similarities in the [Mo/Al]sample / [Mo/Al]Av. Sh. ratios (i.e., ranges and variabilities) among the modern sediments, Phanerozoic shales, and the Archean–Paleoproterozoic shales of this study, implying that the similar mechanisms for the sedimentary enrichment of Mo have operated throughout geologic ages (Fig. 3-6). The observed large variabilities in the degree of Mo enrichment for the Archean- Paleoproterozoic shales of this study can not be explained by rapid changes in the source rock compositions or the distance between the depositional environments and the source areas, because (1) Mo contents of major rock types of the continental crust are almost identical (Table 3-3), (2) such rapid changes in the source rock compositions are very unlikely to have consistently occurred for each formation examined in this study and in the 1.9 Ga Ramah Group, and (3) the and grain sizes of the samples in each formation of this study are almost identical and homogeneous. Since the post-depositional influences (e.g., metamorphism and hydrothermal alteration) are minor for the samples of this study, the most likely explanation for the observed large variations in Mo enrichment for the Archean–Paleoproterozoic samples is the fluctuating redox conditions around the sediment-water interface. Molybdenum fixation into the sediments by OM and/or authigenic sulfide may be enhanced when the sediments are overlain by anoxic / sulfidic bottom water, while the dissolution of Mo in sediments and/or an escape of Mo from sediments may occur when the sediments are occasionally overlain by an oxic bottom water. The latter processes, i.e., the diffusive penetration of dissolved O2 or a downward migration of an oxidation front into sediments, have been suggested to have caused RSM redistribution in the OM-rich sediments of the Mediterranean Sea (e.g., Wilson et al., 1985; Van Santvoort et al., 1996). From a sediment pore water study, Shaw et al. (1990) showed the release of Mo to sediment pore water from 188

solid phases during oxic degradation of OM. There is no significant Mo enrichment in oxic sediments except for Fe/Mn nodules or Fe oxyhydroxide-rich surficial sediments.

Implications of variable Mo/Al ratios of shales Fine-grained clastic sediments (shales) typically deposit (far) below the wave base. Oxygenation of deep water becomes possible only when oxygenated surface water, overlain by an oxic atmosphere, is brought to the deep oceans by the oceanic circulation (much like

the today's situation). Dissolved O2 brought to deep oceans is mainly consumed to oxidize dissolved / particulate OM. Sarmiento (1992) has suggested that a minimum pO2 of 0.1 atm (0.5 PAL) is necessary to maintain the present-day deep water oxygenated. However, such pO2 level partly depends on the distribution and type of OM in the oceans. Modern marine OM exists in particulate and dissolved forms (POM and DOM, respectively). The Archean–Paleoproterozoic oceans would have been dominated by DOM, because POM would not have formed in a bacteria-dominated microbial ecosystem (metazoan was not yet evolved in Archean–Paleoproterozoic). Primary productivity in the Archean–Paleoproterozoic oceans can be assumed to have been similar to that of today, from the inferred operation of the similar geochemical cycle of C through time based on the

Corg burial flux and the Corg isotope records (see chapter 2). Then, more dissolved O2 was necessary to oxidize DOM in the Archean–Paleoproterozoic oceans than that of today (e.g., Logan et al., 1995) and thus the deep oceans could have been driven to be less oxic. However, the observed variations in the shale Mo/Al ratios probably due to fluctuating redox conditions suggest that the Archean–Paleoproterozoic deep oceans were oxic enough to have caused such variations. This further suggests that the Archean–Paleoproterozoic deep oceans were oxic enough to allow ventilation with O2 and remobilization of Mo (i.e., pO2 > 0.1 atm). In Archean–Paleoproterozoic, the minimum pO2 to maintain the oxygenated deep water could have been lower than 0.1 atm (Sarmiento, 1992) if the 189

Archean–Paleoproterozoic primary productivity was lower than that of today despite the similar geochemical cycle of C though time.

The similarities in the [Mo/Al]sample / [Mo/Al]Av. Sh. ratios among the Archean–Paleoproterozoic shales of this study and the 1.9 Ga shales (Table 3-4; Hayashi et al., 1997) should be emphasized, because these samples are before and after the inferred GOE at 2.2 - 2.0 Ga of Holland (1994).

3-5-5. Redox controls on the relationship among Mo, organic C, and pyrite S contents

To examine the degree of fixation of Mo by OM, the Mo/Corg ratios are examined in this section. Later, relationships between Mo and Corg and Mo and S for modern carbonaceous sediments and Phanerozoic black shales are examined and compared to those for the Archean–Paleoproterozoic shales of this study.

3-5-5-1. Archean–Paleoproterozoic

Figure 3-7 shows histograms of the Mo/Corg ratios of the studied samples normalized against the Mo/Corg ratio of the average shale (6.5; Wedepohl, 1991; Warning and Brumsack, 2000). Large variations in the [Mo/Corg]sample / [Mo/Corg]Av. Sh. ratios of

up to a factor of 10 are observed within each formation. The ranges of the [Mo/Corg]sample /

[Mo/Corg]Av. Sh. ratios are different between different formations (e.g., some are 0.1 ~ 1 while the others are 1 ~ 10). These ranges for the Archean–Paleoproterozoic shales of this study overlap with those for modern (carbonaceous) sediments and Phanerozoic (black) shales (Fig. 3-7). It should be noted that the samples with higher [Mo/Corg]sample /

[Mo/Corg]Av. Sh. ratios in the Archean–Paleoproterozoic shales of this study are not the results of Mo enrichment but the results of low Corg contents (< ~0.5 wt.%), and that those

with lower [Mo/Corg]sample / [Mo/Corg]Av. Sh. ratios are the result of Corg enrichment (> ~0.5 wt.%). 190

Despite the similarities mentioned above, the Archean–Paleoproterozoic shales of this study are generally low in Mo contents (< 10 ppm; Table 3-2) when compared to the modern carbonaceous sediments and Phanerozoic black shales whose Mo contents are often on the order of tens of ppm or higher. We suggest that this is because the amounts of

Corg and S in the Archean–Paleoproterozoic shales of this study are generally much lower (see chapter 2) than those in the modern carbonaceous sediments and Phanerozoic black shales (sometimes reaching as much as 40 wt.% Corg), and therefore significant enrichment of Mo (e.g., > 100 ppm) is not expected for the Archean–Paleoproterozoic shales of this study; these are discussed in the next section.

3-5-5-2. Modern and Phanerozoic

Figure 3-8 and 3-9 show the relationships between Mo and Corg, and Mo and S, respectively, for the modern carbonaceous sediments in the Black Sea (Hirst, 1974), Mediterranean (Warning and Brumsack, 2000), Oslo Fjord (Calvert, 1976), and Gulf of California (Brumsack, 1989) and black shales in Cretaceous (Brumsack, 1980; Holland, 1984; Arthur et al., 1990), Devonian (Leventhal et al., 1983; Holland, 1984; Shaffer et al., 1984; Robl and Barron, 1988; Leventhal, 1993), and Cambrian (Armands, 1972; Leventhal, 1990). Emphasis is placed on the Black Sea sediments and the Devonian black shales because these are frequently used as representative carbonaceous materials (e.g., Holland, 1984).

In Fig. 3-8 and 3-9, samples with Mo contents of ≤ 1 ppm or below detection are plotted at Mo = 1 ppm; the shaded areas represent where such low Mo samples may be plotted. Regardless of sample ages, good positive correlations between Mo and Corg and Mo and S are observed, suggesting efficient fixation of dissolved Mo by OM and S. Sulfur in those carbonaceous materials are most likely produced by bacterial sulfate reduction. These observations are not new; however, some important features appear that are often 191

neglected by previous investigators. Therefore we want to emphasize that (1) there is almost

no Mo enrichment in the samples with Corg contents of ≤ 0.5 wt.% and/or S contents ≤ 1 wt.%, and that (2) there is large (two orders of magnitude) variation in Mo contents of

samples with 0.5 ~ 2 wt.% Corg and 1 ~ 3 wt.% S (between the dashed lines in Fig. 3-8 and

Fig. 3-9). In these ranges of Corg and S contents, Mo content ranges between 0 ~ 40 ppm in the Mo-Corg plot (Fig. 3-8) and between 0 ~ 300 ppm in the Mo-S plot (Fig. 3-9).

These wide variation in Mo content of samples with similar Corg and S contents suggest that (1) the efficiency of sedimentary Mo fixation at given Corg and S contents depends on conditions other than Corg and S contents of sediments, and/or, as previously mentioned, that (2) remobilization of Mo and/or S may be significant in environments where the bottom water redox conditions fluctuated in various time scales. For (1), Crusius et al. (1996) and Nameoff (1996) have suggested that the development of not only anoxic conditions but also (biogenic) H2S is important to efficiently remove Mo from seawater. For example, Arabian Sea sediments deposited in the OMZ are not accompanied by the development of H2S necessary for Mo fixation, and lack Mo enrichment (Nameoff, 1996). For (2), such redox fluctuations may be caused by the seasonal development of anoxia at depth due to changes in the primary productivity in the surface water and/or by changes in the pattern of water circulations due to tectonic / climatic changes.

3-5-5-3. Schematic model Based on the observations in Fig. 3-8 and Fig. 3-9, we have developed a schematic model, modifying after Lyons et al. (2000), for the environmental changes for the development of conditions for efficient sedimentary Mo fixation by OM or S (Fig. 3-10).

The schematic Mo-Corg, S plot in Fig. 3-10 is divided into four areas. 192

(A) The upper left area (high Mo + low Corg / low S: not shown) is kept blank because there are no samples to be plotted in this area (see Fig. 3-8, 3-9).

(B) The lower left area (low Mo + low Corg + low S) represents oxic marine sediments overlain by oxic bottom water. There is no significant Mo enrichment due to the oxic conditions. Freshwater environments are included in this area.

(C) The lower right area (low Mo + high Corg + high / low S) represents anoxic marine sediments overlain by oxic bottom water. The bottom water may become anoxic due to the increased consumption of dissolved O2 by increased flux of OM from higher primary productivity in the surface water (e.g., Calvert and Pedersen, 1993) or other possible causes (e.g., Arthur and Sageman, 1994). Supply of sulfate to the sediments may be limited by a high rate of sedimentation (i.e., fast consumption of sulfate followed by no replenishment). Freshwater environments may be placed into this area, if sulfate concentration is appreciable due to weathering of sulfate deposits nearby.

(D) The upper right area (high Mo + high Corg + high S) represents anoxic marine sediments overlain by anoxic bottom water where extensive bacterial sulfate reduction take place in sediments and euxinic water column. Preservation of OM and formation of bacteriogenic sulfide are so enhanced under anoxic conditions that dissolved Mo is very efficiently fixed by OM and S in the sediments. Saline lakes (e.g., due to the weathering of evaporites nearby) with high primary productivity may be placed into this area. Syn- and/or post-depositional hydrothermal activity may add Mo and S to sediments by precipitating Mo sulfide minerals. 193

The cases (B), (C), and (D) may change from one to another due to environmental fluctuations as indicated by the arrow in Fig. 3-10. Especially, fluctuation between the cases (C) and (D) would redistribute sedimentary Mo, resulting in either more enrichments or depletions at certain depths.

3-5-5-4. Significance of low Mo contents of the Archean–Paleoproterozoic black shales The general feature of low Mo contents (< 10 ppm) for the Archean-

Paleoproterozoic black shale of this study can be explained by their low Corg and S contents. Their sedimentary environments would have been oxic with intermittent local anoxia, but would not proceed to euxinic conditions (e.g., Black Sea). In other words, the

Mo-Corg-S data set for the Archean–Paleoproterozoic black shales of this study can be explained by the case (B) and (C) in Fig. 3-10. There seems to be no essential difference in the Mo-Corg-S systematics among Archean, Proterozoic, Phanerozoic, and modern sediments / sedimentary rocks.

Although extreme enrichments of Mo, Corg, and S are found in Phanerozoic black shales, they are rather special phenomena developed in special environments and obviously

rare overall in the Phanerozoic. Such extreme enrichment of Mo, Corg, and S in carbonaceous sediments should not be used as an indicator of an oxic atmosphere.

Generally low Corg nature of the Archean–Paleoproterozoic shales might be due to a biased preservation; i.e., sediments deposited in special environment are not likely to be preferentially preserved. Lack of less degradable OM (e.g., lignin) in shales before the

Paleozoic might be another factor to explain generally low Corg in Archean–

Paleoproterozoic shales; however, marine OM dominates in marine black shales with Corg contents > 5 wt.% (pers. comm., Michael A. Arthur, 2002). There is a very interesting relationship between the Mo contents of various modern

carbonaceous marine sediments and the dissolved O2 content in their overlying bottom 194

waters (Fig. 3-11). The obvious negative correlation between them suggests that the low Mo

(< 5 ppm) sediments are accumulated in oxic environments where dissolved O2 concentration in bottom water is higher than ~5 µM (up to 200 µM). These observations support the above suggestions that the low Mo Archean black shales have deposited in oxygenated environments.

3-6. Implications

Implications obtained from the results of this study are significant in that they are related to the evolution of atmosphere, oceans, and biosphere. The operation of the present-day style Mo geochemical cycle in the Archean–Paleoproterozoic (~3.3 to ~2.2 Ga period in this study) surface environments suggests that the Archean–Paleoproterozoic atmosphere and oceans were already oxygenated enough to allow the redox-cycling of Mo (i.e., continental weathering of Mo- bearing minerals, accumulation of dissolved Mo in the oxygenated oceans, and fixation of dissolved Mo by OM and sulfide minerals in local anoxic environments). Microorganisms in the Archean–Paleoproterozoic oceans utilized dissolved bioessential trace metals such as Mo for their living (e.g., operation of Mo-containing enzymes such as nitrogenase and nitrogenase reductase; Madigan et al., 2000) and played important roles to fix them in sediments (c.f., chapter 4 of this thesis). Sulfate-reducing bacteria would have been particularly important to form authigenic sulfide as a sink of Mo. Operation of bacterial sulfate reduction for the fixation of metals in shales requires sufficient level of dissolved sulfate in the deep oxic oceans. Such sulfate

could have been provided into the Archean oceans by hydration of volcanic SO2 and /or by oxidative weathering of sulfides on the continents. The latter process suggests the 195

development of an oxygenated surface water and atmosphere at least in the Mesoarchean (Ohmoto et al., 1993; Kakegawa and Ohmoto, 1999).

3-7. Conclusions

(1) The continental weathering flux of Mo, estimated from the kinetic data on pyrite

oxidation rates under varying pO2 conditions, has been essentially the same since at least -6 ~3.3 Ga ago, if pO2 was higher than 10 atm or 0.002 % PAL.

(2) The observed positive correlations between Mo and Corg contents of the Archean–Paleoproterozoic shales suggest an important role of sedimentary OM to fix Mo during diagenesis, where bacterial sulfate reduction is active to form authigenic sulfide minerals as another important sink of Mo.

(3) Mechanisms of Mo fixation in the oceans have been the same, i.e., fixation by sedimentary OM and authigenic sulfides.

(4) The observed non-positive correlations between Mo contents and Zn/Al2O3 ratios suggest that the observed Mo enrichments in the Archean–Paleoproterozoic shales are not the results of precipitation of minerals from syn- and post-depositional hydrothermal fluids but the results of authigenic accumulation mediated by OM and S.

(5) The observed wide, within-formation variations in the Mo/Al ratios for the shales of this study are better explained by non-detrital, authigenic control for sedimentary distribution of 196

Mo in the oxygenated Archean–Paleoproterozoic environments. Redistribution of Mo within marine sediments under fluctuating redox conditions has been important.

(6) Generally low Mo contents of the Archean–Paleoproterozoic shales may be due to their

generally low Corg and S contents, not due to a result of a globally anoxic environment. Low Mo shales are typically observed in younger sedimentary rocks and sediments.

(7) There are not essential differences in the Mo-Corg-S systematics among Archean, Proterozoic, Phanerozoic, and modern sediments, suggesting that the present-day style Mo geochemical cycle has operated since at least ~3.3 Ga, and the pO2 was higher than 0.1 atm (0.5 PAL) since at least ~3.3 Ga, and that the redox structure of the oceans has been the same since at least ~3.3 Ga.

(8) Operation of present-day style Mo geochemical cycle in Archean–Paleoproterozoic surface environments suggest early evolution of atmosphere, oceans, and biosphere. 197

References

Adelson, J.M., Helz, G.R., and Miller, C.V. (2001) Reconstructing the rise of recent coastal anoxia: molybdenum in Chesapeake Bay sediments. Geochim. Cosmochim. Acta 65, 237-252.

Alberdi-Genolet, M. and Tocco, R. (1999) Trace metals and organic geochemistry of the Machiques Member (Aptian-Albian) and La Luna Formation (Cenomanian- Campanian), Venezuela. Chem. Geol. 160, 19-38.

Armands, G. (1972) Geochemical studies of uranium, molybdenum, and vanadium in a Swedish alum shale. Stockholm Contr. Geol. 27, 1-148.

Arthur, M.A., Jenkyns, H.C., Brumsack, H. -J., and Schlanger, S.O. (1990) Stratigraphy, geochemistry, and paleoceanography of organic carbon-rich Cretaceous sequences. In Cretaceous Resources, Events and Rhythms (eds. R.N. Ginsburg and B. Beaudoin), Kluwer Academic Publ., Netherlands, 75-119.

Arthur, M.A. and Sageman, B.B. (1994) Marine black shales: Depositional mechanism and environments of ancient deposits. Ann. Rev. Earth Planet. Sci. 22, 499-551.

Berner, R.A. (1981) A new geochemical classification of sedimentary environments. J. Sed. Petrol. 51, 359-365.

Bailey, L.K. and Peters, E. (1976) Decomposition of pyrite in acids by pressure leaching and anodization: the case for an electrochemical mechanism. Can. Metall. Q. 15, 333-344.

Baldi, F., Clark, T., Pollack, S.S., and Olson, G.J. (1992) Leaching of pyrites of various reactivities by Thiobacillus ferrooxidans. Appl. Environ. Microbiol. 58, 1853-1856.

Bertine, K.K. (1972) The deposition of molybdenum in anoxic waters. Mar. Chem. 1, 43- 53.

Bertine, K.K. and Turekian, K.K. (1973) Molybdenum in marine deposits. Geochim. Cosmochim. Acta 37, 1415-1434.

Bishop, J.L., Lougear, A., Newton, J., Doran, P., Froeschl, H., Trautwein, A.X., Kröner, W., and Loeberl, C. (2001) Mineralogical and geochemical analyses of Antarctic lake sediments: A study of reflectance and Mössbauer spectroscopy and C, N, and S 198 isotopes with applications for remote sensing on Mars. Geochim. Cosmochim. Acta 65, 2875-2897.

Borda, M.J., Elsetinow, A.R., Schoonen, M.A., and Strongin, D.R. (2001) Pyrite-induced hydrogen peroxide formation as a driving force in the evolution of photosynthetic organisms on an early Earth. Astrobiology 1, 283-288.

Brookins, D.G. (1988) Eh-pH diagrams for geochemistry. Springer-Verlag, Berlin, 176p.

Brown, E.T., Callonnec, L.L., and German, C.R. (2000) Geochemical cycling of redox- sensitive metals in sediments from Lake Malawi: A diagnostic paleotracer for episodic changes in mixing depth. Geochim. Cosmochim. Acta 64, 3515-3523.

Brumsack, H.J. (1980) Geochemistry of Cretaceous black shales from the Atlantic Ocean (DSDP Legs 11, 14, 36, and 41). Chem. Geol. 31, 1-25.

Brumsack, H.J. (1986a) The inorganic geochemistry of Cretaceous black shales (DSDP Leg 41) in comparison to modern upwelling sediments from the Gulf of California. In North Atlantic Paleoceanography (eds. C.P. Summerhayes and N.J. Shackleton), Geol. Soc. Spec. Pub. No. 21, 447-462.

Brumsack, H.J. (1986b) Trace metal accumulation in black shales from the Cenomanian / Turonian boundary event. In Global Bio-Events (ed. O. Walliser), Lecture Notes in Earth Sciences, Vol. 8, Springer-Verlag, Berlin, 337-343.

Brumsack, H.J. (1989) Geochemistry of recent TOC-rich sediments from the Gulf of California and the Black Sea. Geologische Rundschau 78, 851-882.

Brumsack, H.J. and Gieskes, J.M. (1983) Interstitial water trace-metal chemistry of laminated sediments from the Gulf of California, Mexico. Mar. Chem. 14, 89-106.

Calvert, S.E. (1976) The mineralogy and geochemistry of near-shore sediments. In Chemical Oceanography, 2nd ed., vol. 6 (eds., J.P. Riley and R. Chester), Ch. 33, Academic Press, New York.

Calvert, S.E. (1990) Geochemistry and origin of the Holocene sapropel in the Black Sea. In Facets of Modern Biogeochemistry (eds., V. Ittekkot, S. Kempe, W. Michaelis, and A. Spitzy), Springer, Berlin, 326-352.

Calvert, S.E. and Morris, R.J. (1977) Geochemical studies of organic rich sediments from the Namibian Shelf. II. Metal-organic associations. In A Voyage of Discovery (ed. M.V. Angel), Pergamon Press, New York. 199 Calvert, S.E. and Batchelor, C.H. (1978) Major and minor element geochemistry of sediments from the hole 379A, Leg 42B, Deep Sea Drilling Project. In Initial Reports of the Deep Sea Drilling Project (eds. D.A. Ross and Y.P. Neprochnov, et al.), 527-539.

Calvert, S.E., and Price, N.B. (1983) Geochemistry of Namibian shelf sediments. In Coastal upwelling: its sediment record, part A (J.Thiede and E.Suess, eds.), Plenum press, New York, 337-375.

Calvert, S.E. and Pedersen, T.F. (1993) Geochemistry of recent oxic and anoxic marine sediments: Implications for the geological record. Mar. Geol. 113, 67-88.

Calvert, S.E., Pedersen, T.F., and Karlin, R.E. (2001) Geochemical and isotopic evidence for post-glacial paleoceanographic changes in Saanich Inlet, British Columbia. Mar. Geol. 174, 287-305.

Caplan, M.L. and Bustin, R.M. (1996) Factors governing organic matter accumulation and preservation in a marine petroleum source rock from the Upper Devonian to Lower Carboniferous Exshaw Formation, Alberta. Bull. Can. Petro. Geol. 44, 474-494.

Caplan, M.L. and Bustin, R.M. (1998) Paleoceanographic controls on geochemical characteristics of organic-rich Exshaw mudrocks: role of enhanced primary production. Org. Geochem. 30, 161-188.

Chen, J.H., Wasserburg, G.J., von Damm, K.L., and Edmond, J.M. (1986) The U-Th-Pb systematics in hot springs on the East Pacific Rise at 21 ˚N and Guaymas Basin. Geochim. Cosmochim. Acta 50, 2467-2479.

Clemmey, H. and Badham, N. (1982) Oxygen in the Precambrian atmosphere: An evolution of the geological evidence. Geology 10, 141-146.

Coveney, R.M.Jr. and Martin, S.P. (1983) Molybdenum and other heavy metals of the Mecca Quarry and Logan Quarry Shales. Econ. Geol. 78, 132-149.

Coveney, R.M.Jr., Leventhal, J.S., Glascock, M.D., and Hatch, J.R. (1987) Origins of metals and organic matter in the Mecca Quarry shale member and stratigraphically equivalent beds across the Midwest. Econ. Geol. 82, 915-933.

Coveney, R.M.Jr. and Glascock, M.D. (1989) A review of the origins of metal-rich black shales, central U.S.A., with an inferred role for basinal brines. Applied. Geochem. 4, 347-368. 200 Coveney, R.M.Jr., Watney, W.L., and Maples, C.G. (1991) Contrasting depositional models for Pennsylvanian black shale discerned from molybdenum abundances. Geology 19, 147-150.

Crusius, J. and Thomson, J. (2000) Comparative behavior of authigenic Re, U, and Mo during reoxidation and subsequent long-term burial in marine sediments. Geochim. Cosmochim. Acta 63, 2233-2242.

Crusius, J., Calvert, S., Pedersen, T., and Sage, D. (1996) and molybdenum enrichments in sediments as indicators of oxic, suboxic, and sulfidic conditions of deposition. Earth Planet. Sci. Lett. 145, 68-78.

Davy, R. (1983) A geochemical study of the Mount McRae shale and the upper part of the Mount Sylvia Formation in Core RD1, Rhodes Ridge, Western Australia. Geol. Surv. W. Aust. Record 1983/3.

Davy, R. and Hickman, A.H. (1988) The transition between the Hamersley and Fortescue Groups as evidenced in a drill core. Geol. Surv. W. Aust, Prof. Papers, Report 23, 85-97.

Dean, W.E. and Arthur, M.A. (1986) Inorganic and organic geochemistry of Eocene to Cretaceous strata recovered from the Lower Continental Rise, North American Basin site 603, Deep Sea Drilling Project Leg 93. In Init. Repts. DSDP, 93 (eds. J.E. van Hinte and S.W. Wise, Jr. et al.), Washington (US Government Printing Office), 1093-1137.

Dean, W.E., Gardner, J.V., Piper, D.Z. (1997) Inorganic geochemical indicators of glacial- interglacial changes in productivity and anoxia on the California continental margin. Geochim. Cosmochim. Acta 61, 4507-4518.

Dean, W.E., Piper, D.Z., and Petersen, L.C. (1999) Molybdenum accumulation in Cariaco basin sediment over the past 24 k.y.: A record of water-column anoxia and climate. Geology 27, 507-510.

Desborough, G.A., Hatch, J.R., and Leventhal, J.S. (1991) Geochemical and mineralogical comparison of the upper Pennsylvanian Stark Shale member of the Dennis Limestone, East-Central Kansas, with the Middle Pennsylvanian Mecca Quarry Shale Member of the in Illinois and of the Linton Formation in Indiana. In Metalliferous black shales and related ore deposits Proceedings, 1989 U.S. working group meeting, Int'l. Geol. Correl.. Prog. Project 254, 12-30.

Dimroth, E. and Kimberley, M.M. (1976) Precambrian atmospheric oxygen: evidence in the sedimentary distributions of carbon, sulfur, uranium, and iron. Can. J. Earth Sci. 13, 1161-1185. 201 Dorta, C.C. and Rona, E. (1971) Geochemistry of uranium in the Cariaco Trench. Bull. Mar. Sci. 21, 754-765.

Edwards, K.J., Schrenk, M.O., Hamers, R., and Banfield, J.F. (1998) Microbial oxidation of pyrite: Experiments using microorganisms from an extreme acidic environment. Am. Mineral. 83, 1444-1453.

Emelyanov, E.M., Lisitzin, A.P., Shimkus, K.M., Trimonis, E.S., Lukashev, V.K., Lukashin, V.N., Mitropolskiy, A.Y., and Pilipchuk, M.F. (1978) Geochemistry of late Cenozoic sediments of the Black Sea, Leg 42B. In Initial Reports of the Deep Sea Drilling Project (eds. D.A. Ross and Y.P. Neprochunov et al.), 42(2), 543-605.

Emerson, S.R. and Huested, S.S. (1991) Ocean anoxia and the concentrations of molybdenum and vanadium in seawater. Mar. Chem. 34, 177-196.

Francois, R. (1988) A study on the regulation of the concentrations of some trace metals (Rb, Sr, Zn, Pb, Cu, V, Cr, Ni, Mn, and Mo) in Saanich Inlet sediments, British Columbia, Canada. Mar. Geol. 83, 285-308.

Gross, M.G., Gucluer, S.M., Craeger, J.S., and Dawson, W.A. (1963) Varved marine sediments in a stagnant fjord. Science 141, 918-919.

Hatch, J.R. and Leventhal, J.S. (1992) Relationship between inferred redox potential of the depositional environment and geochemistry of the Upper Pennsylvanian (Missourian) Stark Shale Member of the Dennis Limestone, Wabaunsee County, Kansas, U.S.A. Chem. Geol. 99, 65-82.

Hayashi, K., Fujisawa, H., Holland, H.D., and Ohmoto, H. (1997) Geochemistry of ~1.9 Ga sedimentary rocks from northeastern Labrador, Canada. Geochim. Cosmochim. Acta 61, 4115-4137.

Helz, G.R., Miller, C.V., Charnock, J.M., Mosselmans, J.F.W., Pattrick, R.A.D., Garner, C.D., and Vaughan, D.J. (1996) Mechanism of molybdenum removal from the sea and its concentration in black shales: EXAFS evidence. Geochim. Cosmochim. Acta 60, 3631-3642.

Hirst, D. (1974) Geochemistry of sediments from Eleven Black Sea cores. In The Black Sea: Geology, Chemistry, and Biology (eds. E.T. Degens and D.A. Ross), Amer. Assoc. Petrol. Geol. Memoir 20, 430-455.

Holland, H.D. (1979) Metals in black shales - A reassessment. Econ. Geol. 74, 1676-1680.

Holland, H.D. (1984) The chemical evolution of the atmosphere and oceans. Princeton Univ. Press, Princeton. 202 Holland, H.D. (1994) Early Proterozoic atmospheric change. In Early Life on Earth. Novel Symposium No. 84 (S. Bengston, ed.), Columbia Univ. Press, New York.

Holland, H.D. (1999) When did the Earth's atmosphere become oxic? A reply. Geochemical News 100, 20-22.

Holmes, P.R. and Crundwell, F.K. (2000) The kinetics of the oxidation of pyrite by ferric ions and dissolved oxygen: An electrochemical study. Geochim. Cosmochim. Acta 64, 263-274.

Hoppie, B. and Garrison, R.E. (2001) Miocene phosphate accumulation in the Cuyama Basin, Southern California. Mar. Geol. 177, 353-380.

Huerta-Diaz, M.A. and Morse, J.W. (1992) Pyritization of trace metals in anoxic marine sediments. Geochim. Cosmochim. Acta 56, 2681-2702.

Jacobs, L., Emerson, S., and Huested, S.S. (1987) Trace metal geochemistry in the Cariaco Trench. Deep-Sea Res. 34, 965-981.

James, R. and Elderfield, H. (1996) Chemistry of ore-forming fluids and mineral formation rates in an active hydrothermal sulfide deposit on the Mid-Atlantic Ridge. Geology 24, 1147-1150.

Joachimski, M.M., Ostertag-Henning, C., Pancost, P.D., Strauss, H., Freeman, K.H., Littke, R., Sinninghe Damsté, J.S., and Racki, G. (2001) Water column anoxia, enhanced productivity and concomitant changes in δ13C and δ34S across the - boundary (Kowka-Holy Corss Mountains/Poland). Chem. Geol. 175, 109-131.

Kakegawa, T. and Ohmoto, H. (1999) Sulfur isotope evidence for the origin of 3.4 to 3.1 Ga pyrite at the Princeton gold mine, Barberton Greenstone Belt, South Africa. Precam. Res. 96, 209-224.

Kamei, G. and Ohmoto, H. (2000) The kinetics of reactions between pyrite and O2-bearing water revealed from in situ monitoring of DO, Eh and pH in a closed system. Geochim. Cosmochim. Acta 64, 2585-2601.

Karavaiko, G.I., Smolskaja, L.S., Golyshina, O.K., Jagovkina, M.A., and Egorova, E.Y. (1994) Bacterial pyrite oxidation - influence of morphological, physical and chemical properties. Fuel Process Technol. 40, 151-165.

Karhu, J. and Epstein, S (1986) The implication of the oxygen isotope records in coexisting cherts and phosphates. Geochim. Cosmochim. Acta 50, 1745-1756. 203 Kasting, J.F. (1987) Theoretical constraints on oxygen and carbon dioxide concentrations in the Precambrian atmosphere. Precam. Res. 34, 205-228.

Knauth, L.P. and Epstein, S. (1976) Hydrogen and oxygen isotope ratios in nodular and bedded cherts. Geochim. Cosmochim. Acta 40, 1095-1108.

Knauth, L.P. and Lowe, D.R. (1978) Oxygen isotope geochemistry of cherts from the Onverwacht Group (3.4 billion years), Transvaal, South Africa, with implications for secular variations in the isotopic composition of cherts. Earth Planet. Sci. Lett. 41, 209-222.

Koide, M., Hodge, V.F., Yang, J.S., Stallard, M., and Goldberg, E.G. (1986) Some comparative marine chemistries of rhenium, gold, , and molybdenum. Appl. Geochem. 1, 705-714.

Lange, J., Wedepohl, K.H., Heinrichs, H., and Gohn, E. (1977) Notes about the specific chemical composition of "black shales" from site 367 (Leg 41). In Deep Sea Drilling Project 41 (eds., Y. Lancelot and E. Seibold et al.), Washington (US Government Printing Office), 875-877.

Lasaga, A.C. and Ohmoto, H. (2002) The oxygen geochemical cycle: dynamics and stability. Geochim. Cosmochim. Acta 66, 361-381.

Leventhal, J.S., Briggs, P.H., and Baker, J.W. (1983) Geochemistry of the Chattanooga Shale, Dekalb County, central Tennessee. Southeast. Geol. 24, 101-116.

Leventhal, J. (1990) Comparative geochemistry of metals and rare earth elements from the Cambrian Alum shale and Kolm of Sweden. Spec. Publ. int. Ass. Sediment. 11, 203- 216.

Leventhal, J. (1993) Metals in Black Shales. In Organic Geochemistry: Principles and Applications (eds. M.H. Engel and S.A. Macko), 581-592.

Leventhal, J.S. and Hosterman, J.W. (1982) Chemical and mineralogical analysis of Devonian black-shale samples from Martin County, Kentucky; Caroll and Washington Counties, Ohio; Wise County, Virginia; and Overton County, Tennessee, U.S.A. Chem. Geol. 37, 239-264.

Logan, G.A., Hayes, J.M., Hieshima, G.B., and Summons, R.E. (1995) Terminal Proterozoic reorganization of biogeochemical cycles. Nature 376, 53-56.

Loukola-Ruskeeniemi, K. (1991) Geochemical evidence for the hydrothermal origin of sulfur, base metals and gold in Proterozoic metamorphosed black shales, Kainuu and Outokumpu areas, Finland. Mineral. Deposita 26, 152-164. 204 Loukola-Ruskeeniemi, K. (1999) Origin of black shales and the Serpentinite-associated Cu- Zn-Co ores at Outkumpu, Finland. Econ. Geol. 94, 1007-1028.

Loukola-Ruskeeniemi, K., Heino, T., Talvitie, J., and Vanne, J. (1991) Base-metal-rich metamorphosed black shales associated with Proterozoic in the Kainuu schist belt, Finland: a genetic link with the Outokumpu rock assemblage. Mineral. Deposita 26, 143-151.

Loukola-Ruskeeniemi, K. and Heino, T. (1996) Geochemistry and genesis of the black shale-hosted Ni-Cu-Zn deposit at Talvivaara, Finland. Econ. Geol. 91, 80-110.

Lyons, T.W., Luepke, J.J., Schreiber, M.E., and Zieg, G.A. (2000) Sulfur geochemical constraints on Mesoproterozoic restricted marine deposition: lower Belt Supergroup, northwestern United States. Geochim. Cosmochim. Acta 64, 427-437.

Malcolm, S.J. (1985) Early diagenesis of Molybdenum in estuarine sediments. Mar. Chem. 16, 213-225.

Madigan, M.T., Martinko, J.M., and Parker, J. (2000) Brook biology of microorganisms, 8th ed. Prentice Hall, Upper Saddle River.

Manheim, F.T. and Landergren, S. (1978) Molybdenum. In Handbook of Geochemistry (ed. K.H. Wedepohl), 42, B-O. Springer-Verlag, Berlin.

Martin, J.M. and Meybeck, M. (1979) Elemental mass-balance of material carried by major world rivers. Mar. Chem. 7, 173-206.

Mathews, C.T. and Robins, R.G. (1974) Aqueous oxidation of iron disulfide by molecular oxygen. Austral. Chem. Eng. 15, 19-24.

McKibben, M.A. and Barnes, H.L. (1988) Oxidation of pyrite in low temperature acidic solutions: Rate laws and surface textures. Geochim. Cosmochim. Acta 50, 1509- 1520.

Meyer, F.M. and Robb, L.J. (1996) The geochemistry of black shales from the Chuniespoort Group, Transvaal Sequence, Eastern Transvaal, South Africa. Econ. Geol. 91, 111-121.

Migdisov, A.A., Girin, Y.P., Galimov, E.M., Grinenko, V.A., Barskaya, N.V., Krivitsky, V.A., Sobornov, O.P., and Cherkovsky, S.L. (1980) Major and minor elements and sulfur isotopes of the Mesozoic and Cenozoic sediments at sites 415 and 416, Deep Sea Drilling Project. In Deep Sea Drilling Project 50 (eds., Y. Lancelot and E.L. Winterer, et al.), Washington (US Government Printing Office), 675-689. 205 Morford, J.L. and Emerson, S. (1999) The geochemistry of redox sensitive trace metals in sediments. Geochim. Cosmochim. Acta 63, 1735-1750.

Morford, J.L., Russell, A.D., and Emerson, S. (2001) Trace metal evidence for changes in the redox environment associated with the transition from terrigenous clay to diatomaceous sediment, Saanich Inlet, BC. Mar. Geol. 174, 355-369.

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

Mustin, C., de Donato, P., Berthelin, J., and Marion, P. (1993) Surface sulphur as promoting agent for pyrite leaching by Thiobacillus ferrooxidans. FEMS Microbiol. Rev. 11, 71-78.

Nameoff, T.J. (1996) Suboxic trace metal geochemistry and paleorecord in continental margin sediments of the Eastern Tropical North Pacific. Ph.D. Thesis, University of Washington, 224 p.

Naraoka, H., Ohmoto, H., and others (submitted) Redox fluctuations in the late Archean ocean (2.55 Ga ago): Evidence from carbon-sulfur isotopes and Mn-Mo geochemistry. Geochim. Cosmochim. Acta

Nicholson, R.V., Gillham, R.W., and Reardon, E.J. (1988) Pyrite oxidation in carbonate- buffered solution: 1. Experimental kinetics. Geochim. Cosmochim. Acta 52, 1077- 1085.

Nicholson, R.V., Gillham, R.W., and Reardon, E.J. (1988) Pyrite oxidation in carbonate- buffered solution: 2. Rate control by oxide coatings. Geochim. Cosmochim. Acta 54, 395-402.

Nijenhuis, I.A., Brumsack, H-.J., and De Lange, G.J. (1998) The trace element budget of the Eastern Mediterranean during Pliocene sapropel formation. In Proc. ODP, Sci. Results 160 (eds. A.H.F. Robertson, K-.C. Emeis, C. Richter, and A. Camerlenghi), College Station, Texas (Ocean Drilling Program).

Nordstrom, D.K. and Southam, G. (1997) Geomicrobiology of sulfide mineral oxidation. In Geomicrobiology (eds., J.F. Banfield and K.H. Nealson), Reviews in Mineralogy, 35, Mineral. Soc. Am., Washington, D.C., 361-390.

Ohmoto, H. (1996) Evidence in pre-2.2 Ga paleosols for the early evolution of atmospheric oxygen and terrestrial biota. Geology 24, 1135-1138. 206 Ohmoto, H. (1997) When did the Earth's atmosphere become oxic? Geochemical News 93. 12-12 and 26-27.

Ohmoto, H., Kakegawa, T., and Lowe, D.R. (1993) 3.4-billion-year-old biogenic pyrites from Barberton, South Africa: sulfur isotope evidence. Science 262, 555-557.

Ohmoto, H. (1999) Redox state of the Archean atmosphere: Evidence from detrital heavy minerals in ca. 3250-2750 Ma sandstones from the Pilbara Craton, Australia: Comment. Geology 27, 1151-1152.

Pailler, D., Bard, E., Rostek, F., Zheng, Y., Mortlock, R., and van Geen, A. (2002) Burial of redox-sensitive metals and organic matter in the equatorial Indian Ocean linked to precession. Geochim. Cosmochim. Acta 66, 849-865.

Pedersen, T.F., Pickering, M., Vogel, J.S., Southon, J.N., and Nelson, D.E. (1988) The response of benthic foraminifera to productivity cycles in the eastern equatorial Pacific: Faunal and geochemical constraints on glacial bottom water oxygen levels. Paleoceanography 3, 157-168.

Phillps, G.N., Law, J.D.M., and Myers, R.E. (2001) In the redox state of the Archean atmosphere constrained? Soc. Econ. Geol. Newsletter 47.

Pilipchuk, M.F. and Volkov, I.I. (1974) Behavior of molybdenum in processes of sediment formation and diagenesis in Black Sea. In The Black Sea: geology, Chemistry, and Biology (eds. E.T. Degens and D.A. Ross), Amer. Assoc. Petrol. Geol. Memoir 20, 542-553.

Piper, D.Z. and Isaacs, C.M. (1994) Instability of bottom-water redox conditions during accumulation of Quaternary sediments in the Japan Sea. Paleoceanography 11, 171-190.

Piper, D.Z. and Isaacs, C.M. (1995) Geochemistry of minor elements in the Monterey Formation, California: seawater chemistry of deposition. U.S. Geol. Surv. Prof. Paper 1566.

Piper, D.Z. and Isaacs, C.M. (2001) The Monterey Formation: bottom-water redox conditions and photic-zone primary productivity. In The Monterey Formation: from rocks to molecules (C.M. Isaacs and J. Rullkötter, eds.), Columbia Univ. Press, New York, 31-58.

Rappiza, G., Turekian, K.K., and Hay, B.J. (1991) The geochemistry of rhenium and in recent sediments from the Black Sea. Geochim. Cosmochim. Acta 55, 3741-3752. 207 Robl, T.L. and Barron, L.S. (1988) The geochemistry of Devonian black shales in central Kentucky and its relationship to inter-basinal correlation and depositional environment. In Devonian of the World Vol. II (N.J. McMillan, A.F. Embry, D.J. Glass, eds.) Canadian Soc. of Petroleum Geologists, Calgary. 377-392.

Russell, A.D. and Morford, J.L. (2001) The behavior of redox-sensitive metals across a laminated-massive-laminated transition in Saanich Inlet, British Columbia. Mar. Geol. 174, 341-354.

Rye, R. and Holland, H.D. (1998) Paleosols and the evolution of atmospheric oxygen: a critical review. Am. J. Sci. 88, 621-672.

Salonen, V.P., Grönlund, T., Itkonen, A., Sturm, M., and Vuorinen, I. (1995) Geochemical record on early diagenesis of recent Baltic Sea sediments. Mar. Geol. 129, 101-109.

Sarmiento, J. (1992) Biogeochemical Ocean Models. In Climate System Modeling (K.E.Trenberth, ed.), Cambridge Univ. Press, 519-564.

Shaffer, N., Leninger, R., and Gilstrao, M. (1984) Composition of in southeastern Indiana. 1983 Eastern Oil Shale Symposium, Univ. Kentucky, Inst. Mining Minerals Res., Lexington, 195-205.

Schalleer, T., Morford, J., Emerson, S.R., and Feely, R.A. (2000) Oxyanions in metalliferous sediments: Tracers for paleoseawater metal concentrations? Geochim. Cosmochim. Acta 63, 2243-2254.

Seward, T.M. and Barnes, H.L. (1997) Metal transport by hydrothermal fluids. In Geochemistry of Hydrothermal Ore Deposits, 3rd ed. (ed. H.L. Barnes), John Wiley and Sons, Inc. New York, p 435-486.

Shaw, T.J., Gieskes, J.M., and Jahnke, R.A. (1990) Early diagenesis in differing depositional environments: The response of transition metals in pore water. Geochim. Cosmochim. Acta 54, 1233-1246.

Shimmield, G.B. and Price, N.B. (1986) The behavior of molybdenum and manganese during early sediment diagenesis-offshore Baja California, Mexico. Mar. Chem. 19, 261-280.

Skei, J. (1981) Et biogeokjemist studium av en permanent anoksik fjord - Framvaren ved Farsund. Norway institutt for vannforskning (NIVA). Rapportnumber F-80400 (in Norwegian). 208 Sohlenius, G., Emeis, K.C., Andrén, E., Andrén, T., and Kohly, A. (2001) Development of anoxia during the Holocene fresh-brackish water transition in the Baltic Sea. Mar. Geol. 177, 221-242.

Sternbeck, J., Sohlenius, G., and Hallberg, R.O. (2000) Sedimentary trace elements as proxies to depositional changes induced by a Holocene fresh-brackish water transition. Aq. Geochem. 6, 325-345.

Stumm, W. and Morgan, J.J. (1996) Aquatic Chemistry. 3rd. ed., John Wiley & Sons, 1022 pp.

Taylor, S.R. and McLennan, S.M. (1985) The Continental Crust: its Composition and Evolution. Blackwell Scientific Publications, 311p.

Thomson, J., Higgs, N.C., Croudace, I.W., Colley, S., and Hydes, D.J. (1993) Redox zonation of elements at an oxic / post-oxic boundary in deep-sea sediments. Geochim. Cosmochim. Acta 57, 579-595.

Thomson, J., Higgs, N.C., Wilson, T.R.S., Croudace, I.W., de Lange, G.J., and van Santvoort, P.J.M. (1995) Redistribution and geochemical behavior of redox- sensitive elements around S1, the most recent eastern Mediterranean sapropel. Geochim. Cosmochim. Acta 59, 3487-3501.

Thomson, J., Higgs, N.C., and Colley, S. (1996) Diagenetic redistributions of redox- sensitive elements in northeast Atlantic glacial/interglacial transition sediments. Earth Planet. Sci. Lett. 139, 365-377.

Thomson, J., Nixin, S., Croudace, I.W., Pedersen, T.F., Brown, L., Cook, G.T., and MacKenzie, A.B. (2001) Redox-sensitive element uptake in north-east Atlantic Ocean sediments (Benthic Boundary Layer Experiment sites). Earth Planet. Sci. Lett. 184, 535-547.

Towe, K.M. (1990) Aerobic respiration in the Archean? Nature 348, 54-56.

Towe, K.M. (1991) Aerobic carbon cycling and cerium oxidation: significance for Archean oxygen levels and banded iron formation deposition. Paleogeogr., Paleoclimatol., Paleoecol. (Global Planet. Change Sect.) 97, 113-123.

Towe, K.M. (1994) Earth’s early atmosphere: Constraints and opportunities for early evolution. In Early Life on Earth. Nobel Symposium No. 84. (ed., S. Bengston) Columbia University Press, New York, 36-47.

Van Santvoort, P.J.M., de Lange, G.J., Thomson, J., Cussen, H., Wilson, T.R.S., Krom, M.D., and Ströhle, K. (1996) Active post-depositional oxidation of the most recent 209 sapropel (S1) of the Eastern Mediterranean. Geochim. Cosmochim. Acta 60, 4007- 4024.

Vine, J. D. and Tourtelot, E. B. (1970) Geochemistry of black shale deposits: A summary report. Econ. Geol. 65, 253-272.

Volkov, I.I. and Fomina, L.S. (1974) Influence of organic material and processes of sulfide formation on distribution of some trace elements in deep-water sediments of Black Sea. In The Black Sea: Geology, Chemistry, and Biology (E.T. Degens and D.A. Ross, eds.), Amer. Assoc. Petrol. Geol. Memoir 20, 456-476.

Warning, B. and Brumsack, H-.J. (2000) Trace metal signatures of eastern Mediterranean sapropels. Paleogeogr. Paleoclimatol. Paleoecol. 158, 293-309.

Wedepohl, K.H. (1991) The composition of the upper Earth's crust and the natural cycles of selected metals. Metals in natural raw materials. Natural resources. In Metals and their compounds in the Environment (ed., E. Merian), VCH, Weinheim. 3-17.

Wignall, P.B. (1994) Black Shales. Oxford University Press, Oxford, 127 pp.

Williamson, M.A. and Rimstidt, J.D. (1994) The kinetics and electrochemical rate- determining step of aqueous pyrite oxidation. Geochim. Cosmochim. Acta 58, 5443-5454.

Wilson, T.R.S., Thomson, J., Colley, S., Hydes, D.J., Higgs, N.C., and Sørensen, J. (1985) Early organic diagenesis: the significance of progressive subsurface oxidation fronts in pelagic sediments. Geochim. Cosmochim. Acta 49, 811-822.

Yu, J.Y., McGenity, T.J., and Coleman, M.L. (2001) Solution chemistry during the lag phase and exponential phase of pyrite oxidation by Thiobacillus ferrooxidans. Chem. Geol. 175, 307-317.

Zheng, Y., Anderson, R.F., van Geen, A., and Kuwabara, J. (2000) Authigenic molybdenum formation in marine sediments: A link to pore water sulfide in the Santa Barbara Basin. Geochim. Cosmochim. Acta 64, 4165-4178. 210

(a)10 (b) 10 shale 3.25 Ga shale 3.25 Ga siderite-rich shale siderite-rich shale cherty shale Sheba Fm cherty shale Sheba Fm 8 (shales) 8 (shales)

6 6 Mo Mo [ppm] [ppm] 4 4

2 2

0 0 012345 012345

Corg [wt.%] S [wt.%]

(c)10 (d) 10

8 8

6 6 Mo Mo [ppm] [ppm] 4 4

2 2 3.25 Ga Sheba Fm 3.25 Ga Sheba Fm (graywackes) (graywackes) 0 0 012345 012345

Corg [wt.%] S [wt.%]

Fig. 3-1. Plot of the Mo vs. Corg and Mo vs. S contents of the Archean– Paleoproterozoic samples of this study. Slope of solid line represents the Mo/Corg ratio of the average shale (Wedepohl, 1991; Warning and Brumsack, 2000). The dashed line represents the Mo/Corg ratio of the average black shale (Vine and Tourtelot, 1970). 211

(e)4 (f) 4

3 3

Mo Mo [ppm] 2 [ppm] 2

1 1

2.96 Ga Parktown Fm 2.96 Ga Parktown Fm 0 0 012012

Corg [wt.%] S [wt.%]

(g)4 (h) 4

3 3

Mo Mo [ppm] 2 [ppm] 2

1 1

2.71 Ga Rietgat Fm 2.71 Ga Rietgat Fm 2.72 Ga Pillingini Tuff Fm 2.72 Ga Pillingini Tuff Fm 0 0 012012

Corg [wt.%] S [wt.%]

Fig. 3-1. Plot of the Mo vs. Corg and Mo vs. S contents of the Archean– Paleoproterozoic samples of this study. Slope of solid line represents the Mo/Corg ratio of the average shale (Wedepohl, 1991; Warning and Brumsack, 2000). The dashed line represents the Mo/Corg ratio of the average black shale (Vine and Tourtelot, 1970). 212

(i)10 (j) 10

8 8

6 6 Mo Mo [ppm] [ppm] 4 4

2 2 2.69 Ga Jeerinah Fm 2.69 Ga Jeerinah Fm 2.69 Ga Lewin Shale Fm 2.69 Ga Lewin Shale Fm 0 0 03691215 03691215

Corg [wt.%] S [wt.%]

(k)5 (l) 2.60 Ga Carawine Dolomite Fm 5 2.60 Ga Carawine Dolomite Fm 2.60 Ga Wittenoom Dolomite Fm 2.60 Ga Wittenoom Dolomite Fm >2.60 Ga Marra Mamba Iron Fm >2.60 Ga Marra Mamba Iron Fm 4 4

3 3 Mo Mo [ppm] [ppm] 2 2

1 1

0 0 012345 012345

Corg [wt.%] S [wt.%]

Fig. 3-1. Plot of the Mo vs. Corg and Mo vs. S contents of the Archean– Paleoproterozoic samples of this study. Slope of solid line represents the Mo/Corg ratio of the average shale (Wedepohl, 1991; Warning and Brumsack, 2000). The dashed line represents the Mo/Corg ratio of the average black shale (Vine and Tourtelot, 1970). 213

(m) (n) 5 5 2.56 Ga Oak Tree Fm 2.56 Ga Oak Tree Fm 2.64 Ga Black Reef Fm 2.64 Ga Black Reef Fm 4 4

3 3 Mo Mo [ppm] [ppm] 2 2

1 1

0 0 0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5

Corg [wt.%] S [wt.%]

(o)10 (p) 10 2.22 Ga Timeball Hill Fm, Eastern 2.22 Ga Timeball Hill Fm, Eastern 2.22 Ga Timeball Hill Fm, Central 2.22 Ga Timeball Hill Fm, Central ~2.2 Ga Mapedi Fm ~2.2 Ga Mapedi Fm 8 8

6 6 Mo Mo [ppm] [ppm] 4 4

2 2

0 0 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1

Corg [wt.%] S [wt.%]

Fig. 3-1. Plot of the Mo vs. Corg and Mo vs. S contents of the Archean– Paleoproterozoic samples of this study. Slope of solid line represents the Mo/Corg ratio of the average shale (Wedepohl, 1991; Warning and Brumsack, 2000). The dashed line represents the Mo/Corg ratio of the average black shale (Vine and Tourtelot, 1970). 214

(a) Globally anoxic world

Mo transport by detrital (4+) process (Mo S2)

anoxic atmosphere coarse fraction entirely anoxic ocean

redeposition

• Mo content proportional to fine fraction grain sizes of host rocks. fine fraction

(b) Globally oxic world with local anoxic environments

Mo transport in dissolved (6+) 2- forms (e.g., Mo O4 )

oxic atmosphere oxic ocean with anoxic basins redeposition

anoxic basin • Mo content proportional to sulfide and organic matter content of host rocks. • Great heterogeneity in Mo content.

Fig. 3-2. Molybdenum distribution in sediments depending on the redox state of the environments. (a) Globally anoxic world (anoxic atmosphere and entirely anoxic ocean). (b) Globally oxic world with local anoxic environments (like today's). 215

1.0

P O2 = 1 atm

0.5 - MoO + HMoO4 P 2 O2 = 0.2 P Eh [v] O2 = 10 -20 2- MoO4 Mo3O8 0 P O2 = 10 MoS2 -40 molybdenite P O2 = 10 -60 25 ˚C P -0.5 [∑Mo] = 10-8 M H2 = 1 atm [∑S] = 10-3 M

02468101214 pH

Fig. 3-3. Eh-pH diagram for aqueous species and solids in the system Mo-S-O2-H2O at 25 ˚C and 1 bar total pressure. The diagram is drawn for ∑Mo (total molybdenum) = 10-8 M and ∑S (total sulfur) = 10-3 M (Brookins, 1988). Contours for various pO2 are also indicated. The diagram does not significantly change even under T = 60 ˚C, considering the inferred higher heat flux and higher pCO2 level in the Archean than in modern (x 100 ~ 1000 PAL; Kasting, 1987; PAL: present atmospheric level, = 350 ppm). 216

pO2 [PAL] 10-9 10-6 10-3 1

> 2.2 Ga ? < 2.0 Ga ? 3 10 [cm ] 105 yr 3 1 [cm ] =1000 104 yr 100 ] 3 10 105 yr 1 3 104 yr 0.1 Pyrite [mm surface area x10 0.01

0.001

10-4

10-5

10-6 10-10 10-8 10-6 10-4 0.01 1 pO2 [atm]

Fig. 3-4. Relationship between rate of pyrite oxidation and pO2 values. Rate equation adapted from Williamson and Rimstidt (1994). Conditions used for calculation: pH = 5, T = 40 ˚C, density of pyrite = 5 g/cm3. The vertical dashed line at pO2 = 0.2 atm or 1 PAL (present atmospheric level) represents present condition. The lower shaded area represents the region sandwiched by 105 yr line and 104 yr line (typical soil retention time). The upper shaded area represent the case when surface area for pyrite oxidation increased by a factor of 103 by fractures developed during oxidation. For example, the point at the filled circle represents an case where pyrite crystal of 1 mm3 in size is completely dissolved -6 within 20,000 years when pO2 is higher than 10 atm and fractures developed during oxidation increased surface area by a factor of 103. Also shown at the upper X axis are the inferred rise of atmospheric O2 level during 2.2-2.0 Ga from < 0.01 PAL to > 0.15 PAL (Rye and Holland, 1998). 217

10 10 (a)shale 3.25 Ga (b) 3.25 Ga siderite-rich Sheba Fm Sheba Fm silica-rich (shales) (graywackes)

Mo Mo [ppm] 1 * [ppm] 1 *

0.1 0.1 110100 1000 110100 1000

Zn/Al2O3 Zn/Al2O3 [ppm/wt.%] [ppm/wt.%]

10 10 (c)2.96 Ga Parktown Fm (d)

Mo Mo 1 1 [ppm] * [ppm] *

2.71 Ga Rietgat Fm 2.72 Ga Pillingini Tuff Fm 0.1 0.1 110100 1000 110100 1000

Zn/Al2O3 Zn/Al2O3 [ppm/wt.%] [ppm/wt.%]

Fig. 3-5. Plot of the Mo contents vs. Zn/Al2O3 ratios. Star symbol indicate average shale composition (Turekian and Wedepohl, 1961; Taylor and McLennan, 1985; Wedepohl, 1991). Other symbols are explained in the figure. Note that both X and Y axes are in log scale. 218

(e)10 (f) 10

Mo 1 Mo 1 [ppm] * [ppm] *

2.60 Ga Carawine Dolomite Fm 2.69 Ga Jeerinah Fm 2.60 Ga WittenoomDolomite Fm 2.69 Ga Lewin Shale Fm >2.60 Ga Marra Mamba Iron Fm 0.1 0.1 110100 1000 110100 1000

Zn/Al2O3 Zn/Al2O3 [ppm/wt.%] [ppm/wt.%]

(g)10 (h) 10

Mo 1 * Mo 1 * [ppm] [ppm]

2.22 Ga Timeball Hill Fm (eastern) 2.22 Ga Timeball Hill 2.56 Ga Oak Tree Fm Fm (central) 2.64 Ga Black Reef Fm ~2.2 Ga Mapedi Fm 0.1 0.1 110100 1000 110100 1000

Zn/Al2O3 Zn/Al2O3 [ppm/wt.%] [ppm/wt.%]

Fig. 3-5. Plot of the Mo contents vs. Zn/Al2O3 ratios. Star symbol indicate average shale composition (Turekian and Wedepohl, 1961; Taylor and McLennan, 1985; Wedepohl, 1991). Other symbols are explained in the figure. Note that both X and Y axes are in log scale. 219

(a) ~2.2 Ga Mapedi Fm central Transvaal eastern Transvaal 2.22 Ga Timeball Hill Fm (b)

(c) 2.56 Ga Oak Tree Fm (d) 2.60 Ga Wittenoom Dolomite Fm (e) 2.60 Ga Carawine Dolomite Fm (f) >2.60 Ga Marra Mamba Iron Fm (g) 2.64 Ga Black Reef Fm (h) 2.69 Ga Lewin Fm (i) 2.69 Ga Jeerinah Fm

(j) 2.71 Ga Rietgat Fm (k) 2.72 Ga Pillingini Tuff Fm (l) 2.96 Ga Parktown Fm

(m) 3.25 Ga Sheba Fm (graywacke) shale siderite-rich silica-rich (n) 3.25 Ga Sheba Fm

0.1 1 10 100

(o) NE Atlantic n < 130 Antarctic Lake n = 17 Chesapeake Bay n = 75 California Margin n = 260 Gulf of California n > 60 Baja California n = 100 Mediterranean n > 400 Saanich Inlet n > 220 East Pacific Rise n = 60 Pakistan Margin n = 17 Washington Coast n =29 NW African margin n > 10 Lake Marawi n > 50 Modern and Japan Sea n = 44 Phanerozoic Mexican Margin n > 100 sediments Arabian Margin n = 12 Cariaco Trench n > 90 Black Sea n > 140 Framvaren Cretaceous SW African Shelf n > 140 Jurassic n = 25 n > 40 Devonian Carboniferous n > 1870 Cambrian n = 6

0.1 1 10 100 1000

[Mo/Al]sample

[Mo/Al]Av. Sh. 220

Fig. 3-6. Enrichment factors (EF) for the Mo contents relative to the Al contents of the Archean–Paleoproterozoic samples of this study (3-6-a through -n). Each data point is shown. Those of the modern and Phanerozoic sediments are also shown using literature data (3-6-o). The ranges and the number of data are shown. EF is defined as a ratio of [Mo/Al]sample to [Mo/Al]average shale, where Mo/Al = ppm/wt.% ratio. Data for modern sediments are from NE Atlantic (Thomson et al., 1996), Antarctic lake (Bishop et al., 2001), the Chesapeake Bay (Adelson et al., 2001), the California Margin (Dean et al., 1997), Gulf of California (Brumsack, 1986a, 1989), Baja California (Shimmield and Price, 1986), the Saanich Inlet (Gross et al., 1963; Emerson and Huested, 1991; Calvert and Pedersen, 1993; Crusius et al., 1996; Calvert et al., 2001; Morford et al., 2001; Russell and Morford, 2001), the Mediterranean Sea (Thomson et al., 1995; Nijenhuis et al., 1998; Warning and Brumsack, 2000), East Pacific Rise (Schaller et al., 2000), Pakistan Margin (Crusius et al., 1996), Washington Coast (Morford and Emerson, 2001), NW Atlantic Margin (Morford and Emerson, 2001), Lake Malawi (Brown et al., 2000), the Japan Sea (Crusius et al., 1996), the Mexican Margin (Nameoff, 1996), the Arabian Margin (Morford and Emerson, 2001), the Cariaco Trench (Dorta and Rona, 1971; Jacob et al., 1987; Emerson and Huested, 1991; Calvert and Pedersen, 1993; Adelson et al., 2001), the Black Sea (Pilipchuk and Volkov, 1974; Brumsack, 1989; Calvert, 1990; Emerson and Huested, 1991; Crusius et al., 1996; Warning and Brumsack, 2000), the Framvaren Fjord (Skei, 1981; Jacob et al., 1987; Emerson and Huested, 1991), and SW African shelf (Calvert and Price, 1983; Emerson and Huested, 1991). Data for Phanerozoic sediments are from Cretaceous (Brumsack, 1986a, b; Dean and Arthur, 1986; Arthur et al., 1990; Warning and Brumsack, 2000), Jurassic (Brumsack, 1986), Carboniferous (Coveney et al., 1987), Devonian (Leventhal et al., 1983; Robl and Barron, 1988; Caplan and Bustin, 1996, 1998; Joachimski et al., 2001), and Cambrian (Leventhal, 1990). 221

(a) ~2.2 Ga Mapedi Fm central Transvaal eastern Transvaal (b) 2.22 Ga Timeball Hill Fm (c) 2.56 Ga Oak Tree Fm (d) 2.60 Ga Wittenoom Dolomite Fm

(e) 2.60 Ga Carawine Dolomite Fm (f) >2.60 Ga Marra Mamba Iron Fm (g) 2.64 Ga Black Reef Fm

(h) 2.69 Ga Lewin Fm (i) 2.69 Ga Jeerinah Fm

(j) 2.71 Ga Rietgat Fm

(k) 2.72 Ga Pillingini Tuff Fm (l) 2.96 Ga Parktown Fm (m) 3.25 Ga Sheba Fm (graywacke) siderite-rich silica-rich (n) shale 3.25 Ga Sheba Fm

California Margin n = 260 (o) NE Atlantic n < 130 Antarctic lake n = 17 Black Sea n > 140 Baltic Sea n = 21 Saanich Inlet n > 130 Mexican Margin n > 100 N Atlantic n = 37 Modern and Eocene Phanerozoic Ucluelet Inlet n = 15 sediments Panama Basin n = 40 Gulf of California n = 50 East Pacific Rise n = 60 Mediterranean n > 430

Eocene Cretaceous n > 150 Jurassic n = 2 Carboniferous n > 40 Devonian n > 1850 Cambrian n = 6

0.01 0.1 1 10 100

[Mo/Corg]sample

[Mo/Corg]Av. Sh. 222

Fig. 3-7. Enrichment factors (EF) for the Mo contents relative to the Corg contents of the Archean–Paleoproterozoic samples of this study (3-7-a through -n). Each data point is shown. Those of the modern and Phanerozoic sediments are also shown using literature data (3-7-o). The ranges and the number of data (where available) shown. EF is defined as a ratio of [Mo/Corg]sample to [Mo/Corg]average shale, where Mo/Corg = ppm/wt.% ratio. Data for modern sediments are from the California Margin (Dean et al., 1997), NE Atlantic (Thomson et al., 1996, 2001), Antarctic lake (Bishop et al., 2001), the Black Sea (Pilipchuk and Volkov, 1974; Volkov and Fomina, 1974; Lange et al., 1977; Holland, 1984; Brumsack, 1989; Warning and Brumsack, 2000), the Baltic Sea (Sternbeck et al., 2000), the Saanich Inlet (Calvert et al., 2001; Russell and Morford, 2001), the Mexican Margin (Nameoff, 1996), N. Atlantic (Crusius and Thomson, 2000), the Ucluelet Inlet (Calvert and Pedersen, 1993), Panama Basin (Pedersen et al., 1988), Gulf of California (Brumsack, 1986, 1989), East Pacific Rise (Schaller et al., 2000), and the Mediterranean Sea (Nijenhuis et al., 1998; Warning and Brumsack, 2000). Data for Phanerozoic sediments are from Eocene (Desborough et al., 1976), Cretaceous (Brumsack, 1980, 1986a, b; Alberdi-Genolet and Tocco, 1999; Warning and Brumsack, 2000), Jurassic (Brumsack, 1986), Devonian (Leventhal and Hosterman, 1982; Leventhal et al., 1983; Robl and Barron, 1988; Caplan and Bustin, 1996, 1998; Joachimski et al., 2001), and Cambrian (Leventhal, 1990). 223

400 : Black Sea sediments : Mediterranean sediments : Oslo Fjord anoxic sediments : Gulf of California sediments 100 : Cretaceous black shale : Devonian black shales : Cambrian Alum Shale

10 Mo [ppm] Avg. Shale

1

0.1

0.01 0.1 1 10 20

Corg [wt.%]

Fig. 3-8. Plot of the Mo vs. Corg contents for modern marine carbonaceous sediments and Phanerozoic black shales. Data are from the Black Sea (Hirst, 1974), the Mediterranean Sea (Warning and Brumsack, 2000), Oslo Fjord (Calvert, 1976), Gulf of California (Brumsack, 1989), Cretaceous (Brumsack, 1980; Holland, 1984; Arthur et al., 1990), Devonian (Leventhal et al., 1983; Holland, 1984; Shaffer et al., 1984; Robl and Barron, 1988; Leventhal, 1993), and Cambrian (Armands, 1972; Leventhal, 1990). Average shale data is from Wedepohl (1991) and Warning and Brumsack (2000). Samples with Mo concentration less than 1 ppm are plotted at Mo = 1 ppm. Shaded area represent Mo < 1 ppm and Corg < 1.5 wt.%. Note the overall positive correlation between Mo and Corg, and variabilities of Mo content with Corg content between 0.5 - 2.0 wt.%. 224

400 : Black Sea sediments : Mediterranean sediments : Oslo Fjord anoxic sediments : Gulf of California sediments 100 : Cretaceous black shale : Devonian black shales : Cambrian Alum Shale

10 Mo [ppm] Avg. Shale

1

0.1 0.01 0.1 13 10 S [wt.%]

Fig. 3-9. Plot of the Mo vs. S contents for modern marine carbonaceous sediments and Phanerozoic black shales. Data are fromthe Black Sea (Hirst, 1974), the Mediterranean Sea (Warning and Brumsack, 2000), Oslo Fjord (Calvert, 1976), Gulf of California (Brumsack, 1989), Cretaceous (Brumsack, 1980; Holland, 1984; Arthur et al., 1990), Devonian (Leventhal et al., 1983; Holland, 1984; Shaffer et al., 1984; Robl and Barron, 1988; Leventhal, 1993), and Cambrian (Armands, 1972; Leventhal, 1990). Average shale data is from Wedepohl (1991) and Warning and Brumsack (2000). Samples with Mo concentration less than 1 ppm are plotted at Mo = 1 ppm. Shaded area represent Mo < 1 ppm and S < 3 wt.%. Note the overall positive correlation between Mo and S, and variabilities of Mo content with S content between 1 - 3 wt.%. 225

(A) (D) High Mo + High Corg + High S • euxinic marine condition: sulfidic bottom water & Mo anoxic sediments • saline lake? [ppm] • hydrothermal overprint ~10 (B)Low Mo + Low Corg (C) Low Mo + High Corg + Low S + Low / High S

• oxic marine condition • anoxic sediments overlain • freshwater condition by oxic bottom water, 2- (SO4 -limited) high sedimentation rate 2- • pyrite-reoxidation? (SO4 -limited) • freshwater condition 2- (SO4 -limited)

~1

Corg [wt.%], S [wt.%]

Fig. 3-10. Schematic diagram showing relationship among Mo, Corg, and S content of sediments / sedimentary rocks. Environmental changes occur among the cases (B), (C), and (D). Modified after Lyons et al. (2000). 226

200 California Margin 100 Mexican Margin Oman Margin Santa Barbara California Borderland Basins Cariaco Trench & Framvaren Fjord Sannich Inlet Black Sea

[µM] Japan Sea 2 Pakistan Margin 10 Dissolved O

Oxic seawater

Dissolved O > 1 [µM]: Berner (1981) 1 2 Anoxic seawater 0 0102030405060140

Average shale Sediment Mo [ppm] Mo content: ~1 [ppm]

Fig. 3-11. Relationship between Mo contents in modern sediments and dissolved O2 contents in their overlying bottom waters. Dashed line at dissolved O2 content of 1 [µM] separates between oxic and anoxic condition (Berner, 1981). Modified after Zheng et al. (2000), and references for modern sediments are therein. 227 Table 3-1. Summary of the geological settings and samples of this study.

Drillcore Drillhole locality Supergroup (SG) Age Formation (Fm) N Group (G) [Ga]

South Africa PU1308 Agnes gold mine, Swaziland SG 3.25 Sheba Fm 17 Barberton Fig Tree G MRE10 Sheba gold mine, Swaziland SG 3.25 Sheba Fm 10 Barberton Fig Tree G DRH13 26˚52'S, 26˚23'E, Witwatersrand SG 2.96 Parktown Fm 15 near West Rand G Witwatersrand MSF6 26˚33'S, 27˚12'E Ventersdorp SG 2.71 Rietgat Fm 7 near Klerksdorp Platberg G Witwatersrand Transvaal SG Chuniespoort G 2.56 Oak Tree Fm 3 Pretoria G 2.22 Timball Hill Fm 4

JPBR near Klerksdorp Transvaal SG 2.64 Black Reef Fm 4 Witwatersrand Wolkberg G PTB3 24˚55' S, 30˚44'E Transvaal SG 2.22 Timeball Hill Fm 12 Pilgrim's Rest Pretoria G SA1677 Sishen iron mine, Griqualand Seq. ~2.2 Mapedi Fm 8 Griqualand West Olifantshoek G Australia WRL1 East of Wittenoom, Mt. Bruce SG 2.72 Pillingini Tuff Fm 3 Hamersley Fortescue G 2.69 Jeerinah Fm 6 >2.60 Marra Mamba Iron Fm 4 2.60 Wittenoom Dol. Fm 3 RHDH2A Ripon Hill, Mt. Bruce SG 2.69 Lewin Shale Fm 8 Northeastern Fortescue G Hamersley Mt. Bruce SG 2.60 Carawine Dol. Fm 6 Hamersley G

N: number of samples 228 Table 3-1. (continued)

Drillcore Dominant lithology of Tectonic settings of the groups the studied samples

South Africa PU1308 Black shales (some are rich Foreland basin, fore deep basin, in siderite and silica) evolving back-arc, passive continental margin, early rift to MRE10 Graywackes continental shelf, and shelf-rise

DRH13 Black shales Shallow marine or tidal marine with minor alluvial depositional environments

MSF6 Black shales

MSF6 Black shales Shallow marine cratonic MSF6 Black shales environments

JPBR Carbonate-rich black shales PTB3 Black shales Shallow marine cratonic environments

SA1677 Red shales

Australia WRL1 Black shales, carbonate rich Shallow marine WRL1 black shales Deep marine WRL1 Deep marine WRL1 Deep marine RHDH2A Black shales Deep marine

RHDH2A Carbonate-rich black shales Shallow marine or intertidal environments

See chapter 2 for the references of the tectonic settings of the samples. 229 Table 3-2. Geochemical data for the Archean–Paleoproterozoic samples of this study.

Samples Mo Zn Al2O3 ∑Fe2O3 Corg Ccarb Stotal [ppm] [ppm] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Carbonate-rich black shales, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) PU1308-01 0.7 47 3.87 43.50 0.68 8.07 0.15 PU1308-02 0.8 109 5.54 42.94 0.94 8.20 0.00 PU1308-03 0.5 130 1.91 36.50 0.43 6.91 0.16 PU1308-04 0.0 93 5.06 37.92 0.94 7.27 0.67 PU1308-05 1.2 139 8.15 31.62 1.43 6.25 0.00 PU1308-06 0.6 133 3.21 43.46 1.02 8.38 0.00 PU1308-07 1.8 200 17.46 13.25 2.97 2.19 0.00 PU1308-08 3.2 0.0 5.84 40.21 1.29 7.83 0.00 PU1308-09 1.2 124 0.98 11.73 0.46 3.24 0.00 PU1308-10 2.0 98 19.16 10.59 1.87 1.84 0.00 PU1308-11 0.2 139 20.77 17.55 0.22 2.78 0.00 PU1308-12 1.5 1014 6.47 38.36 1.31 6.71 2.15 PU1308-13 6.6 511 18.19 25.32 2.28 3.88 2.70 PU1308-14 2.1 214 3.48 49.31 1.08 9.22 0.00 PU1308-15 0.3 103 1.46 42.63 0.49 8.14 0.00 PU1308-16 0.9 97 5.83 43.11 0.93 8.07 0.00 PU1308-17 0.5 471 1.77 41.14 0.42 7.80 0.00

Average 1.4 213 7.60 33.48 1.10 6.28 0.34 S.D. 1.6 246 6.77 12.76 0.72 2.46 0.81

Graywackes, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) MRE10-01 1.0 80 9.48 6.08 0.16 1.77 0.00 MRE10-02 1.3 133 13.42 9.89 0.52 1.52 0.00 MRE10-03 2.8 98 18.13 10.84 0.43 0.91 0.00 MRE10-04 1.1 93 11.35 9.13 0.21 1.21 0.00 MRE10-05 5.2 214 14.76 9.11 0.39 0.21 0.00 MRE10-06 2.4 96 15.29 9.09 0.12 2.17 0.00 MRE10-07 1.4 93 10.41 7.17 0.16 1.51 0.00 MRE10-08 0.9 125 8.10 6.15 0.17 2.83 0.00 MRE10-09 1.8 126 13.39 10.56 0.28 1.63 0.00 MRE10-10 1.9 91 10.88 9.01 0.29 2.91 0.21

Average 2.0 115 12.52 8.70 0.27 1.67 0.02 S.D. 1.3 39 3.04 1.69 0.13 0.82 0.07 230 Table 3-2. (continued)

Samples Mo Zn Al2O3 ∑Fe2O3 Corg Ccarb Stotal [ppm] [ppm] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Iron-rich shales, Parktown Fm, West Rand G, Witwatersrand SG (2.96 Ga) DRH13-01 1.6 55 8.94 24.68 0.23 0.04 0.08 DRH13-02 0.8 54 6.95 25.55 0.08 0.15 0.08 DRH13-03 0.7 81 6.16 28.96 0.11 0.28 0.00 DRH13-04 1.2 76 10.51 25.47 0.06 0.02 0.00 DRH13-05 0.9 65 9.27 24.59 0.17 0.07 0.00 DRH13-06 0.9 72 10.41 23.92 0.07 0.02 0.69 DRH13-07 0.8 77 8.00 27.47 0.04 0.02 0.46 DRH13-08 1.0 51 4.64 41.72 0.14 0.08 0.00 DRH13-09 1.7 75 2.77 36.14 0.25 0.09 0.12 DRH13-10 2.9 57 13.70 17.17 0.51 0.00 0.30 DRH13-11 1.3 94 13.90 15.83 0.07 0.02 0.00 DRH13-12 0.9 124 12.93 14.88 0.70 0.42 0.00 DRH13-13 1.4 104 13.32 12.03 0.18 0.05 0.41 DRH13-14 2.5 126 17.72 8.00 0.27 0.01 0.35 DRH13-15 2.4 211 17.88 7.85 0.22 0.00 0.18

Average 1.4 88 10.47 22.28 0.21 0.08 0.18 S.D. 0.7 41 4.46 9.70 0.18 0.12 0.22

Shales, Pillingini Tuff Fm, Fortescue G, Mt. Bruce SG (2.72 Ga) WRL1-01 1.3 99 15.79 16.96 0.21 0.01 0.09 WRL1-02 2.6 107 14.07 8.34 0.26 0.12 0.07 WRL1-03 1.1 66 11.01 7.48 0.14 0.42 0.13

Average 1.6 90 13.62 10.93 0.20 0.18 0.10 S.D. 0.8 22 2.42 5.24 0.06 0.21 0.03

Black shales, Rietgat Fm, Platberg G, Ventersdorp SG (2.71 Ga) MSF6-01 1.2 66 8.92 4.61 0.32 0.24 0.13 MSF6-02 1.3 104 9.74 6.48 0.27 0.30 0.00 MSF6-03 1.4 92 11.96 6.46 0.59 0.50 0.00 MSF6-04 2.8 71 13.92 6.78 0.95 0.00 0.64 MSF6-05 1.3 111 17.90 7.94 1.30 0.03 0.38 MSF6-06 1.8 95 19.94 8.18 1.59 0.00 0.67 MSF6-07 3.5 111 16.64 9.61 1.10 0.04 0.53

Average 1.9 93 14.14 7.15 0.87 0.16 0.34 S.D. 0.9 18 4.19 1.59 0.50 0.19 0.29 231 Table 3-2. (continued)

Samples Mo Zn Al2O3 ∑Fe2O3 Corg Ccarb Stotal [ppm] [ppm] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Black shales, Lewin Shale Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) RHDH2A-01 2.1 33 13.26 4.25 2.67 0.34 0.21 RHDH2A-02 3.2 527 15.37 4.12 6.83 0.32 1.44 RHDH2A-03 2.4 40 16.16 4.76 1.56 0.19 0.07 RHDH2A-04 3.6 0 15.84 5.74 2.66 0.03 0.07 RHDH2A-05 3.1 117 15.69 7.52 1.83 0.04 0.59 RHDH2A-06 4.2 200 14.49 8.63 2.50 0.44 0.58 RHDH2A-07 4.2 146 14.21 10.60 2.05 0.10 4.04 RHDH2A-08 5.3 0 16.29 4.77 2.18 0.85 0.28

Average 3.5 133 15.16 6.30 2.78 0.29 0.91 S.D. 1.0 175 1.07 2.37 1.68 0.27 1.34

Black shales, Jeerinah Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) WRL1-04 2.8 816 15.55 0.17 5.52 0.14 1.55 WRL1-05 3.6 593 13.79 2.59 7.39 2.32 1.84 WRL1-06 3.9 0 12.70 0.43 12.04 0.38 2.98 WRL1-07 8.3 0 12.92 0.42 8.19 0.34 3.03 WRL1-08 3.0 293 13.87 0.48 2.73 0.31 1.95 WRL1-09 4.2 150 11.23 0.21 3.87 0.10 0.74

Average 4.3 309 13.34 0.72 6.62 0.60 2.02 S.D. 2.0 333 1.44 0.93 3.36 0.85 0.88

Carbonate-rich black shales, Black Reef Fm, Wolkberg G, Transvaal SG (2.64 Ga) JPBR-01 0.4 0 0.77 1.03 0.09 12.02 0.00 JPBR-02 0.4 0 0.86 1.56 0.09 11.71 0.64 JPBR-03 0.5 0 3.26 2.68 0.33 11.83 0.04 JPBR-04 2.1 36 26.57 0.03 1.11 0.07 0.00

Average 0.8 9 7.87 1.32 0.41 8.91 0.17 S.D. 0.8 18 12.52 1.10 0.48 5.90 0.31 232 Table 3-2. (continued)

Samples Mo Zn Al2O3 ∑Fe2O3 Corg Ccarb Stotal [ppm] [ppm] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Black shales, Marra Mamba Iron Fm, Hamersley G, Mt. Bruce SG (>2.60 Ga) WRL1-10 0.5 107 7.14 28.95 0.19 0.01 0.00 WRL1-11 0.5 0 2.47 34.33 0.37 1.51 0.37 WRL1-12 0.8 55 4.46 32.38 0.93 1.40 3.16 WRL1-13 1.2 36 4.00 34.70 0.74 3.80 0.14

Average 0.8 50 4.52 32.59 0.56 1.68 0.92 S.D. 0.3 45 1.94 2.63 0.34 1.57 1.50

Black shales, Wittenoom Dolomite Fm, Hamersley G, Mt. Bruce SG (2.60 Ga) WRL1-14 0.7 19 1.57 10.62 0.74 10.59 0.08 WRL1-15 3.1 0 14.28 13.01 2.78 0.13 1.64 WRL1-16 2.8 66 12.24 17.08 2.67 0.04 0.14

Average 2.2 28 9.36 13.57 2.06 3.59 0.62 S.D. 1.3 34 6.83 3.27 1.15 6.06 0.88

Carbonate-rich black shales, Carawine Dolomite Fm, Hamersley G, Mt. Bruce SG (2.60 Ga) RHDH2A-09 1.1 0 5.87 15.19 1.41 7.21 0.44 RHDH2A-10 0.7 0 4.99 4.07 1.02 9.22 0.08 RHDH2A-11 1.5 0 7.32 7.73 2.91 6.10 0.73 RHDH2A-12 1.6 0 9.50 5.44 1.91 6.68 0.14 RHDH2A-13 2.9 97 10.76 6.74 4.24 2.15 0.36 RHDH2A-14 2.4 93 15.77 4.20 4.45 7.29 1.23

Average 1.7 32 9.04 7.23 2.66 6.44 0.50 S.D. 0.8 49 3.95 4.15 1.45 2.35 0.43

Black shales, Oak Tree Fm, Chuniespoort G, Transvaal SG (2.56 Ga) MSF6-08 1.2 97 17.05 1.99 2.30 0.15 0.27 MSF6-09 1.2 49 13.74 2.60 1.64 0.05 0.94 MSF6-10 1.2 83 15.53 1.49 1.66 0.00 0.00

Average 1.2 76 15.44 2.03 1.87 0.07 0.40 S.D. 0.0 25 1.66 0.56 0.37 0.07 0.49 233 Table 3-2. (continued)

Samples Mo Zn Al2O3 ∑Fe2O3 Corg Ccarb Stotal [ppm] [ppm] [wt.%] [wt.%] [wt.%] [wt.%] [wt.%]

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) [Eastern Transvaal] PTB3-01 2.0 82 24.69 7.44 0.41 0.00 0.00 PTB3-02 3.4 13 21.62 4.85 0.35 0.88 0.32 PTB3-03 7.7 74 24.88 8.85 0.46 0.00 0.00 PTB3-04 0.8 53 25.95 8.74 0.39 0.00 0.00 PTB3-05 1.0 58 24.73 10.18 0.39 0.00 0.00 PTB3-06 1.1 81 25.09 9.42 0.39 0.40 0.06 PTB3-07 1.3 66 24.91 9.96 0.36 0.28 0.00 PTB3-08 1.0 16 29.44 2.87 0.50 0.38 0.42 PTB3-09 1.3 50 26.36 9.04 0.43 0.01 0.00 PTB3-10 1.1 36 25.30 10.63 0.47 0.45 0.00 PTB3-11 0.8 76 25.54 8.61 0.43 0.00 0.00 PTB3-12 1.0 46 24.36 9.15 0.30 0.45 0.00

Average 1.9 54 25.24 8.31 0.41 0.24 0.07 S.D. 2.0 23 1.77 2.28 0.06 0.28 0.14

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) MSF6-11 0.9 111 18.43 9.63 0.28 0.00 0.00 MSF6-12 2.4 117 20.60 7.68 0.71 0.00 0.00 MSF6-13 2.6 90 18.85 6.96 0.55 0.00 0.00 MSF6-14 2.2 87 20.07 6.74 0.71 0.00 0.00

Average 2.0 101 19.48 7.75 0.56 0.00 0.00 S.D. 0.8 15 1.02 1.31 0.20 0.00 0.00

Red shales, Mapedi Fm, Pretoria G, Transvaal SG (~2.2 Ga) SA1677-01 1.0 70 16.98 10.91 0.10 0.00 0.00 SA1677-02 0.4 54 16.98 10.76 0.08 0.09 0.00 SA1677-03 0.3 27 19.47 9.52 0.09 0.00 0.00 SA1677-04 0.3 63 19.78 10.24 0.07 0.00 0.00 SA1677-05 0.8 19 18.79 13.26 0.07 0.00 0.00 SA1677-06 0.8 65 12.37 18.54 0.06 0.20 0.00 SA1677-07 0.5 77 14.50 14.75 0.06 0.72 0.00 SA1677-08 1.1 26 14.97 18.43 0.06 0.00 0.00

Average 0.7 50 16.73 13.30 0.07 0.13 0.00 S.D. 0.3 23 2.63 3.62 0.02 0.25 0.00

Fm: Formation; G: Group; SG: Supergroup; b.d.: below detection, S.D.: Standard deviation. 234 Table 3-3. Molybdenum concentration of various rocks.

Rocks Mo content [ppm]

Igneous rocks Granite 1.1 Granitoid 1.8 Intermediate 0.8 Basalt 1.2 Ultramafic 0.3

Metamorphic rocks 0.9 Slate- 0.2 Phyllite 0.2 ~ 1.2 Schist 0.3 Granitegneiss 0.5

Sedimentary materials Recent sediments Oxic sediments 1.5 Anoxic sediments 1.0 ~ 100 Deep Sea Ferromanganese nodule 380 Sedimentary rocks Shale 0.7 ~ 2 Sandstone 0.3 Graywacke 0.7 Carbonate 0.4 Black shale 3.0 ~ 1000 Phosphorites 11

Continental crust 1.2 Earth's crust 1

Data from Manheim and Landergren (1978) 235 Table 3-4. Comparison of the average Mo to organic C and Mo to Al ratios among Archean–Paleoproterozoic sedimenary rocks

Samples Mo/Corg Mo/Al [Mo/Al]sample N References

Fm: Formation; SG: Supergroup [ppm/% ± 1σ] [ppm/% ± 1σ] [Mo/Al]Av. Sh.

Proterozoic Ramah Group, Labrador 1.9 Ga Nullataktok Fm (black shales) 0.67 ± 0.76 0.22 ± 0.24 1.47 19 1 Pretoria Group, Transvaal SG ~2.2 Ga Mapedi Fm 9.24 ± 4.79 0.08 ± 0.04 0.53 8 2 2.22 Ga Timeball Hill Fm 3.58 ± 0.74 0.19 ± 0.07 1.29 4 2 2.22 Ga Timeball Hill Fm, Pilgrim's R 4.54 ± 4.36 0.14 ± 0.15 0.95 12 2 2.56 Ga Oak Tree Fm 0.65 ± 0.09 0.15 ± 0.01 0.97 3 2 2.90 ± 1.42 0.55 ± 0.39 3.67 4 2 Ha2.60 Ga Wittenoom Dolomite Fm >2.60 Ga Marra Mamba Iron Fm 1.03 ± 0.09 0.56 ± 0.24 3.72 3 2 2.60 Ga Carawine Dolomite Fm 1.68 ± 0.82 0.36 ± 0.17 2.41 4 2 0.68 ± 0.13 0.36 ± 0.08 2.38 8 2

Arch 2.64 Ga Black Reef Fm Fortescue Group, Mt. Bruce SG 2.69 Ga Lewin Shale Fm 1.50 ± 0.63 0.44 ± 0.12 2.91 6 2 2.69 Ga Jeerinah Fm 0.75 ± 0.35 0.62 ± 0.32 4.17 6 2 2.72 Ga Pillingini Tuff Fm 7.93 ± 2.01 0.23 ± 0.11 1.50 3 2 Platberg Group, Ventersdorp SG 2.71 Ga Rietgat Fm 2.72 ± 1.34 0.26 ± 0.10 1.72 7 2 West Rand Group, Witwatersrand SG 2.96 Ga Parktown Fm 9.79 ± 5.52 0.30 ± 0.25 1.99 15 2 Fig Tree Group, Swaziland SG 3.25 Ga Sheba Fm (graywackes) 8.09 ± 5.10 0.29 ± 0.14 1.93 10 2 3.25 Ga Sheba Fm (shales) 1.24 ± 0.77 0.53 ± 0.54 3.52 17 2

Average Shale 6.5 0.15 1.00 3 Average Black Shale 3.1 1.4 9.5 4

N: Number of samples References: 1: Hayashi et al . (1997); 2: This study; 3: Wedepohl (1991) and Warning and Brumsack (2000); 4: Vine and Tourtelot (1970). 236 Table 3-5. Source and sink fluxes for Mo.

Fluxes Mo References

Source fluxes [108 mol/yr] River 1.8 1 Eolian dust negligible Continental margin sediments 0.08 ~ 0.17 2 Hydrothermal (low and high T) negligible

Source total 1.88 ~ 1.97

Sink fluxes [108 mol/yr]

Sediments Oxic 0.9 3 Anoxic 0.2 ~ 0.8 3, 4 Continental margin sediments negligible Mn nodule and metalliferous sediments negligible 4

Sediment total 1.1 ~ 1.7 Hydrothermal (low and high temperature) 0.04 ~ 0.17 5

Sink total 1.14 ~ 1.87

Oceanic residence time with respect to river water input [kyr] 800

References 1:Martin and Meybeck (1979) 2:Morford and Emerson (1999) 3:Bertine and Turekian (1973) 4:Manheim and Landergren (1978) 5:Chen et al. (1986) Chapter 4

Geochemistry of Archean–Paleoproterozoic black shales:

III. N2-fixation and redox-cycling of nitrogen by microorganisms

Abstract

Isotopic compositions and contents of organic-N (Norg), inorganic-N (Nclay), organic

carbon (Corg), and Mo were investigated in the Mesoarchean–Paleoproterozoic black shales in South Africa and Australia. Samples are from fresh drillcores of the 3.25 Ga Fig Tree Group (17 shales and 12 graywacke), the 2.6 Ga Chuniespoort Group (3 shales), the 2.2 Ga Pretoria Group (12 shales), the 2.6 Ga Hamersley (13 shales), and the 2.7 Ga Fortescue Group (8 shales). This study is different from the recent studies on Archean N isotope geochemistry (Beaumont and Robert, 1999, Precambrian Research 96, 63-82) in which hydrothermally influenced chemical sediments such as chert and banded iron-formations were used.

13 A mean δ Corg value of -28.7 ± 1.0 ‰ for the Fig Tree Group samples suggests

oxygenic photosynthetic (probably cyanobacterial) fixation of atmospheric CO2 whose partial pressure was probably higher than today. The distribution pattern and magnitude of

15 15 the δ Norg and δ Nclay values (0.2 ‰ and 4.3 ‰ in average, respectively) of the 3.25 Ga black shales and other Archean–Paleoproterozoic samples are found to be very similar to those of Cretaceous black shales. Such similarity suggests that certain common 238

mechanisms to control N biogeochemical cycling have been operating since at least the Mesoarchean. As such, we suggest two mechanisms that are not mutually exclusive. One is the microbially mediated redox-cycling of N, possibly involving denitrification and

microbial (possibly cyanobacterial) N2-fixation from the atmosphere in the ancient sedimentary environments. The atmosphere-hydrosphere system in the Archean would have been sufficiently oxidized to allow redox-cycling of N. The microbial fixation of atmospheric N2 probably is suggested to have its deep-root into the evolution of microbial biosphere of the early Earth. The other mechanism is the microbially mediated non-redox

cycling of NH3 in sediments. In either case, this study suggests the operation of the microbial N cycling at least since the Mesoarchean. 239

4-1. Introduction

4-1-1. Nitrogen isotope as an indicator of past biological activity In the Earth's history, major biological and environmental changes occurred in the Precambrian. They include the origin of life and evolution of biosphere (e.g., Schopf, 1983; Schopf and Klein, 1992; Schidlowski et al., 1992; Bengston, 1994), the chemical evolution of atmosphere-oceans-crust-mantle system (e.g., Holland, 1984, Taylor and McLennan,

1985), and especially, the rise of atmospheric O2 (e.g., Holland, 1984, 1994; Kasting, 1993; Ohmoto, 1997; Holland, 1999). Almost all aspects of those evolutionary changes have been debated and very little consensus has been reached. Recently, Beaumont and Robert (1999) showed that secular variations of N isotopic compositions of kerogen in Precambrian chemical sediments might have reflected redox evolution of the surface environments. Nitrogen is potentially one of the most useful markers of past biological activity and environmental changes. The nitrogen biogeochemical cycle in the oceans today, and probably in the past, has been controlled by processes such as biological fixation, assimilation, nitrification, and denitrification. These processes are usually accompanied by kinetic fractionations of N isotopes. Their isotopic fingerprints are recorded in sedimentary organic matter (OM) that constitutes only a very minor portion of primary production due to organic and inorganic remineralization. Despite the possibility of some modifications during diagenesis and metamorphism, N isotopic compositions of marine sedimentary OM have been regarded to reflect those of their source organisms influenced by the surrounding environments (e.g., in water column, in sediments) (Macko et al., 1993). Therefore, N isotopic compositions of sedimentary OM have been used to elucidate sources of OM whether marine or terrestrial (e.g., Peters et al., 1978; Sweeney and Kaplan, 1980; Minagawa and Wada, 1986; Ohkouchi 240

et al., 1997), degree of nitrate utilization in surface waters (e.g., Altabet and Francois, 1994), and paleoenvironmental biogeochemistry (e.g., Rau et al., 1987).

4-1-2. Nitrogen biogeochemistry in Precambrian Compared to the modern N biogeochemical cycles, very little is known about the Precambrian N biogeochemical cycles. Studies have been hampered by a paucity of very old, relatively unmetamorphosed and undeformed rock samples and their generally very low N contents. Only a few studies of Precambrian sedimentary N isotopes have been published (e.g., Hayes et al., 1983; Schidlowski et al., 1983; Sano and Pilinger, 1990; Beaumont and Robert, 1999; Pinti and Hashizume, 2001; Pinti et al., 2001). However, such studies typically focused on the OM in chemical sediments, rather than the OM in normal clastic sediments, to elucidate information on the ancient biogeochemical cycles of N. Beaumont and Robert (1999) used Precambrian cherts and BIFs (banded iron-formations) and discussed their results for the kerogen N isotopic compositions in terms of environmental redox evolution in the Paleoproterozoic (e.g., Holland, 1994). However, Pinti and Hashizume (2001) pointed out that the N isotopic record of Beaumont and Robert (1999) was likely influenced by submarine hydrothermal vent organisms, suggesting the limitation of chemical sediments as recorders of a global environmental change. Furthermore, the data set of Beaumont and Robert (1999) exhibits considerable inter-subsample variability both in N isotopic compositions and N contents, casting some doubt on their conclusions. Key questions remain concerning how N isotope records can be reliably used to constrain the evolution of the atmosphere and biosphere, especially when and how N biogeochemical cycles involving microbial redox-reactions and N2-fixation began. 241

4-1-3. Objectives To obtain evolutionary insights into these questions and to constrain the environmental redox evolution in the Archean from a nitrogen perspective, we performed stable isotope analyses of N and Corg using the carbonaceous shales of the Mesoarchean Swaziland Supergroup and Neoarchean–Paleoproterozoic Transvaal Supergroup in South Africa and the Neoarchean–Paleoproterozoic Mt. Bruce Supergroup in Western Australia. We used modern weathering-free, low metamorphic grade drillcore samples of shales, rather than outcrop samples of chemical sediments such as cherts and BIFs studied by Beaumont and Robert (1999). The main reason for avoiding the OM in cherts and BIFs is because such OM may represent the remnants of chemoautotrophic microorganisms that lived near submarine hydrothermal vents (c.f., Pinti and Hashizume, 2001).

Our study provides new data for OM-bound N (Norg), clay-bound N (Nclay), and

Corg elemental and stable isotopic compositions of the shales, and incorporates new data on the Mo content of the shales from Chapter 3 of this thesis. Our study aims to understand (1) if microbially-mediated redox-cycling of N was already operating in the Archean oceans, especially to constrain the time when microbial N2-fixation and denitrification began to operate, and (2) the connection between the redox-cycling of N to that of O2. The results of this study have implications for the evolution of the biosphere and the chemical evolution of the atmosphere and oceans. This work is one of a series of studies designed to constrain the paleoredox environments from the geochemistry of Archean black shales. 242

4-2. Background

4-2-1. Nitrogen biogeochemical cycles Nitrogen has been critical to all living organisms ever since their origin(s) in the distant past, because N resides in protein, cell membrane, DNA, etc. The largest N reservoir in the surface environment of the Earth is the atmosphere. The average crustal abundance of N is very low (~50 ppm by mass; e.g., Chameides and Perdue, 1997). Nitrogen occurs in

different chemical oxidation states (-3 to +5), varying from its most reduced form of NH3,

to NH2OH (hydroxylamine), N2 (dinitrogen ), N2O (nitrous oxide), NO (nitric oxide), - - NO2 (nitrite), and to its most oxidized form NO3 (e.g., Chameides and Perdue, 1997). All of these oxidation states are biologically significant and microbes are capable of carrying out redox reactions transforming one form into another. The element N therefore is subject to microbial cycling in nature. Nitrogen enters the ocean by atmospheric deposition, riverine input, and biological/lightning fixation, and exits the ocean by denitrification and / or burial in the sediments (e.g., Chameides and Perdue, 1997). Figure 4-1 shows schematic biogeochemical pathways of N in the ocean. Some microbes such as cyanobacteria can fix atmospheric N2 into bioavailable forms (Eq. 4-1; Fig. 4-1). Biologically fixed N, such as amino acids,

protein, and nucleic acids (all in the form of amino functional group: -NH2), is released into + the ocean as ammonium NH4 (ammonification) when microbes are degraded principally + by heterotrophic bacteria (i.e., mineralization) (Fig. 4-1). The released NH4 is either taken - up by other organisms as nutrient (i.e., ammonia assimilation) or oxidized to NO3 (i.e., nitrification) (Eq. 4-2 and 4-3; Fig. 4-1). Nitrification proceeds in two step processes, mediated by two kinds of nitrifying bacteria: Nitrosomonas (converting ammonia to nitrite;

- Eq. 4-2) and nitrobacteria (converting nitrite to nitrate; Eq. 4-3). NO3 is either reduced to

N2 and N2O (i.e., denitrification or dissimilatory reduction) (Eq. 4-4; Fig. 4-1) in an O2- 243

deficient environment or assimilated by other organisms (i.e., assimilatory reduction) (Eq. 4-5; Fig. 4-1). Pseudomonas denitrificans is capable of denitrification.

N2-fixation: + - N2 + 5 H2O ---> 2 NH4 + 2 OH + 1.5 O2...... (4-1) Nitrification:

+ - + NH4 + 1.5 O2 ---> NO2 + H2O + 2 H ...... (4-2) - - NO2 + 0.5 O2 ---> NO3 ...... (4-3) Denitrification: (dissimilartory nitrate reduction)

- + CH2O + 0.8 NO3 + 0.8 H ---> CO2 + 1.4 H2O + 0.4 N2 ...(4-4) Assimilatory nitrate reduction

- + + NO3 + H2O + 2 H ---> NH4 + 2 O2 ...... (4-5)

Cyanobacteria are the most important fixers of atmospheric N2 in the oceans today (Jaffe, 1992) and probably in the Precambrian. Cyanobacteria can use a variety of inorganic and organic sources of N to fulfill their nutritional requirements; however, like other

- + diazotrophs, cyanobacteria prefer fixed N sources such as NO3 and NH4 . N2-fixation is the last resort since it requires a high level of both energy and reductant (Eq. 2-1).

4-2-2. Nitrogen isotopic fractionations

During biological N2-fixation, there is little N isotopic fractionation between atmospheric N2 and the fixed form of N (e.g., Hoering and Ford, 1960; Wada et al., 1975; Wada and Hattori, 1991; Minagawa and Wada, 1986). During microbial mineralization of OM, little overall isotopic fractionation occurs (e.g., Hoering and Ford, 1960; Wada and Hattori, 1991). During denitrification, the lighter isotope (14N) is preferentially processed

15 - over the heavier isotope ( N) and returned to the atmosphere. The remaining NO3 pool 244

15 - 15 then becomes enriched in N. As a result, dissolved NO3 in the oceans has a mean δ N - value of +6 % (e.g., Wada et al., 1975). Marine organisms utilize NO3 as the main N- bearing nutrient; their N isotopic compositions are typically found to be, on average, +6 ‰.

15 + The δ N values of present marine inorganic NH4 vary from -3.5 ‰ to +7.5 ‰ (Sweeney et al., 1978). This may be due to biological N2-fixation from the atmosphere and + decomposition of Norg including a complex nutrient process. Bacterial uptake of NH4 from pore fluids can cause significant isotopic fractionation (≈ -14 ‰, as high as -22 ‰ + depending on the NH4 concentration; pers. comm. Michael Arthur). N2-fixation, nitrification, and denitrification are redox reactions and mediated by organisms.

4-2-3. Previous studies of N in the Archean There are two main paths of Archean N studies using geological samples. One uses the N content in metasedimentary rocks, or sometimes even in igneous rocks, as a biomarker (Honma and Itihara, 1981; Itihara and Suwa, 1985; Haendel et al., 1986; Itihara et al., 1986; Itihara and Tainosho, 1989; Honma, 1996; Hall, 1999; Sadofsky and Bebout, 2000), and the second deals with N isotopes in attempts to understand the biological evolution in mostly chemical sediments (Sano and Pilinger, 1990; Beaumont and Robert, 1999; Pinti and Hashizume, 2001; Pinti et al., 2001). There are very few papers regarding the biogeochemistry of Archean N using data from clastic sedimentary rocks (e.g., Schidlowski et al., 1983). Very recent review papers by Boyd (2001a, b) emphasized an

+ importance of NH4 as a biomarker in Precambrian metasediments and N in future astrobiological studies. The present study focuses on the biogeochemical aspects of N in Archean black shales, and relevant studies are only briefly reviewed here.

Hayes et al. (1983) reported twenty-two δ15N values of kerogen (ranging from +0.8 to +9.9 ‰, with a mean of +4.2 ‰) extracted from Precambrian shales, sandstones, carbonates, and cherts (their ages ranging from 0.8 Ga to 3.5 Ga). For the Archean samples, 245

Hayes et al. (1983) reported four δ15N values of kerogen extracted from the 3.4 Ga shales of the Gorge Creek Group, Pilbara, Western Australia (1.6 ‰ and 3.0 ‰) and from the 3.5 Ga cherts of the Swaziland Supergroup (2.9 ‰ and 3.2 ‰). The results of Hayes et al. (1983) are used in later discussion. Schidlowski et al. (1983) reviewed the results of earlier studies by PPRG (Precambrian Palobiology Research Group) and others, and suggested that an active biological N cycle existed as early as ~2.5 Ga ago based on the data set presented by Hayes et al. (1983). Sano and Pilinger (1990) utilized the stepped heating method combined with static

mass spectrometry to investigate the evolution of an atmospheric δ15N value. Sano and Pilinger (1990) analyzed six chert samples of various ages (3.5 Ga to 15 Ma) including one in the 3.5 Ga Onverwacht Group (+1.5 ‰), and one in the 2.6 Ga Marra Mamba Iron

Formation (-0.1 ‰). They suggested that the atmospheric δ15N value has been constant since 3.5 Ga, assuming that N released at the lower temperatures during stepped heating was reflecting trapped N from the coeval atmosphere and that there was no contamination of present-day air into samples. Beaumont and Robert (1999) extracted kerogen from Precambrian cherts and BIFs

(32 samples ranging from 3.5 to 0.7 Ga in age), and reported their δ15N values. They range from -6.2 to +13 ‰ for the Early Archean samples (number of analysis: n = 33), -4.1 to +10.6 ‰ for the Late Archean samples (n = 19), +1.6 to +10.1 ‰ for the Early Proterozoic samples (n = 18), and +0.9 to +3.4 ‰ for the Late Proterozoic samples (n = 3). From these results, Beaumont and Robert (1999) showed that the secular variations of N isotopic compositions of kerogen in cherts and BIFs might reflect the redox evolution of the surface environments. A critical examination of the data by Beaumont and Robert (1999) is discussed later in this paper. Pinti and Hashizume (2001), commenting on Beaumont and Robert (1999), noted that the isotopically light N reflects the N fixation induced by chemosynthetic bacteria at deep-sea hydrothermal vents. Pinti et al. (2001) expanded the 246 previous work by Sano and Pilinger (1990) and suggested the activity of chemosynthetic bacteria lived near deep-sea hydrothermal vents.

4-3. Geologic settings and samples

Geological settings of and samples from the Swaziland, Transvaal, and Mt. Bruce Supergroups are described in chapter 2, more detail in the Appendix A and B, and briefly summarized in Table 4-1. All the samples used in this study were collected from drillcores to avoid the possible effects of modern weathering on geochemical signatures on the samples. In total, 65 samples were processes in this study, and 49 samples were used for the determination of N isotopic compositions. Focus of later discussion on N isotope is placed on the oldest shale samples of the 3.25 Ga Sheba Formation of the Fig Tree Group.

4-4. Analytical methods

Methods for pulverization, determination of the Corg and Mo contents, and the Corg isotopic compositions are already described in the preceding chapters 2 and 3. Detailed descriptions of the analytical methods utilized in this study were also given in the Appendix C. In this section, only brief explanations of the analytical methods for N contents and N isotopic compositions are presented. A portion of powdered samples was treated with 2 N HCl overnight (or until bubbling ceased) at room temperature to remove the carbonate fraction. To extract kerogen, another portion of the powdered sample was first treated with an organic solvent and then with HCl and HF in teflon bottles at 60 ~ 80 ˚C. An elemental analyzer (EA), either at Penn 247

State (CE Instruments NA2500) or at Tohoku University (Carlo Erba 1108 EA) in Japan, was used to measure N content of bulk samples, decarbonated samples, and kerogen samples. The detection limit was ~0.01 wt.%, and the reproducibilities were better than ±0.3 %. Nitrogen isotopic compositions of kerogen and decarbonated samples were measured with a CF-IRMS system at Tokyo Metropolitan University and/or with a conventional method at Penn State. Isotopic compositions of N are expressed with standard

δ (delta) notations in per mil (‰) relative to air N2:

15 3 δ N = (Rsample/Rstandard -1) • 10 ...... (4-6)

where R = 15N/14N. The reproducibility of the δ15N values was better than ± 0.2 ‰, respectively.

4-5. Results

4-5-1. Organic C contents

The measured Corg contents of the studied samples are presented in Table 4-2, and described in the chapter 2. See chapter 2 for more detailed description of the results of Corg contents. In this section only a brief summary is given.

The Corg contents of a majority of the studied samples fall within a range between

~0 and ~3 wt.%. Some samples have elevated Corg contents of up to ~7 wt.% (one sample

of the 2.69 Ga Lewin Shale Formation). The Corg contents of selected Phanerozoic shales and Archean–Paleoproterozoic samples are complied from literature and presented in Table 4-3. 248

The preserved Corg contents of black shales of this study are comparable to those of other Precambrian and Phanerozoic carbonaceous shales (Table 4-3). For example, the

average Corg contents of shales of the 3.25 Ga Sheba Formation (1.10 wt.%) are similar to those of Pennsylvanian bioturbated shales (1.23 wt.%; Ingall et al., 1993) and that of the 2.64 Ga Black Reef Formation (1.39 wt.%; Watanabe et al., 1997). Shales of the 2.60 Ga Wittenoom Dolomite Formation, Carawine Dolomite Formation, and the 2.69 Ga Lewin

Shale Formation have higher average Corg contents (2.06 ~ 2.79 wt.%) compared to the

other shales of this study; however, such high Corg content is less than the Corg content of average black shales (3.2 wt.%; Vine and Tourtelot, 1970).

The average Corg content of shales of the Timeball Hill Formation (0.41 wt.%) is slightly higher than that of the Bothaville Formation of the ~2.9 Ga Witwatersrand Supergroup and the K8, Booysens, and Formations of the ~2.7 Ga Ventersdorp Supergroup (0.14 ~ 0.32 wt.%; Watanabe et al., 1997) and the Silverton and Strubenkop Formations of the Transvaal Supergroup (0.12 ~ 0.26 wt.%; Watanabe et al.,

1997). These rather low Corg contents are close to the Corg content of the world average shale (Turekian and Wedepohl, 1961).

4-5-2. Nitrogen contents

4-5-2-1. Bulk nitrogen contents The total N contents of the black shales of this study are presented in Fig. 4-2-a through -h, Table 4-2, and Table 4-3. The N contents of the Archean–Paleoproterozoic black shales of this study are generally very low (< 0.1 wt.%; mostly < 0.05 wt.%) compared to those of the Phanerozoic age, such as Cretaceous Corg-rich shales (0.50 wt.% on average: Rau et al., 1987), the Jurassic Corg-rich shales (0.15 wt.% on average: Ingall et

al., 1993), the Devonian laminated (Corg-rich) shale (0.34 wt.% on average: Ingall et al., 249

1993), and Pennsylvanian laminated shale (0.53 wt.% on average: Ingall et al., 1993). The

black shales of the 3.25 Ga Sheba Formation have Ntotal contents ranging from 0.01 to 0.10 wt.% with a mean of 0.04 ± 0.03 wt.% (Table 4-2, 4-3). This range, and those of other shales, overlap with the typical N contents of recent marine sediments (0.01 ~ 0.10 wt.%; Waples and Sloan, 1980; Rau et al., 1987; Fig. 4-2-l) and the N contents for ~100 Neoarchean and Paleoproterozoic shales from South Africa (< 0.01 to 0.09 wt.% with a mean of 0.02 wt.%: Watanabe et al., 1997; Fig. 4-2-l and Table 4-3).

4-5-2-2. Nitrogen contents of organic matters and clays

We estimated the Nclay contents from the N contents of bulk rock and kerogen samples, assuming that the N contents of other constituents (e.g., carbonates and sulfides) are negligible (Table 4-2). This method to estimate the Nclay content is different from the one used by Compton et al. (1992) in which Nclay ("fixed-N" in Compton et al., 1992) was measured by the Kjedahl digestion and distillation technique (Keeney and Nelson, 1982). Using the data published by Watanabe et al. (1997) for Neoarchean and Paleoproterozoic shales (Silverton, Timeball Hill, Black Reef, and K8 Formations in South

Africa), we calculated their Norg and Nclay contents (Table 4-4). For comparison, Fig. 4-3 and Table 4-4 include published data of Norg and Nclay contents for some modern and ancient sediments; surface (~30 cm) sediments of the eastern subtropical Atlantic Ocean near Moroccan coast from Freudenthal et al. (2001), surface (~4.5 m) sediments of the Central Pacific Ocean from Müller (1977), relatively deep sediments of the Shikoku Basin from Schorno (1980), lagoon sediments of the Nakaumi Lagoon (coastal brackish lake) from Sampei and Matsumoto (2001), and the Miocene Monterey Formation in California from Compton et al. (1992). The Norg and Nclay contents of black shales of the Oak Tree

Formation and the Jeerinah Formation are also included in Table 4-3. The Norg/Ntotal ratios 250

of the samples are summarized in Table 4-4 and plotted with their ages in Fig. 4-3, together with the literature data mentioned above.

The black shales of the 3.25 Ga Sheba Formation have the Norg contents ranging from nearly zero (i.e., below the detection limit of 0.005 wt.%) to 0.02 wt.% with a mean

value of 0.01 wt.%; the Nclay contents range from 0.01 to 0.08 wt.% with a mean value of 0.03 wt.% (Table 4-2, 4-4).

The Norg/Ntotal ratio of the studied samples range from 0.01 to 0.26 with an average

of 0.13 ± 0.09 (Table 4-2, 4-4). Among samples from the same formations, the Norg/Ntotal ratios range from 0.07 to 0.19 in the Timeball Hill Formation and from 0.07 to 0.26 in the

Sheba Formation (Fig. 4-3). The average Norg/Ntotal ratios of the Sheba Formation (0.13)

and the Black Reef Formation (0.18) are close to the lower Norg/Ntotal ratio for the Phanerozoic sediments (~0.1).

4-5-3. Nitrogen to organic C ratios

The mean values of atomic Ntotal/Corg ratios for the Archean–Paleoproterozoic shales are variable: ranging from 0.01 to 0.09 (n = 12) for the Timeball Hill Formation, 0.01 (n = 3) for the Oak Tree Formation, 0.01 (n = 3) for the Wittenoom Dolomite Formation, 0.01 (n = 6) for the Carawine Dolomite Formation, 0.02 (n = 4) for the Marra Mamba Iron Formation, 0.02 (n = 8) for the Lewin Shale Formation, 0.05 (n = 13: siderite-poor shales are not counted; see discussion 2-6-1) for the Fig Tree Group (shales), and 0.05 (n = 12) for the Fig Tree Group (graywackes) (Table 4-2, 4-4).

For comparison, the relationship between Ntotal and Corg contents of some

Phanerozoic black shales are presented in Fig. 4-4 and their Ntotal/Corg ratios are listed in

Table 4-3. Cretaceous Corg-rich shales (Brumsack, 1980) and Devonian laminated (Corg- rich) shales (Ingall et al., 1993) have nearly identical low atomic Ntotal/Corg ratios (0.04).

On the other hand, the Cretaceous Corg-poor shales (Brumsack., 1980), Jurassic Corg-poor 251

shales (Ingall et al., 1993), Devonian bioturbated (Corg-poor) shales (Ingall et al., 1993), and Pennsylvanian bioturbated shales (Ingall et al., 1993) display variable Ntotal/Corg ratios (0.04, 0.07, 0.40, and 0.10, respectively). Except for the Cretaceous shales, all the

Phanerozoic shales in Table 4-3 display lower Ntotal/Corg ratios for the Corg-rich shales

(0.04 ~ 0.05) than those for the Corg-poor shales (0.07 ~ 0.40). The variable Ntotal/Corg ratios for the Archean and Paleoproterozoic shales nearly fall within the range for those of

the Phanerozoic Corg-poor shales.

The N/Corg ratios of shales of the 3.25 Ga Sheba Formation (0.05) are comparable to those of black shales of the Devonian (0.04), Pennsylvanian (0.05), Jurassic (0.04), and Cretaceous (0.04) (Table 4-3; Ingall et al., 1993; Rau et al., 1987). With respect to the slopes of the regression lines (see Fig. 4-4, although such lines are not shown), it is

noteworthy that (1) the average N/Corg ratios of the 3.25 Ga Sheba Formation (shales) (0.03) is almost identical to those of Cretaceous (0.04), Devonian (0.03), and

Pennsylvanian (0.04) Corg-rich shales, (2) the average N/Corg ratios of the 2.56 Ga Oak

Tree Formation (0.004) is close to those of the Jurassic Corg-rich shales (0.01), and (3) the

2.22 Ga Timeball Hill Formation (0.07) is close to the Jurassic Corg-poor shales (0.11) (Ingall et al., 1993).

4-5-4. Carbon isotopic compositions of organic C

13 Carbon isotopic compositions of OM (δ Corg values) of the Archean–Paleoproterozoic shales of this study are presented in Fig. 2-15, Table 4-2, and

13 Table 4-3. See chapter 2 for detailed description of the δ Corg results. For comparison, the 13 δ Corg values of Cretaceous black shales (Rau et al., 1987) and those of the Archean–Paleoproterozoic shales in South Africa used by Watanabe et al. (1997) are presented in Table 4-3. All the Archean and Paleoproterozoic samples of this study have the 252

13 13 δ Corg values (-48.9 to -25.1 ‰) nearly equal or less than the δ Corg values of the Cretaceous black shales (-27.6 to -24.4 ‰; Fig. 2-15, Table 4-2).

4-5-5. Nitrogen isotopic compositions

Sedimentary N is mostly composed of OM-bound N (Norg) and clay-bound N 15 15 (Nclay). We measured the δ N values of kerogen (δ Norg) and decarbonated bulk rock 15 15 (δ Ntotal). The δ Nclay values were computed using the following equation:

15 15 15 15 δ Nclay = δ Ntotal + Norg/(1 - Nclay) x (δ Ntotal - δ Norg)(4-7)

The results are presented in Fig. 4-5 and Table 4-2.

15 The δ Norg values are obtained from shales of the 2.56 Ga Oak Tree Formation (Fig. 4-5-j; +0.7 to 1.1 ‰ with a mean of +0.9 ± 0.2 ‰ (n = 3)), and shales (Fig. 4-5-p; - 0.8 to 0.7 ‰ with a mean of +0.2 ± 0.6 ‰ (n = 6)) and graywackes (Fig. 4-5-o; -0.2 to 2.7 ‰ with a mean of +0.9 ± 1.3 ‰ (n = 5)) of the 3.25 Ga Sheba Formation. For the remaining samples, sufficient amount of N2 gas for the stable isotope analysis was not obtained during combustion of the kerogen because of the very low Norg contents. 15 The δ Ntotal values of shales range from +5.2 to +6.9 ‰ with a mean of +5.9 ± 0.5 ‰ (n = 12) for the 2.22 Ga Timeball Hill Formation (Fig. 4-5-h), +1.6 to +2.6 ‰ with a mean of +2.1 (n = 2) for the 2.56 Ga Oak Tree Formation (Fig. 4-5-j), +5.7 to +6.1 ‰ with a mean of +5.8 ± 0.3 ‰ (n = 3) for the 2.60 Ga Wittenoom Dolomite Formation (Fig. 4-5-k), 0.0 to +0.4 ‰ with a mean of +0.2 (n = 2) ‰ for the 2.60 Ga Carawine Dolomite Formation (Fig. 4-5-l), +0.1 to +4.7 ‰ with a mean of +2.2 ± 2.0 (n = 4) ‰ for the >2.60 Ga Marra Mamba Iron Formation (Fig. 4-5-m), +0.0 to +6.7 ‰ with a mean of +3.4 ± 3.1 (n = 5) ‰ for the 2.69 Ga Lewin Shales Formation (Fig. 4-5-n), and +2.5 to +4.9 ‰ with a mean of +3.4 ± 0.7 (n = 15) ‰ for the 3.25 Ga Sheba Formation (Fig. 4-5-p). 253

15 The calculated δ Nclay values of shales range from +3.3 to +4.8 ‰ with a mean of +4.3 ± 0.6 (n = 6) ‰ for the 3.25 Ga Sheba Formation (Fig. 4-5-p). This indicates that the

15 15 δ Nclay values are about 4 ‰ higher than the δ Norg values in these samples. Such differences in the δ15N values are also seen from Fig. 4-5-c for the Cretaceous black shales between Corg-rich shale and Corg-poor shales. Here it is assumed that N in Corg-rich shale

is mostly Norg, while N in Corg-poor shale is mostly Nclay, as Rau et al. (1987) originally claims. Furthermore, modern marine sediments (Peters et al., 1978) also exhibit comparable

15 differences of about 4 ‰ in the δ N values between kerogen (Norg) and bulk samples (a 15 mixture of Norg and Nclay), although their absolute δ N values are different (shifted toward more positive values) from those of Cretaceous black shales. We speculate that there are

15 15 mechanisms to cause differences in the δ Norg and δ Nclay values regardless of geologic age, which will be discussed in detail later. For the comparison of the Archean–Paleoproterozoic samples with the Phanerozoic

and modern samples, the δ15N values of Cretaceous black shales (Rau et al., 1987; Ohkouchi et al., 1997) and those of modern marine sediments are presented in Fig. 4-5-a

15 through -c. Rau et al. (1987) reported the δ N values for Corg-rich Cretaceous shales

ranging from -1.2 to -2.7 ‰ with a mean of -1.9 ± 0.6 (1σ) ‰ (n = 6) and those for Corg- poor shales ranging from +1.6 to +3.3 ‰ with a mean of +2.3 ± 0.7 ‰ (n = 5: Fig. 4-5-c).

Ohkouchi et al. (1997) reported δ15N values of the black shales for the OAE2 (Oceanic Anoxic Event 2; Schlanger et al., 1987) during the Cenomanian-Turonian ranging from -0.6 to -2.0 ‰ with a mean of -1.3 ± 0.6 ‰ (n = 7) and those for the adjacent rocks ranging from -1.4 to +9.0 ‰ with a mean of +2.8 ± 2.7 ‰ (n = 24: Fig. 4-5-c). Peters et al. (1978)

report the δ15N values of kerogen extracted from marine surface sediments of the modern oceans ranging from +0.3 to +9.9 ‰ with a mean value of +4.7 ± 2.4 ‰ (n = 25) (Fig. 4-

15 5-a) and the δ Ntotal values of decarbonated modern marine sediments ranging from +2.5 to +9.4 ‰ with a mean value of +6.2 ± 1.7 ‰ (n = 47) (Fig. 4-5-b). These studies on the 254

15 15 younger samples indicate that the younger samples have a similar δ Norg - δ Nclay relationship to the Archean samples.

15 The δ Norg values of kerogen extracted from the Archean–Paleoproterozoic shales (studied by Hayes et al., 1983) and cherts / BIFs (studied by Beaumont and Robert, 1999) are also presented in Fig. 4-5. They include shales of the 1.4 Ga Ropert Group (Fig. 4-5-e), 1.6 Ga Bungle Bungle Dolomite (Fig. 4-5-f), 1.8 Ga (Fig. 4-5-g), 2.5 Ga Hamersley Group (Fig. 4-5-h), and 3.4 Ga Gorge Creek Group (Fig. 4-5-q) (Hayes et al., 1983) and cherts / BIFs of Early Proterozoic, Late Archean, and Early Archean (Fig. 2-5-r).

15 15 The δ Norg and δ Ntotal values of all Precambrian samples (Fig. 4-5) are within the range of those of Cretaceous shales and modern sediments.

4-5-6. Molybdenum contents The results of Mo analyses for the Archean–Paleoproterozoic samples of this study are already given in the chapter 3. In this chapter, Mo data of the 3.25 Ga Sheba Formation is used. The range of Mo contents of the samples from the 3.25 Ga Sheba Formation is from 0.3 to 6.6 ppm with a mean of 1.6 ± 1.5 ppm (Table 4-2). This value is about the same as the Mo content of the world average shales (1.3 ppm; Wedepohl, 1991) but lower than that of world average black shales (10 ppm; Vine and Tourtelot, 1970). However, the average Mo content of 1.6 ppm should not be directly compared to that of other samples because of the significant dilution by carbonate minerals in the shales of the Sheba

Formation (Ccarb content sometimes reaches to ~6 wt.%, corresponding to the CO2 content of ~22 wt.% or the CaCO3 content of 50 wt.%; see Table 4-2 or Fig. 2-13). Detailed discussion of Mo is given in the chapter 3 of this thesis. 255

4-6. Discussion

In our attempt to constrain the evolution of the microbial biosphere and the

atmosphere based on the N and Corg geochemistry of the Archean–Paleoproterozoic black shales, we suggest that microbially mediated redox-cycling of N involving biological N2- + fixation, nitrification, and denitrification and / or microbially mediated cycling of NH4 in the Archean oceanic environments. In this section, we discuss (1) the factors controlling N

and Corg contents of shales, (2) the microbial activity inferred from Corg and N isotopes, and (3) the availability of Mo to microorganisms. The effects of post-depositional overprints on the contents and isotopes of N and Corg are evaluated. Then we provide the implications for the Archean biosphere and oceans. Finally, we critically evaluate the validity of conclusions given by Beaumont and Robert (1999) for the Archean N biogeochemical cycles.

4-6-1. Factors controlling the N and organic C contents of shales

The accumulation of N and Corg in shales is controlled by many factors including the type of microorganisms. The contents and isotopic ratios of N and Corg in shales are modified by post-depositional processes such as diagenesis and metamorphism. In this section, various factors and effects are evaluated, and the original N and Corg contents of shales at the time of deposition are estimated to better understand the processes responsible for the formation of sedimentary OM.

4-6-1-1. Type of organisms: Terrestrial vs. Marine?

Generally, N/Corg ratios of terrigenous OM are low (<0.1) and those of marine OM are high (>0.1) (e.g., Premuzic et al., 1982; Emerson and Hedges, 1988; Meyers, 1994). Varying mixing ratios of these two types of OM influence the composition of OM in 256

Phanerozoic and modern sediments. However, during the Precambrian there were no vascular land plants, except for locally developed biomats (e.g., Rye and Holland, 2000; Watanabe et al., 2000). Therefore, this factor is considered not to be relevant to the Archean–Paleoproterozoic shales of this study.

4-6-1-2. Change in the N and organic C contents during diagenesis and metamorphism The N content of marine sediments generally decreases with increasing depth of burial (Müller, 1977; Waples and Sloan, 1980; Compton et al., 1992), but the magnitude of decrease in N content is generally less than that in Corg. For example, in a 500 m depth

profile of pelagic sediments at a north-western Pacific Ocean site, the Corg content decreases about 4-fold from the surface to the bottom (from ~0.4 wt.% to ~0.1 wt.%) while the N content decrease is 2-fold (from ~0.04 wt.% to ~0.02 wt.%) (Waples and Sloan, 1980). This is probably because of the different pathways for the sedimentary fixation of N and

Corg during diagenesis. During diagenesis, sedimentary OM is partially or fully decomposed by microbial (and thermal) degradation (c.f., Fig. 4-1). Nitrogen is released in

+ pore fluids in the form of NH4 , which is either utilized (recycled) by heterotrophic microorganisms (assimilation and nitrification), sorbed on mineral surfaces, or fixed in clays such as illite and smectite (occupying K+-site because of the identical charge and the

+ + similar ionic radius of NH4 = 1.66 Å and K = 1.59 Å, depending on the structural coordination), or escapes from the system by diffusion. On the other hand, Corg released in pore fluid is not fixed in clay. Therefore, N is more likely to be preserved in sediments than

Corg. During regional and contact metamorphism, N content decreases with the increasing grade of metamorphism (Haendel et al., 1986; Bebout and Fogel, 1992; Mingram and Bräuer, 2001). However, some N remains even at high-grade metamorphism in the host minerals, such as biotite, muscovite, and feldspar (Itihara and Honma, 1979; Honma and 257

Itihara, 1981). The remaining N in metamorphic rocks has been used as a biomarker for pre-metamorphic biological activity (e.g., Itihara and Suwa, 1985; Itihara et al., 1986; Haendel et al., 1986; Bebout and Fogel, 1992; Boyd et al., 1993; Honma, 1996; Hall, 1999;

+ Mingram and Bräuer, 2001). The NH4 decrease may reflect a loss of N by thermal decomposition, dehydration, or cation exchange. Nitrogen is released from silicates as volatiles, such as N2 and NH3 (i.e., metamorphic volatilization; Valley, 1986), depending on the redox state in the rocks (Haendel et al., 1986). From the typical depth profile of N content in marine sediments (e.g., Müller, 1977;

Waples and Sloan, 1980), the original total N (Ntotal) content at the time of deposition may

be approximated to be twice the remaining (observed) Ntotal content. During metamorphism

of up to the greenschist facies, the Ntotal content appears to decrease by up to about three

fold (Bebout and Fogel, 1992). Therefore, it is not easy to estimate the original Norg and

Nclay contents at the time of deposition from the observed Norg and Nclay contents in metamorphosed sedimentary rocks. However, as a first-order approximation, we may

attempt to estimate the original Ntotal contents from the observed Ntotal contents of mildly

metamorphosed rocks (up to greenschist facies) as follows. The original Ntotal contents probably lie between two times and six times (i.e., x 2 (diagenesis) ~ x 6 (diagenesis + greenschist metamorphism)) of the measured Ntotal values. This results in the original N values to be between > 0.02 and < 0.60 wt.% corresponding to the measured values of 0.01 ~ 0.10 wt.% for the shales of the 3.25 Ga Sheba Formation. This range overlaps with the

typical Ntotal content of modern marine sediments, Ntotal = 0.01 ~ 0.10 wt.% (Fig. 4-2;

Waples and Sloan, 1980; Rau et al., 1987), and brackish lake Corg-rich sediments, Ntotal =

0.04 ~ 0.25 wt.% and Corg = 0.0 ~ 3.0 wt.% (Sampei and Matsumoto, 2001). The similarity in N contents between Archean-Paleoproterozoic shales and modern sediments suggests that the N in the Archean–Paleoproterozoic shales are the products of similar biological activity. 258

4-6-1-3. Total N, organic-bound N, clay-bound N contents of the Archean shales

The contrasting Ntotal contents between Precambrian and Phanerozoic shales may be attributed to either (1) the effect of regional metamorphism which is generally higher in

Precambrian shales, or (2) the large difference in Corg contents that often covaries with

Ntotal contents. As shown in Table 4-2 and Fig. 4-4, among the black shales of Cretaceous (Rau et al., 1987), Devonian (Ingall et al., 1993), and Pennsylvanian (Ingall et al., 1993), N contents of Corg-poor shales are almost always lower than those of Corg-rich shales.

Contents of N in shales appear to be dependent on their Corg contents.

The intra- and inter-formations variabilities in the Ntotal content may suggest: (1) the

variable degrees of the initial Ntotal content of the sedimentary OM (i.e., variable mode of primary production and preservation of OM) and (2) the variable degrees of post-

depositional alteration leading to transformation of Norg into Nclay or removal of N from the rocks.

Our results show that there are differences in the Norg/Ntotal ratios between Phanerozoic / modern sediments (0.07 ~ 1.00) and Precambrian sediments (0.01 ~ 0.26) (Fig. 4-3 and Table 4-4). Such difference can be attributed to the effect of post-depositional effects such as metamorphism. As explained in the previous section, N released during diagenetic degradation of OM is partially fixed into the lattice structure of clay minerals. During metamorphism, N fixed into clays are preferentially preserved because of their stability even in high temperature-pressure conditions compared to N in OM (kerogen).

Therefore, the Norg/Ntotal ratios of sediments would decrease with increasing metamorphism. Indeed, Phanerozoic metasediments that suffered a higher grade of

metamorphism exhibit lower Norg/Ntotal ratios (0.07; Chugoku Belt; see Fig. 4-3) comparable to those for the Precambrian (Fig. 4-3 and Table 4-4, Itihara and Itihara, 1979).

A common mechanism to partition sedimentary N into Norg and Nclay seems to have been 259

operating since the Archean, emphasizing the importance of the active role of microorganisms in the degradation of sedimentary OM and the inorganic fixation of N in clay minerals since the Archean.

4-6-1-4. Relationship between the N and organic C contents

The similarities in the N/Corg ratios (and/or slopes of the regression lines) among the Archean black shales and the Phanerozoic black shales suggest that (1) the types of microorganisms that contributed to the sedimentary OM of the black shales, (2) their biochemistry for Corg and N, and (3) the diagenetic history of OM degradation may have been similar between the Archean and Phanerozoic OM. In Fig. 4-4, while the majority of

the samples show positive correlation between N and Corg with their slopes of the regression lines similar to those for the Phanerozoic shales, some samples whose N contents are below detection limit (plotted on the x-axis in Fig. 4-4) and 4 samples of the 3.25 Ga Sheba Formation (shales) with their N content between 0.08 and 0.10 do not display such positive correlation. The former (N = 0 wt.% samples) is interpreted to have resulted from extensive OM degradation coupled with the inefficient N fixation into clays

(i.e., no Nclay preserved), and the latter (4 samples in the Sheba Fm.) is interpreted to have resulted from an extensive OM degradation coupled with a limited amount of clays capable of N fixation (i.e., excess N not fixed into clays was lost). The latter interpretation is supported by the observation that the N content of 4 samples from the Sheba Formation

(0.08 - 0.10 wt.%) are close to the minimum N content of the Corg-rich shales of the Devonian (0.090 wt.%), Pennsylvanian (0.075 wt.%), and Jurassic (0.095 wt.%) (Fig. 4-4),

and that of the Corg-poor shales of the Devonian (0.08 ~ 0.12 wt.%; Ingall et al., 1993) and Cretaceous (0.07 wt.%; Brumsack, 1980). 260

4-6-2. Microbial activity inferred from the C isotopes of organic C

The discussion of the microbial activity inferred from the Corg isotopes of the Archean–Paleoproterozoic shales of this study are already given in the previous chapter 2

(section 2-5-3-2). In the present section, we focus on the environmental factors on the Corg isotopic compositions by considering the two contrasting environments, whether or not there was an influence of syndepositional hydrothermal activity in the depositional environments.

13 According to Beaumont and Robert (1999), the δ Corg values of kerogen extracted from a chert of the 3.5 Ga Onverwacht Group range from -25 to -35 ‰ with a mean of -30

‰. Robert (1988) previously reported a δ13C value of -33 ‰ to best represent the original δ13C value for the Barberton cherts. In the Onverwacht Group, there are remarkable abiogenic mudpool structures formed by fluid emission in the mineralized hydrothermal vent system (de Wit et al., 1982; Paris et al., 1985; de Ronde et al., 1993). It is noted by Pinti and Hashizume (2001) that the ~3.5 Ga old chert samples used by Beaumont and Robert (1999) are likely to have deposited in environments influenced by hydrothermal

13 activity. Therefore, by using the δ Corg values of cherts and shales, it is possible to compare the types of microorganisms that lived in hydrothermally-influenced sedimentary environments (cherts) and that lived in rather 'normal' sedimentary environments (shales).

13 The average δ Corg value for the 3.5 Ga chert (-30 or -33 ‰: Beaumont and Robert, 1999 and Robert, 1988, respectively) is isotopically lighter by 1~4 ‰ than the

13 δ Corg value for the shales of the 3.25 Ga Sheba Formation of this study. Such a difference can be attributed to a microbial community having different biochemical pathways to cause greater isotopic discrimination of 12C compared to surface-living normal microorganisms such as cyanobacteria. Indeed, Brooks et al. (1987) and Kennicutt et al.

(1985) reported a similarly depleted δ13C value for organic compounds associated with modern hydrothermal vents and seeps (e.g., -36 ‰ for seep organisms: n = 53; -33 ‰ for 261

vent organisms: n = 96). The above suggestions imply a diversity (in microbial niche or biochemical pathways) of a Mesoarchean microbial biosphere.

4-6-3. Microbial activity inferred from the N isotopes

4-6-3-1. Effects of metamorphism and hydrothermal alteration From the study of metasedimentary rocks in Germany (Variscides), Haendel et al.

(1986) showed that an increase in the δ15N values is accompanied by a decrease in the

Ntotal content with increasing metamorphic grade. Bebout and Fogel (1992) also showed a similar result from the study of subduction-related metasedimentary rocks in California (Catalina schist). For example, rocks of a metamorphic grade lower than greenschist facies

15 appear to retain their original Ntotal content (0.01 ~ 0.1 wt.%) and δ Ntotal values (1 ~ 3

‰), while those higher than greenschist facies have consistently lower Ntotal content (< 0.03 wt.%) and elevated δ15N values (3.5 ~ 6 ‰) (Fig. 4-6; Bebout and Fogel, 1992). From the study of low-grade metamorphic rocks of carbonaceous sediments, Ader et al. (1998)

15 showed that the Norg content decreases with increasing maturity of the OM, while its δ N values essentially remain unchanged. Therefore the metamorphic effects on the sedimentary

N content can be summarized as follows: only a small amount of Ntotal (about 5 ~ 7 %; Haendel et al., 1986) is mobilized during low-grade metamorphism (greenschist facies), a substantial amount of Ntotal is lost during the transition to medium-grade metamorphism

(lower facies), and significant amount of Ntotal (up to about 90 % of their original sedimentary N; Haendel et al., 1986) is lost during high-grade metamorphism (upper amphibolite - granulite facies). For the metamorphic effects on the sedimentary N

isotope, low grade metamorphism has little effect in modifying the δ15N values while high grade metamorphism has significant effects in modifying them (Haendel et al., 1986; Juster 262

et al., 1987; Bebout and Fogel, 1992; Williams et al., 1995; Ader et al., 1998; Mingram and Bräuer, 2001). Metamorphism and/or hydrothermal alteration are not likely to have made the

15 15 difference between the δ Norg and δ Nclay values for the black shales of the 3.25 Ga Sheba Formation. Their N content up to 0.1 wt.% (Bebout and Fogel, 1992) and low-grade metamorphism (Anhaeusser, 1986), as also evidenced by well-preserved (Fig. 2-18), suggest a negligible metamorphic modification on the δ15N values of shales of the 3.25 Ga Sheba Formation (Fig. 4-4). Although localized hydrothermal Au mineralization in the Barberton Greenstone Belt (Anhaeusser, 1986) has been recognized, the studied samples include neither Au-mineralized parts, vein-rich parts, fractured parts, nor pyrite bands of probable hydrothermal in origin. Furthermore, it is not currently known, theoretically or experimentally, if metamorphism lower than the greenschist facies and/or

15 15 hydrothermal alteration can selectively modify only δ Norg or δ Nclay. Therefore, we 15 15 believe that the observed differences between the δ Norg and δ Nclay values, as well as their absolute values, mostly reflect the original values at the time of deposition.

4-6-3-2. N2-fixation There is little N isotopic fractionation between the source (the present-day

15 atmospheric N2: δ N = 0 ‰ by definition) and the product (Norg) during biological N2- fixation (Hoering and Ford, 1960; Minagawa and Wada, 1986; Wada and Hattori, 1991).

Cyanobacterial N2-fixation has been suggested to have contributed to the formation of the OM in Cretaceous OAE black shales, based on the δ15N values of OM in shales (near 0 ‰) and the biomarker evidence suggestive of prokaryotes (Rau et al., 1987; Ohkouchi et

15 al., 1997). Our results of the δ Norg values of shales for the 3.25 Ga Sheba Formation (+0.2 ± 0.6 ‰; Fig. 4-5-p and Table 4-2) suggest that (1) direct biological fixation of

15 atmospheric N2 could have occurred as early as 3.25 Ga ago (assuming that the δ N 263

values of N2 in the Archean atmospheric was the same as today, 0 ‰) and that (2) the OM

from the N2-fixers significantly contributed to the formation of sedimentary OM (kerogen).

This suggestion is consistent with the previous assertion that biological N2-fixation had already emerged in the Mesoarchean (Beaumont and Robert, 1999). Analyses of the gene

sequences encoding the catalytic subunits for the nitrogenase (the enzyme for N2-fixation) have shown that it was highly conserved in cyanobacteria and other eubacteria. This strongly suggests that nitrogenase has an ancient, common ancestral origin (Zehr et al., 1995).

15 As previously mentioned, the δ Norg values of modern marine sediments (+4.7 ± 2.4 ‰, Fig. 4-5-a) are shifted toward more positive directions compared to those of the Cretaceous and Archean black shales (near 0 ‰, Fig. 4-5-c through -p). This is probably because the OM in the modern sediments (Fig. 4-5-a) studied by Peters et al. (1978) had a

significantly lower contribution of N2-fixers compared to the OM in the Cretaceous and 15 15 Archean black shales of this study. The similarity in the δ Norg and δ Nclay (or 15 δ Ntotal) relationship among the modern, Cretaceous and Archean sediments suggests not only that biological N2-fixation was important both in Cretaceous and Archean, but also that 15 there might have been a common mechanism to have created the difference between δ Norg 15 and δ Nclay values since the Archean (see the following two sections).

4-6-3-3. Redox cycling of N We suggest that the redox-cycling of N involving nitrification and denitrification during syndepositional and / or early diagenetic processes may have been responsible for

15 15 the important differences between the δ Norg and δ Nclay values, not only in the modern and Cretaceous, but also in the Archean (Fig. 4-1 and 4-5). In an environment without nitrification and denitrification (Fig. 4-1-a; i.e., entirely

+ - anoxic environments where there is no O2 to oxidize NH4 (nitrification) and no NO3 to 264

+ be reduced to N2 (denitrification)), NH4 would be the dominant N species in the oceans 15 15 (Beaumont and Robert, 1999). The δ Norg values would reflect the δ N values of + 15 15 15 dissolved NH4 (≈ δ N values of N2 in the atmosphere) and the δ Nclay ≈ δ Norg values (Fig. 4-1-a). There would result in no significant redox cycling of N. On the other hand, in an environment with nitrification and denitrification, like today,

- NO3 would be the dominant N species in the oceans (Fig. 4-1-b and -c). Significant isotope fractionation occurs during denitrification (∆ ≈ -6 ‰: Wada and Hattori, 1991) in 14 which N is preferentially returned to the atmosphere as N2 and/or N2O and the remaining - 15 NO3 pool becomes enriched in N. In modern oceans, this process occurs in a mid-depth oxygen-minimum zone of a water column, in an anoxic bottom water in a stratified ocean /

15 - lake, and in sediments near the sediment-water interface. This N-enriched NO3 is utilized + 15 again (recycled) as a nutrient for organisms. The NH4 released by the decay of N- 15 - enriched organisms which utilized N-enriched NO3 as a result of denitrification becomes 15 15 + progressively enriched in N. In a sediment column, such N-enriched NH4 is partially 15 15 fixed into clays during early diagenesis, and the δ Nclay values inherit the N-enriched signatures. The sedimentary OM that escaped microbial degradation would keep its original

15 15 δ N values formed in the surface water. In this way, the difference between the δ Norg 15 and δ Nclay values is created. 15 When N2-fixers are not active, then the differences between the δ Norg and 15 δ Nclay values would be small (Peters et al., 1978) [the case (c) is dominant compared to the case (b) in Fig. 4-1]. When microbial N2-fixation becomes active in the surface ocean and the preservation of OM created by N2-fixers is increased (due to enhanced productivity 15 and consumption of dissolved O2 by degradation), the δ N values of sedimentary OM approach 0 ‰ or even negative values because of a greater contribution of OM from N2- fixers [the case (b) is dominant compared to the case (c) in Fig. 4-1]. Thus, the differences 265

15 15 between the δ Norg and δ Nclay values become increased (Rau et al., 1987; Ohkouchi et al., 1997).

15 In the Santa Barbara Basin, there is a difference between pore water δ NNH4 and 15 Corg-rich sediment δ Ntotal values (+10.2 ‰ and +6.8 ‰, respectively, ∆ = 3.4 ‰) (pers. comm. Michael Arthur). These values appear to have been created by non N2-fixing

organisms. When N2-fixers become a dominant contributor of the sedimentary OM in the 15 15 Santa Barbara Basin, it is expected that the pore water δ NNH4 and sediment δ Ntotal 15 values decrease; the δ Norg values would approach ~ 0 ‰. In soils, similar a difference of 15 + ~ 3.5 ‰ has been observed in the δ N values between OM and hydrolyzable NH4 15 15 (δ Norg < δ NNH4) partly because of denitrification (pers. comm. Michael Arthur). Such 15 + difference of ~3.5 ‰ in the δ N values between OM (or bulk) and NH4 is similar to that 15 15 between δ Norg and δ Nclay values observed in this study. Schidlowski et al. (1983) noted that where cyanobacteria were predominant in the

Archean, the dominant process of cellular N incorporation could conceivably have been N2- - fixation. This suggestion is probably based on the assumed scarcity of NO3 in the anoxic Archean environments. Schidlowski et al. (1983) also noted that if denitrifying

- microorganisms were recycling oxidized N species such as NO3 , then the occurrence of 15N-enriched OM in Precambrian kerogen could be regarded as indicative of the existence of autotrophic N bacteria. Our finding suggests such redox-cycling of N probably involved

15 15 nitrification-denitrification based not on the N-enriched absolute δ Norg values, but, 15 15 rather based on the difference between δ Norg and δ Nclay values.

4-6-3-4. Non-redox cycling of ammonia Another mechanism that may have created the observed N isotopic signatures for

+ organic-bound N and clay-bound N is the bacterial cycling of NH4 in sediments. During + bacterial uptake of pore water NH4 which is produced by microbial OM decay, there is 266

+ significant isotopic fractionations (~ -14 ‰) depending on pore water NH4 concentration + (enhanced fractionation with elevated NH4 concentration; pers. comm. Michael Arthur). + 15 The remaining NH4 pool that is not utilized by bacteria becomes enriched in N. Such 15 + N-enriched NH4 may be sorbed on and fixed into clay minerals (e.g., illite, smectite). 15 15 15 Therefore, the difference in the δ N values between OM and clays (δ Norg < δ Norg) can be created.

4-6-4. N2-fixation as a source of atmospheric O2 ?

As Boyd (2001a, b) notes, N2-fixation acts like a biogeochemical pump in which N2

is extracted from the atmosphere and fixed into the crust. All of the NH3 in the crust has been ultimately derived from the atmosphere through biological activity. In the net reaction of N2-fixation, 1 mole of N2 and 3 moles of water produce 1.5 moles of O2 as shown in the

overall reaction (Eq. 4-1). Although the majority of NH3 is oxidized by the reverse reaction, some fraction is buried in the sediments and eventually subducted into the crust. Nitrogen burial to the net production of atmospheric O2 (Eq. 4-1), just like Corg burial (Eq. 4- 8).

N2 + 3 H2O ---> 2 NH3 + 1.5 O2...... (4-1) ...... CO2 + H2O ---> CH2O + O2 (4-8)

The quantitative importance of N burial as a source of atmospheric O2 is uncertain; however, Boyd (2001a, b) roughly estimated that at present, 5 % of the O2 in the Earth's atmosphere may have resulted from N2-fixation. We also roughly estimate the importance of the net production of O2 induced from the sedimentary burial of biologically fixed N.

Assuming that the N and Corg in the shales of the 3.25 Ga Sheba Formation were first-cycle

N and Corg (no inheritance of prefixed N and Corg in their source rocks introduced by 267

continental weathering), the sedimentary atomic N/Corg ratio of ~0.05 for the 3.25 Ga shales leads to an estimation that O2 from N burial is only a factor of 0.05 [N/Corg] x 1.5

[O2/N2] ÷ 2 [N2/N] = ~0.04 [O2/Corg], or 4 % of O2 from Corg burial. This value (4 %) is close to the value (5 %) estimated by Boyd (2001a, b) for the modern environments.

Although it appears that the net O2 production from sedimentary burial of newly fixed N is quantitatively insignificant (only 4 % of Corg burial) compared to that of newly fixed Corg,

the role of biological N2-fixation as an O2-source in the Archean should be further considered because of its biochemical antiquity.

4-6-5. Molybdenum availability to microorganisms

Availability of Mo and Fe to N2-fixers is essential to operate their enzyme

complexes nitrogenase facilitating N2-fixation. The complex is composed of two enzymes: dinitrogenase reductase (Fe-protein) and dinitrogenase. Dinitrogenase is also known as the Mo-Fe protein (Postgate, 1992). Not only N2-fixers, but also denitrifying bacteria require Mo and Fe in their enzyme nitrate reductase, a polypeptide containing Mo and Fe-S clusters (Madigan et al., 2000). Both Mo and Fe should be in the bio-available forms, i.e.,

dissolved forms rather than solid, for microorganisms that operate N2-fixation and denitrification. Most likely, Fe seems to have been available to the Archean microorganisms as a bioessential element, whether or not the Archean environment is oxic or anoxic. This is because Fe has been sufficiently supplied to the oceans via submarine hydrothermal venting and by rivers carrying dissolved and/or particulate Fe, as evidenced by occurrence of BIFs and Fe-rich shales / carbonates in the Archean. Solubility of Fe in the oxic oceans is generally kept extremely low (~2 ppb in the modern oceans), but increases in anoxic water bodies as the solubility of Fe2+ increases. 268

The supply of Mo from the continents to the oceanic environments would not have been significantly different under a globally anoxic environment or a globally oxic

-6 environment if pO2 is higher than 10 atm (see chapter 3, section 3-5-2). Mo-bearing (6+) 2- minerals are well oxidized into solution to form molybdate (Mo O4 ) and accumulate in the oceans. In local anoxic basins overlain by oxic surface seawater, dissolved Mo is reduced and fixed in the sediments by OM and/or S (e.g., Holland, 1984, 1994; see also chapter 3 of this thesis and references therein). In order to examine Mo availability to microorganisms, we investigated the Mo

concentration and its relationship to the Corg content of the shales of the 3.25 Ga Sheba

Formation, together with some younger Corg-rich sediments and sedimentary rocks.

Relationships between the Mo and Corg contents for the Archean black shales, together with Devonian, Cretaceous OAE, and modern Black Sea sediments for comparison, are shown in Fig. 4-7.

The Mo and Corg contents all display positive correlations independent of the ages of the samples. The correlation coefficients (R) are 0.63 for the 3.25 Ga shales, 0.66 for the Devonian shales, 1.00 for the Cretaceous OAE shales, and 0.70 for the modern Black Sea sediments. Their consistent positive correlations, although their slopes of regression lines are different (Fig. 4-7), suggest that Mo, once in a dissolved form, was fixed by OM and / or S (Corg and S often show positive correlation; see chapter 2 of this thesis), and therefore suggest availability of Mo for microorganisms even in the Archean. However, previous studies (Holland, 1994) suggested that such a correlation was not present in the Archean due to an inferred globally anoxic condition where the Mo weathering flux into the oceans might have been very low. Observation of the positive correlation between Mo and Corg in the 3.25 Ga shales suggest an important role of microorganisms in the Archean geochemical cycle of Mo. Microorganisms could have utilized bioavailable form of Mo in the Archean oceans for living, possibly for operation of 269

the enzymes nitrogenase for N2-fixation and nitrate-reductase for denitrification. Demonstrated availability of Mo (and Fe) to the Archean microorganisms can not be used as evidence for the operation of the above enzymes, but Mo and Fe were, at any rate, available if such mechanisms functioned. Detailed discussion of the relationship between

Corg and redox-sensitive metals such as Mo and U in the Archean and Paleoproterozoic black shales are presented in chapters 3 and 5 of this thesis.

4-6-6. Implications for the Archean atmosphere, oceans, and biosphere

4-6-6-1. Environmental prerequisites for the early evolution of biological N2-fixation + In the early oceans, NH4 was probably generated from atmospheric N2 via + lightning (Mancinelli and McKay, 1988) and contributed to the oceanic NH4 pool. In such a world, microorganisms did not have to develop an energy-demanding mechanism to fix

N2 from the atmosphere (to break the triple bond of N2 requires more energy compared to + other biochemical processes). A decrease in the oceanic NH4 concentration due to a pO2 + increase that enhances oxidation of NH4 (Towe, 1994) and/or pCO2 decrease (Kasting,

1993) that discourages the production of NO from CO2 and N2 via lightning (Navarro-

González et al., 2001) could have led to the evolution of nitrogenase for N2 fixation. Since nitrogenase is phylogenetically deep-rooted, unique to prokaryotes, and found among

Archaebacteria as well as cyanobacteria (Postgate, 1982), the evolution of N2-fixation could have been very early in the Archean.

4-6-6-2. Source of nitrate and the evolution of the atmosphere

+ - Microbial nitrification of NH4 to NO3 requires O2 (Eq. 4-2 and 4-3). However, - NO3 in the Archean could have been made through disproportionation of photochemically produced NO (Mancinelli and McKay, 1988). Its yield is low according to Kasting (1990). 270

We suggest microbial nitrification using dissolved O2 in the oceans could have been important in the redox-cycling of N in the Archean for the following reasons: (1) the Mo-

Corg positive correlations in the Archean shales suggests that the microorganisms utilized dissolved Mo in the oceans (Fig. 4-7); (2) the discovery of Fe-oxide-rich hydrothermal discharge vents (ironstone pod) in the Fig Tree Group suggests an oxygenated deep water (de Ronde et al., 1994, 1997); and (3) according to the recent simulation of oxygen geochemical cycles (Lasaga and Ohmoto, 2002a), O2 produced by cyanobacteria could have already transformed the world oxic in less than 20 million years since its first appearance, possibly earlier than 3.8 Ga ago. The oxygenated deep water in the Archean, which requires an oxygenated atmosphere (e.g., Sarmiento, 1992; Lasaga and Ohmoto, 2002a), contrasts to the previous suggestion of entirely sulfidic deep oceans in a globally anoxic world (Canfield, 1998; Canfield et al., 2000). As previously noted, nitrification normally requires molecular O2 for its reactions. Hence, nitrification must have evolved after the formation of

free O2 in the oceans by oxygenic photosynthesizers (Falkowski, 1997). This is consistent - with our suggestion of redox-cycling of N in the Archean. The NO3 converted biologically + from NH4 would have contributed to photosynthesizers for their nutrient requirement and also to a diverse group of heterotrophs, including anaerobic bacteria such as denitrifiers, as an electron acceptor.

4-6-6-3. Oceanic environments leading to biological N2-fixation Even if the Archean atmosphere was oxidized to the present level, some parts of the oceans could very likely have been anoxic like today's environments (e.g., redox-stratified oceans in the Black Sea, Cariaco Trench, Saanich Inlet, Framvaren Fjord, etc.). With this in

mind, we propose a possible scenario leading to localized N2-fixation and accumulation of black shales in the Archean oceans (Fig. 4-8-a through -d). 271

Reduced ocean circulation (whether local or global), possibly due to climatic and/or tectonic reasons, leads to (density and/or thermal) stratification and increased denitrification

in the deep O2-depleted zones in the water columns and/or sediments (Fig. 4-8-b). The - 3- oceans become depleted in NO3 due to loss via denitrification. Phosphate (PO4 ) is

regenerated into the stratified O2-depleted bottom water from sediments under an anoxic condition (Fig. 4-8-c; Van Cappellen and Ingall, 1996). When such stratified oceans

3- - (periodically?) overturn, deep-water PO4 is upwelled into the NO3 -depleted surface

waters, which leads cyanobacteria to flourish and to fix N2 (Fig. 4-8-d). This cyanobacterial bloom may be similar to that in the modern-day dinoflagellate-dominated red tide in which cyanobacteria Trichodesmium are active (Minagawa and Wada, 1986). Increased primary

productivity in the surface oceans leads to consumption of dissolved O2 in the deep oceans due to degradation of sinking OM (Fig. 4-8-b, -d). The bottom water becomes depleted in

O2 and leads to an enhanced preservation of OM in sediments (Fig. 4-8-b; Pedersen and Calvert, 1990). Methanogens, methanotrophs, and sulfate-reducing bacteria (SRB) become active with abundant OM as an energy source and with a sufficient concentration of dissolved sulfate. Molybdenum originally fixed in OM is released into pore waters and/or bottom waters upon bacterial degradation of OM, and is again fixed in authigenic pyrite formed during bacterial sulfate reduction. The activity of SRB in the black shales of the 3.25 Ga Sheba Formation has been supported by S isotope studies of disseminated pyrite (chapter 3 and 4 of this thesis).

4-6-6-4. Diverse microbial biosphere in the Archean This study and many other previous studies have shown that (cyanobacterial) oxygenic photosynthetic activity in the Archean surface (photic) oceans was probably as vigorous as today. Previous studies have shown that sulfate reduction by bacteria was in operation in the ~3.25 Ga deep oceans and/or sediments (Ohmoto et al., 1993; Kakegawa 272

and Ohmoto, 1999), which suggests the existence of at least an appreciable concentration of

2- SO4 in the oceans as electron donors for SRB. Such suggestions have been very recently supported by Shen et al. (2001) from a study of S isotopes of sulfate and sulfide minerals formed in the 3.5 Ga surface oceans. The existence of sulfate in the 3.6 ~ 3.25 Ga oceans has also been suggested by Huston et al. (2001) from a comprehensive study of volcanogenic massive sulfide deposit. Another recent study (de Ronde and Ebbesen, 1996) has shown that microbiological activity and the productivity by chemosynthesizers associated with seafloor hydrothermal environments were already vigorous as early as 3.25

Ga. Furthermore, this study has shown that all N2-fixers, nitrifiers, and denitrifiers could have been important for the microbially-mediated N redox-cycling in the 3.25 Ga oceans. These suggestions, when combined, allow us to infer the existence of a diverse and complex microbial ecosystem in Mesoarchean oceanic environments. Such a microbial ecosystem could have likely existed much earlier than the middle Archean, and even possibly into the early Archean or earlier.

4-6-7. Critical examination of Beaumont and Robert (1999)

15 13 Beaumont and Robert (1999) published δ N, δ C, Corg, and N data using kerogen extracted from Precambrian cherts and banded iron-formations. The authors noted that there were unusually large ranges in δ15N values (e.g., -1.9 ~ 13 ‰; the sample "Boudou7025" from the Onverwacht Group) and C/N atomic ratios (e.g., 35.0 ~ 593.7, the sample "PPRG182" from the Onverwacht Group) among subsamples (different aliquots) that were made from the same rock chip/powders. The authors raised several possibilities to explain these large variations: (1) procedural effects during acid treatment to extract kerogen, (2) contamination by young organic compounds, (3) metamorphic effects, (4) contamination of

modern atmospheric N2, and (5) original heterogeneity. The authors rejected the 273

possibilities (1), (2), (3), and (4); and claim (5), that their data represent the original heterogeneity maintaining the primary signatures of ancient organisms. We think that possibility (5) is also very unlikely for the following reasons. Although a small portion of a rock sample may not represent the total signature of the rock sample or geological unit from which it is taken, powdered samples are expected, in principle, to have been very well-homogenized during powdering involving extensive mixing. However, the data of Beaumont and Robert (1999) display large differences as

large as 19 ‰ in their δ15N values (e.g., the sample "Boudou7025") and as large as a factor of 17 in their C/N atomic ratios (e.g., the sample "PPRG182") among different fractions of the same powdered samples. Reproducibilities of reliable N isotopic analyses are typically within ± 0.2 ‰ and those of N and C content analyses are typically within 1~2 relative

wt.%. Therefore, such an unusually large subsample heterogeneity in the δ15N and C and N contents of Beaumont and Robert (1999) should not be expected if the analyses are properly performed. To test possibility (1), Beaumont and Robert (1999) used pure natural OM instead of Archean OM to determine possible effects of chemical treatment during kerogen extraction. Although they found no significant effect, such a test using pure natural OM may not be readily applied to Archean OM. If we accept that there existed a very small scale (microscale?) subsample heterogeneity in the δ15N values of OM in Archean chert / BIF samples as Beaumont and Robert (1999) claim, it would imply that there existed a very small scale redox heterogeneity likely caused by diverse microbial activity (of coexisting aerobic and anaerobic microorganisms forming a community) to cause a very large microscale heterogeneity in the

δ15N values. Without redox processes such as nitrification-denitrification, it is very difficult to have N isotopic fractionation as large as 19 ‰ on a very small spatial (mm ~ cm) scale. Such a large isotopic heterogeneity on a very small scale produced by microbiological redox 274

gradients would further suggest that Archean oceanic environments where chert/BIF were deposited with significant influences of hydrothermal activity (de Wit et al., 1992; de Ronde et al., 1993, 1997) were at least partially oxygenated. This scenario is inconsistent with the view that Archean oceanic environments were globally anoxic before the proposed rise of atmospheric O2 at 2.2 ~ 2.0 Ga (e.g., Holland, 1994). The N isotope data and their interpretation presented by Beaumont and Robert (1999) are internally inconsistent.

Therefore, we do not agree that the negative δ15N values reported by Beaumont and Robert (1999) for early Archean samples represent a global oceanic signature (Pinti and Hashizume, 2001; Boyd, 2001a, b). However, we acknowledge the suggestion by Pinti et al.

+ (2001) that chemosynthetic bacteria using NH4 contained in hydrothermal fluids could 15 produce a δ Norg value of -7 ‰. Anoxygenic photosynthesizers produce chlorophyll with 15 15 a δ Norg value of -10 ‰. Such negative δ Norg values in the Archean are probably restricted to environments influenced by hydrothermal activity or photic zone anoxia, both of which are not uncommon phenomena in modern environments.

4-7. Conclusions

Through an isotopic and geochemical study of the Archean and Paleoproterozoic black shales, and with an emerging line of evidence for the early (pre 2.2 Ga) rise of

atmospheric pO2, we reached to the following conclusions:

(1) OM in the Archean–Paleoproterozoic black shales reflects biological production by oxygenic photosynthesizers such as cyanobacteria and by methanogens and methanotrophs. Such microorganisms existed in the Archean–Paleoproterozoic oceans (at least 3.25 Ga)

under an atmosphere with elevated pCO2. 275

(2) Cyanobacteria could have played an important role as oxygenic photosynthesizers and

N2-fixers in the Archean–Paleoproterozoic oceans.

(3) Redox-cycling of N by microorganisms through their N2-fixation, nitrification, denitrification were already operating at least 3.25 Ga ago. "Redox-cycling" implies the

existence of O2 or other form of oxidants to allow operation of the N biogeochemical cycle like today's. Microbial recycling of ammonia in sediments was also already operating in the Mesoarchean oceans.

(4) Microbial N2-fixation was an important process for the N budget in the Archean oceans,

for the formation of Archean–Paleoproterozoic black shales, and for the role as a net O2- source when combined with OM burial.

(5) An operation of the enzymes responsible for N-involving biochemical processes such as

microbial N2-fixation by Archean microorganisms is supported by the availability of bio- essential metals (Mo and Fe). However, such metal availability is not evidence for an operation of the enzymes.

(6) Early development of N2-fixers in the Archean oceans supports the early evolution of complex biosphere.

(7) Previous results of N isotopic composition of kerogen in Archean–Paleoproterozoic chemical sediments (Beaumont and Robert, 1999) do not suggest a change in N biogeochemical cycles linked to the inferred redox evolution of the atmosphere at around 276

Paleoproterozoic (Great Oxidation Event; Holland, 1994) because of the irreproducible and inaccurate data. 277

References

Ader, M., Boudou, J.P., Javoy, M., Goffe, B., and Daniels, E. (1998) Isotope study on organic nitrogen of Westphalian anthracites from the Western Middle field of Pennsylvania (U.S.A.) and from the Bramsche Massif (Germany). Geochim. Cosmochim. Acta 43, 315-323.

Altabet, M.A. and Francois, R. (1994) Sedimentary nitrogen isotopic ratio as a recorder for surface ocean nutrient utilization. Global Biogeochem. Cycles 8, 103-116.

Anhaeusser, C.R. (1986) Archean gold mineralization in the Barberton Mountain Land. in Mineral Deposits of Southern Africa (ed. by Anhaeusser, C. R. and Maske, S.), Geol. Soc. S. Afr., Johannesburg, vol. 1, 113-154.

Awramik, S.M., Schopf, J.W. and Walter, M.R. (1988) Carbonaceous filaments from North Pole, Western Australia: Are they fossil bacteria in Archaean stromatolites? A discussion. Precam. Res. 39, 303-309.

Beaumont, V. and Robert, F. (1999) Nitrogen isotope ratios of kerogens in Precambrian cherts: a record of the evolution of atmosphere chemistry? Precam. Res. 96, 63-82.

Bebout, G.E. and Fogel, M.L. (1992) Nitrogen-isotope compositions of metasedimentary rocks in the Catalina Schist, California: Implications for metamorphic devolatilization history. Geochim. Cosmochim. Acta 56, 2839-2849.

Bengston, S. (1994) Early Life on Earth. Novel Symposium No. 84. Columbia Univ. Press, New York.

Bian, L., Hinrichs, K.U., Xie, T., Brassell, S.C., Iversen, H., Fossing, H., Jørgensen, B.B., and Hayes, J.M. (2001) Algal and archaeal polyisoprenoids in a recent : Molecular isotopic evidence for anaerobic oxidation of methane. Geochem. Geophys. Geosyst. 2, 2000GC000112.

Bidigare, R.R., Flugge, A., Freeman, K.H., Hanson, K.L., Hayes, J.M., Hollander, D., Jasper, J.P., King, L.L., Laws, E.A., Milder, J., Millero, F.J., Pancost, R.P., Popp, B.N., Steinberg, P.A., and Wakeham, S.G. (1997) Consistent fractionation of 13C in nature and in the laboratory: growth-rate effects in some haptophyte algae. Global Biogeochem. Cycles 11, 279-292.

Boetius, A., Ravenschlag, K., Schubert, C., Rickert, D., Widdel, F., Gieseke, A., Amann, R., Jørgensen, B.B., Witte, U., and Pfannkuche, O. (2000) A marine microbial 278 consortium apparently mediating anaerobic oxidation of methane. Nature 407, 623- 626.

Boyd, S.R. (2001a) Ammonium as a biomarker in Precambrian metasediments. Precam. Res. 108, 159-173.

Boyd, S.R. (2001b) Nitrogen in future biosphere studies. Chem. Geol. 176, 1-30.

Brocks, J.J., Logan, G.A., Buick, R., and Summons, R.E. (1999) Archean Molecular Fossils and the Early Rise of Eukaryotes. Science 285, 1033-1036.

Brooks, J.M., Kennicutt, M.C., Fisher, C.R., Macko, S.A., Cole, K., Childress, J.J., Bidigare, R.R., and Vetter, R.D. (1987) Deep-sea hydrocarbon seep communities: Evidence for energy and nutritional carbon sources. Science 238, 1138-1142.

Brumsack, H.J. (1980) Geochemistry of Cretaceous black shales from the Atlantic Ocean (DSDP Legs 11, 14, 36, and 1). Chem. Geol. 31, 1-25.

Canfield, D.E. (1998) A new model for Proterozoic ocean chemistry. Nature 396, 450-453.

Canfield, D.E., Habicht, K.S., and Thamdrup, B. (2000) The Archean sulfur cycle and the early history of atmospheric oxygen. Science 288, 658-661.

Chameides, W.L. and Perdue, E.M. (1997) Biogeochemical cycles, A computer-interactive study of Earth system science and global change. Oxford Univ. Press, Oxford, 224p

Compton, J.S., Williams, L.B., and Ferrell, R.E. (1992) Mineralization of organogenic ammonium in the Monterey Formation, Santa Maria and San Joaquin basins, California, USA. Geochim. Cosmochim. Acta 56, 1979-1991.

DeLong, E.F. (2000) Resolving a methane mystery. Nature 407, 577-578.

de Ronde, C.E.J., Kamo, S., Davis, D.W., de Wit, M.J. and Spooner, E.T.C. (1991) Field, geochemical and U/Pb isotopic constraints from hypabyssal felsic intrusions within the Barberton greenstone belt, South Africa: Implications for tectonics and the timing of gold mineralization. Precam. Res. 49, 261-280.

de Ronde, C.E.J., Spooner, E.T.E., de Wit, M.J., and Bray, C.J. (1992a) Shear zone-related, Au quartz vein deposits in the Barberton Greenstone Belt, South Africa: Field petrographic characteristics, fluid properties, and light stable isotope geochemistry. Econ. Geol. 87, 366-402. 279 de Ronde, C.E.J., Spooner, E.T.E., and de Wit, M.J. (1992b) Geologic and carbon isotopic evidence for pelagic, photosynthetic marine biota at ~3.23 to 3.47 × 109 yr, Barberton greenstone belt, South Africa. In Frontiers of life: Proceedings (eds. J. Tran Thanh Van et al.), 3rd Recontres of Blois, Blois, France, October 14-19, 1991. Paris, Edintons Frontieres, 435-436.

de Ronde, C.E.J., de Wit, M.J. and Spooner, E.T.C. (1993) Early Archean (>3.2 Ga) Fe- oxide-rich, hydrothermal discharge vents in the Barberton greenstone belt, South Africa. Geol. Soc. Am. Bull. 106, 86-104.

de Ronde C.E.J. and de Wit, M.J. (1994) Tectonic history of the Barberton greenstone belt, South Africa: 490 million years of Archean crustal evolution. Tectonics 13, 983- 1005.

de Ronde, C.E.J. and Ebbesen, T.W. (1996) 3.2 b.y. of organic compound formation near sea-floor hot springs. Geology 24, 791-794.

de Ronde, C.E.J., Channer, D.M.deR, Faure, K., Bray, C.J., and Spooner, E.T.C. (1997) Fluid chemistry of Archean seafloor hydrothermal vents: Implications for the composition of circa 3.2 Ga seawater. Geochim. Cosmochim. Acta 61, 4025-4042.

de Wit, M.J., Hart, R., Martin, A., and Abbot, P. (1982) Archean abiogenic and probable biogenic structures associated with mineralized hydrothermal vent systems and regional metasomatism, with implications for greenstone belt studies. Econ. Geol. 77, 1783-1801.

Emelyanov, E.M., Lisitzin, A.P., Shimkus, K.M., Trimonis, E.S., Lukashev, V.K., Lukashin, V.N., Mitropolskiy, A.Y., and Pilipchuk, M.F. (1978) Geochemistry of late Cenozoic sediments of the Black Sea, Leg 42B. In Initial Reports of the Deep Sea Drilling Project (eds. D.A. Ross and Y.P. Neprochunov et al.), 42(2), 543-605.

Emerson, S. and Hedges, J.I. (1988) Processes controlling the organic carbon content of open ocean sediments. Paleoceanogr. 3, 621-634.

Falkowski, P.G. (1997) Evolution of the nitrogen cycle and its influence on the biological sequestration of CO2 in the ocean. Nature 387, 272-275.

Freeman, K.H. and Hayes J.M. (1992) Fractionation of carbon isotopes by phytoplankton and estimates of ancient pCO2 levels. Global Biogeochem. Cycles 6, 629-644.

Freudenthal, T., Wagner, T., Wenzhöfer, F., Zabel, M., and Wefer, G. (2001) Early diagenesis of organic matter from sediments of the eastern subtropical Atlantic: Evidence from stable nitrogen and carbon isotopes. Geochim. Cosmochim. Acta 65, 1795-1808. 280 Haendel, D., Mühle, K., Nitzsche, H.M., Stiehl, G., and Wand, U. (1986) Isotopic variations of the fixed nitrogen in metamorphic rocks. Geochim. Cosmochim. Acta 50, 749- 758.

Hall, A. (1999) Ammonium in granites and its petrogenic significance. Earth Sci Rev. 45, 145-165.

Hayes, J.M. (1994) Global methanotrophy at the Archean-Proterozoic transition. In Early Life on Earth. Nobel Symposium No. 84 (ed. S. Bengston), pp. 220-236. Columbia Univ. Press, New York.

Hayes, J.M., Kaplan, I.R., and Wedeking, K.W. (1983) Precambrian organic geochemistry, preservation of the record. In Earth's Earliest Biosphere: Its Origin and Evolution.(ed. J.W. Schopf), pp. 92-134, Cambridge Univ. Press, Cambridge.

Heubeck, C. and Lowe, D.R.(1994) Depositional and tectonic setting of the Archean Moodies Group, Barberton Greenstone Belt, South Africa. Precam. Res. 68, 257- 290.

Hinga, K.R., Arthur, M.A., Pilson, M.E.Q., Whitaker, D. (1994) Carbon isotope fractionation by marine phytoplanktons in culture: the effects of CO2 concentration, pH, temperature, and species. Global Biogeochem. Cycles 8, 91-102.

Hinrichs, K.U., Hayes, J.M., Sylva, S.P., Brewer, P.G., and DeLong, E.F. (1999) Methane- consuming archaebacteria in marine sediments. Nature 398, 802-805.

Hoehler, T.M., Alperin, M.J., Albert, D.B., and Martens, C.S. (1994) Field and laboratory studies of methane oxidation in an anoxic marine sediment: Evidence for a methanogen-sulfate reducer consortium. Global Biogeochem. Cycles 8, 451-463.

Hoehler, T.M. and Alperin, M.J. (1996) Anaerobic methane oxidation by a methanogen- sulfate reducer consortium: geochemical evidence and biochemical considerations. In Microbial growth on C1 compounds (eds., M.E. Lidstrom and F.R. Tabita), Kluwer, San Diego.

Hoering, T.C. and Ford, H.T. (1960) The isotope effect in the fixation of nitrogen by Azotobacter. J. Am. Chem. Soc. 82, 376-378.

Holland, H.D. (1984) The chemical evolution of the atmosphere and oceans. Princeton Univ. Press, Princeton.

Holland, H.D. (1994) Early Proterozoic atmospheric change. In Early Life on Earth. Novel Symposium No. 84 (S. Bengston, ed.), Columbia Univ. Press, New York. 281 Holland, H.D. (1999) When did the Earth's atmosphere become oxic? A reply. Geochemical News 100, 20-22.

Honma, H. and Itihara, Y. (1981) Distribution of ammonium in minerals of metamorphic and granitic rocks. Geochim. Cosmochim. Acta 45, 983-988.

Honma, H. (1996) High ammonium contents in the 3800 Ma Isua supracrustal rocks, central West Greenland. Geochim. Cosmochim. Acta 60, 2173-2178.

Huston, D.L., Brauhart, C.W., Drieberg, S.L., Davidson, G.J., and Groves, D.I. (2001) Metal leaching and inorganic sulfate reduction in volcanic-hosted massive sulfide mineral systems: Evidence from the paleo-Archean Panorama district, Western Australia. Geology 29, 687-690.

Ingall, E.D., Bustin, R.M., and Van Cappellen, P. (1993) Influence of water column anoxia on the burial and preservation of carbon and phosphorus in marine sediments. Geochim. Cosmochim. Acta 57, 303-316.

Itihara, Y. and Suwa, K. (1985) Ammonium contents of biotites from Precambrian rocks in + Finland: The significance of NH4 as a possible chemical fossil. Geochim. Cosmochim. Acta 49, 145-151.

Itihara Y. and Tainosho, Y. (1989) Ammonium and insoluble nitrogen in Precambrian rocks from the Gawler Craton, Australia: Inference of life activity. J. Geol. Soc. Japan 95, 439-445.

Itihara, Y. and Honma, H. (1979) Ammonium in biotite from metamorphic and granitic rocks of Japan. Geochim. Cosmochim. Acta 43, 503-509.

Itihara, Y., Suwa, K., and Hoshino, M. (1986) Organic matter in the Kavirondian sedimentary rocks of Archaean period in . Geochem. J. 20, 201-207.

Jackson, M.P.A., Eriksson, K.A., and Harris, C.W. (1987) Early Archean foredeep sedimentation related to crustal shortening: a reinterpretation of the Barberton Sequence, southern Africa. Tectonophyscis 136, 360-366.

Jaffe, D.A. (1992) The nitrogen cycle. In Global Biogeochemical Cycles (eds. S.S.Butcher, R.J.Charlson, G.H.Orians, and G.V.Wolfe), Academic Press, London, 263-284.

+ Juster, T.C., Brown, P.E., and Bailey, S.W. (1987) NH4 -bearing illite in very low grade metamorphic rocks associated with , northeastern Pennsylvania. Am. Mineral. 72, 555-565. 282 Kakegawa, T. and Ohmoto, H. (1999) Sulfur isotope evidence for the origin of 3.4 to 3.1 Ga pyrite at the Princeton gold mine, Barberton Greenstone Belt, South Africa. Precam. Res. 96, 209-224.

Kasting, J.F. (1987) Theoretical constraints on oxygen and carbon dioxide concentrations in the Precambrian atmosphere. Precam. Res. 34, 205-228.

Kasting, J.F. (1990) Bolide impacts and the oxidation state of carbon in the Earth's early atmosphere. Origins Life Evol. Biosph. 20, 199-231.

Kasting, J.F. (1993) Earth's early atmosphere. Science 259, 920-926.

Keeney, D.R. and Nelson, D.W. (1992) Nitrogen-inorganic forms. In Methods of Soil Analysis, Part 2, Chemical and Microbiological Properties (eds. A.L. Page, R.H.Miller, and D.R.Keeney), Amer. Soc. Agronomy, Soil Sci. Soc. Amer., 643- 698.

Kennicutt II, M.C., Burke, R.A. Jr., MacDonald, I.R., Brooks, J.M., Denoux, G.J., and Macko, S.A. (1992) Stable isotope partitioning in seep and vent organisms: chemical and ecological significance. Chem. Geol. 101, 293-310.

Kump, L.R. and Arthur, M.A. (1999) Interpreting carbon-isotope excursions: carbonates and organic matter. Chem. Geol. 161, 181-198.

Lasaga, A.C. and Ohmoto, H. (2002) The oxygen geochemical cycle: dynamics and stability. Geochim. Cosmochim. Acta 66, 361-381.

Leventhal, J. (1993) Metals in Black Shales. In Organic Geochemistry: Principles and Applications (eds. M.H. Engel and S.A. Macko), 581-592.

Lowe, D.R. and Byerly, G.R. (1999) Geologic Evolution of the Barberton Greenstone Belt, South Africa. Geol. Soc. Amer. Spec. Paper 329. 319p.

Macko, S.A., Engel, M.H., and Parker, P.L. (1993) Early diagenesis of organic matter in sediments. Assessment of mechanisms and preservation by the use of isotopic molecular approaches. In Organic Geochemistry. Principles and Applications (eds. M.H. Engel and S.A. Macko), Plenum Press.

Madigan, M.T., Martinko, J.M., and Parker, J. (2000) Brook biology of microorganisms, 8th ed. Prentice Hall, Upper Saddle River.

Mancinelli, R.L. and McKay, C.P. (1988) The evolution of nitrogen cycling. Origins Life Evol. Biosphere 18, 311-325. 283 Maslin, M.A., Hall, M.A., Shackleton, N.J., and Thomas, E. (1996) Calculating surface 13 water pCO2 from foraminiferal organic δ C. Geochim. Cosmochim. Acta 60, 5089-5110.

Minagawa, M. and Wada, E. (1986) Nitrogen isotope ratios of red tide organisms in the East China Sea: a characterization of biological nitrogen fixation. Mar. Chem. 19, 245-259.

Mingram, B. and Bräuer, K. (2001) Ammonium concentration and nitrogen isotope composition in metasedimentary rocks from different tectonometamorphic units of the European Variscan Belt. Geochim. Cosmochim. Acta 65, 273-287.

Mojzsis, S.J., Arrhenius, G., McKeegan, K., Harrison, T.M., Nutman, A.P., and Friend, C.R.L. (1996) Evidence for life on Earth before 3,800 million years ago. Nature 384, 55-59.

Müller, P.J. (1977) C/N ratios in Pacific deep-sea sediments: Effect of inorganic ammonium and organic nitrogen compounds sorbed by clays. Geochim. Cosmochim. Acta 41, 765-776.

Meyers, P.A. (1994) Preservation of elemental and isotopic source identification of sedimentary organic matter. Chem. Geol. 144, 289-302.

Navarro-González, R., McKay, C.P., and Mvondo, D.N. (2001) A possible nitrogen crisis for Archaean life due to reduced nitrogen fixation by lightening. Nature 412, 61-64.

Ohkouchi, N., Kawamura, K., Wada, E., and Taira, A. (1997) High abundances of hopanols and hopanoic acids in Cretaceous black shales. Ancient Biomolecules 1, 183-192.

Ohmoto, H. (1997) When did the Earth's atmosphere become oxic? Geochemical News 93. 12-12 and 26-27.

Ohmoto, H., Kakegawa, T., and Lowe, D.R. (1993) 3.4-billion-year-old biogenic pyrites from Barberton, South Africa: sulfur isotope evidence. Science 262, 555-557.

Orphan, V.J., House, C.H., Hinrichs, K.U., McKeegan, K.D., and DeLong, E.F. (2001) Methane-consuming archaea revealed by directly coupled isotopic and phylogenetic analysis. Science 293, 484-487.

Pavlov, A.A., Kasting, J.F., Eigenbrode, J.L., and Freeman, K.H. (2001) Organic haze in Earth's early atmosphere: source of low-13C Late Archean kerogens? Geology 29, 1003-1006. 284 Pedersen, T.F. and Calvert, S.E. (1990) Anoxic vs. Productivity: What controls the formation of organic-carbon-rich sediments and sedimentary rocks? Am. Assoc. Petrol. Bull. 74, 454-466.

Peters, K.E., Sweeney, R.E., and Kaplan, I.R. (1978) Correlation of carbon and nitrogen stable isotope ratios in sedimentary organic matter. Limnol. Oceanogr. 23, 598-604.

Pinti, D.L. and Hashizume, K. (2001) 15N-depleted nitrogen in Early Archean kerogens: clues on ancient marine chemosynthetic-based ecosystem? A comment to Beaumont, V., Robert, F., 1999. Precam. Res. 105, 85-88.

Pinti, D.L., Hashizume, K., and Matsuda, J. (2001) Nitrogen and argon signatures in 3.8 to 2.8 Ga metasediments: Clues on the chemical state of the Archean ocean and the deep biosphere. Geochim. Cosmochim. Acta 65, 2301-2315.

Postgate, J. (1987) Nitrogen Fixation 2nd eds, 73, Edward Arnold London.

Premuzic, E.T., Benkovitz, C.M., Gaffney, J.S., and Walsh, J.J. (1982) The nature and distribution of organic matter in the surface sediments of world oceans and seas. Org. Geochem. 4, 63-77.

Rau, G.H., Arthur, M.A., and Dean, W.E. (1987) 15N/14N variations in Cretaceous Atlantic sedimentary sequences: implications for past changes in marine N biogeochemistry. Earth Planet Sci. Lett. 82, 269-279.

Redfield, A.C., Ketchum, B.H., and Richards, F.A. (1963) The influence of organisms on the composition of sea water. In The Sea (ed. M.N.Hill), Wiley, New York, 26-77.

Robert, F. (1988) Carbon and oxygen isotope variation in Precambrian chert. Geochim. Cosmochim. Acta 52, 1473-1488.

Robl, T.L. and Barron, L.S. (1988) The geochemistry of Devonian black shales in central Kentucky and its relationship to inter-basinal correlation and depositional environment. In Devonian of the World Vol. II (eds. N.J. McMillan, A.F. Embry, D.J. Glass) Canadian Soc. of Petroleum Geologists, Calgary. 377-392.

Rosing, M.T. (1999) 13C-depleted carbon microparticles in >3700-Ma sea-floor sedimentary rocks from West Greenland. Science 283, 674-676.

Sadofsky, S. and Bebout, G.R. (2000) Ammonium partitioning and N-isotope fractionation among coexisting micas during high-temperature fluid-rock interactions: Examples from the New England Appalachians. Geochim. Cosmochim. Acta 64, 2835-2849. 285 Sampei, Y. and Matsumoto, E. (2001) C/N ratios in a sediment core from Nakaumi Lagoon, southwest Japan - usefulness as an organic source indicator. Geochem J. 35, 189- 205.

Sano, Y. and Pilinger, C.T. (1990) Nitrogen isotopes and N2/Ar ratios in cherts: An attempt to measure time evolution of atmospheric δ15N value. Geochem. J. 24, 315-325.

Sarmiento, J. (1992) Biogeochemical Ocean Models. In Climate System Modeling (ed. K.E.Trenberth), Cambridge Univ. Press, 519-564.

Schidlowski, M. (1988) A 3,800-million-year isotopic record of life from carbon in sedimentary rocks. Nature 333, 313-318.

Schidlowski, M. (2001) Carbon isotopes as biogeochemical recorders of life over 3.8 Ga of Earth history: evolution of a concept. Precam. Res. 106, 117-134.

Schidlowski, M., Hayes, J.M., and Kaplan, I.R. (1983). Isotopic inferences of ancient biochemistries: carbon, sulfur, hydrogen, and nitrogen. In Earth’s Earliest Biosphere: Its Origin and Evolution (ed. J.W. Schopf), Princeton Univ. Press, Princeton, 149-186.

Schidlowski, M., Golubic, S., Kimberley, M.M., McKirdy, D.M., and Trudinger, P.A. (1992) Early Organic Evolution: Implications for Mineral and Energy Resources. Springer-Verlag, Berlin.

Schlanger, S.O., Arthur, M.A., Jenkyns, H.C., and Scholle, P.A. (1987) The Cenomanian- Turonian oceanic anoxic event, I. Stratigraphy and distribution of organic carbon- rich beds and the marine δ13C excursion. In Marine Petroleum Source Rocks (eds. J. Brooks and A. Fleet), Geol. Soc. London Spec. Pub. 26, 371-399.

Schopf, J.W. and Packer, B.M. (1987). Early Archean (3.3 billion to 3.5 billion-years-old) microfossils from the Warrawoona Group, Western Australia. Science 237, 70-73.

Schopf, J.W. (1983) Earth’s Earliest Biosphere: Its Origin and Evolution. Princeton Univ. Press, Princeton.

Schopf, J.W. (1993) Microfossils of the early Archean Apex chert: new evidence of the antiquity of life. Science 260, 640-646.

Schopf, J.W. and Klein, C. (1992) The Proterozoic Biosphere: A Multidisciplinary Study. Cambridge Univ. Press, Cambridge. 286 Schorno, K.S. (1980) Geochemistry of carbon: Deep Sea Drilling Project Legs 58 and 59. In Initial Reports of the Deep Sea Drilling Project v.58 (eds. K.C. deVries and K. Kobayashi), Washington (US Gov. Print. Office), 641-646.

Schouten, S., Wakeham, S.G., and Sinninghe-Damsté, J.S. (2001) Evidence for anaerobic methane oxidation by archaea in euxinic waters of the Black Sea. Org. Geochem. 32, 1277-1281.

Schubert, C.J. and Calvert, S.E. (2001) Nitrogen and carbon isotopic composition of marine and terrestrial organic matter in Arctic Ocean sediments: implications for utilization and organic matter composition. Deep-Sea Res. I 48, 789-810.

Sclater, J.G., Jaupart, C., and Galson, D. (1980) The heat flow through oceanic and continental crust and the heat loss of the Earth. Rev. Geophys. Space Phys. 18, 269- 311.

Shen, Y., Buick, R., and Canfield, D.E. (2001) Isotopic evidence for microbial sulphate reduction in the early Archaean era. Nature 410, 77-81.

Simoneit, B.R.T., Brenner, S., Peters, K.E., and Kaplan, I.R. (1981) Thermal alteration of Cretaceous black shale by diabase intrusions in the Eastern Atlantic: II. Effects on bitumen and kerogen. Geochim. Cosmochim. Acta 45, 1581-1602.

Strauss, H., DesMarais, D.J., Hayes, J.M. and Summons, R.E. (1992) The Carbon-Isotopic Record. In The Proterozoic Biosphere: A multidisciplinary study (eds. J.W. Schopf and C. Klein), Cambridge University Press, Cambridge, UK, 117-127.

Sweeney, R.E., Liu, K.K., and Kaplan, I.R. (1978) Oceanic nitrogen isotopes and their uses in determining the source of sedimentary nitrogen. In Stable isotopes in the Earth Sciences, Proc. Int. Symp., New Zealand Dep. Sci. Ind. Res., 599-604.

Sweeney, R.E. and Kaplan, I.R. (1980) Natural abundance of 15N as a source indicator for near shore marine sedimentary and dissolved nitrogen. Mar. Chem. 9, 81-94.

Taylor, S.R. and McLennan, S.M. (1985) The Continental Crust: its Composition and Evolution. Blackwell Scientific Publications, 311p.

Towe, K.M. (1994) Earth’s early atmosphere: Constraints and opportunities for early evolution. In Early Life on Earth. Nobel Symposium No. 84. (ed., S. Bengston) Columbia University Press, New York, 36-47.

Turekian, K. and Wedepohl, K. (1961) Distribution of elements in some major units of the earth's crust. Geol. Soc. Am. Bull. 72, 175-192. 287 Valley, J.W. (1986) Stable isotope geochemistry of metamorphic rocks. In Stable isotopes in high temperature geological processes (eds. J.M. Valley, H.P. Taylor, Jr., and J.R. O'Neil), Reviews in Mineralogy 16, Mineral. Soc. Am., 445-489.

Van Cappellen, P. and Ingall, E.D. (1996) Redox stabilization of the atmosphere and oceans by phosphorus-limited marine production. Science 271, 493-496.

Viljoen, M.J. and Viljoen, R.P. (1969) An introduction to the geology of the Barberton granite-greenstone terrain. geol. Soc. South Afr. Spec. Pub. 2, 9-28.

Vine, J. and Tourtelot, E. (1970) Geochemistry of black shale deposits. Econ. Geol. 65, 253-272.

Wada, E., Kadonaga, T., and Matsuo, S. (1975) 15N abundance in nitrogen of naturally occurring substances and global assessment of denitrification from isotopic viewpoint. Geochem. J. 9, 139-148.

Wada, E. and Hattori, A. (1991) Nitrogen in the Sea: Forms, Abundances, and Rate Processes. CRC Press, Florida.

Walsh, M.M. (1992) Microfossil and possible microfossils from the Early Archean Onverwacht Group, Barberton Mountain Land, South Africa. Precam. Res. 54, 271- 293.

Waples, D.W. and Sloan, J.R. (1980) Carbon and nitrogen diagenesis in deep sea sediments. Geochim. Cosmochim. Acta 44, 1463-1470.

Ward, J.H.W. (1995) Geology and metallogeny of the Barberton greenstone belt: a survey. J. Afr. Earth Sci. 21, 213-240.

Watanabe, Y., Naraoka, H., Wronkiewicz, D.J., Condie, K.C., and Ohmoto, H. (1997) Carbon, nitrogen, and sulfur geochemistry of Archean and Proterozoic shales from the Kaapvaal Craton, South Africa. Geochim. Cosmochim. Acta 61, 3441-3459.

Wedepohl, K.H. (1991) The composition of the upper Earth's crust and the natural cycles of selected metals. Metals in natural raw materials. Natural resources. In Metals and their compounds in the Environment (ed., E. Merian), VCH, Weinheim. 3-17.

Williams, L.B., Ferrell, R.E., Hutcheon, I., Bakel, A.J., Walsh, M.M., and Krouse, H.R. (1995) Nitrogen isotope geochemistry of organic matter and minerals during diagenesis and hydrocarbon maturation. Geochim. Cosmochim. Acta 59, 765-779. 288 Wronkiewicz, D.J. and Condie, K.C. (1990) Geochemistry and mineralogy of sediments from Ventersdorp and Transvaal Supergroup, South Africa: Cratonic evolution during the early Proterozoic. Geochim. Cosmochim. Acta 54, 343-354.

Yamaguchi, K.E. Ohkouchi, N., and Ohmoto, H. (in prep) Factors controlling the low N/Corg ratios of the Phanerozoic black shales: Implications for Precambrian.

Zehr, J.P., Mellon, M., Braun, S., Litaker, W., Steppe, T., and Paerl, H.W. (1995) Diversity of heterotrophic nitrogen-fixation genes in a marine cyanobacterial mat. Appl. Environ. Microbiol. 61, 2527-2532. 289

(a) Before O2-rise, (b) After O2-rise, (c) After O2-rise, No biological N 2-fixation N2-fixation mode Normal mode

N N CO2 N2 2 2 Lightning ~ 0 ‰

NH3 -fixation 2 Atmosphere N ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Seawater N2-fixers Non N2-fixers Organisms Organisms Organisms D D A ~ 0 ‰ ~ 6 ‰ A

D Denitrification + A NH4 + NO - NH4 Nitrification 3

A A ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Sediments Organisms Organisms Organisms A A A D D D Sed. D + Sed. D + Sed. D + NH4 NH4 NH4 OM OM OM Fixation Fixation Fixation

Kerogen Clay Kerogen Clay Kerogen Clay ~ 0 ‰ ~ 0 ‰ ~ 0 ‰ ~ 0 ‰ ~ 6 ‰ ~ 6 ‰

Fig. 4-1. Schematic diagram showing the N biogeochemical cycles. (a) Before the O2-rise and the onset of biological N2-fixation (i.e., Primitive Earth). Lightning was probably the major process to form NH3 in the atmosphere. (b) 15 After the O2-rise, N2-fixation mode. The δ Norg values are close to 0 ‰. (c) 15 After the O2-rise, normal mode. The δ Norg values are ~+6 ‰ due to redox- 15 cycling of N involving denitrification and nitrification. The δ Nclay values are also ~+6 ‰. In Phanerozoic and Proterozoic (and probably Archean) oceanic environments, sedimentary kerogen and clays are the product of the case (b) and 15 15 (c), and their δ Norg and δ Nclay values become between ~0 and ~+6 ‰, depending on the degree of contribution from N2-fixers to sedimentary OM. A: assimilation (assimilatory or dissimilatory); D: decomposition (including ammonification). 290

(a) 8 6 2.22 Ga Timeball Hill Fm, 4 Eastern Transvaal 2 0 3 (b) 2 2.56 Ga Oak Tree Fm 0

(c) 2 2.60 Ga Wittenoom Dolomite Fm 0

(d) 6 4 2.60 Ga Carawine Dolomite Fm 2 0 (e) 4 2 >2.60 Ga Marra Mamba Iron Fm 0

(f) 4 2.69 Ga Lewin Shale Fm Number of samples 2 0

(g) 12 3.25 Ga Sheba Fm, Graywacke 0

(h) 8 6 3.25 Ga Sheba Fm, 4 Shale 2 0

Modern sediments (Müller, 1977) (i) Modern sediments (Waples and Sloan, 1980) Modern sediments (Rau et al., 1987) Neoarchean-Paleoproterozoic shales (Watanabe et al., 1997)

0 0.1 0.14 N [wt.%]

Fig. 4-2. Histogram of the bulk N contents of the Archean– Paleoproterozoic samples of this study (a through h), and of the modern sediments and Neoarchean– Paleoproterozoic shales from literature. 291

1

0.9

0.8

0.7

0.6 Norg 0.5 ? Devonian ∑N 0.4

0.3 ? Range of modern sediments

0.2

0.1

0 3.5 3 2.5 2 1.5 1 0.5 0

Age [Ga]

Fig. 4-3. Relationship between the Norg/∑N ratios and ages in Ga of the Archean– Paleoproterozoic samples of this study and some Phanerozoic shales (Compton et al., 1992; Itihara and Honma, 1979; Caplan and Bustin, 1996) and modern sediments (Freudenthal et al., 2001; Schubert and Calvert, 2001; Schorno, 1980; Sampei and Matsumoto, 2001) from literature. 292

0.2 3.25 Ga Sheba Fm (shale) 3.25 Ga Sheba Fm (graywacke) 2.69 Ga Lewin Shale Fm Devonian >2.60 Ga Marra Mamba Iron Fm Pennsylvanian 2.60 Ga Wittenoom Dolomite Fm 2.60 Ga Carawine Dolomite Fm 0.15 2.56 Ga Oak Tree Fm Cretaceous 2.22 Ga Timeball Hill Fm Jurassic -rich Corg

N 0.1 [wt.%]

-poor

Jurassic org C

0.05

0 0 0.5 1 1.5 2 2.5 3 3.5 4

Corg [wt.%]

Fig. 4-4. Cross plot showing relationship between the N and Corg contents of the Archean–Paleoproterozoic samples of this study. Lines represent regression lines for some Phanerozoic shales. Phanerozoic data are from Ingall et al. (1993) for the Jurassic Corg-rich and Corg-poor shales, the Pennsylvanian black shales, and the Devonian black shales; Brumsack (1980) for the Cretaceous black shales. Fm: Formation. 293

N2-fixers

Vent and Seep organisms (n = 110; Kennicutt II et al., 1992)

10 Modern sediments kerogen (a) 5

0 20 Modern sediments 15 bulk

(b) 10

5

0 8 Cretaceous (c) 5 black shale

0 1 Triassic oil shale (d) 0 (e) 1 1.4 Ga Roper Group 0 (f) 1 1.6 Ga Bungle Bungle Dolomite 0 (g) 1 1.8 Ga Rove Fm 0 8 2.22 Ga 4 Timeball Hill Fm, (h) E. Transvaal 0 1 2.5 Ga Hamersley Group (i) 0 Number of analyses 4 (j) 2 2.56 Ga Oak Tree Fm 0 2.60 Ga Wittenoom (k) 2 0 Dolomite Fm 2 2.60 Ga Carawine Dolomite Fm (l) 0 (m) 2 >2.60 Ga Marra Mamba Iron Fm 0 3 (n) 2 2.69 Ga Lewin Shale Fm 0 2 3.25 Ga Sheba Fm, graywacke (o) 0 10 3.25 Ga Sheba Fm, (p) 5 shale

0 2 3.4 Ga Gorge Creek Group (q) 0

(r) Ealry Proterozoic Late Archaean Ealry Archaean

-10 0 10 20 δ15N [‰] 294

Fig. 4-5. Histogram showing the δ15N values of the Archan–Paleoproterozic samples of this study and the various samples from literature. (a) kerogen in modern marine sediments (Peters et al., 1978); (b) whole rock samples of modern marine sediments (Peters et al., 1978); (c) Cretaceous marine shales. Organic carbon-rich (Corg > 1 wt.%) (filled columns) and Corg-poor (Corg < 1 wt.%) (open columns) shales (Ohkouchi et al., 1997; Rau et al., 1987); (d) Triassic oil shale (Chicaarelli et al., 1993) (e) 1.4 Ga shales from the Roper Group (kerogen: filled columns) (Hayes et al., 1983); (f) 1.6 Ga shales from the Bungle Bungle Dolomite (kerogen: filled columns) (Hayes et al., 1983); (g) 1.8 Ga shales from the Rove Formation (kerogen: filled column) (Hayes et al., 1983); (h) 2.22 Ga black shales from the Timeball Hill Formation (whole rock: open columns); (i) 2.5 Ga shales from the Hamersley Group (kerogen: filled columns) (Hayes et al., 1983); (j) 2.56 Ga black shales from the Oak Tree Formation: kerogen (filled columns) and whole rock (open columns) samples; (k) 2.60 Ga black shales from the Wittenoom Dolomite Formation (whole rock: open columns); (l) 2.60 Ga black shales from the Carawine Dolomite Formation (whole rock: open columns); (m) >2.60 Ga black shales from the Marra Mamba Iron Formation (whole rock: open columns); (n) 2.69 Ga black shales from the Jeerinah Formation (whole rock: open columns); (o) 3.25 Ga graywackes from the Sheba Formation (kerogen: filled columns); (p) 3.25 Ga black shales from the Sheba Formation (kerogen: filled columns, whole rock: open columns, and clay: hatched columns); (q) 3. 4 Ga shales of the Gorge Creek Group (kerogen: filled column) (Hayes et al., 1983); and (r) Early Proterozoic, Late Archean, and Early Archean chemical sediments (cherts and banded iron-formations) (kerogen) (Beaumont and Robert, 1999). Instead of histogram, the ranges and averages are indicated by bars and triangles, respectively. Fm: Formation. 295

0.12

Lawsonite-Albite Blueschist 0.1 Greenschist -Amphibolite Amphibolite

N [wt%] Increasing metamorphic 0.05 grade

0 123456

δ15N [‰]

Fig. 4-6. Relationship between the N contents and the δ15N values for the Catalina Schist metasedimentary rocks of various metamorphic grades (from lawsonite-albite facies to amphibolite facies). Data from Bebout and Fogel (1992). The rocks subjected to metamorphism associated with subduction occurred in early Cretaceous (metamorphism at temperature: 350-750 ˚C and pressure depth: 15-45 km). The vertical dashed line at δ15N = 3.5 ‰ indicates separation between low grade (left of the line) and high grade (right of the line) metamorphism. The horizontal dashed line at N = 0.03 wt.% indicates separation between low grade (above of the line) and high grade (down of the line) metamorphism. Thick arrows represents the direction of increasing metamorphic grade. 296

500

100 Devonian R= 0.70

Cretaceous OAE R= 1.00 10 Mo Black Sea [ppm] R= 0.66

Fig Tree R= 0.64 1

R: correlation coefficient 0.1 0.1 1 10 20

Corg [wt.%]

Fig. 4-7. Plot of the Mo vs. Corg contents of the shales of the Fig Tree Group (filled circle). Both axes are in logarithmic scale. For comparison, Devonian black shales (open diamond; Robl and Barron, 1988; Leventhal, 1993; compilation by Holland, 1984), Cretaceous OAE sediments (open triangle; Brumsack, 1980), and modern Black Sea sediments (open square; Emelyanov et al., 1978) are also shown. Correlation coefficients (R) and slopes for the regression lines are also shown. Note positive correlations between Mo and Corg content independent of the depositional ages. Cretaceous data represent average content from 88 samples (38 samples with Corg > 1 wt.%, 11 samples with Corg 0.5-1 wt.%, and 39 samples with Corg < 0.5 wt.%: Brumsack, 1980). 297

N2 Photosynthesizers Oxic Oxic Denitrification

Anoxic

(a) (b)

Stratification

N N N2-fixation

P P P P P P (d) (c)

Overturn 3- PO4 regeneration

Fig. 4-8. Schematic illustrations of the Archean oceanic environments proposed in this study. Globally oxic atmosphere and oceans with localized anoxic basins are suggested. (a) Normal condition. Active photosynthetic activity. (b) Stratified ocean. Enhanced preservation of organic matter. (c) Phosphorus regeneration in an anoxic condition. (d) Overturn and collapse of stratification. Upwelling of P-enriched deep water into surface water. N2-fixers become active. See text for the detailed explanation. 298 Table 4-1. Summary of the geological settings and samples of this study.

Drillcore Drillhole locality Supergroup (SG) Age Formation (Fm) N Group (G) [Ga]

South Africa PU1308 Agnes gold mine, Swaziland SG 3.25 Sheba Fm 17 Barberton Fig Tree G MRE10 Sheba gold mine, Swaziland SG 3.25 Sheba Fm 12 Barberton Fig Tree G

MSF6 26˚33'S, 27˚12'E Transvaal SG 2.56 Oak Tree Fm 3 near Klerksdorp Chuniespoort G Witwatersrand

PTB3 24˚55' S, 30˚44'E Transvaal SG 2.22 Timeball Hill Fm 12 Pilgrim's Rest Pretoria G Australia WRL1 East of Wittenoom, Mt. Bruce SG >2.60 Marra Mamba Iron Fm 4 Hamersley Fortescue G 2.60 Wittenoom Dol. Fm 3 RHDH2A Ripon Hill, Mt. Bruce SG 2.69 Lewin Shale Fm 8 Northeastern Fortescue G Hamersley Mt. Bruce SG 2.60 Carawine Dol. Fm 6 Hamersley G

N: number of samples 299 Table 4-1. (continued)

Drillcore Dominant lithology of Tectonic settings of the groups the studied samples

South Africa PU1308 Black shales (some are rich Foreland basin, fore deep basin, in siderite and silica) evolving back-arc, passive continental margin, early rift to MRE10 Graywackes continental shelf, and shelf-rise

MSF6 Black shales Shallow marine cratonic environments

PTB3 Black shales Shallow marine cratonic environments Australia WRL1 Black shales, carbonate rich Shallow marine WRL1 black shales Deep marine RHDH2A Black shales, carbonate-rich Deep marine black shales

RHDH2A Shallow marine or intertidal environments

See chapter 2 for the references of the tectonic settings of the samples. 300 Table 4-2. Geochemical data of the Archean–Paleoproterozoic samples of this study.

1 2 3 4 5 6 13 7 15 7 15 8 15 Samples Corg Ntot Norg Nclay Ntot/Corg Mo δ Corg δ Norg δ Nclay δ Ntotal [wt.%] [wt.%] [wt.%] [wt.%] [ppm] [‰] [‰] [‰] [‰]

Transvaal Supergroup, South Africa 2.22 Ga Timeball Hill Formation, Pretoria Group 1 0.41 0.06 n.d. n.d. 0.12 n.d. -29.3 n.d. n.d. 6.6 2 0.35 0.03 n.d. n.d. 0.08 n.d. -27.5 n.d. n.d. 5.2 3 0.46 0.04 n.d. n.d. 0.07 n.d. -29.3 n.d. n.d. 6.5 4 0.39 0.04 n.d. n.d. 0.09 n.d. -28.8 n.d. n.d. 5.9 5 0.39 0.05 n.d. n.d. 0.10 n.d. -28.4 n.d. n.d. 5.6 6 0.39 0.04 n.d. n.d. 0.09 n.d. -28.5 n.d. n.d. 5.7 7 0.36 0.04 n.d. n.d. 0.10 n.d. -28.3 n.d. n.d. 5.6 8 0.50 0.06 n.d. n.d. 0.11 n.d. -28.6 n.d. n.d. 5.7 9 0.43 0.04 n.d. n.d. 0.09 n.d. -28.4 n.d. n.d. 5.7 10 0.47 0.04 n.d. n.d. 0.07 n.d. -28.1 n.d. n.d. 5.8 11 0.43 0.05 n.d. n.d. 0.10 n.d. -29.4 n.d. n.d. 6.9 12 0.30 0.04 n.d. n.d. 0.11 n.d. -27.8 n.d. n.d. 6.0

Average 0.41 0.04 - - 0.09 - -28.5 - - 5.9 S.D. 0.06 0.01 - - 0.01 - 0.6 - - 0.5

2.56 Ga Oak Tree Formation, Chuniespoort Group 1 2.30 0.02 n.d. n.d. 0.01 n.d. -33.1 1.0 n.d. 1.6 2 1.64 0.01 n.d. n.d. 0.01 n.d. -34.1 1.1 n.d. 2.6 3 1.66 0.01 n.d. n.d. 0.01 n.d. -33.3 0.7 n.d. n.d.

Average 1.87 0.01 - - 0.01 - -33.5 0.9 - 2.1 S.D. 0.38 0.00 - - 0.00 - 0.5 0.2 - 0.7 301 Table 4-2. (continued)

1 2 3 4 5 6 13 7 15 7 15 8 15 Samples Corg Ntot Norg Nclay Ntot/Corg Mo δ Corg δ Norg δ Nclay δ Ntotal [wt.%] [wt.%] [wt.%] [wt.%] [ppm] [‰] [‰] [‰] [‰]

Mt. Bruce Supergroup, Western Australia 2.60 Ga Wittenoom Dolomite Formation, Hamersley Group 1 0.74 0.01 n.d. n.d. 0.01 n.d. -40.4 n.d. n.d. 5.7 2 2.78 0.01 n.d. n.d. 0.00 n.d. -39.1 n.d. n.d. 6.1 3 2.67 0.05 n.d. n.d. 0.02 n.d. -32.0 n.d. n.d. 5.7

Average 2.06 0.02 - - 0.01 - -37.2 - - 5.8 S.D. 1.15 0.02 - - 0.01 - 4.5 - - 0.3

2.60 Ga Carawine Dolomite Formation, Hamersley Group 1 1.41 b.d. n.d. n.d. - n.d. -41.2 n.d. n.d. n.d. 2 1.02 b.d. n.d. n.d. - n.d. -45.3 n.d. n.d. n.d. 3 2.91 b.d. n.d. n.d. - n.d. -48.9 n.d. n.d. n.d. 4 1.91 b.d. n.d. n.d. - n.d. -43.7 n.d. n.d. n.d. 5 4.24 0.06 n.d. n.d. 0.01 n.d. -47.6 n.d. n.d. 0.4 6 4.45 0.05 n.d. n.d. 0.01 n.d. -43.1 n.d. n.d. 0.0

Average 2.66 0.06 - - 0.01 - -45.0 - - 0.2 S.D. 1.45 0.01 - - 0.00 - 2.9 - - 0.3

>2.60 Ga Marra Mamba Iron Formation, Hamersley Group 1 0.19 0.01 n.d. n.d. 0.05 n.d. n.d. n.d. n.d. 4.7 2 0.37 0.01 n.d. n.d. 0.02 n.d. -37.7 n.d. n.d. 2.9 3 0.93 0.01 n.d. n.d. 0.01 n.d. -40.4 n.d. n.d. 1.2 4 0.74 0.01 n.d. n.d. 0.01 n.d. -43.3 n.d. n.d. 0.1

Average 0.56 0.01 - - 0.02 - -40.4 - - 2.2 S.D. 0.34 0.00 - - 0.02 - 2.8 - - 2.0 302 Table 4-2. (continued)

1 2 3 4 5 6 13 7 15 7 15 8 15 Samples Corg Ntot Norg Nclay Ntot/Corg Mo δ Corg δ Norg δ Nclay δ Ntotal [wt.%] [wt.%] [wt.%] [wt.%] [ppm] [‰] [‰] [‰] [‰]

Mt. Bruce Supergroup, Western Australia 2.69 Ga Lewin Shale Formation, Fortescue Group 1 2.67 0.09 n.d. n.d. 0.03 n.d. -39.4 n.d. n.d. 0.1 2 6.83 0.12 n.d. n.d. 0.02 n.d. -42.8 n.d. n.d. 0.0 3 1.56 0.06 n.d. n.d. 0.03 n.d. -39.1 n.d. n.d. 4.6 4 2.66 b.d. n.d. n.d. - n.d. -42.2 n.d. n.d. n.d. 5 1.83 b.d. n.d. n.d. - n.d. -42.0 n.d. n.d. n.d. 6 2.50 0.06 n.d. n.d. 0.02 n.d. -45.4 n.d. n.d. 5.4 7 2.05 0.01 n.d. n.d. 0.00 n.d. -43.7 n.d. n.d. n.d. 8 2.18 0.07 n.d. n.d. 0.03 n.d. -42.4 n.d. n.d. 6.7

Average 2.79 0.07 - - 0.02 - -42.1 - - 3.4 S.D. 1.68 0.04 - - 0.01 - 2.1 - - 3.1

Swaziland Supergroup, South Africa 3.25 Ga Sheba Formation, Fig Tree Group (graywackes) 1 0.16 b.d. n.d. n.d. - n.d. -25.0 n.d. n.d. n.d. 2 0.52 0.01 n.d. n.d. 0.02 n.d. -22.6 -0.2 n.d. n.d. 3 0.43 0.02 n.d. n.d. 0.03 n.d. -24.4 -0.2 n.d. n.d. 4 0.21 b.d. n.d. n.d. - n.d. -25.7 n.d. n.d. n.d. 5 0.39 0.01 n.d. n.d. 0.03 n.d. -24.5 0.7 n.d. n.d. 6 0.12 b.d. n.d. n.d. - n.d. -25.5 n.d. n.d. n.d. 7 0.17 b.d. n.d. n.d. - n.d. -26.2 n.d. n.d. n.d. 8 0.16 b.d. n.d. n.d. - n.d. -26.2 n.d. n.d. n.d. 9 0.17 b.d. n.d. n.d. - n.d. -26.0 1.6 n.d. n.d. 10 0.17 b.d. n.d. n.d. - n.d. -25.7 n.d. n.d. n.d. 11 0.28 0.01 n.d. n.d. 0.04 n.d. -24.8 n.d. n.d. n.d. 12 0.29 0.02 n.d. n.d. 0.05 n.d. -24.4 2.7 n.d. n.d.

Average 0.26 0.01 - - 0.03 - -25.1 0.9 - - S.D. 0.13 0.00 - - 0.01 - 1.0 1.3 - - 303 Table 4-2. (continued)

1 2 3 4 5 6 13 7 15 7 15 8 15 Samples Corg Ntot Norg Nclay Ntot/Corg Mo δ Corg δ Norg δ Nclay δ Ntotal [wt.%] [wt.%] [wt.%] [wt.%] [ppm] [‰] [‰] [‰] [‰]

Swaziland Supergroup, South Africa 3.25 Ga Sheba Formation, Fig Tree Group (black shales) 1 0.68 0.02 n.d. n.d. 0.02 0.7 -29.3 n.d. n.d. 3.0 2 0.94 0.03 0.01 0.02 0.03 0.8 -29.9 0.4 4.6 3.8 3 0.43 0.01 n.d. n.d. 0.02 0.5 -27.6 n.d. n.d. n.d. 4 0.94 0.03 0.01 0.02 0.02 b.d. -28.6 0.4 4.8 3.9 5 1.43 0.04 n.d. n.d. 0.03 1.2 -27.7 n.d. n.d. 4.9 6 1.02 0.02 0.00 0.02 0.02 0.6 -29.2 0.7 4.7 4.4 7 2.97 0.08 n.d. n.d. 0.02 1.8 -29.0 n.d. n.d. 4.0 8 1.29 0.03 0.01 0.02 0.02 3.2 -29.9 0.6 4.7 3.7 9 0.46 0.01 n.d. n.d. 0.02 1.2 -29.2 n.d. n.d. n.d. 10 1.87 0.10 n.d. n.d. 0.04 2.0 -27.6 n.d. n.d. 3.3 11 0.22 0.09 n.d. n.d. 0.36 0.2 -27.4 n.d. n.d. 2.8 12 1.31 0.04 n.d. n.d. 0.03 1.5 -31.2 n.d. n.d. 2.9 13 2.28 0.10 0.02 0.08 0.04 6.6 -29.3 n.d. n.d. 3.8 14 1.08 0.03 0.00 0.03 0.03 2.1 -28.6 n.d. n.d. 2.5 15 0.49 0.01 0.00 0.01 0.02 0.3 -28.1 -0.2 3.7 2.7 16 0.93 0.03 0.00 0.03 0.03 0.9 -28.6 n.d. n.d. 2.9 17 0.42 0.01 0.00 0.01 0.02 0.5 -27.5 -0.8 3.3 2.9

Average 1.10 0.04 0.01 0.03 0.05 1.5 -28.7 0.2 4.3 3.4 S.D. 0.72 0.03 0.00 0.02 0.08 1.6 1.1 0.6 0.6 0.7

S.D.: Standard deviation (1s)

1: Measured on decarbonated samples (converted to bulk wt.%) 2: Measured on bulk samples 3: Calculated from measured value on kerogen (converted to bulk wt.%)

4: Calculated: Nclay = Ntotal - Norg 5: Atomic ratios 6: Measured on bulk samples 7: Measured on kerogen 8: Calculated using mass balance (see text for explanation) n.d.: not determined b.d.: below detection 304 Table 4-3. Comparison of the organic C and N contents, N to organic C ratios, and C isotopic compositions of organic C among the Archean, Paleoproterozoic, and Phanerozoic shales.

Samples Number Corg [wt.%] N [wt.%] (Fm: Formation) of samples Range mean ± 1σ Range mean ± 1σ

Average shale - - 0.2 - - Average black shale - - 3.2 - -

Cretaceous black shales Organic-carbon rich 6 2.95 ~ 19.00 11.34 ± 6.69 0.12 ~ 0.82 0.50 ± 0.29 Organic-carbon poor 5 1.58 ~ 2.74 1.96 ± 0.48 0.07 ~ 0.11 0.08 ± 0.02

Jurassic shale (inhospitable) 6 2.04 ~ 6.59 4.19 ± 2.22 0.12 ~ 0.18 0.15 ± 0.03 Jurassic shale (restricted) 3 1.47 ~ 1.78 1.62 ± 0.16 0.11 ~ 0.12 0.12 ± 0.01 Jurassic shale (aerobic) 3 0.47 ~ 0.81 0.68 ± 0.18 0.03 ~ 0.07 0.05 ± 0.02 Devonian shales (laminated) 48 2.09 ~ 12.80 8.22 ± 2.84 0.14 ~ 0.50 0.34 ± 0.09 Devonian shales (bioturbated) 28 0.03 ~ 2.05 0.52 ± 0.49 0.02 ~ 0.17 0.10 ± 0.03

Pennsylvanian shale (laminated) 10 2.01 ~ 19.51 10.34 ± 6.74 0.16 ~ 0.96 0.53 ± 0.30 Pennsylvanian shale (bioturbated) 7 0.93 ~ 1.44 1.23 ± 0.19 0.12 ~ 0.14 0.13 ± 0.01

Transvaal Supergroup Silverton Fm 14 0.07 ~ 0.90 0.26 ± 0.26 0.01 ~ 0.09 0.03 ± 0.03 Strubenkop Fm 4 0.07 ~ 0.20 0.12 ± 0.06 0.02 ~ 0.05 0.03 ± 0.01 Timeball Hill Fm 12 0.30 ~ 0.50 0.41 ± 0.06 0.03 ~ 0.06 0.04 ± 0.01 Oak Tree Fm 3 1.64 ~ 2.30 1.87 ± 0.38 0.01 ~ 0.02 0.01 ± 0.00 Black Reef Fm 12 0.08 ~ 1.90 1.39 ± 0.50 0.00 ~ 0.02 0.01 ± 0.00

Mt. Bruce Supergroup Wittenoom Dolomite Fm 3 0.74 ~ 2.78 2.06 ± 1.15 0.00 ~ 0.05 0.02 ± 0.03 Carawine Dolomite Fm 6 1.02 ~ 4.45 2.66± 1.45 0.00 ~ 0.06 0.02 ± 0.03 Marra Mamba Iron Fm 4 0.19 ~ 0.93 0.56 ± 0.34 0.01 ~ 0.05 0.02 ± 0.02 Lewin Shale Fm 8 1.56 ~ 6.83 2.79 ± 1.68 0.00 ~ 0.12 0.05 ± 0.04

Ventersdorp Supergroup Bothaville Fm 3 0.16 ~ 0.25 0.20 ± 0.05 0.00 0.00

Witwatersrand Supergroup K8 Fm 16 0.07 ~ 0.23 0.14 ± 0.05 0.00 ~ 0.05 0.02 ± 0.01 Booysens Fm 17 0.07 ~ 0.29 0.16 ± 0.08 0.00 ~ 0.01 0.01 ± 0.00 Roodepoort Fm 4 0.26 ~ 0.36 0.32 ± 0.04 0.01 ~ 0.01 0.01 ± 0.00

Swaziland Supergroup Sheba Fm (graywackes) 12 0.12 ~ 0.52 0.26 ± 0.13 0.00 ~ 0.02 0.01 ± 0.01 Sheba Fm (shales) 17 0.22 ~ 2.97 1.10 ± 0.72 0.01 ~ 0.10 0.04 ± 0.03 305

Table 4-3. (continued)

13 Samples Ntotal/Corg δ Corg [‰] References (Fm: Formation) atomic ratio Range mean ± 1σ

Average shale - - - Wedepohl (1991) Average black shale - - - Vine and Tourtelot (1970)

Cretaceous black shales Organic-carbon rich 0.038 -26.3 ~ -27.6 -27.1 ± 0.4 Rau et al . (1987) Organic-carbon poor 0.036 -24.4 ~ -25.3 -25.0 ± 0.4 Rau et al . (1987)

Jurassic shale (inhospitable) 0.037 - - Ingall et al . (1993) Jurassic shale (restricted) 0.062 - - Ingall et al . (1993) Jurassic shale (aerobic) 0.065 - - Ingall et al . (1993) Devonian shales (laminated) 0.038 - - Ingall et al . (1993) Devonian shales (bioturbated) 0.395 - - Ingall et al . (1993)

Penn. shale (laminated) 0.049 - - Ingall et al . (1993) Penn. shale (bioturbated) 0.095 - - Ingall et al. (1993)

Transvaal Supergroup Silverton Fm 0.099 -29.1 ~ -32.9 -30.7 ± 2.0 Watanabe et al . (1997) Strubenkop Fm 0.214 -24.3 ~ -25.9 -25.3 ± 0.9 Watanabe et al. (1997) Timeball Hill Fm 0.094 -27.5 ~ -29.4 -28.5 ± 0.6 this study Oak Tree Fm 0.006 -33.1 ~ -34.1 -33.5 ± 0.5 this study Black Reef Fm 0.006 -29.7 ~ -35.5 -32.8 ± 2.3 Watanabe et al . (1997)

Mt. Bruce Supergroup Wittenoom Dolomite Fm 0.010 -32.0 ~ -40.4 -37.4 ± 4.5 this study Carawine Dolomite Fm 0.011 -41.2 ~ -48.9 -45.0 ± 2.9 this study Marra Mamba Iron Fm 0.022 -37.7 ~ -43.3 -40.4 ± 2.8 this study Lewin Shale Fm 0.022 -39.1 ~ -45.4 -42.1 ± 2.1 this study

Ventersdorp Supergroup Bothaville Fm 0.000 -30.1 ~ -30.6 -30.3 ± 0.4 Watanabe et al . (1997)

Witwatersrand Supergroup K8 Fm 0.122 -25.0 ~ -32.7 -28.8 ± 2.8 Watanabe et al. (1997) Booysens Fm 0.054 -26.2 ~ -34.5 -29.4 ± 3.2 Watanabe et al . (1997) Roodepoort Fm 0.027 -29.0 ~ -31.5 -30.3 ± 1.8 Watanabe et al . (1997)

Swaziland Supergroup Sheba Fm (graywackes) 0.034 -22.6 ~ -26.2 -25.1 ± 1.0 this study Sheba Fm (shales) 0.046 -27.4 ~ -31.2 -28.7 ± 1.1 this study 306 Table 4-4. Comparison of total N, organic N, and clay N contents among the modern sediments, Phanerozoic shales, and Precambrian shales.

Location / Formation Age Samples Number Analytical Ref. of samples method

Modern sediments Moroccan Coast, E. subtropical Atlantic modern 30cm core 32 KOBr, HCl 1 Central Arctic Ocean, N. Atlantic modern Surf. Sed, POM 18 KOBr, KOH 2 Yermak Plateau, Arctic Ocean, N. Atlantic modern Surf. Sed, POM 19 KOBr, KOH 2 Shikoku Basin, NW Pacific modern 300m core 20 Kjeldalh 3 Parece Vela Basin, NW Pacific modern 300m core 10 Kjeldalh 3

Nakaumi Lagoon, Japan modern 20m core 80 HCl, KCl, H2O2 4

Phanerozoic shales Monterey Fm, California Miocene shale 40 Kjeldahl 5 Chugoku Belt, Japan Permian metasediment 1 Kjeldahl 6 Exshaw Fm, Alberta Devonian shale 14 Not Available 7

Proterozoic shales (Transvaal SG, South Africa) Silverton Fm, Pretoria Gp 2.2 Ga shale 1 HCl-HF 8 Timeball Hill Fm, Pretoria Gp 2.3 Ga shale 4 HCl-HF 8 Oak Tree Fm, Chuniespoort Gp 2.56 Ga shale 2 HCl-HF 9 Black Reef Fm, Wolkberg Gp 2.6 Ga shale 2 HCl-HF 8

Archean shales Kavirondian System, Kenya 2.5 ~ 2.8 Ga shale 3 Kjeldahl 6 K8 Fm, Witwatersrand SG, South Africa 2.7 Ga shale 1 HCl-HF 8 Jeerinah Fm, Fortescue G, Mt.Bruce SG, Aus. 2.69 Ga shale 1 HCl-HF 9 Sheba Fm, Fig Tree G, Swaziland SG, S.Afr. 3.25 Ga shale 9 HCl-HF 9

References 1: Freudenthal et al . (2001), 2: Schubert and Calvert (2001), 3: Schorno (1980), 4: Sampei and Matsumoto (2001), 5: Compton et al . (1992), 6: Itihara and Honma (1979), 7: Caplan and Bustin (1996), 8: Watanabe et al . (1997), 9: This study Fm: Formation; Gp: Group; SG: Supergroup

307

Table 4-4. (continued)

Location / Formation Ntotal Norg Nclay Norg [wt.%] [wt.%] [wt.%] Ntotal

Modern sediments Moroccan Coast 0.03 ~ 0.11 0.02 ~ 0.10 0.01 ~ 0.02 0.67 ~ 0.90 Central Arctic Ocean 0.05 ~ 0.18 0.03 ~ 0.13 0.03 ~ 0.07 0.45 ~ 0.72 Yermak Plateau 0.05 ~ 0.22 0.02 ~ 0.18 0.02 ~ 0.06 0.40 ~ 0.81 Shikoku Basin < 0.00 ~ 0.07 0.02 ~ 0.07 < 0.00 ~ 0.01 0.71 ~ 0.97 Parece Vela Basin < 0.00 ~ 0.01 < 0.00 < 0.00 0.19 ~ 0.82 Nakaumi Lagoon 0.05 ~ 0.20 0.01 ~ 0.16 0.02 ~ 0.08 0.10 ~ 0.80

Phanerozoic shales Monterey Fm 0.03 ~ 0.73 0.03 ~ 0.73 < 0.00 ~ 0.21 0.26 ~ 1.00 Chugoku Belt 0.04 < 0.00 0.04 0.07 Exshaw Fm 0.18 ~ 0.68 0.09 ~ 0.58 0.09 ~ 0.10 0.50 ~ 0.85

Proterozoic shales Silverton Fm, Pretoria Gp 0.05 < 0.00 0.05 0.01 Timeball Hill Fm, Pretoria Gp 0.05 0.00 ~ 0.01 0.02 ~ 0.06 0.07 ~ 0.19 Oak Tree Fm, Chuniespoort Gp 0.01 ~ 0.02 < 0.00 0.01 ~ 0.02 0.04 ~ 0.12 Black Reef Fm, Wolkberg Gp 0.01 ~ 0.02 < 0.00 0.01 ~ 0.02 0.16 ~ 0.20

Archean shales Kavirondian System, Kenya < 0.00 < 0.00 < 0.00 0.04 ~ 0.05 K8 Fm 0.01 < 0.00 0.01 0.01 Jeerinah Fm 0.02 < 0.00 0.02 0.06 Sheba Fm 0.01 ~ 0.10 < 0.00 ~ 0.02 0.01 ~ 0.08 0.07 ~ 0.26 Chapter 5

Geochemistry of Archean–Paleoproterozoic black shales: IV. Geochemical cycle of uranium

Abstract

The geochemical cycle of U in Archean–Paleoproterozoic surface environments was influenced by the redox conditions of the atmosphere and oceans and the chemical composition and volume of the continental crust. Relationships among U, Th, Al, and organic carbon (Corg) contents in >100 drillcore samples of the Mesoarchean-Neoarchean (3.2 Ga Fig Tree Group, 2.9 Ga West Rand Group, 2.7 Ga Fortescue Group, 2.6 Ga Hamersley Group, 2.6 Ga Wolkberg Group, and 2.5 Ga Chuniespoort Group) and Paleoproterozoic (2.2 Ga Pretoria Group) black shales were investigated to constrain the Archean–Paleoproterozoic geochemical cycle of U. The continental weathering flux of U in the Archean was probably the same as today in spite of the possibility that the chemical composition of the upper continental crust was less felsic and lower in U content than that of today, because of enhanced weathering rate of

U-bearing silicate minerals under high pCO2 atmosphere. Feldspars are quantitatively important U-bearing minerals in the upper crust and uraninite, contrary to popular view, is considered to be minor for the total continental weathering flux of U. 309

Generally low U contents of Archean-Paleoproterozoic shales (< 10 ppm) with the continental weathering flux of U similar to today are mainly explained by the following: more extensive submarine hydrothermal activity at mid-oceanic ridges as a major sink for oceanic U and the lack of significant enrichment of organic matter in the shales.

The observed considerable variations in the U/Corg, U/Al, and Th/U ratios of the Archean–Paleoproterozoic shales within each formation, regardless of its depositional age and geologic setting, are similar to those found in modern sediments and Phanerozoic sedimentary rocks and likely to have resulted from the varying redox conditions of sedimentary environments.

The observed positive correlations between U and Corg contents of the Archean–Paleoproterozoic shales suggest active roles for sedimentary organic matter in the fixation of dissolved U in the oceans. The Th/U ratios of the Archean–Paleoproterozoic shales are generally lower than those of the upper continental crusts because of the sedimentary enrichments of U by organic matter in the shales. Decoupling of U and Th has been commonly recorded in the shales of this study. The observed secular increase in the Th/U ratios of the shales combined with their considerable variations are consistent with the previously published model for the effective sedimentary recycling of U. We suggest that the geochemical cycle of U in the Archean–Paleoproterozoic surface environments was essentially the same as today, and that the sedimentary recycling of U had been an important process at least since the Mesoarchean. 310

5-1. Introduction

When did the Earth’s atmosphere become oxic? The rise of atmospheric O2 has far- reaching implications for the chemical evolution of atmosphere and biosphere. However, the timing and mechanisms of the rise of atmospheric O2 (e.g. 2.2-2.0 Ga GOE: Great Oxidation Event; see Holland, 1984, 1994) have been intensely debated for decades (see Ohmoto, 1997; Holland, 1999; Phillips et al., 2001 for a summary of recent discussions). Because of a lack of direct samples of a Precambrian atmosphere, we have to use indirect approaches to constrain its redox state. Geological records such as paleosols, BIFs (banded iron-formations), uranium ores, and red beds have provided useful but inconclusive evidence for that subject (Fig. 5-1). Holland (1994) suggests the use of the U contents of shales as a proxy for the paleoredox environments. This suggestion is based on the assumptions that the chemical weathering flux of the U-bearing minerals on the continent and thus the U contents in the oceans should be higher in an oxic world than in an anoxic world, because of the increased

6+ 6+ solubility of UO2 as U -complexes with increasing pO2. U -complexes in seawater are fixed by organics in sediments. Therefore, a positive correlation is expected between U and

Corg (organic carbon) contents of shales deposited under an oxic atmosphere. Uranium in the shales accumulated in an anoxic world would be dominated by a component of detrital

minerals. Therefore, a correlation is not expected between the U and Corg contents.

However, the data from numerous studies on the U-Corg relationships in modern marine sediments and Phanerozoic shales (see examples given in section 5-2-4) do not fit into such a simple prediction. Questions remain concerning whether the trend of the U enrichment in marine black shales is really a function of the redox state of the atmosphere, and, if not, which factors control U enrichments in shales. 311

From a study of U and Th contents in fine-grained sedimentary rocks (i.e., shales, mudstones, and siltstones), McLennan and Taylor (1980) found secular trends of increasing Th/U ratios and decreasing U contents toward younger geologic age, and interpreted the trend by sedimentary recycling in an oxic world in which U and Th are decoupled (see the Background section below). Furthermore, Taylor and McLennan (1985) have recognized that the abundance of U and Th in the sedimentary rocks increased at the Archean–Paleoproterozoic boundary (~2.5 Ga), and interpreted it to indicate the response to the inferred episodic change in the chemical composition (from mafic to felsic) and the volume of the upper continental crust. Despite the usefulness of U and Th in shales as an indicator of a paleoredox environment, McLennan and Taylor’s (1980) conclusion is based on the samples with very limited number, locality, and quality. For example, they analyzed only 7 samples of highly metamorphosed rocks from Greenland to represent the early Archean samples (redistribution of U and Th may have been significant), while 21 samples of low metamorphic grade to represent the Phanerozoic sedimentary rocks. Such selections of samples may weaken the validity of their conclusions and necessitate more detailed studies of U and Th in shales, of various ages and localities, with minor metamorphic grades. In order to constrain the paleoredox surface environments of the Archean- Paleoproterozoic atmosphere and hydrosphere and the evolution of the continental crust, we have performed a geochemical investigation of 100 shale and 12 graywacke samples of Archean–Paleoproterozoic age. The samples were collected from fresh drillcores from 13 formations of the Swaziland, Witwatersrand, Ventersdorp, and Transvaal Supergroups in South Africa, and Mt. Bruce Supergroup in Australia. Their depositional ages range from

~3.3 to ~2.2 Ga. Based on the U/Al, U/Corg, and Th/U ratios and U contents of these shales and also on the Th/U ratios and U contents of igneous rocks from literature, we suggest that 312 the globally oxygenated environments with localized anoxic basins were already established in the Archean, at least by ~3.3 Ga ago.

5-2. Background

5-2-1. Uranium and Thorium in rocks and minerals Both U and Th are lithophile elements. They are strongly concentrated into the earth's crust compared to mantle materials. Typical U and Th contents of various igneous, metamorphic, and sedimentary rocks are summarized in Table 5-1. The average content of U in the upper continental crusts is 2.8 ppm while that of Th is 10.7 ppm (Table 5-1; Taylor and McLennan, 1985). Felsic rocks have much higher U and Th contents compared to mafic rocks. For example, granitic rocks have 4.8 ppm U and 21.5 ppm Th, whereas mafic and ultramafic rocks have considerably low U (0.01 ~ 1 ppm) and Th (0.05~ 4 ppm); there is a difference of one to three orders of magnitude in the U and Th contents between felsic and mafic rocks (Table 5-1). Contents of U and Th show large variations among various types of sedimentary rocks (Table 5-1). The average contents of U and Th in shales (3.2 ppm and 11.7 ppm, respectively) reflect those of the upper continental crust. Sandstones and carbonates exhibit much lower contents of U and Th compared to shales: 1.4 ppm U and 3.9 ppm Th in sandstones, and 1.9 ppm U and 1.2 ppm Th in carbonates. Typical U and Th contents of both major and minor rock-forming minerals are also presented in Table 5-1. In contrast to major rock-forming minerals, some of the minor rock- forming minerals can accommodate significant quantities of both U and Th (e.g., apatite and sphene), or either U or Th (e.g., U > Th: zircon and xenotime; U < Th: allanite and monazite) (Table 5-1). The Th/U ratio falls in a narrow range, typically between 1-6, in major rock-forming minerals, while minor minerals exhibit a much broader range (Table 5- 313

(4+) (4+, 4+, 6+) 1). The major U-bearing ore mineral is uraninite (U O2 to U 3O8), and the

major Th-bearing ore minerals are thorite (ThSiO4) and thorianite (ThO2); with significant amounts of Th also occur in monazite and zircon (Table 5-1).

5-2-2. Decoupling of U and Th Uranium and thorium are geochemical twins because of their similar ionic radius (U: 1.05 Å, and Th: 1.10 Å; Faure, 1986) and identical charge (U4+ and Th4+), and behave coherently in geological environments. However, they behave differently in aqueous environments. Under oxic conditions, U4+ is very easily oxidized to highly soluble U6+. In oxidizing waters of neutral to alkaline pH, typical of open-ocean seawater, U exists as a

6+ 3 4- 4+ soluble stable U -carbonate complex (UO2(CO )3 ). In contrast, Th is not easily oxidized and in nature occurs only in the Th4+ valence state. Th is not particularly soluble in aqueous solutions. These different properties lead to geochemical decoupling of U and Th in oxidizing environments. Decoupling of U and Th in geologic environments has been widely used in many studies dealing with Holocene to Archean sediments / sedimentary rocks to infer their paleoredox conditions of the depositional environments (see examples given below).

5-2-3. Uranium as a proxy for sedimentary redox environments

5-2-3-1. Recent sediments In present-day environments, most authigenic U precipitates in an anoxic condition at just below the sediment-water interface (SWI) where the dissolved U6+ species, transported by diffusion from the overlying water column, are reduced to highly insoluble U4+ or U5+ species (e.g., Langmuir, 1978; Kniewald and Branica, 1988; Klinkhammer and Palmer, 1991; Fisher and Wignall, 2001). Although many factors seem to control the 314

efficiency of authigenic U enrichments in sediments (e.g., the rate of sedimentation, the rate of U reduction, the rate of U replenishment to sediments, the amount of time a sediment layer spends near the SWI, and the stability of anoxic condition in the overlying water column and/or sediments, etc.), it is generally accepted that sedimentary redox condition has a major influence on the enrichment of authigenic U (e.g., Jacobs and Emerson, 1982; Barnes and Cochran, 1990; Shaw et al., 1990; Leventhal, 1993; Piper, 1994). Jones and Manning (1994) have demonstrated, using statistical approaches, that U/Th ratios and authigenic U content of sediments are useful indicators of paleoredox sedimentary environments. Numerous studies on U as a redox proxy have been done on recent sediments of the Black Sea (Calvert, 1990; Barnes and Cochran, 1991; Colodner et al., 1995), the Cariaco Trench (Anderson, 1987; Calvert and Pedersen, 1993), the Saanich Inlet (Anderson et al., 1989; Calvert and Pedersen, 1993; Morford et al., 2001; Russell and Morford, 2001), the Baltic Sea (Skei et al., 1988; Sternbeck, 2000), the Mediterranean (Thomson et al., 1995), West Africa (Veeh et al., 1974; Calvert and Price, 1983), various sites in the Pacific, Atlantic, and Indian Oceans (Bertine and Turekian, 1973; McKee et al., 1987; Thomson et al., 1993; Piper and Isaacs, 1994; Colodner et al., 1995; Nameoff, 1996; Thomson et al., 1996; Morford and Emerson, 1999; Zheng, 1999; Crusius and Thomson, 2000; Schaller et al., 2000; Chase et al., 2001; Thomson et al., 2001; Pailler et al., 2002), and lakes (Brown et al., 2000; Bishop et al. 2001).

5-2-3-2. Phanerozoic black shales Although relatively fewer in numbers compared to modern sediments, some studies on U have been done for Phanerozoic shales of the Miocene (Piper and Isaacs, 1995, 2001), Cretaceous (Brumsack, 1989; Arthur et al., 1990), Jurassic (Wignall, 1990; Jones and Manning, 1994), Permian (Wolkowicz, 1990), Carboniferous (Coveney et al., 1987; 315

Coveney and Glascock, 1989; Hatch and Leventhal, 1992; Fisher and Wignall, 2001), Devonian (Leventhal and Hosterman, 1982), and Cambrian (Leventhal, 1990).

5-2-3-3. Precambrian black shales The studies focused on U in Precambrian shales as a proxy for paleoenvironmental redox conditions are even more scarce. According to Holland (1994), there are only three

Archean Corg-rich shales have been studied for enrichments of redox-sensitive metals (RSM), including U. The first example is the 2.7 Ga low metamorphic-grade Jeerinah Formation in Western Australia (Davy and Hickman, 1988). The second example is the 2.5 Ga low metamorphic-grade Mt. McRae shale, again in Western Australia (Davy, 1983). The third example is the ~2.0 Ga highly metamorphosed (mineralized ore) in Outokumpu, Finland (Loukola-Ruskeeniemi, 1991, 1999). Meyer and Robb (1996) studied the 2.5-2.6 Ga old black shales of the Chuniespoort Group, South Africa, and found widely ranging U contents from 0.2 to 9.3 ppm. Meyer and Robb (1996), however, did not pay particular attention to U in their discussion. A serious lack of study on U for Precambrian black shales becomes apparent, in spite of its great potential as a proxy for the paleoredox environments.

5-3. Geological settings and samples

All of the samples used in this study are from unweathered, low metamorphic grade drillcores. They belong to the Swaziland Supergroup, the Witwatersrand Supergroup, the Ventersdorp Supergroup, and the Transvaal Supergroup in South Africa and the Mt. Bruce Supergroup in Australia. The South African samples include the ~3.3 Ga Fig Tree Group (the 3.25 Ga Sheba Formation), the ~3.0 Ga West Rand Group (the 2.96 Ga Parktown Formation), the ~2.7 Ga Platberg Group (the 2.71 Ga Rietgat Formation), the ~2.6 Ga 316

Wolkberg Group the (2.64 Ga Black Reef Formation), the ~2.6 Ga Chuniespoort Group (the 2.56 Ga Oak Tree Formation), the ~2.2 Ga Pretoria Group (the 2.22 Ga Timeball Hill Formation), and ~2.2 Ga Olifantshoek Group (the ~2.2 Ga Mapedi Formation). The Australian samples include the ~2.7 Ga Fortescue Group (the 2.69 Ga Jeerinah Formation and the 2.69 Ga Lewin Shale Formation) and the ~2.6 Ga Hamersley Group (the >2.60 Ga Marra Mamba Iron Formation, the 2.60 Ga Wittenoom Dolomite Formation, and the 2.60 Ga Carawine Dolomite Formation). More information on stratigraphy, locality, ages, geological / tectonic settings, and the number of the samples are summarized in Table 5-2.

5-4. Analytical methods

The analytical methods utilized in this study for the determination of the Al and Corg contents are already presented in the chapter 2. The analytical methods for U and Th are the same as those for Mo described in the chapter 3. Their described descriptions are also presented in the Appendix C. The detection limit for U and Th abundance is 0.1 ppm, and the reproducibility was better than ± 5 % of the samples.

5-5. Results

5-5-1. Contents of U, Th, Al, organic C, and carbonate C.

Concentrations of U, Th, Al2O3, Corg, and Ccarb in the samples of this study are summarized in Table 5-3, and relationships of U versus Corg and U versus Al2O3 are shown in Fig. 5-2 (-a through -p). The total ranges of the U, Th, Al2O3, Corg, and Ccarb contents are 0.1 ~ 8.0, 0.5 ~ 31.6, 0.8 ~ 29.4, 0.1 ~ 12.0, and 0 ~ 12.0 wt.%, respectively. 317

The results of Al2O3, Corg, and Ccarb elemental analyses are already described in the chapter 2 (see the sections 2-4-5 and 2-4-7).

For the Ccarb-rich (and/or Al2O3-poor) samples, U and Th contents are generally

low, typically less than 2 ppm. Samples relatively rich in Corg are not always rich in U;

however, positive correlations between U and Corg are found in shales of the 3.25 Ga Sheba Formation (Fig. 5-2-a), the 2.71 Ga Rietgat Formation (Fig. 5-2-g), the 2.60 Ga Carawine Dolomite Formation (Fig. 5-2-k), and the 2.22 Ga Timeball Hill Formation (Fig. 5-2-o).

The slopes of the linear correlation lines between the U and Corg contents (or the U/Corg ratios when the positive linear correlations exist) for the studied samples are rarely exceed

that of the average shale (U/Corg = 15; Wedepohl , 1971, 1991, Taylor and McLennan,

1985). Corg-poor (< 0.5 wt.%) samples, such as the 3.25 Ga Sheba Formation (graywackes, Fig. 5-2-c), the 2.96 Ga Parktown Formation (Fig. 5-2-e), and the 2.72 Ga Pillingini Tuff

Formation (Fig. 5-2-i), exhibit U-Corg distribution close to the slope of the average shale, which is also poor in Corg (0.2 wt.%). Red shale samples show their U-Corg-Al2O3 contents clustered around the those of the average shale (Fig. 5-6-o). Positive correlations

between the U and Al2O3 contents are often observed for the samples of this study; however, the slopes of the correlation lines vary from ~0.1 to ~0.3 , depending on each set of samples.

5-5-2. Corrections for the U and Th contents with ages Both U and Th are radioactive elements and decay with time. Therefore, the original contents of U and Th in sedimentary rocks at the time of deposition must have been higher than the measured values. In order to compare the original contents of U and Th in rocks of different ages, we used the following two simple equations for age-correction :

(0.155125 x t)...... Ucorr = Umeas x e (Eq. 5-1) 318

(0.049475 x t) ...... Thcorr = Thmeas x e (Eq. 5-2) where the subscript "corr" and "meas" designate age-corrected contents and measured contents, respectively, and "t" designates the age of a sample in the unit of Ga. Only major isotopes of 238U (> 99 %) and 232Th (~100 %) are considered, and their decay constants are adapted from Faure (1986) (original references therein). Age-corrected concentrations of U and Th for each sample are not shown in the Table 5-3; however, they are collectively used in Tables 5-4, 5-5, and 5-6 and Figures 5-4, 5-6, and 5-14.

5-5-3. Uranium to organic C ratios Carbonaceous sediments are often the sink of RSM, and an association of U with

Corg has been widely observed in modern Corg-rich sediments. To examine the degree of U

enrichment with respect to Corg contents in the samples of this study, their U/Corg ratios are

calculated and normalized to the U/Corg wt. (ppm / wt.%) ratios of the average shale (15;

Wedepohl, 1971, 1991; Taylor and McLennan, 1985). The [(U/Corg)sample/(U/Corg)Av. Sh.] ratios are presented in the histograms (Fig. 5-3). The Ucorr/Corg ratios are presented in Fig. 5-4.

The samples from the same formation exhibit a large variation in the U/Corg ratios, in many cases by a factor of 10 or more. For example, the shales samples from the 3.25 Ga Sheba Formation (Fig. 5-3-n) and those of the 2.96 Ga Parktown Formation (Fig. 5-3-l) exhibit about 2 orders of magnitude variation in the U/Corg ratios.

A significant enrichment of U relative to Corg contents (i.e., U[ppm] / Corg [wt.%] > 100) is not observed among the samples of this study. The carbonaceous samples of 3.25 -

2.56 Ga (Fig. 5-3-b, -c, -d, -e, -g, -h, -i, -j, -n) display significantly smaller Ucorr/Corg ratios,

typically 1/10 and sometimes 1/100 of the U/Corg ratio of the average shale (Fig. 5-3-i and

5-3-n). The Corg-poor dark-colored samples, such as the 3.25 Ga Sheba Formation 319

(graywacke; Fig. 5-3-m), the 2.96 Ga Parktown Formation (Fig. 5-3-l), the 2.72 Ga Pillingini Tuff Formation (Fig. 5-3-k), and the >2.60 Ga Marra Mamba Iron Formation

(Fig. 5-3-f), exhibit the Ucorr/Corg ratios similar to or slightly higher than the U/Corg ratio of the average shale. The ~2.2 Ga red shales show their average shale-normalized Ucorr/Corg

ratios greater than 2 (Fig. 5-3-a). This is due to their very low content of Corg (0.07 wt.% in average; Table 5-3) with modest content of U (2.2 ppm in average; Table 5-3) and not due to significant U enrichment.

The observed within-formation variations in the U/Corg ratios and the very low

U/Corg ratios compared to the U/Corg ratio of the average shale (15; Wedepohl, 1971, 1991;

Taylor and McLennan, 1985) are typically found in the modern Corg-rich sediments (e.g., Leventhal et al., 1983; Brumsack, 1989; Warning and Brumsack, 2000) and Phanerozoic

Corg-rich sedimentary rocks (e.g., Eocene Black Shales: Desborough et al., 1976; Pennsylvanian black shales: Coveney et al., 1987; Devonian black shales: Holland, 1984; Leventhal, 1993; Cambrian black shale: Armands, 1972; Leventhal, 1990) (Fig. 5-3).

5-5-4. Uranium to aluminum ratios Al and Ti-bearing minerals are the basic building blocks of crustal rocks. In sediments, Al resides in an aluminosilicate fraction (i.e., clays), and is commonly used as an indicator of detrital materials. Metal contents of sediments are often normalized to the Al content in order to examine the metal enrichments relative to detrital fractions. For this purpose, U/Al wt. (ppm / wt.%) ratios of the samples are used in this study. To compare the U/Al ratios of the Archean–Paleoproterozoic samples with the U/Al ratio of the average shale (0.33; Wedepohl, 1971, 1991; Taylor and McLennan, 1985), the U/Al ratios of the Archean–Paleoproterozoic samples are normalized to the U/Al ratios of the average shale

and presented in the Fig. 5-5 (U not age-corrected). The Ucorr/Al ratios are presented in Fig. 5-6 against ages. 320

Similar to the U/Corg ratios, significant variations in the U/Al ratios are observed among the carbonaceous samples from the same formations, although the magnitude of

variation is smaller than that for the U/Corg ratios. One sample from the 2.64 Ga Black Reef Formation displays significantly depleted U/Al ratios (Fig. 5-5-g). This is because the

sample has a high Al2O3 content (26.57 wt.%; whereas the average shale's Al2O3 content is 18 ~ 19 wt.%: Wedepohl, 1971; Taylor and McLennan, 1985) with modest U content (2.9 ppm). The Ucorr/Al ratios of 2.2 Ga red shale samples narrowly cluster around the U/Al ratio of the average shale. The degree of U enrichment relative to Al (terrigenous fraction) appears to be typically a factor of 2 to 3 and up to a factor of 5 (the >2.60 Ga Marra Mamba Iron Formation; Fig. 5-5-f) compared to the average shale. While nearly a 2 log unit depletion of

U relative to Corg is observed (see previous section), only minor depletions of U relative to Al is found (up to a factor of ~0.5 in the 3.25 Ga siderite-rich shale in the Sheba Formation; except for the above-mentioned one sample in the 2.64 Ga Black Reef Formation). There is no notable secular trend for the U/Al ratios of the samples of this study. The observed within-formation variations in the U/Al ratios are typically found in

modern Corg-rich sediments (e.g., Skei, 1980; Brumsack, 1989; Calvert, 1990; Calvert and Pedersen, 1993; Nameoff, 1996; Thomson et al., 1996; Morford and Emerson, 1999; Warning and Brumsack, 2000; Brown et al., 2000; Russell and Morford, 2001) and

Phanerozoic Corg-rich sedimentary rocks (e.g., Cretaceous black shales: Arthur et al., 1990; Devonian black shales: Leventhal, 1993).

5-5-5. Thorium to uranium ratios The Th/U wt. ratios of the studied samples are compared to the Th/U ratio of the average shale (3.8; McLennan and Taylor, 1985; Wedepohl, 1971, 1991) to examine the 321

decoupling of U and Th (see the section 5-2-2). The Th/U ratios for the samples of this study are presented in the histograms in Fig. 5-7. The distribution of Th/U ratios of the samples of this study is highly variable. The Th/U ratios of the carbonaceous samples of this study range from ~0 to ~7 (Fig. 5-7). In contrast to the carbonaceous samples, the Th/U ratios of the red shales from the ~2.2 Ga Mapedi Formation (~8) are substantially higher than that of the average shale.

5-6. Discussion

Based on the U/Corg, U/Al, and Th/U ratios of the Archean–Paleoproterozoic shales obtained in this study, we will examine the validity of various models proposed by previous researchers on the chemical evolution of the atmosphere, oceans, and continental crust. We first examine the possible causes for the observed variations and correlations among those ratios, then we examine the geochemical cycle of U emphasizing the similarities and differences between modern and Archean, and we provide possible explanations for generally low U contents in Archean–Paleoproterozoic shales. Lastly, we discuss the importance of sedimentary recycling of U and of the decoupling of U and Th in understanding the evolution of the atmosphere, oceans, and continental crust.

5-6-1. Post-depositional effects on the fractionation of the U/Corg, U/Al, and Th/U ratios

The U/Corg, U/Al, and Th/U ratios are used here to evaluate different aspects of sedimentary U enrichment / depletion during geological processes. The results of this study show that there are substantial variabilities in the U/Corg, U/Al, and Th/U ratios of the Archean–Paleoproterozoic black shales. Such variabilities in shales could have been created by one or more reasons including (1) metamorphic effects, (2) change in the source rock 322

compositions, and (3) early diagenetic effects and/or changes in the depositional environments in terms of redox conditions. Among these possibilities, we suggest that (3) is

the major process to cause fractionations of the U/Corg, U/Al, and Th/U ratios of the studied samples for the following reasons.

5-6-1-1. Metamorphic effects High-grade metamorphic rocks (e.g., granulite facies) typically exhibit loss of U and Th (e.g., McLennan and Taylor, 1980). However, McLennan and Taylor (1980) have suggested that U and Th (probably Al, too) are essentially immobile during metamorphism at grades below the amphibolite facies. Thermal decomposition of OM during metamorphism may release U from OM, and the released U may be lost from the system or

fixed in diagenetic minerals. These processes may lead to fractionations of the U/Corg, U/Al,

and Th/U ratios; the U loss will decrease the initial U/Corg and U/Al ratios and increase the initial Th/U ratios (Fig. 5-11). Such metamorphic modification of the U/Corg, U/Al, and Th/U ratios will be effective over all the samples in the individual sequence (formation).

However, the observed wide variations in the U/Corg, U/Al, and Th/U ratios for the Archean–Paleoproterozoic shales of this study are not supported by such nearly homogeneous metamorphic effects. Therefore, the variable U/Corg, U/Al, and Th/U ratios of the studied samples are not likely to have been created by minor grade of metamorphism (equal or lower than the greenschist facies) from which the samples suffered.

5-6-1-2. Source rocks effects If the detrital transport of physically weathered materials on the continents would dominate over that solution transport of chemically weathered materials, then the Th/U and U/Al ratios of sediments would inherit those of the source materials. The variations in the Th/U and U/Al ratios of sediments are created by variations in the proportion of two or 323 more different types of source rocks, such as metamorphic rocks and granitic rocks (Table 5-1). However, shales are well-mixed, well-sorted, and fine-grained terrigenous sedimentary rocks. To explain the observed large fractionations of the Th/U and U/Al ratios of the Archean–Paleoproterozoic shales solely by this mixing scenario, repeated and rapid changes are required in the mixing ratio of the different source rocks for the shales in each formation of this study. Such case is considered here to be unlikely.

5-6-1-3. Early diagenetic effects and/or changes in the redox depositional environments Early diagenesis may have caused redistribution of some elements, especially of the

RSM such as U and Mo, and thus fractionation of the U/Corg, U/Al, and Th/U ratios in the shales of this study. In modern marine sediments where anoxic conditions develop due to the abundance of Corg, a fluctuating redox boundary within sediments and/or overlying bottom water (e.g., by seasonal change and climatic change) would cause redistribution, by repeated dissolution and precipitation, of the RSM that were once fixed in unconsolidated sediments. Such diagenetic redistribution of the RSM in sediments occurs in ~mm to ~cm scale, as revealed by numerous studies of modern marine sediments (e.g., in the Black Sea, Cariaco Trench, Saanich Inlet, Framvaren Fjord, off shore California, etc. see examples listed in the section 5-2-4). In a sedimentary environment, the coexistence of both oxic and anoxic conditions is necessary for the redistribution of the RSM in sediments. Either of them alone is not capable of redistributing RSM. The size of drillcore samples of this study is typically several cm in length perpendicular to bedding plane. This is large enough to average out the possible diagenetic enrichment and/or redistribution of the RSM in ~mm to ~cm scale before consolidation / compaction. Therefore, the U data obtained for each sample of this study probably represent a temporal record where possible diagenetic enrichment and/or remobilization of the RSM is time-integrated. 324

5-6-2. Geochemical cycle of U Uranium-bearing minerals in the continental crust are subject to the chemical and physical weathering on the continents and transported by rivers to the oceans. The present- day oceanic concentration of U is 3.2 ppb (Chen et al., 1986; Cochran, 1992). The depth profile of U in oxygenated seawater is uniform because of a conservative biogeochemical behavior of U due to the formation of stable and soluble U6+-carbonate complex (e.g., Ku et al., 1977; Langmuir, 1978; Cochran, 1992; see section 5-2-2). The residence time of dissolved U in the modern oceans is estimated to be 0.25 to 0.50 Ma (Cochran, 1992; Morford and Emerson, 1999). Dissolved U in the ocean is precipitated on the seafloor or fixed into mid-oceanic ridges (MOR) systems through hydrothermal circulation of seawater (e.g., Klinkhammer and Palmer, 1991; Cochran, 1992). The sources and sinks of U in the modern oceans are summarized in Table 5-7. It has been demonstrated that the can enrich U up to a factor of 40 compared to its original U content, and a half to two-thirds of U delivered to the oceans are fixed in oceanic basalt through hydrothermal alteration or submarine weathering (Table 5-7; McLennan and Taylor, 1980). In the modern oceans, the sediment sink flux is ~0.3 x 108 mol/yr and the MOR sink flux is ~0.2 x 108 mol/yr (Table 5-7). The geochemical cycle of U could have been much different in the distant past if the

Earth's surface environments was anoxic (i.e., before the rise of pO2 at ~4 Ga or ~2 Ga). This is because U is a RSM and its geochemical behavior depends on the surrounding redox conditions. However, it is not clear if such a difference existed in the geochemical cycle of U between an oxic world and an anoxic world. Therefore each process in the course of the U geochemical cycling needs to be examined. In this section, we critically evaluate (1) the source fluxes of U from the continental weathering under different redox 325 environments and (2) the sink fluxes of U into sediments and MOR under different thermal regime between the cooler modern environments and the warmer Archean environments.

5-6-2-1. Source of U: weathering of U-bearing minerals As briefly reviewed in the preceding background section (5-2-1), U resides in a variety of minerals in the continental crust and the U concentration has considerable variations among the various types of U-bearing minerals and rock types (Table 5-1). Figure 5-8 shows the U contents of major U-bearing minerals (feldspars and zircons) in the three major types of igneous rocks; basalt, andesite, and granite. The dependence on the pO2 level of the dissolution rates of U-bearing minerals such as feldspar and uraninite are critically important to estimate the continental weathering flux of U at a given pO2 condition.

Uraninite dissolution rate

The dissolution rate of uraninite (UO2) has been investigated by many researchers (e.g., Grandstaff, 1976; Ono, 2001), and shown to be redox-dependent; i.e., slow in reducing conditions and fast in oxic conditions. This observation has been used to argue for the reducing Archean–Paleoproterozoic environments because of the presence of inferred detrital uraninite in the quartz-pebble conglomerates (e.g., Witwatersrand deposits, Elliot Lake deposits: see chapter 1 for review). However, U-bearing silicate minerals (mainly feldspars) are more abundant than uraninite in terms of the total U budget in the igneous rocks (mainly granite). Therefore, the dissolution of U-bearing feldspars seems more important for the continental weathering flux of U rather than that of uraninite.

To examine the solubility of UO2 and the stability fields of major U-bearing aqueous species, an Eh-pH diagram is constructed for aqueous species and solid phase of

U in a U-O2-CO2-H2O system (Fig. 5-9). To approximate conditions in the Archean, 326

- 2- higher ∑C (= CO2[aq] + HCO3 + CO3 ) content is set at 0.1 M, corresponding to ~100 to

~ 1000 times higher pCO2 in the Archean than the present-day level of 350 ppm (e.g.,

Kasting, 1987). Superimposed are the contours for various levels of pO2 levels (Fig. 5-9).

There is a considerable variations in the estimates of pO2 levels in the Archean, ranging from as low as 10-13 atm (Kasting, 1993), < 0.002 atm (Holland, 1994), <0.0002 atm (Rye and Holland, 1998), and ~0.2 atm (Ohmoto, 1997; Lasaga and Ohmoto, 2002a). In a acidic pH range of ~2.5 to ~6, the U solubility contour of 10-6 M (= ~0.2 ppm) coincides with the

-60 pO2 contour of ~10 atm. Even at such a low pO2 condition, the dominant U-bearing 6+ aqueous species is a U -complex. The dominant reaction controlling the solubility of UO2 is:

2- + UO2 + 2 CO2 + H2O + 0.5 O2 ---> UO2(CO3)2 + 2 H ...... (Eq. 5-3)

6+ According to the stoichiometry of this reaction, the solubility of UO2 as U - complexes increases by 0.5 log unit with an increase of pO2 value by 1 log unit. The -48 solubility exceeds 1 M (= 23.8 wt.% U) even at pO2 as low as ~10 atm. This implies that the aerated surface waters (i.e., river water, ocean surface water) never become saturated with

-48 UO2 even when the pO2 is as low as ~10 atm. The U content of such waters is controlled by the availability of U. The U content in water probably does not exceed ~1 ppm in most geologic conditions (c.f., Langmuir, 1997). Uranium in the surface waters may precipitate as

-60 UO2 if a conditions of pO2 < 10 atm is created by OM in sediments.

Feldspar dissolution rate

Contrary to a common belief among geologists, UO2 is not a major U-bearing mineral in normal igneous rocks. Most of U in normal igneous rocks occurs as a trace constituent of silicate minerals, most importantly in feldspars, and also as a constituent in 327

insoluble heavy minerals (e.g., zircons, monazite) (see Fig. 5-8). This suggests that the rate of U release during weathering of igneous rocks is essentially the same as the dissolution rates of feldspars. The dissolution rates of various feldspar minerals have been demonstrated by many studies conducted in field and laboratory to be quite fast (e.g., Blum

and Stillings, 1995), much faster than uraninite at any pO2 level. Actually, the dissolution rates of feldspar minerals are more a function of pH and pCO2 (increasing dissolution rates

with decreasing pH), not a function of pO2 (Blum and Stillings, 1995). Because the solubility of UO2 in the surface water is so high (see above), essentially all the U leached out from rocks during weathering would be transported by rivers and shallow groundwater to the oceans. In the Archean where pCO2 was probably much greater than today (e.g., Kasting, 1987) and thus the oceanic / rainwater pH was lower than today, the dissolution rate of feldspar minerals in the Archean continental crust was most likely much faster than today.

5-6-2-2. Sink of U: sediments vs. mid-oceanic ridges The two major sinks of oceanic U are sediments and MOR (Fig. 5-10). Uranium is incorporated into marine sediments both by (1) uptake in marine organisms and OM formed in the surface water and deposited at the seafloor and by (2) diffusion into sediments followed by reduction, removal from pore waters, and fixation (e.g., Cochran, 1992). Compared to the suboxic sediments, anoxic sediments, and metalliferous sediments , the oxic sediments play a minor role as a sink of oceanic U (Table 5-7). Another important uptake of U occurs in MOR through seawater hydrothermal circulation, where dissolved U6+ is reduced to U4+ and fixed. In the present-day oceans, the ratio of U sink flux into sediments to the U sink flux into MOR is ~1.5 (Table 5-7). The relative importance of sediments and MOR as a sink of oceanic U would have been different in the Archean compared to the present-day ocean (Fig. 5-10). As previously 328

mentioned, the heat flux in the Archean was almost certainly higher than modern, and the submarine hydrothermal activity was also more active than today. The depth of altered seafloor basalts by hydrothermal circulation was probably greater in the Archean than today. Archean seawater temperature was probably higher than modern due to the higher

heat flux and the enhanced greenhouse warming by elevated pCO2 (e.g., Kasting, 1987). These estimates suggest that the efficiency of the U uptake in MOR through hydrothermal circulation was most likely higher in the Archean than in modern. Therefore, the ratio of U flux into sediments to the U sink flux into MOR is less than 1.5 and probably less than 1 in the Archean, based on the first-order approximation that the rate of U uptake in MOR is proportional to the heat flux (Fig. 5-10). A quantitative treatment of this matter will be give in later discussion (see the section 5-6-4).

5-6-3. Sedimentary fixation of U by organic matter through geologic time

5-6-3-1. Archean-Paleoproterozoic shales

Association of U with Corg-rich sediments and black shales has been demonstrated for the Archean–Paleoproterozoic black shales of this study by their first-order positive correlations between U and Corg (Fig. 5-2). A dilution by clastic component for shales with nearly constant U/Corg ratios may explain positive correlations between U and Corg. However, such possibility is unlikely because of the non-zero intercept of regression lines

between U and Corg and the modest scattering of the data points shown in Fig. 5-2. The

positive correlations between U and Corg contents created during sedimentation and/or early diagenesis may be disturbed by post-depositional processes such as later stage diagenesis and metamorphisms (Fig. 5-11), although the metamorphic effects are considered to be minor for the samples of this study (see the section 5-6-1-1). Despite such possible disturbance, the observed first-order positive correlations between U and Corg suggest that 329

the sedimentary (syn-sedimentary and/or diagenetic) fixation of U by OM could have been

operating in the oceans of Archean–Paleoproterozoic in age. However, there are some Corg- rich samples that lack U enrichments.

5-6-3-2. Modern sediments and Phanerozoic shales

Positive correlations between U and Corg contents are typically found in the modern marine carbonaceous sediments and Phanerozoic black shales (see examples listed in the

section 5-2-4). However, the Corg-rich shales / sediments are not necessarily rich in U.

Shales with low U and low Corg contents exist throughout geologic time. Furthermore,

shales with low U and high Corg contents also exist throughout geologic time (see examples

listed in the section 5-2-4). These two types (low-U and low-Corg, and low-U and high-

Corg) are rather typical features of the modern marine fine-grained sediments, especially those deposited under oxygenated bottom water with low / high rain rate of OM. According to the data compiled by Zheng (1999) of mass accumulation rates (MAR) of authigenic U, which are proportional to the U contents of modern marine

sediments and the dissolved O2 (DO) contents of the overlying bottom water, a nearly hyperbolic relationship between U-MAR and DO is observed (Fig. 5-12). It is tempting to relate the contents of U in the sediments to those of Corg because the O2-depleted environments enhance the preservation of OM (e.g., Arthur and Schlanger, 1979). However, significant amount of OM can accumulate in sediments overlain by O2-rich bottom water when, for example, combined with a high sedimentation rate (i.e., rapid burial of OM).

Therefore, it is possible for U to be enriched in sediments overlain by O2-rich bottom water if the buried OM contains significant amounts of U.

On the other hand, sedimentary enrichment of U does not necessarily occur in Corg- rich sediments. Figure 5-12 shows that a sample from Mid-Atlantic Bight deposited under

O2-rich bottom water but has an elevated U content (corresponding to 60 ~ 70 U-MAR), 330

which is comparable to the U contents of several samples deposited under O2-depleted bottom water at 23 ˚N North Mexican Margin. A large variability in the U contents (~20 to

~130 U-MAR) is observed among the sediments overlain by O2-depleted bottom water (Fig. 5-12). It is emphasized here that the low-U sediments with variable DO contents in their overlying bottom water (i.e., variable U/Corg ratios) are indeed observed in modern sediments (Fig. 5-12). Such features of U and Corg (i.e., low U contents and variable

U/Corg ratios) are typically found in the Archean–Paleoproterozoic shales of this study, suggesting that their sedimentary environments are similar to those of the typical modern sediments.

5-6-4. Low U contents of the Archean–Paleoproterozoic black shales Although the sedimentary diagenetic processes involving redox reactions are proposed for the enrichment of U with OM, the U contents of Archean–Paleoproterozoic

shales are generally lower than that of Phanerozoic Corg-rich shales. While all the samples of this study (~3.3 to ~2.2 Ga) and the 1.9 Ga shales from Labrador (Table 5-6; Hayashi et al., 1997) have U contents less than 10 ppm, the U contents of Phanerozoic Corg-rich shales sometimes reach tens of ppm or, although rarely, over 100 ppm.

Holland (1994) has proposed that generally low U contents of the Archean Corg- rich shales are because of the reduced oceanic concentration of U due to the reduced chemical weathering of U-bearing mineral (uraninite) under an anoxic atmosphere before the inferred GOE at 2.2 ~ 2.0 Ga. Although Holland has not explicitly stated, his interpretation of the cause of low U contents of Archean shales was based on the following three important assumptions: (1) U4+-complexes, rather than U6+-complexes, become important aqueous species under a reducing atmosphere; (2) U in rivers and oceans comes mostly from weathering of UO2; and (3) the concentrations of U in the surface waters 331

(rivers and oceans) are controlled by the solubility of UO2. These three assumptions are already shown to be most likely invalid in the previous section (5-6-2-1) using the Eh-pH diagram for U-O2-CO2-H2O system (Fig. 5-9). To further examine the possible causes of the low U content of the Archean–Paleoproterozoic shales, we discuss the influence of (1) the chemical composition and the volume (area) of the upper continental crust and the pCO2 level on the weathering flux of U, (2) the extensive submarine hydrothermal activity on the uptake of U in the oceans, and (3) the OM enrichment on the accumulation of U. Then we suggest that the U geochemical cycle among the surface reservoirs has been basically the same throughout the geologic time, and that a lack of U enrichment in the Archean shales can not be used as an evidence for an anoxic atmosphere.

5-6-4-1. Continental weathering flux of U

The weathering rate of silicates (Fw, sil) depends on temperature and pH, which in turn depend on pCO2 (Berner et al., 1983; Lasaga et al., 1985; Brady, 1991), and also on the total continental area of soil formation (Acont; Lasaga and Ohmoto, 2002b). This relationship is expressed as follows:

0 0.25 ...... Fw, sil = Fw, sil • [pCO2, PAL] • Acont (Eq. 5-4)

where [pCO2, PAL] and Acont values are relative to the present values. The value of 0.25 is taken from Lasaga and Ohmoto (2002b) which agrees well with the BLAG model by Berner et al. (1983).

The higher pCO2 level in the Archean atmosphere may have increased the weathering rates of silicate minerals compared to today. The Archean pCO2 level has been estimated to be ~100 to ~ 1000 times higher than the present-day level of 350 ppm (e.g.,

Kasting, 1987). This estimation of the Archean pCO2 corresponds to a factor of ~3.2 to 332

0.25 ~5.6 for the term [pCO2, PAL] . The volume (area) of the continents in the Archean has not been well constrained, but it could have been nearly the same as today (e.g., Armstrong, 1981; Reymer and Schubert, 1984) or smaller than that of today (a factor of ~0.2 to ~0.5: e.g., Taylor and McLennan, 1985). The episodic change in the chemical composition of the upper continental crusts at around Archean–Proterozoic boundary (2.5 Ga) (e.g., Taylor and McLennan, 1985) would influence the continental weathering flux of U, because the U (and Th) contents are markedly different between mafic and felsic rocks (Table 5-1). Figure 5-8 illustrates the distinctive difference in the U contents among basalt, andesite, and granite, together with U content of major rock-forming minerals. From a simple mass balance calculation, the U contents of the average Archean upper continental crust may have been about a half of that of the Phanerozoic upper continental crust (see Table 5-1). The combination of the factors discussed above (two opposing effects) may have resulted that the continental weathering flux of U to the oceans during Archean was more or less similar to that in the Phanerozoic.

5-6-4-2. Extensive submarine hydrothermal activity As stated above, the continental weathering flux of U to the oceans during Archean may have been similar to that in the Phanerozoic:

in in ...... F [Archean] ≈ F [today] (Eq. 5-5)

in in where F [Archean] and F [today] refer to the continental weathering flux of U into the Archean ocean and that into the modern ocean, respectively. The U influx to the oceans must have been balanced by the U outflux from the ocean (Fout) throughout the geologic time: 333

in out...... F = F (Eq. 5-6) out out ...... F [Archean] = F [today] (Eq. 5-7)

Uranium in the ocean is removed by precipitation to the ocean floors (mostly OM in

out out shales), Fppt , and the other is by submarine hydrothermal process in MOR, FMOR , where U6+ in the seawater is probably reduced to U4+ in MOR (see Fig. 5-10; e.g., Cochran, 1992).

out out out out ...... Fppt + FMOR [Archean] = Fppt + FMOR [today] (Eq. 5-8)

According to Table 5-7,

out out ...... Fppt [today] = 1.5 • FMOR [today] (Eq. 5-9)

As previously stated, the submarine hydrothermal activity was probably more extensive in Archean than in modern because of the inferred higher heat flux. Thus,

out out ...... FMOR [Archean] = C1 • FMOR [today], C1 > 1 (Eq. 5-10)

where C1 is a dimension-less constant which represents an intensity of the submarine hydrothermal activity in the Archean. When the Eq. 5-8 through 5-10 are combined, the following equation can be drawn to compare outflux of U into sediments between Archean and modern:

out out ...... Fppt [Archean] = {(2.5 - C1) ÷ 1.5} • Fppt [today] (Eq. 5-11) out out ...... Fppt [Archean] < Fppt [today], when C1 > 1 (Eq. 5-12) 334

The relationship in Eq. 5-12 explains why the U contents of the Corg-rich Archean

(–Paleoproterozoic) shales are lower than those of the modern Corg-rich sediments and Phanerozoic black shales (see also Fig. 5-13). The equation 5-11 further suggests that the

Archean MOR will uptake all the dissolved U in the oceans if C1 ≥ 2.5 (Fig. 5-13). out On the first-order approximation, if FMOR is proportional to the heat flux (i.e., out activity of submarine hydrothermal activity), then FMOR [Archean] can be estimated to be out 2 ~ 3 times as high as FMOR [today] (i.e., C1 = 2 ~ 3; the heat flux in the Archean has been estimated to be 2 ~ 3 times higher than that in modern; Sclater et al., 1980; Kasting, per. comm. 2001). Therefore,

out out ...... FMOR [Archean] = 2 ~ 3 • FMOR [today] (Eq. 5-13) out out ...... Fppt [Archean] ≤ 0.00 ~ 0.33 • Fppt [today] (Eq. 5-14)

out out The Eq. 5-14 indicates that the Fppt [Archean] was less than a third of the Fppt [today] (Fig. 5-13). This relationship may explain why the Archean–Paleoproterozoic shales of this study and of Hayashi et al. (1997) have generally low U contents compared to the younger sediments that were accumulated in the relatively low temperature regime.

5-6-4-3. Lack of significant Corg enrichment The removal flux of U by OM in shales may be expressed as follows (e.g., Lasaga, 1998; Lasaga and Ohmoto, 2002a):

out ...... Fppt = k • Aorg • [U]ocean (Eq. 5-15) 335

where k is the rate constant (mass U per unit surface area of OM per time), Aorg is the total

surface of reactive OM in the global marine sediments, and [U]ocean is the average oceanic concentration of U. A combination of Eq. 5-14 and Eq. 5-15 yields the following relationship:

Aorg • [U]ocean [Archean] ≤ 0.00 ~ 0.33 • Aorg • [U]ocean [today] ...... (Eq. 5-16)

The Eq. 5-16 implies that either (1) the total amount of OM in the global marine sediments and/or (2) the oceanic concentration of U was much less in the Archean compared to today. The latter case is already discussed in the preceding section. Below we discuss the former case. But in either case, the lower U contents of the Archean shales, compared to the younger ones, was not likely related to the atmospheric pO2 level. An extreme enrichment of U (i.e., >100 ppm) in shales is typically accompanied by an extreme enrichment of Corg in shales. For example, from a study of Upper Carboniferous black shales, Fisher and Wignall (2001) have found that U enrichments of

>30 ppm are always associated with elevated Corg content of > 6 wt.% (reaching ~200 ppm

U with >30 wt.% Corg). Although extreme enrichments of U and Corg in shales are often found in the Phanerozoic shales such as the Cambrian Alum shales (Leventhal, 1990), the Pennsylvanian Mecca Quarry shales (Coveney et al., 1987), and the Cretaceous (Cenomanian–Turonian) shales (e.g., Arthur et al., 1990), they are rather the products of local environments and not the results of a global anoxic event. Such local environments where U enrichment occurs are also found in the present-day oceanic environments (e.g., Black Sea, Mediterranean, Cariaco Trench, Saanich Inlet, etc.); however, such local anoxic environment are minor in the global context. Fisher and Wignall (2001) have suggested that the U enrichment occurs only under specific conditions of low, but fluctuating, oxygen 336

regime and extremely slow sedimentation rates. In order for shales to be extremely enriched in U, a special set of environmental conditions appears to be necessary. Therefore, a general lack of extreme enrichment of U (> 100 ppm) in the Archean shales can be explained, at

least partly, by their general lack of extreme enrichment of Corg. Such absence of extreme

enrichment of Corg in the Archean shales may be due to either or combination of, but not

limited to, (1) a higher probability of tectonic survival of Corg-poor sediments: Corg-rich shales deposited in small, local, special environments may well be destroyed by tectonic processes (e.g., subduction); (2) the labile nature of sedimentary OM: the Archean OM was more labile than Phanerozoic OM because of the absence of detrital OM from land plant including refractory lignin; (3) the low primary productivity: it is difficult to estimate the Archean primary productivity from the preserved geological record; and (4) fluctuating redox conditions of bottom water: invasions of oxic fluids into anoxic sediments could have decomposed OM and leached authigenic U (McKee et al., 1987; Fisher and Wignall, 2001). It has been experimentally demonstrated that even a brief (~minutes) exposure of anoxic sediments to oxygen can result in the rapid dissolution of authigenic U (Cochran et al., 1986; Anderson et al., 1989). Factors controlling preservation of OM in shales have been extensively discussed by Pedersen and Calvert (1990).

5-6-5. Estimation of U concentration in the Archean oceans From the equations developed above, we estimate the concentration of dissolved U in the Archean oceans. We introduce a new parameter FWR that represents hydrothermal (water-rock interaction) flux through MOR. Then,

out ...... FMOR = FWR • [U]ocean (Eq. 5-17)

From the Eq. 5-15 and Eq. 5-17, 337

out out out F = Fppt + FMOR ...... = k • Aorg • [U]ocean + FWR • [U]ocean (Eq. 5-18)

out ...... [U]ocean = F / [k • Aorg + FWR] (Eq. 5-19)

From Eq. 5-9 and 5-15,

out out Fppt [today] = k • Aorg • [U]ocean [today] = 1.5 • FMOR [today] ...... = 1.5 • FWR • [U]ocean [today] (Eq. 5-20) <=>

...... k • Aorg [today] = 1.5 • FWR [today] (Eq. 5-21)

Then Eq. 5-21 becomes

out ...... [U]ocean [today] = F / [2.5 • FWR] [today] (Eq. 5-22)

Using the values for present-day environments, i.e., [U]ocean = 3.2 ppb (Chen et al., 1986; Cochran, 1992) and Fout = Fin = 0.46 x 108 [mol/yr] (Table 5-7),

8 ..... 3.2 ppb [today] = 0.46 x 10 (mol/yr) / [2.5 • FWR] [today] (Eq. 5-23) 8 ...... FWR [today] = 0.058 x 10 [mol/yr/ppb] [today] (Eq. 5-24) 8 ...... k • Aorg [today] = 0.086 x 10 [mol/yr/ppb] [today] (Eq. 5-25)

If FWR [Archean] = 2 ~ 3 x FWR [today] (see the discussion above), Eq. 5-19 becomes as follows (Eq. 5-7 is also used): 338

out [U]ocean [Archean] = F [Archean] / [k • Aorg + FWR [today]] out = F [today] / [k • Aorg + 2 ~ 3 • FWR [today]] [Archean] = 0.46 x 108 / [0.086 x 108 + 2 ~ 3 • 0.058 x 108]

≈ 1.7 ~ 2.3 ppb...... (Eq. 5-26)

Therefore, we suggest that the U concentration in the Archean oceans was ~2 ppb, i.e., ~50 to ~70 % of today's value (3.2 ppb). The above argument may be applied to other elements dissolved in the oceans; the enhanced submarine hydrothermal activity in the Archean would have increased the role of MOR as an efficient sink of some elements and therefore may explain their lower contents in the Archean sedimentary rocks compared to those of modern sediments and sedimentary rocks.

We then estimate the value of the parameter Aorg in the Archean with respect to U from the Eq. 5-16:

Aorg [Archean] • 1.7 ~ 2.3 ppb ≤ 0.00 ~ 0.33 • Aorg [today] • 3.2 ppb ...... (Eq. 5-27) <=>

...... Aorg [Archean] ≤ 0.00 ~ 0.62 • Aorg [today] (Eq. 5-28)

At a first order approximation, the total surface of reactive OM in the global marine sediments in the Archean was smaller by at least a factor of ~0.6 than that of present-day environments. 339

5-6-6. Tectonic recycling of U and decoupling of U and Th through geologic time Secular changes in the U and Th contents of sedimentary rocks have been examined by many researchers (e.g., Ronov and Migdisov, 1971; Veizer, 1973; McLennan and Taylor, 1980; Taylor and McLennan, 1985). However, there exist some serious problems in the data sets used by earlier workers. McLennan and Taylor (1980) have questioned the data quality of some earlier studies because of the absence of (1) the standard materials for inter- laboratory calibrations and (2) the confirmation of the results from the studies of the other continents where Archean-Proterozoic sedimentary rocks are preserved (i.e., Africa, Australia, and India). Therefore McLennan and Taylor (1980) used advanced analytical techniques and drillcores free of modern weathering to produce reliable data sets. However, their selections of samples have not been adequate to extract reliable geochemical information of unaltered original signatures (see below). McLennan and Taylor (1980) used samples younger than 2.5 Ga that have been subjected only to minor metamorphic grade (lower than greenschist facies). However, McLennan and Taylor (1980) used 7 samples from Akilia in Greenland to represent Early Archean (> 3.7 Ga) and 5 samples from Malene Supracrustal in Greenland to represent Middle Archean (> 3.0 Ga) crust. The Akilia samples yield very low contents of U (average 0.38 ppm) and Th (average 1.4 ppm). The Malene samples also yield very low contents of U (average 0.73 ppm) and Th (average 1.6 ppm). Because these samples have been metamorphosed to amphibolite and granulite facies, the possibility of U remobilization (loss) can not be excluded. Nonetheless, McLennan and Taylor (1980) have suggested secular 'increase' in the U and Th contents during Archean. In order to test their suggestion, shales of similar metamorphic grades must be compared. Uranium and Th are radiogenic elements and decay with time. To accurately compare the U and Th contents of rocks of different ages at the time of formation, the 340

measured U and Th contents must be corrected for their ages. McLennan and Taylor (1980) did not employ age-corrections for their samples. This study takes advantages of the low-metamorphic grade samples that were systematically collected from 13 formations in South Africa and Australia ranging in ages from 3.25 to ~2.2 Ga and analyzed for U and Th abundance by ICP-MS. Together with the previously published data of shales (McLennan and Taylor, 1980; Taylor and McLennan,

1985) and depleted mantle (Collerson and Kamber, 1999), the Thcorr/Ucorr ratios of the shales of this study are plotted, against ages, in Fig. 5-14.

Secular change in the age-corrected Th/U ratios

Important observations in Fig. 5-14 include: (1) the variability of the Thcorr/Ucorr ratios for the samples within each formation of this study is significant when compared to the Condie's evolution curve (see Fig. 5-14) (this is already discussed in the section 5-6-1);

(2) the lower Thcorr/Ucorr ratios compared to the Condie's evolution curve is particularly notable; (3) the scattering of the data by McLennan and Taylor (1980) is significant (see the above section); (4) there is a converging trend with the decrease in age between the Condie's evolution curve and the hypothetical evolution curve of the PAAS (calculated backward in time from the present value: see Fig. 5-14; Taylor and McLennan, 1985); (5) there is a

complementary trend of decreasing Thcorr/Ucorr ratios in the mantle-derived volcanic rocks (depleted mantle) by Collerson and Kamber (1999) (Fig. 5-14); and (6) there is no significant change in the Thcorr/Ucorr ratio of shales at the Archean–Proterozoic boundary at 2.5 Ga.

The generally lower Thcorr/Ucorr ratios of this study compared to those of the previous studies (Condie, 1993; see Fig. 5-14) are probably due to the nature of samples

(black shales) with generally higher U and Corg contents compared to the sample sets (non- black shales) used by the previous studies. The variabilities of the Th/U ratios depending on 341

the Corg contents are significant because they suggests, as previously discussed (see the section 5-6-3), the fixation of dissolved U in the oceans by OM that requires a redox gradient near SWI.

Secular increase in the Thcorr/Ucorr ratios for the post-Archean shales (Fig. 5-14) has been found and interpreted as due to an incremental loss of U to mantle during each cycle of sedimentary processes (weathering and sedimentation) under oxygenated environments (Taylor and McLennan, 1985). The 'lost U' would be enriched in the subducted altered oceanic crusts (Fig. 5-10). The efficiency and importance of sedimentary recycling in environments where U is decoupled from Th are supported by (i) the agreements between the results of this study and those of previous studies, (ii) the secular

decrease in the difference of the Thcorr/Ucorr ratios between the Condie's evolution curve and

the PAAS evolution curve, and (iii) the complimentary decrease in the Thcorr/Ucorr ratios of the depleted mantle (Fig. 5-14). The inferred change in the volume and the chemical compositions of the continental

crust at ~2.5 Ga (e.g., Taylor and McLennan, 1985) are not recorded in the Thcorr/Ucorr ratios of shales. This result is not unexpected because the volumetric (and areal) evolution of the continental crust would not leave its signatures in the chemical compositions of shales and because the Th/U ratios of mafic (U-poor) and felsic (U-rich) rocks are not much

different (Table 5-1). The observed increase in the Ucorr/Al ratios of the shales of this study in the Neoarchean (Fig. 5-6) may suggest an increase of the felsic component in the upper continental crust; however, the U enrichment of the Neoarchean samples of this study are

probably due to the Corg enrichment (Fig. 5-4). Therefore, from the results of this study, it appears premature to argue the crustal evolution. 342

5-7. Conclusions

(1) The continental weathering flux of U-bearing minerals in the Archean would have been similar to that of today. This is because the major U-bearing minerals in continents in terms of abundance and weathering flux are feldspars (not uraninite), the dissolution of feldspar is

quite fast, and its dissolution kinetics is dependent of pCO2 level (faster dissolution under

high pCO2 in the Archean) but independent of pO2 level.

(2) The variabilities in the U contents and the U/Corg, U/Al, and Th/U ratios observed for the samples in the same formations resulted from the weathering and sedimentary processes involving redox reactions during diagenesis. Such variabilities are due to neither the minor grade of metamorphism nor the changes in source rock compositions.

(3) The positive correlation between U and Corg contents and the low Th/U ratios in Corg- rich shales suggest an active role of OM in fixing the dissolved U in the Archean–Paleoproterozoic oceans. The existence of redox gradient is probably a prerequisite for the sedimentary fixation of dissolved U in the oceans by OM.

(4) Compared to the Phanerozoic black shales, the generally lower U contents of the Archean–Paleoproterozoic shales may be explained by either or combination of the more extensive submarine hydrothermal activity and their generally lower Corg contents. Shales

and sediments of low U and low Corg contents are typically found in Phanerozoic and modern geological records. Therefore, contrary to the previous suggestion by Holland (1994), the low U Archean shales can not be used as evidence for a globally anoxic environment. 343

(5) The secular increase in the sedimentary Th/U ratios throughout geologic time suggests the operation and its efficiency of the sedimentary recycling of U combined with the decoupling of U and Th.

(6) All of the above conclusions are fully consistent with the globally oxygenated atmosphere and oceans with localized anoxic environments in the Archean (e.g., Lasaga and Ohmoto, 2002a). The Archean geochemical cycle of U was essentially the same as today. Major difference in the geochemical cycles of U between Archean and modern environments are the higher heat flux (thus submarine hydrothermal activity) and the more mafic continental crust in the Archean. 344

References

Anderson, R.F. (1987) Redox behavior of uranium in an anoxic marine basin. Uranium 3, 145-164.

Anderson, R.F., LeHuray, A.P., Fleisher, M.Q., and Murray, J.W. (1989) Uranium deposition in Saanich Inlet sediments, Vancouver Island. Geochim. Cosmochim. Acta 53, 2205-2213.

Armands, G. (1972) Geochemical studies of uranium, molybdenum, and vanadium in a Swedish alum shale. Stockholm Contr. Geol. 27, 1-148.

Armstrong, R.L. (1981) Radiogenic isotopes: the case for crustal recycling on a near- steady-state no-continental-growth Earth. Phil. Trans. Royal Soc. Lond., A 301, 443-472.

Arthur, M.A. and Schlanger, S.O. (1979) Cretaceous "Oceanic Anoxic Events" as casual factors in development of reef-reservoired giant oil fields. Am. Assoc. Petrol. Geol. Bull. 63, 870-885.

Arthur, M.A., Jenkyns, H.C., Brumsack, H. -J., and Schlanger, S.O. (1990) Stratigraphy, geochemistry, and paleoceanography of organic carbon-rich Cretaceous sequences. In Cretaceous Resources, Events and Rhythms (eds. R.N. Ginsburg and B. Beaudoin), Kluwer Academic Publ., Netherlands, 75-119.

Barnes, C.E. and Cochran, J.K. (1990) Uranium removal in oceanic sediments and the oceanic U balance. Earth Planet Sci. Lett. 97, 94-101.

Barnes, C.E. and Cochran, J.K. (1991) Geochemistry of uranium in Black Sea sediments. Deep-Sea Res 38, S1237-S1254.

Berner, R.A., Lasaga A.C., and Garrels R.M. (1983) The carbonate-silicate geochemical cycle and its effect on the atmospheric carbon dioxide over the past 100 million years. Am. J. Sci., 283, 641-683.

Bertine, K.K. and Turekian, K.K. (1973) Molybdenum in marine deposits. Geochim. Cosmochim. Acta 37, 1415-1434.

Bethke, C.M. (1994) The geochemist's workbench, Ver. 2.0, A user's guide to Rxn, Act2, Tact, React, and Gtplot. Hydrology Program. Urabana, IL, Univ. Illinois. 345 Bethke, C.M. (1996) Geochemical reaction modeling. New York, Oxford Univ. Press.

Bishop, J.L., Lougear, A., Newton, J., Doran, P., Froeschl, H., Trautwein, A.X., Kröner, W., and Loeberl, C. (2001) Mineralogical and geochemical analyses of Antarctic lake sediments: A study of reflectance and Mössbauer spectroscopy and C, N, and S isotopes with applications for remote sensing on Mars. Geochim. Cosmochim. Acta 65, 2875-2897.

Blum, A.E. and Stillings, L.L. (1995) Feldspar dissolution kinetics. In Chemical weathering rates of silicate minerals (eds., A.F. White and S.L. Brantley), Reviews in Mineralogy 31, Mineral. Soc. America, 291-351.

Brady, P.V. (1991) The effect of silicate weathering on global temperature and atmospheric pCO2. J. Geophys. Res. 96, 18101-18106.

Brown, E.T., Callonnec, L.L., and German, C.R. (2000) Geochemical cycling of redox- sensitive metals in sediments from Lake Malawi: A diagnostic paleotracer for episodic changes in mixing depth. Geochim. Cosmochim. Acta 64, 3515-3523.

Brumsack, H.J. (1989) Geochemistry of recent TOC-rich sediments from the Gulf of California and the Black Sea. Geologische Rundschau 78, 851-882.

Burke, K., Kidd, W.S.F., and Kushy, T.M. (1985) Is the Ventersdorp rift system of southern Africa related to a continental collision between the Kaapvaal and Zimbabwe Craton at 2.64 Ga ago? Tectonophysics 115, 1-24.

Burke, K., Kidd, W.S.F., and Kusky, T.M. (1986) Archean foreland basin tectonics of the Witwatersrand, South Africa. Tectonics 5, 436-456.

Calvert, S.E. (1990) Geochemistry and origin of the Holocene sapropel in the Black Sea. In Facets of Modern Biogeochemistry (eds., V. Ittekkot, S. Kempe, W. Michaelis, and A. Spitzy), Springer, Berlin, 326-352.

Calvert, S.E. and Pedersen, T.F. (1993) Geochemistry of recent oxic and anoxic marine sediments: Implications for the geological record. Mar. Geol. 113, 67-88.

Calvert, S.E., and Price, N.B. (1983) Geochemistry of Namibian shelf sediments. In Coastal upwelling: its sediment record, part A (eds., J. Thiede and E. Suess), Plenum press, New York, 337-375.

Chase, Z., Anderson, R.F., and Fleisher, M.Q. (2001) Evidence from authigenic uranium for increased productivity of the glacial Subantarctic Ocean. Paleoceanography 16, 468-478. 346 Chen, J.H., Wasserberg, G.J., von Damm, K.L., and Edmond, J.M. (1986) The U-Th-Pb systematics in hot spring on the East Pacific Rise at 21 ˚N and Guaymas Basin. Geochim. Cosmochim. Acta 50, 2467-2479.

Cochran, J.K., Carey, A.E., Sholkowicz, E.R., and Suprenant, L.D. (1986) The geochemistry of uranium and thorium in coastal marine sediments and sediments pore waters. Geochim. Cosmochim. Acta 50, 663-680.

Cochran, J.K. (1992) The oceanic chemistry of the U- and Th-series nuclides. In Uranium Series Disequilibrium: Application to Environmental Problems (eds. M. Ivanovich and R.S. Harmon), 2nd Ed., Clarendon Press, pp 334-395.

Collerson, K.D. and Kamber, B.S. (1999) Evolution of the continents and the atmosphere inferred from Th-U-Nb systematics of the depleted mantle. Science 283, 1519-1522.

Colodner, D., Edmond, J., and Boyle, E. (1995) Rhenium in the Black Sea: A comparison with molybdenum and uranium. Earth Planet. Sci. Lett. 131, 1-15.

Condie, K.C. (1993) Chemical compositions and evolution of the upper continental crusts: Contrasting results from surface samples and shales. Chem. Geol. 104, 1-37.

Coveney, .R.M.Jr. and Glascock, M.D. (1989) A review of the origins of metal-rich Pennsylvanian black shales, central U.S.A., with an inferred role for basinal brines. Applied Geochem. 4, 347-368.

Coveney, .R.M.Jr., Leventhal, J.S., Glascock, M.D., and Hatch, J.R. (1987) Origins of metals and organic matter in the Mecca Quarry shale member and stratigraphically equivalent beds across the Midwest. Econ. Geol. 82, 915-933.

Crusius, J. and Thomson, J. (2000) Comparative behavior of authigenic Re, U, and Mo during reoxidation and subsequent long-term burial in marine sediments. Geochim. Cosmochim. Acta 13, 2233-2242.

Davy, R. (1983) A geochemical study of the Mount McRae shale and the upper part of the Mount Sylvia Formation in Core RD1, Rhodes Ridge, Western Australia. Geol. Surv. W. Aust. Record 1983/3.

Davy, R. and Hickman, A.H. (1988) The transition between the Hamersley and Fortescue Groups as evidenced in a drill core. Geol. Surv. W. Aust, Prof. Papers, Report 23, 85-97.

Desborough, G., Pitman, J., and Huffman, C. (1976) Concentration and mineralogical residence of elements in the Green River shale. Chem. Geol. 17, 13-26. 347 Eriksson, K.A. (1983) A paleohydrologic model for early Proterozoic dolomitization and silification. Precam. Res. 21, 299-321.

Faure, G. (1986) Principles of Isotope Geology, 3rd Edition, John Wiley & Sons, 589p.

Fisher, Q.J. and Wignall, P.B. (2001) Paleoenvironmental controls on the uranium distribution in an Upper Carboniferous black shale (Gastrioceras listeri Marine Band) and associated strata; England. Chem. Geol. 175, 605-621.

Grandstaff, D.E. (1976) A kinetic study of the dissolution of uraninite. Econ. Geol. 71, 1493 - 1506.

Hart, S.R. and Staudigel, H. (1982) The control of alkalis and uranium in seawater by ocean crust alteration. Earth Planet. Sci. Lett. 58, 202-212.

Hatch, J.R. and Leventhal, J.S. (1992) Relationship between inferred redox potential of the depositional environment and geochemistry of the Upper Pennsylvanian (Missourian) Stark Shale Member of the Dennis Limestone, Wabaunsee County, Kansas, U.S.A. Chem. Geol. 99, 65-82.

Hayashi, K., Fujisawa, H., Holland, H.D., and Ohmoto, H. (1997) Geochemistry of ~1.9 Ga sedimentary rocks from northeastern Labrador, Canada. Geochim. Cosmochim. Acta 61, 4115-4137.

Hemming, S.R. and McLennan, S.M. (2001) Pb isotope compositions of modern deep sea . Earth Planet. Sci. Lett. 184, 489-503.

Holland, H.D. (1984) The chemical evolution of the atmosphere and oceans. Princeton Univ. Press, Princeton.

Holland, H.D. (1994) Early Proterozoic atmospheric change. In Early Life on Earth. Novel Symposium No. 84 (S. Bengston, ed.), Columbia Univ. Press, New York.

Holland, H.D. (1999) When did the Earth's atmosphere become oxic? A reply. Geochemical News 100, 20-22.

Jacobs, L. and Emerson, S. (1982) Trace metal solubility in an anoxic fjord. Earth Planet. Sci. Lett. 60, 237-252.

Jones, B. and Manning, D.A.C. (1994) Comparison of geochemical indices used for the interpretation of paleoredox conditions in ancient mudstones. Chem. Geol. 111, 111-129.

Kasting, J.F. (1993) Earth's early atmosphere. Science 259, 920-926. 348 Klinkhammer, G.P. and Palmer, M.R. (1991)) Uranium in the oceans: where it goes and why. Geochim. Cosmochim. Acta 55, 1799-1806.

Kniewald, G. and Branica, M. (1988) Role of uranium (IV) in marine sedimentary environments: a geochemical possibility. Mar. Chem. 24, 1-12.

Ku, T.L., Knauss, K.G., and Mathieu, G.G. (1977) Uranium in open ocean: concentration and isotopic composition. Deep-Sea Res. 24, 1005-1017.

Langmuir, D (1978) Uranium solution-mineral equilibria at low temperatures with applications to sedimentary ore deposits. Geochim. Cosmochim. Acta 42, 1005- 1017.

Langmuir, D. (1997) Aqueous environmental geochemistry. Prentice-Hall, New Jersey. 600p.

Lao, Y. (1991) Transport and burial rates of 10Be and 231Pa in the Pacific Ocean. Unpub. Ph.D. Thesis, Columbia University.

Lasaga, A.C. (1998) Kinetic theory in the Earth sciences. Princeton Univ. Press, Princeton, 811p.

Lasaga, A.C. and Ohmoto, H. (2002a) The oxygen geochemical cycle: dynamics and stability. Geochim. Cosmochim. Acta 66, 361-381.

Lasaga, A.C. and Ohmoto, H. (2002b) Long term evolution of atmospheric oxygen and carbon dioxide. Am. J. Sci., in review.

Lasaga, A.C., Berner, R.A., and Garrels, R.M. (1985) An improved geochemical model of atmospheric CO2 fluctuations over the past 100 million years. In The Carbon Cycle and Atmospheric CO2: Natural Variations Archean to Present, Sundquist, R.E. and Broecker, W.S., eds. Geophys. Monogr., 32, 397-411.

Leventhal, J. (1990) Comparative geochemistry of metals and rare earth elements from the Cambrian Alum shale and Kolm of Sweden. Spec. Publ. Int. Assoc. Sedimentol. 11, 203-216.

Leventhal, J. (1993) Metals in Black Shales. In Organic Geochemistry: Principles and Applications (eds. M.H. Engel and S.A. Macko), Plenum Press, 581-592.

Leventhal, J.S. and Hosterman, J.W. (1982) Chemical and mineralogical analysis of Devonian black-shale samples from Martin County, Kentucky; Caroll and Washington Counties, Ohio; Wise County, Virginia; and Overton County, Tennessee, U.S.A. Chem. Geol. 37, 239-264. 349 Leventhal, J.S., Briggs, P.H., and Baker, J.W. (1983) Geochemistry of the Chattanooga Shale, Dekalb County, central Tennessee. Southeast. Geol. 24, 101-116.

Loukola-Ruskeeniemi, K. (1991) Geochemical evidence for the hydrothermal origin of sulfur, base metals and gold in Proterozoic metamorphosed black shales, Kainuu and Outokumpu areas, Finland. Mineral. Deposita 26, 152-164.

Loukola-Ruskeeniemi, K. (1999) Origin of black shales and the Serpentinite-associated Cu- Zn-Co ores at Outkumpu, Finland. Econ. Geol. 94, 1007-1028.

McKee, B.A., Demaster, D.J., and Nittrouer, C.A. (1987) Uranium geochemistry on the Amazon Shelf - evidence for uranium release from bottom sediments. Geochim. Cosmochim. Acta 51, 2779-2786.

McLennan, S.M. and Taylor, S.M. (1980) Th and U in sedimentary rocks: crustal evolution and sedimentary recycling. Nature 285, 621-624.

Meyer, F.M. and Robb, L.J. (1996) The geochemistry of black shales from the Chuniespoort Group, Transvaal Sequence, Eastern Transvaal, South Africa. Econ. Geol. 91, 111-121.

Morford, J.J. and Emerson, S. (1999) The geochemistry of redox sensitive trace metals in sediments. Geochim. Cosmochim. Acta 63, 1735-1750.

Morford, J.L., Russell, A.D., and Emerson, S. (2001) Trace metal evidence for changes in the redox environment associated with the transition from terrigenous clay to diatomaceous sediment, Saanich Inlet, BC. Mar. Geol. 174, 355-369.

Nameoff, T.J. (1996) Suboxic trace metal geochemistry and paleo-record in continental margin sediments of the Eastern Tropical North Pacific. Unpub. Ph.D. Thesis, University of Washington.

Ohmoto, H. (1997) When did the Earth's atmosphere become oxic? The Geochemical News 93, 12-12 and 26-27.

Ono, S. (2001) Unpublished Ph.D. thesis, Pennsylvania State University.

Pailler, D., Bard, E., Rostek, F., Zheng, Y., Mortlock, R., and van Geen, A. (2002) Burial of redox-sensitive metals and organic matter in the equatorial Indian Ocean linked to precession. Geochim. Cosmochim. Acta 66, 849-865.

Pedersen, T.F. and Calvert, S.E. (1990) Anoxic vs. Productivity: What controls the formation of organic-carbon-rich sediments and sedimentary rocks? Amer. Assoc. Petrol. Bull. 74, 454-466. 350 Phillips, G.N., Myers, R.E., Law, J.D.M., Bailey, A.C., Cadle, A.B., Beneke, S.D., and Giusti, L. (1989) The Witwatersrand gold fields: Part I. Postdepositional history, synsedimentationary processes, and gold distribution. Econ. Geol. Monogr. 6, 585- 597.

Phillps, G.N., Law, J.D.M., and Myers, R.E. (2001) In the redox state of the Archean atmosphere constrained? Soc. Econ. Geol. Newslett. 47.

Piper, D.Z. (1994) Seawater as the source of minor elements in black shales, phosphorites and other sedimentary rocks. Chem. Geol. 114, 95-114.

Piper, D.Z. and Isaacs, C.M. (1994) Instability of bottom-water redox conditions during accumulation of Quaternary sediments in the Japan Sea. Paleoceanography 11, 171-190.

Piper, D.Z. and Isaacs, C.M. (1995) Geochemistry of minor elements in the Monterey Formation, California: seawater chemistry of deposition. U.S. Geol. Surv. Prof. Paper 1566.

Piper, D.Z. and Isaacs, C.M. (2001) The Monterey Formation: bottom-water redox conditions and photic-zone primary productivity. In The Monterey Formation: from rocks to molecules (C.M. Isaacs and J. Rullkötter, eds.), Columbia Univ. Press, New York, 31-58.

Reymer, A. and Schubert, G. (1984) Phanerozoic addition rates to the continental crust and crustal growth. Tectonics 3, 63-77.

Ronov, A.B. and Migdisov, A.A. (1971) Geochemical history of the crystalline basement and the sedimentary cover of the Russian and North American Platforms. Sedimentology 16, 137-185.

Russell, A.D. and Morford, J.L. (2001) The behavior of redox-sensitive metals across a laminated-massive-laminated transition in Saanich Inlet, British Columbia. Mar. Geol. 174, 341-354.

Rye, R. and Holland, H.D. (1998) Paleosols and the evolution of atmospheric oxygen: a critical review. Am. J. Sci. 88, 621-672.

Sarin, M.M., Krishnaswwami, S., Somayajulu, B.L.K., and Moore, W.S. (1990) Chemistry of uranium, thorium, and radium isotopes in the Ganga-Brahmaputra river system: weathering processes and fluxes to the Bay of Bengal. Geochim. Cosmochim. Acta 54, 1387-1396. 351 Sclater, J.G., Jaupart, C., and Galson, D. (1980) The heat flow through oceanic and continental crust and the heat loss of the Earth. Rev. Geophys. Space Phys. 18, 269- 311.

Skei, J.M., Loring, D.H., and Rantala, R.T.T. (1988) Partitioning and enrichment of trace metals in a sediment core from Framvaren, South Norway. Mar. Chem. 23, 269- 281.

Sternbeck, J., Sohlenius, G., and Hallberg, R.O. (2000) Sedimentary trace elements as proxies to depositional changes induced by a Holocene fresh-brackish water transition. Aq. Geochem. 6, 325-345.

Taylor, S.R. and McLennan, S.M. (1985) The Continental Crust: its Composition and Evolution. Blackwell Scientific Publications, 311p.

Thomson, J., Higgs, N.C., Croudace, I.W., Colley, S., and Hydes, D.J. (1993) Redox zonation of elements at an oxic / post-oxic boundary in deep-sea sediments. Geochim. Cosmochim. Acta 57, 579-595.

Thomson, J., Higgs, N.C., Wilson, T.R.S., Croudace, I.W., de Lange, G.J., and van Santvoort, P.J.M. (1995) Redistribution and geochemical behavior of redox- sensitive elements around S1, the most recent eastern Mediterranean sapropel. Geochim. Cosmochim. Acta 59, 3487-3501.

Thomson, J., Higgs, N.C., and Colley, S. (1996) Diagenetic redistributions of redox- sensitive elements in northeast Atlantic glacial/interglacial transition sediments. Earth Planet. Sci. Lett. 139, 365-377.

Thomson, J., Nixin, S., Croudace, I.W., Pedersen, T.F., Brown, L., Cook, G.T., and MacKenzie, A.B. (2001) Redox-sensitive element uptake in north-east Atlantic Ocean sediments (Benthic Boundary Layer Experiment sites). Earth Planet. Sci. Lett. 184, 535-547.

Veeh, H.H., Calvert, S.E., and Price, N.B. (1974) Accumulation of uranium in sediments and phosphorites on the south west African shelf. Mar. Chem. 2, 189-202.

Veizer, J. (1976) Evolution of ores of sedimentary affiliation through geologic history; relations to the general tendencies in evolution of the crust, hydrosphere, atmosphere and biosphere. In Handbook of strata-bound and stratiform ore deposits, I: Principles and general studies; Vol. 3, Supergene and surficial ore deposits; textures and fabrics (eds. K.H. Wolf), Elsevier, Amsterdam, 1-41.

Warning, B. and Brumsack, H-.J. (2000) Trace metal signatures of eastern Mediterranean sapropels. Paleogeogr., Paleoclimatol., Paleoecol. 158, 293-309. 352 Wedepohl, K.H. (1971) Environmental influences on the chemical composition of shales and clays. In Physics and Chemistry of the Earth (eds., L.H. Ahrens, F. Press, S.K. Runcorn, and H.C. Urey), Pergamon Press, Oxford, 8, 307-331.

Wedepohl, K.H. (1991) The composition of the upper Earth's crust and the natural cycles of selected metals. Metals in natural raw materials. Natural resources. In Metals and their compounds in the Environment (ed., E. Merian), VCH, Weinheim. 3-17.

Wignall, P.B. (1990) Benthic palaeoecology of the late Jurassic Kimmeridge Clay of England. Paleontol. Assoc., London, Spec. Paper Paleontol., No. 43, 74 pp.

Wolkowicz, S. (1990) Uranium enrichment in the Permian organic-rich Walchia shale, Intra-Sudetic depression, southwestern Poland. Spec. Pub. Int. Assoc. Sedimentol. 11, 217-224.

Wronkiewicz, D.J. and Condie, K.C. (1990) Geochemistry and mineralogy of sediments from Ventersdorp and Transvaal Supergroup, South Africa: Cratonic evolution during the early Proterozoic. Geochim. Cosmochim. Acta 54, 343-354.

Yamada, M. and Tsunogai, S. (1983/1984) Postdepositional enrichment of uranium in sediment from the Bering Sea. Mar. Geol. 54, 263-276.

Zheng, Y. (1999) The marine geochemistry of germanium, molybdenum, and uranium: the sinks. Unpub. Ph.D. Thesis, Columbia University. 353

pO2 ≤ 1 % PAL pO2 ≥ 15 % PAL

Paleosols ?? Mt.Roe Hekpoort

Red beds ? Shebandowan red bed Uranium ? ? ores Witwatersrand Elliot Lake Oklo

Banded Iron Formations Hamersley

Eukaryotes ? Biomarker Grypania evidence Uranium in black shales Mt. McRae Outokumpu

This study 3.02.62.2 1.8 1.4 1.0 0.6

Time before present (Ga)

Fig. 5-1. Summary of the geological materials bearing on the O2 content of the Precambrian atmosphere (modified from Holland, 1994). The dark columns indicate the geological materials used to infer the low pO2 level and the light columns indicate those used to infer the high pO2 level. 354

(a)5 (b) 5 3.25 Ga Sheba Fm 3.25 Ga Sheba Fm Shales Shales 4 4

3 3 U U [ppm] [ppm] 2 2

1 1 shale siderite-rich shale cherty shale 0 0 012345 051015 20 25

Corg [wt. %] Al2O3 [wt. %]

(c)5 (d) 5 3.25 Ga Sheba Fm 3.25 Ga Sheba Fm Graywackes Graywackes 4 4

3 3 U U [ppm] [ppm] 2 2

1 1

0 0 012345 051015 20 25

Corg [wt. %] Al2O3 [wt. %]

Fig. 5-2. Plots of the U vs. Corg and U vs. Al2O3 contents for (a, b) the 3.25 Ga Sheba Formation (shales, drillcore PU1308) and for (c, d) the 3.25 Ga Sheba Formation (graywackes, drillcore MRE10). The solid line connects the origin and average shale composition, shown by open circles, of Wedepohl (1991). The dashed line is an approximate correlation line. 355

(e)5 (f) 5 2.96 Ga Parktown Fm 2.96 Ga Parktown Fm 4 4

3 3 U U [ppm] [ppm] 2 2

1 1

0 0 012051015 20 25

Corg [wt. %] Al2O3 [wt. %]

(g)10 (h) 10 2.71 Ga Rietgat Fm 2.71 Ga Riegat Fm 8 8

6 6 U U [ppm] [ppm] 4 4

2 2

0 0 012051015 20 25

Corg [wt. %] Al2O3 [wt. %]

Fig. 5-2. Plots of the U vs. Corg and U vs. Al2O3 contents for (e, f) the 2.96 Ga Parktown Formation (drillcore DRH13) and for (g, h) 2.71 Ga Rietgat Formation (drillcore MSF6). The line connects the origin and average shale composition, shown by open circles, of Wedepohl (1991). The dashed line is an approximate correlation line. 356

(i)10 (j) 10 2.72 Ga Pillingini Tuff Fm 2.72 Ga Pillingini Tuff Fm 2.69 Ga Jeerinah Fm 2.69 Ga Jeerinah Fm 2.69 Ga Lewin Shale Fm 2.69 Ga Lewin Shale Fm 8 8

6 6 U U [ppm] [ppm] 4 4

2 2

0 0 051015 051015 20

Corg [wt. %] Al2O3 [wt. %]

10 10 (k)2.60 Ga Carawine Dolomite Fm (l) 2.60 Ga Carawine Dolomite Fm >2.60 Ga Marra Mamba Iron Fm >2.60 Ga Marra Mamba Iron Fm 2.60 Ga Wittenoom Dolomite Fm 2.60 Ga Wittenoom Dolomite Fm 8 8

6 6 U U [ppm] [ppm] 4 4

2 2

0 0 0246810 051015 20

Corg [wt. %] Al2O3 [wt. %]

Fig. 5-2. Plots of the U vs. Corg and U vs, Al2O3 contents for (i, j) the 2.72 Ga Pillingini Tuff, 2.69 Ga Jeerinah, and 2.69 Ga Lewin Shale Formations (drillcores WRL1 and RHDH2A) and for (k, l) the 2.60 Ga Carawine Dolomite, 2.60 Ga Wittenoon Dolomite, and >2.60 Ga Marra Mamba Iron Formations (drillcores RHDH2A and WRL1. The solid line connects the origin and average shale composition, shown by open circles, of Wedepohl (1991). The dashed line is an approximate correlation line. 357

(m)5 (n) 5

4 4

3 3 U U [ppm] [ppm] 2 2

1 1 2.64 Ga Black Reef Fm 2.64 Ga Black Reef Fm 2.56 Ga Oak Tree Fm 2.56 Ga Oak Tree Fm 0 0 012345 01020304050

Corg [wt. %] Al2O3 [wt. %]

(o)10 (p) 10

8 8

6 6 U U [ppm] [ppm] 4 4 2.22 Ga Timeball Hill Fm, Eastern 2 2 2.22 Ga Timeball 2.22 Ga Timeball Hill Fm, Eastern Hill Fm, Central 2.22 Ga Timeball Hill Fm, Central ~2.2 Ga Mapedi Fm ~2.2 Ga Mapedi Fm 0 0 0 0.2 0.4 0.6 0.8 1 01020304050

Corg [wt. %] Al2O3 [wt. %]

Fig. 5-2. Plots of the U vs. Corg and U vs. Al2O3 contents for (m, n) the 2.56 Ga Oak Tree and 2.64 Ga Black Reef Formations (drillcores MSF6 and JPBR) and for (o, p) the 2.22 Ga Timeball Hill and ~2.2 Ga Mapedi Formations (drillcores PTB3, MSF6, and SA1677). The solid line connects the origin and average shale composition, shown by open circles, of Wedepohl (1991). The dashed line is an approximate correlation line. 358

(a) ~2.2 Ga Mapedi Fm

eastern Transvaal (b) central Transvaal 2.22 Ga Timeball Hill Fm

(c) 2.56 Ga Oak Tree Fm

(d) 2.60 Ga Carawine Dolomite Fm

(e) 2.60 Ga Wittenoom Dolomite Fm

(f) >2.60 Ga Marra Mamba Iron Fm

(g) 2.64 Ga Black Reef Fm

(h) 2.69 Ga Lewin Fm

(j) 2.69 Ga Jeerinah Fm (j) 2.71 Ga Rietgat Fm

(k) 2.72 Ga Pillingini Tuff Fm

(l) 2.96 Ga Parktown Fm

(m) 3.25 Ga Sheba Fm (graywackes)

silica-rich siderite-rich shale (n) 3.25 Ga Sheba Fm

0.01 0.1 1 10 7 (o) 6 5 Various modern sediments and 4 Phanerozoic black shales 3 1 2

0.01 0.1 1 10

[U/Corg]sample

[U/Corg]Av. Sh.

Fig. 5-3. Enrichment factor (EF) for U relative to organic C (Corg) in the Archean–Paleoproterozoic shales of this study. EF is defined as a ratio of [U/Corg]sample to [U/Corg]average shale, where U/Corg = ppm / wt.% ratio. Uranium contents are not corrected for the ages of samples. 1. Eocene Green River Formation: Desborough et al. (1976); 2. Gulf of California: Brumsack (1989); 3. Black Sea: Brumsack (1989), Leventhal et al. (1983); 4. Mediterranean sapropel: Warning and Brumsack (2000); 5. Pennsylvanian: Coveney et al. (1987); 6. Devonian: Holland (1984), Leventhal (1993); 7. Cambrian Alum Shale: Armands (1972), Leventhal (1990). 359

100

4 3 1314 9 Av. Sh.

5 Decay corrected 10

2 7 Ucorr 8 11 Corg 1

12

1 10

15

6

0.1 4 3.5 3 2.5 2 1.5 1 0.5 0 Age [Ga]

Fig. 5-4. Variations in the Ucorr/Corg wt. (ppm / wt.%) ratios for the Archean–Paleoproterozoic samples of this study and the black shales of Hayashi et al. (1997). The filled squares and the vertical gray bars represent the averages and ±1σ, respectively, of the data of this study. The open circle at age = 0 represents the average shale composition (15; Wedepohl, 1971, 1991; Taylor and McLennan, 1985; Condie, 1993), and the curve from the open circle represents decay-corrected evolution curve for the average shale(the Corg loss is not corrected). The Ucorr are U contents corrected for their decay (all the data in this plot are corrected). 1: Sheba Fm (shales); 2: Sheba Fm (graywackes); 3: Parktown Fm; 4: Pillingini Tuff Fm; 5: Rietgat Fm; 6: Jeerinah Fm; 7: Lewin Shale Fm; 8: Black Reef Fm; 9: Marra Mamba Iron Fm; 10: Carawine Dolomite Fm; 11: Wittenoom Dolomite Fm; 12: Oak Tree Fm; 13: Timeball Hill Fm; 14: Timeball Hill Fm (eastern Transvaal); 15: Nullataktok Fm (Labrador, Hayashi et al., 1997). 360

(a) ~2.2 Ga Mapedi Fm

central Transvaal (b) eastern Transvaal 2.22 Ga Timeball Hill Fm

(c) 2.56 Ga Oak Tree Fm

(d) 2.60 Ga Carawine Dolomite Fm

(e) 2.60 Ga Wittenoom Dolomite Fm >2.60 Ga Marra Mamba Iron Fm (f) 2.64 Ga Black Reef Fm (g) 2.69 Ga Lewin Fm (h) 2.69 Ga Jeerinah Fm (i) 2.71 Ga Rietgat Fm (j) 2.72 Ga Pillingini Tuff Fm (k) 2.96 Ga Parktown Fm (l) 3.25 Ga Sheba Fm (graywackes) (m) silica-rich siderite-rich shale (n) 3.25 Ga Sheba Fm

0.1 1 10

13 12 11 10 (o) 9 8 6 7 Various modern sediments 4 5 2 3 1

0.1 1 10 100

[U/Al]sample

[U/Al]Av. Sh.

Fig. 5-5. Enrichment factor (EF) for U relative to Al in the Archean– Paleoproterozoic shales of this study. EF is defined as a ratio of [U/Al]sample to [U/Al]average shale, where U/Al = ppm / wt.% ratio. Uranium content is not corrected for the ages of samples. 1. NW African margin (n = 12): Morford and Emerson (1999); 2. Washington Coast (n = ~40): Morford and Emerson (1999); 3. Lake Marawi (n = 50): Brown et al. (2000); 4. NE Atlantic (n = ~100): Thomson et al. (1996); 5. Saanich Inlet (n = 30): Russell and Morford (2001); 6. Arabian margin (n = 16): Morford and Emerson (1999); 7. Tropical Pacific: Nameoff (1996); 8. Saanich Inlet: Calvert and Pederson (1993); 9. Cariaco Trench: Calvert and Pedersen (1993); 10. Black Sea: Calvert (1990); 11. Mediterranean (n = 240): Waring and Brumsack (2000); 12. Framvaren: Skei (1980); 13. Black Sea (n = 12): Waring and Brumsack (2000). 361

2

1.8

1.6 9

1.4

1.2 11

Ucorr 13 7 Al 1 4 0.8 5

14 0.6 10 2 Av. Sh. 0.4 6 Decay corrected 1 3 12 0.2 15 8

0 4 3.5 3 2.5 2 1.5 1 0.5 0 Age [Ga]

Fig. 5-6. Variations in the Ucorr/Al wt. (ppm / wt.%) ratios for the Archean–Paleoproterozoic samples of this study and the black shales of Hayashi et al. (1997). The filled squares and the vertical gray bars represent the averages and ±1σ, respectively, of the data of this study. The open circle at age = 0 represents the average shale composition (0.33; Wedepohl, 1971, 1991; Taylor and McLennan, 1985; Condie, 1993), and the curve from the open circle represents decay-corrected evolution curve for the average shale. The Ucorr are U contents corrected for their decay (all the data in this plot are corrected). 1: Sheba Fm (shales); 2: Sheba Fm (graywackes); 3: Parktown Fm; 4: Pillingini Tuff Fm; 5: Rietgat Fm; 6: Jeerinah Fm; 7: Lewin Shale Fm; 8: Black Reef Fm; 9: Marra Mamba Iron Fm; 10: Carawine Dolomite Fm; 11: Wittenoom Dolomite Fm; 12: Oak Tree Fm; 13: Timeball Hill Fm; 14: Timeball Hill Fm (eastern Transvaal); 15: Nullataktok Fm (Labrador, Hayashi et al., 1997). 362

(a) 3 ~2.2 Ga Mapedi Fm 0 7 (b) 3 central eastern Transvaal Transvaal 2.22 Ga Timeball Hill Fm 0 (c) 2 2.56 Ga Oak Tree Fm 0 (d) 2 2.60 Ga Carawine Dolomite Fm 0 2 (e) 2.60 Ga Wittenoom Dolomite Fm 0 (f) 2 >2.60 Ga Marra Mamba Iron Fm 0 (g) 2 2.64 Ga Black Reef Fm 0 5 (h) 2 2.69 Ga Lewin Fm

Count 0 (i) 2 2.69 Ga Jeerinah Fm 0 (j) 3 2.71 Ga Rietgat Fm 0 2 (k) 2.72 Ga Pillingini Tuff Fm 0 6 (l) 2 5 2.96 Ga Parktown Fm 0 8 (m) 2 3.25 Ga Sheba Fm 0 (graywackes) 5 (n) shale siderite- silica-rich rich 3.25 Ga Sheba Fm 0 0246810

Th/U [wt. ratio]

Fig. 5-7. Histogram of the Th/U ratios for the samples of this study. U contents are not corrected for decay. A vertical line at Th/U = 3.8 represents the average crustal value of today (after Taylor and McLennan, 1985). 363

10 4.8

2.4 1

U 0.43 Total U [ppm]

0.1 U in insoluble accessary minerals (e.g., zircon)

0.01 U in feldspars

Basalt Andesite Granite

Fig. 5-8. Uranium contents of basalt, andesite, and granite, and of their rock-forming minerals. Note the log scale. Data from Faure (1986) and Smith and Williams (1990). 364

1.0 3 2+ 2 P 2- CO O2 = 1 atm 2 2 UO ) 0.5 3 UO P

(CO 4- O2 = 0.2 2 UO2(CO3)3 P Eh [v] O2 = 10 UO -20 U4+ 0 P O2 = 10 Uraninite -40 UO2 P O2 = 10 -60 25 ˚C P -0.5 [∑U] = 10-6 M H2 = 1 atm [∑C] = 0.1 M

02468101214 pH

Fig. 5-9. Eh-pH diagram for aqueous species and solids in the system U-O2-CO2-H2O at 25 ˚C and 1 bar total pressure. The diagram is drawn for ∑U (total uranium) = 10-6 M and ∑C (total carbon) = 0.1 M. The diagram is constructed by the computer program "Geochemist's workbench" (Bethke, 1994, 1996) using the preinstalled thermodynamic data. The uraninite domain decreases with increasing ∑C, increasing T, and decreasing ∑U. Dashed contours showing various pO2 levels are based on the equation: Eh = 1.23 + 0.0148 • log PO2 - 0.0592 • pH. The value ∑C = 0.1 M is used because of the inferred high pCO2 level in the Archean (x 100 ~ 1000 PAL; Kasting, 1987. PAL: present atmospheric level, 350 ppm). The diagram does not significantly change even under T = 60 ˚C, considering the inferred higher heat flux in the Archean than in modern. 365

(a) Phanerozoic U transport in dissolved forms (U 6+)

atmosphere seawater U6+ 6+ 6+ U Felsic continental crust U anoxic basin U6+ High U contents Low heat flux Less active hydrothermal activity MOR

(b) Archean U transport in dissolved forms (U 6+)

atmosphere seawater U6+ 6+ 6+ U Mafic continental crust U U6+ Low U contents anoxic basin High heat flux Active hydrothermal activity MOR

Fig. 5-10. A model for the secular change in the weathering flux of U from the continent and the efficiency of U removal from seawater into mid-oceanic ridges (MOR). The weathering flux of U may not be much different between under an anoxic and an oxic atmosphere. Quantitatively the dissolution of feldspar is more important than that of uraninite for the total weathering flux of U, because feldspar is the major U-bearing minerals in the continental crust and its dissolution rate does not strongly depend on pO2. The dissolution rate of uraninite (UO2) depends on pO2, but it is probably quantitatively minor. The heat flux was very likely higher in the Archean than in modern, and thus the submarine hydrothermal activity was more active in the Archean than in modern. Therefore the removal of U from seawater into MOR was more efficient in the Archean than in modern. The chemical composition of the continental crust was probably more mafic in the Archean than in modern and thus the U content of the continental crust was lower in the Archean than in modern. Combination of these effects may explain generally low U contents of Archean shales. 366

Source Weathering Sedimentation Diagenesis / Rocks Metamorphism

(+ O2) loss 4+ U6+ U OM in feldspars, etc. in seawater + U4+ in OM + U4+ U4+ in diagenetic in OM minerals

V VV detrital U4+ U4+ U4+ in zircons in zircons in zircons

Good U-C Poor U-C correlation correlation

Fig. 5-11. Schematic diagram of the geochemical history of uranium to cause good or poor relationship between organic carbon and uranium in black shales. OM: organic matter. 367

300 California Borderland Basins 42˚N California Margin 35 ˚N California Margin 250 23 ˚N Mexican Margin Arabian Sea Mid-Atlantic Bight Bering Sea 200 Equatorial Pacific [µM] 2

150

Dissolved O 100

50

0 050100 150

Authigenic U mass accumulation rate [µg/cm2/kyr]

Fig. 5-12. A comparison of authigenic U mass accumulation rate (MAR) in a number of sedimentary environments plotted against the bottom water dissolved O2 (DO) contents overlying the sediment cores. Modified after Zheng (1999). The data include California Borderland Basins (Zheng, 1999), the California Margin at 42 ˚N (Lao, 1991), the California Margin at 35 ˚N (Klinkhammer and Palmer, 1991; Zheng, 1999), the Mexican Margin at 23 ˚N (Nameoff, 1996), the Arabian Sea (Borole et al., 1982), the Middle Atlantic Bight (Zheng, 1999), the Bering Sea (Yamaga and Tsunogai, 1983/84), and the Equatorial Pacific (Zheng, 1999). A dashed line at 40 µM DO separates the region where the authigenic U MAR exhibits a strong response to changes in bottom water DO content (< 40 µM), form the region where no relationship is evident (> 40 µM). 368

1

0.66 out Fppt [Archean] 0.5 out Fppt [Today] Uptake by MOR of all oceanic U 0.33

Estimated range of C1

0 1 1.5 2 2.5 3

C1

out out Fig. 5-13. Relationship between Fppt [Archean] / Fppt [Today] ratios and C1 for U. See text for explanation of symbols. 369

8

7

6

5 PAAS 8 Thcorr PAAS Shale U 4 Condie (1993) corr 1

9 3 3 13

12 Depleted Mantle 2 4 10 14 Collerson and Kamber (1999) 2 5 6 11 7

1 Shale McLennan and Taylor (1980)

0 4 3.5 3 2.5 2 1.5 1 0.5 0 Age [Ga]

Fig. 5-14. Evolution of Thcorr/Ucorr wt. ratios for various sedimentary rocks through geologic time. Thcorr and Ucorr are the Th and U contents corrected for their decay (all the data in this plot are corrected). The filled rectangles and the vertical gray bars represent the averages and ± 1σ, respectively, of the data of this study. The open squares connected by the thin line represent shales from Condie (1993). The open diamonds connected by the thick line represent shales from McLennan and Taylor (1980). The dashed line represents the model depleted mantle curve of Collerson and Kamber (1999). The dotted line represents the evolution curve of PAAS (Taylor and McLennan, 1985) which is extended backward in time from the modern value (Th/U = 4.71). 1: Sheba Fm (shales); 2: Sheba Fm (graywackes); 3: Parktown Fm; 4: Pillingini Tuff Fm; 5: Rietgat Fm; 6: Jeerinah Fm; 7: Lewin Shale Fm; 8: Black Reef Fm; 9: Marra Mamba Iron Fm; 10: Carawine Dolomite Fm; 11: Wittenoom Dolomite Fm; 12: Oak Tree Fm; 13: Timeball Hill Fm; 14: Timeball Hill Fm (eastern Transvaal). Fm: Formation. 370 Table 5-1. Typical concentrations of U and Th in various igneous, metamorphic, and sedimentary rocks, upper continental crust, and rock-forming minerals.

Rocks / minerals U [ppm] Th [ppm] Th/U average typical range average typical range wt. ratio

Igneous rocks Granitic rocks 4.8 - 21.5 - 4.5 Andesite 2.4 - 8 - 3.3 Basalt 0.43 - 1.6 - 3.7 Gabbro 0.84 - 3.8 - 4.5 Ultramafic rocks 0.014 - 0.05 - 3.6 Metamorphic rocks Granitic 3.5 - 12.9 - 3.7 Granulite 1.6 - 7.2 - 4.5 Sedimentary rocks Shale 3.2 - 11.7 - 3.7 Sandstone 1.4 - 3.9 - 2.8 Carbonate rocks 1.9 - 1.2 - 0.6

Continental crust 2.8 - 10.7 - 3.8

Major minerals

Quartz 1.7 0.1-10 - 0.5-10 1-5 Feldspar 2.7 0.1-10 - 0.5-10 1-6 Biotite 8.1 1-60 - 0.5-50 0.5-3 7.9 0.2-60 - 5-50 2-4 Pyroxene 3.6 0.1-50 - - Minor minerals Allanite 200 30-1,000 9,100 1,000-20,000 10-20 Apatite 65 10-100 70 15-250 1 Sphene 280 10-700 510 100-1,000 1.7 Zircon 1,330 100-6,000 650 100-10,000 0.4 Monazite 3,000 500-3,000 125,000 2,000-200,000 >25 Xenotime - 300-40,000 - - Magnetite - 1-30 - >1

Data sources: rocks, Faure (1986); upper continental crust, Taylor and McLennan (1985); Minerals, Rogers and Adams (1969a, b).

371 Table 5-2. Summary of the geological settings and samples of this study.

Drillcore Drillhole locality Supergroup (SG) Age Formation (Fm) N Group (G) [Ga]

South Africa PU1308 Agnes gold mine, Swaziland SG 3.25 Sheba Fm 17 Barberton Fig Tree G MRE10 Sheba gold mine, Swaziland SG 3.25 Sheba Fm 12 Barberton Fig Tree G DRH13 26˚52'S, 26˚23'E, Witwatersrand SG 2.96 Parktown Fm 15 near Klerksdorp West Rand G Witwatersrand MSF6 26˚33'S, 27˚12'E Ventersdorp SG 2.71 Rietgat Fm 7 near Klerksdorp Platberg G Witwatersrand Transvaal SG Chuniespoort G 2.56 Oak Tree Fm 3 Pretoria G 2.22 Timball Hill Fm 4

JPBR near Klerksdorp Transvaal SG 2.64 Black Reef Fm 4 Witwatersrand Wolkberg G PTB3 24˚55' S, 30˚44'E Transvaal SG 2.22 Timeball Hill Fm 12 Pilgrim's Rest Pretoria G SA1677 Sishen iron mine, Griqualand Seq. ~2.2 Mapedi Fm 8 Griqualand West Olifantshoek G Australia WRL1 East of Wittenoom, Mt. Bruce SG 2.72 Pillingini Tuff Fm 3 Hamersley Fortescue G 2.69 Jeerinah Fm 6 >2.60 Marra Mamba Iron Fm 4 2.60 Wittenoom Dol. Fm 3 RHDH2A Ripon Hill, Mt. Bruce SG 2.69 Lewin Shale Fm 8 Northeastern Fortescue G Hamersley Mt. Bruce SG 2.60 Carawine Dol. Fm 6 Hamersley G

N: number of samples 372 Table 5-2. (continued)

Drillcore Dominant lithology of Tectonic settings of the groups the studied samples

South Africa PU1308 Black shales (some are rich Foreland basin, fore deep basin, in siderite and silica) evolving back-arc, passive continental margin, early rift to MRE10 Graywackes continental shelf, and shelf-rise

DRH13 Black shales Shallow marine or tidal marine with minor alluvial depositional environments

MSF6 Black shales

MSF6 Black shales Shallow marine cratonic MSF6 Black shales environments

JPBR Carbonate-rich black shales PTB3 Black shales Shallow marine cratonic environments

SA1677 Red shales

Australia WRL1 Black shales, carbonate rich Shallow marine WRL1 black shales Deep marine WRL1 Deep marine WRL1 Deep marine RHDH2A Black shales Deep marine

RHDH2A Carbonate-rich black shales Shallow marine or intertidal environments

See chapter 2 for the references of the tectonic settings of the samples. 373 Table 5-3. Geochemical data for the Archean–Paleoproterozoic samples of this study.

Samples U Th Al2O3 Corg Ccarb [ppm] [ppm] [wt.%] [wt.%] [wt.%]

Carbonate-rich black shales, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) PU1308-01 0.3 1.2 3.87 0.68 8.07 PU1308-02 0.3 1.5 5.54 0.94 8.20 PU1308-03 0.1 0.8 1.91 0.43 6.91 PU1308-04 0.4 1.7 5.06 0.94 7.27 PU1308-05 0.6 2.4 8.15 1.43 6.25 PU1308-06 0.3 1.2 3.21 1.02 8.38 PU1308-07 1.6 4.7 17.46 2.97 2.19 PU1308-08 0.9 1.8 5.84 1.29 7.83 PU1308-09 0.1 0.1 0.98 0.46 3.24 PU1308-10 4.5 15.4 19.16 1.87 1.84 PU1308-11 4.6 15.1 20.77 0.22 2.78 PU1308-12 1.0 2.5 6.47 1.31 6.71 PU1308-13 1.9 7.9 18.19 2.28 3.88 PU1308-14 0.4 1.1 3.48 1.08 9.22 PU1308-15 0.1 0.4 1.46 0.49 8.14 PU1308-16 0.6 1.6 5.83 0.93 8.07 PU1308-17 0.1 0.6 1.77 0.42 7.80

Average 1.0 3.5 7.60 1.10 6.28 S.D. 1.4 4.8 6.77 0.72 2.46

Graywackes, Sheba Fm, Fig Tree G, Swaziland SG (3.25 Ga) MRE10-01 2.3 7.8 9.48 0.16 1.77 MRE10-02 1.5 4.7 13.42 0.52 1.52 MRE10-03 1.9 6.5 18.13 0.43 0.91 MRE10-04 1.9 5.4 11.35 0.21 1.21 MRE10-05 1.7 5.2 14.76 0.39 0.21 MRE10-06 1.7 4.9 15.29 0.12 2.17 MRE10-07 2.8 9.8 9.43 0.17 1.15 MRE10-08 1.5 5.1 10.41 0.16 1.51 MRE10-09 2.1 6.5 10.02 0.17 1.21 MRE10-10 1.2 4.1 8.10 0.17 2.83 MRE10-11 1.3 4.1 13.39 0.28 1.63 MRE10-12 1.2 3.9 10.88 0.29 2.91

Average 1.8 5.7 12.06 0.26 1.59 S.D. 0.5 1.7 2.96 0.13 0.77 374 Table 5-3. (continued)

Samples U Th Al2O3 Corg Ccarb [ppm] [ppm] [wt.%] [wt.%] [wt.%]

Iron-rich shales, Parktown Fm, West Rand G, Witwatersrand SG (2.96 Ga) DRH13-01 1.1 4.0 8.94 0.23 0.04 DRH13-02 0.8 2.7 6.95 0.08 0.15 DRH13-03 0.7 2.5 6.16 0.11 0.28 DRH13-04 1.1 3.9 10.51 0.06 0.02 DRH13-05 1.0 3.1 9.27 0.17 0.07 DRH13-06 1.2 4.4 10.41 0.07 0.02 DRH13-07 1.2 3.2 8.00 0.04 0.02 DRH13-08 0.4 1.6 4.64 0.14 0.08 DRH13-09 0.2 1.1 2.77 0.25 0.09 DRH13-10 1.5 5.0 13.70 0.51 0.00 DRH13-11 1.5 5.4 13.90 0.07 0.02 DRH13-12 0.5 2.1 12.93 0.70 0.42 DRH13-13 1.3 4.3 13.32 0.18 0.05 DRH13-14 1.8 4.3 17.72 0.27 0.01 DRH13-15 1.7 5.2 17.88 0.22 0.00

Average 1.1 3.5 10.47 0.21 0.08 S.D. 0.5 1.3 4.46 0.18 0.12

Shales, Pillingini Tuff Fm, Fortescue G, Mt. Bruce SG (2.72 Ga) WRL1-01 4.6 14.1 15.79 0.21 0.01 WRL1-02 2.6 9.1 14.07 0.26 0.12 WRL1-03 1.8 6.7 11.01 0.14 0.42

Average 3.0 10.0 13.62 0.20 0.18 S.D. 1.5 3.8 2.42 0.06 0.21

Black shales, Rietgat Fm, Platberg G, Ventersdorp SG (2.71 Ga) MSF6-01 1.7 5.7 8.92 0.32 0.24 MSF6-02 2.4 8.6 9.74 0.27 0.30 MSF6-03 1.9 7.2 11.96 0.59 0.50 MSF6-04 3.5 10.3 13.92 0.95 0.00 MSF6-05 3.0 9.7 17.90 1.30 0.03 MSF6-06 3.8 11.3 19.94 1.59 0.00 MSF6-07 5.1 17.9 16.64 1.10 0.04

Average 3.1 10.1 14.14 0.87 0.16 S.D. 1.2 3.9 4.19 0.50 0.19

375 Table 5-3. (continued)

Samples U Th Al2O3 Corg Ccarb [ppm] [ppm] [wt.%] [wt.%] [wt.%]

Black shales, Lewin Shale Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) RHDH2A-01 5.0 11.1 13.26 2.67 0.34 RHDH2A-02 3.0 10.6 15.37 6.83 0.32 RHDH2A-03 6.4 17.3 16.16 1.56 0.19 RHDH2A-04 2.9 7.8 15.84 2.66 0.03 RHDH2A-05 3.6 9.2 15.69 1.83 0.04 RHDH2A-06 4.7 11.4 14.49 2.50 0.44 RHDH2A-07 3.7 10.5 14.21 2.05 0.10 RHDH2A-08 3.2 8.6 16.29 2.18 0.85

Average 4.1 10.8 15.16 2.78 0.29 S.D. 1.2 2.9 1.07 1.68 0.27

Black shales, Jeerinah Fm, Fortescue G, Mt. Bruce SG (2.69 Ga) WRL1-04 2.7 10.4 15.55 5.52 0.14 WRL1-05 1.7 7.1 13.79 7.39 2.32 WRL1-06 2.3 6.6 12.70 12.04 0.38 WRL1-07 2.8 6.9 12.92 8.19 0.34 WRL1-08 2.5 7.4 13.87 2.73 0.31 WRL1-09 2.1 6.3 11.23 3.87 0.10

Average 2.3 7.4 13.34 6.62 0.60 S.D. 0.4 1.5 1.44 3.36 0.85

Carbonate-rich black shales, Black Reef Fm, Wolkberg G, Transvaal SG (2.64 Ga) JPBR-01 0.1 0.5 0.77 0.09 12.02 JPBR-02 0.1 0.8 0.86 0.09 11.71 JPBR-03 0.1 0.5 3.26 0.33 11.83 JPBR-04 2.9 9.0 26.57 1.11 0.07

Average 0.8 2.7 7.87 0.41 8.91 S.D. 1.4 4.2 12.52 0.48 5.90 376 Table 5-3. (continued)

Samples U Th Al2O3 Corg Ccarb [ppm] [ppm] [wt.%] [wt.%] [wt.%]

Black shales, Marra Mamba Iron Fm, Hamersley G, Mt. Bruce SG (>2.60 Ga) WRL1-10 4.1 17.8 7.14 0.19 0.01 WRL1-11 1.1 4.8 2.47 0.37 1.51 WRL1-12 1.4 4.8 4.46 0.93 1.40 WRL1-13 0.8 3.3 4.00 0.74 3.80

Average 1.9 7.7 4.52 0.56 1.68 S.D. 1.5 6.8 1.94 0.34 1.57

Black shales, Wittenoom Dolomite Fm, Hamersley G, Mt. Bruce SG (2.60 Ga) WRL1-14 0.4 1.2 1.57 0.74 10.59 WRL1-15 6.1 9.2 14.28 2.78 0.13 WRL1-16 2.5 7.0 12.24 2.67 0.04

Average 3.0 5.8 9.36 2.06 3.59 S.D. 2.9 4.1 6.83 1.15 6.06

Carbonate-rich black shales, Carawine Dolomite Fm, Fortescue G, Mt. Bruce SG (2.60 Ga) RHDH2A-09 1.4 5.2 5.87 1.41 7.21 RHDH2A-10 1.1 4.1 4.99 1.02 9.22 RHDH2A-11 1.5 6.5 7.32 2.91 6.10 RHDH2A-12 1.8 6.2 9.50 1.91 6.68 RHDH2A-13 2.8 8.1 10.76 4.24 2.15 RHDH2A-14 4.3 13.5 15.77 4.45 7.29

Average 2.1 7.3 9.04 2.66 6.44 S.D. 1.2 3.4 3.95 1.45 2.35

Black shales, Oak Tree Fm, Chuniespoort G, Transvaal SG (2.56 Ga) MSF6-08 2.5 8.8 17.05 2.30 0.15 MSF6-09 1.4 5.9 13.74 1.64 0.05 MSF6-10 1.6 5.2 15.53 1.66 0.00

Average 1.8 6.6 15.44 1.87 0.07 S.D. 0.6 1.9 1.66 0.37 0.07 377 Table 5-3. (continued)

Samples U Th Al2O3 Corg Ccarb [ppm] [ppm] [wt.%] [wt.%] [wt.%]

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) PTB3-01 6.7 24.5 24.69 0.41 0.00 PTB3-02 6.1 22.5 21.62 0.35 0.88 PTB3-03 6.9 23.8 24.88 0.46 0.00 PTB3-04 6.3 27.2 25.95 0.39 0.00 PTB3-05 7.3 25.8 24.73 0.39 0.00 PTB3-06 6.2 26.6 25.09 0.39 0.40 PTB3-07 6.0 25.7 24.91 0.36 0.28 PTB3-08 6.8 31.6 29.44 0.50 0.38 PTB3-09 6.7 25.7 26.36 0.43 0.01 PTB3-10 7.4 26.8 25.30 0.47 0.45 PTB3-11 6.5 29.4 25.54 0.43 0.00 PTB3-12 6.5 25.7 24.36 0.30 0.45

Average 6.6 26.3 25.24 0.41 0.24 S.D. 0.4 2.4 1.77 0.06 0.28

Black shales, Timeball Hill Fm, Pretoria G, Transvaal SG (2.22 Ga) MSF6-11 5.4 20.6 18.43 0.28 0.00 MSF6-12 5.7 21.8 20.60 0.71 0.00 MSF6-13 8.0 23.4 18.85 0.55 0.00 MSF6-14 6.6 20.1 20.07 0.71 0.00

Average 6.4 21.5 19.48 0.56 0.00 S.D. 1.2 1.5 1.02 0.20 0.00

Red shales, Mapedi Fm, Pretoria G, Transvaal SG (~2.2 Ga) SA1677-01 2.5 20.2 16.98 0.10 0.00 SA1677-02 2.0 18.0 16.98 0.08 0.09 SA1677-03 2.5 20.2 19.47 0.09 0.00 SA1677-04 2.4 19.2 19.78 0.07 0.00 SA1677-05 2.6 20.8 18.79 0.07 0.00 SA1677-07 2.0 15.9 14.50 0.06 0.72 SA1677-08 2.1 12.2 14.97 0.06 0.00

Average 2.3 18.1 17.35 0.08 0.12 S.D. 0.3 3.1 2.10 0.02 0.27

Fm: Formation; G: Group, SG: Supegroup; S.D.: standard deviation. 378 Table 5-4. Comparison of the U to organic C ratios for the Archean– Paleoproterozoic sedimentary rocks.

Samples N *1 U/Corg Ucorr/Corg*2 (Fm: Formation; SG: Supergroup) [ppm/wt.% ± 1σ] [ppm/wt.% ± 1σ]

Ramah Group, Labrador *3 1.9 Ga Nullataktok Fm (black shales) 19 0.2 ± 0.2 0.2 ± 0.3 Pretoria Group, Transvaal SG ~2.2 Ga Mapedi Fm 8 30.4 ± 5.2 42.3 ± 7.3 2.22 Ga Timeball Hill Fm 4 12.8 ± 5.1 18.1 ± 7.3 2.22 Ga Timeball Hill Fm, Pilgrim's Rest 12 16.4 ± 2.1 23.2 ± 2.9 2.56 Ga Oak Tree Fm 3 1.0 ± 0.1 1.4 ± 0.2 Hamersley Group, Mt. Bruce SG 2.60 Ga Wittenoom Dolomite Fm 3 1.2 ± 0.9 1.8 ± 1.3 >2.60 Ga Marra Mamba Iron Fm 4 6.8 ± 9.8 10.1 ± 14.6 2.60 Ga Carawine Dolomite Fm 6 0.9 ± 0.2 1.3 ± 0.3 Wolkberg Group, Transvaal SG 2.64 Ga Black Reef Fm 4 1.4 ± 0.9 2.1 ± 1.4 Fortescue Group, Mt. Bruce SG 2.69 Ga Lewin Shale Fm 8 1.8 ± 1.1 2.8 ± 1.6 2.69 Ga Jeerinah Fm 6 0.5 ± 0.3 0.7 ± 0.4 2.72 Ga Pillingini Tuff Fm 3 15.3 ± 6.5 23.4 ± 10.0 Platberg Group, Ventersdorp SG 2.71 Ga Rietgat Fm 7 4.4 ± 2.3 6.6 ± 3.5 West Rand Group, Witwatersrand SG 2.96 Ga Parktown Fm 15 9.3 ± 8.1 14.7 ± 12.8 Fig Tree Group, Swaziland SG 3.25 Ga Sheba Fm (graywackes) 12 8.7 ± 4.8 14.4 ± 8.0 3.25 Ga Sheba Fm (shales) *4 17 0.6 ± 0.5 0.9 ± 0.9

Average Shale *5 - 15 -

*1 Number of samples.

*2 Ucorr is age-corrected U content. *3 Hayashi et al. (1997).

*4 For the calculation of average U/Corg and Ucorr/Corg ratios for shales of Fig Tree Group, one sample with

abnormally low U content (PU1308-11) is excluded. When included, average U/Corg and Ucorr/Corg ratios are 1.8 ± 5.0 and 2.9 ± 8.2, respectively.

*5 Data from Wedepohl (1991). 379 Table 5-5. Comparison of the U/Al ratios and EF (enrichment factor) for the Archean–Paleoproterozoic sedimentary rocks.

Samples U/Al Ucorr/Al *1 [U/Al]sample [Ucorr/Al]sample

(Fm: Formation; SG: Supergroup) [ppm/wt.% ± 1σ] [ppm/wt.% ± 1σ] [U/Al]Av. Sh. [U/Al]Av. Sh.

Ramah Group, Labrador *2 1.9 Ga Nullataktok Fm (black shales) 0.06 ± 0.06 0.07 ± 0.07 0.18 0.21 Pretoria Group, Transvaal SG ~2.2 Ga Mapedi Fm 0.25 ± 0.02 0.35 ± 0.03 0.76 1.06 2.22 Ga Timeball Hill Fm 0.63 ± 0.13 0.88 ± 0.18 1.91 2.67 2.22 Ga Timeball Hill Fm, Pilgrim's Rest 0.50 ± 0.04 0.70 ± 0.05 1.52 2.12 2.56 Ga Oak Tree Fm 0.22 ± 0.05 0.33 ± 0.07 0.67 1.00 Hamersley Group, Mt. Bruce SG 2.60 Ga Wittenoom Dolomite Fm 0.54 ± 0.23 0.82 ± 0.35 1.64 2.48 >2.60 Ga Marra Mamba Iron Fm 0.73 ± 0.30 1.10 ± 0.45 2.21 3.33 2.60 Ga Carawine Dolomite Fm 0.43 ± 0.06 0.65 ± 0.09 1.30 1.97 Wolkberg Group, Transvaal SG 2.64 Ga Black Reef Fm 0.21 ± 0.10 0.32 ± 0.15 0.64 0.97 Fortescue Group, Mt. Bruce SG 2.69 Ga Lewin Shale Fm 0.51 ± 0.16 0.78 ± 0.25 1.55 2.36 2.69 Ga Jeerinah Fm 0.33 ± 0.06 0.51 ± 0.09 1.00 1.55 2.72 Ga Pillingini Tuff Fm 0.41 ± 0.13 0.62 ± 0.20 1.24 1.88 Platberg Group, Ventersdorp SG 2.71 Ga Rietgat Fm 0.41 ± 0.10 0.62 ± 0.16 1.24 1.88 West Rand Group, Witwatersrand SG 2.96 Ga Parktown Fm 0.19 ± 0.04 0.31 ± 0.07 0.58 0.94 Fig Tree Group, Swaziland SG 3.25 Ga Sheba Fm (graywackes) 0.29 ± 0.12 0.49 ± 0.20 0.88 1.48 3.25 Ga Sheba Fm (shales) 0.20 ± 0.10 0.33 ± 0.17 0.61 1.00

Average Shale *3 0.33 - 1.00 -

*1 Ucorr represents age-corrected U content. *2 Hayashi et al . (1997). *3 U data from PAAS (Taylor and McLennan, 1985). 380 Table 5-6. Comparison of the U and Th contents and the Th/U ratios for the Archean–Paleoproterozoic sedimentary rocks.

Samples U Th Th/U Thcorr/Ucorr *1 (Fm: Formation; SG: Supergroup) [ppm] [ppm] [ppm/ppm] [ppm/ppm]

Ramah Group, Labrador *2 1.9 Ga Nullataktok Fm (black shales) 0.4 ± 0.5 0.8 ± 0.4 1.1 ± 0.6 0.9 ± 0.5 Pretoria Group, Transvaal SG ~2.2 Ga Mapedi Fm 2.2 ± 0.4 17.3 ± 3.6 7.8 ± 0.9 6.4 ± 0.7 2.22 Ga Timeball Hill Fm 6.4 ± 1.2 21.5 ± 1.5 3.4 ± 0.5 2.7 ± 0.4 2.22 Ga Timeball Hill Fm, Pilgrim's Rest 6.6 ± 0.4 26.3 ± 2.4 4.0 ± 0.4 3.2 ± 0.3 2.56 Ga Oak Tree Fm 1.8 ± 0.6 6.6 ± 1.9 3.7 ± 0.6 2.8 ± 0.4 Hamersley Group, Mt. Bruce SG 2.60 Ga Wittenoom Dolomite Fm 3.0 ± 2.9 5.8 ± 4.1 2.6 ± 1.0 2.0 ± 0.7 >2.60 Ga Marra Mamba Iron Fm 1.9 ± 1.5 7.7 ± 6.8 4.0 ± 0.4 3.0 ± 0.3 2.60 Ga Carawine Dolomite Fm 2.1 ± 1.2 7.3 ± 3.4 3.6 ± 0.5 2.7 ± 0.4 Wolkberg Group, Transvaal SG 2.64 Ga Black Reef Fm 0.8 ± 1.4 2.7 ± 4.2 4.5 ± 1.5 3.4 ± 1.1 Fortescue Group, Mt. Bruce SG 2.69 Ga Lewin Shale Fm 4.1 ± 1.2 10.8 ± 2.9 2.7 ± 0.4 2.0 ± 0.3 2.69 Ga Jeerinah Fm 2.3 ± 0.4 7.4 ± 1.5 3.2 ± 0.7 2.4 ± 0.5 2.72 Ga Pillingini Tuff Fm 3.0 ± 1.5 10.0 ± 3.8 3.4 ± 0.3 2.6 ± 0.2 Platberg Group, Ventersdorp SG 2.71 Ga Rietgat Fm 3.1 ± 1.2 10.1 ± 3.9 3.3 ± 0.3 2.5 ± 0.3 West Rand Group, Witwatersrand SG 2.96 Ga Parktown Fm 1.1 ± 0.5 .3.5 ± 1.3 3.5 ± 0.6 2.6 ± 0.4 Fig Tree Group, Swaziland SG 3.25 Ga Sheba Fm (graywackes) 1.8 ± 0.5 5.7 ± 1.7 3.2 ± 0.2 2.3 ± 0.2 3.25 Ga Sheba Fm (shales) 1.0 ± 1.4 3.5 ± 4.8 3.7 ± 1.3 2.6 ± 0.9

Average Shale *3 3.1 14.6 ~4 -

*1 Thcorr and Ucorr represents age-corrected Th and U content, respectively. See text for method. *2 Hayashi et al. (1997). *3 U and Th data from PAAS (Taylor and McLennan, 1985). Th/U ratio of 4 is an approximate value between 4.71 (Taylor and McLennan, 1985) and 3.24 (Turekian and Wedepohl, 1961). 381 Table 5-7. Source and sink fluxes for U.

Fluxes U References

Source fluxes [108 mol/yr] River 0.4 1 Eolian dust negligible Continental margin sediments 0.06 2 Hydrothermal (low and high T) negligible

Source total 0.46

Sink fluxes [108 mol/yr]

Sediments Oxic 0.03 3 Anoxic 0.06 4, 5 Continental margin sediments 0.12 ~ 0.25 5, 6 Mn nodule and metalliferous sediments 0.06 3

Sediment sink total 0.27 ~ 0.40 Hydrothermal (low and high temperature) 0.12 ~ 0.27 7, 8

Sink total 0.39 ~ 0.67

Oceanic residence time with respect to river water input [kyr] 250 ~ 500 9

References 1: Sarin et al. (1990) 2: McKee et al. (1987) 3: Cochran (1982) 4: Emerson and Huested (1991) 5: Barnes and Cochran (1990) 6: Klinkhammer and Palmer (1991) 7: Hart and Staudigel (1982) 8: Trefry et al. (1994) 9: Morford and Emerson (1999) Appendices

Introduction

The Appendixes have the following four sections (A, B, C, and D).

In the Appendix A, geologic settings of South Africa and Australia during the periods of Archean-Paleoproterozoic which are relevant to this study are provided.

In the Appendix B, various information on the drillcore samples used in this study are provided.

In the Appendix C, a variety of analytical methods used in this study is described more in detail than those appear in each chapter of the thesis.

In the Appendix D, a published paper in Science is provided. It is about mass-independent S isotope fractionation. It’s draft was first written by the author in September 2000 based on his discovery, and it was rewritten by his advisor for submission in January 2001. Appendix A

Geological settings

A-1. Introduction

The Archean–Paleoproterozoic samples of this study were taken from modern weathering-free drillcores in South Africa and Australia. In Appendix A, geologic settings of Archean-Paleoproterozoic South Africa and Australia are described in more detail than those in preceding each chapter.

A-2. South Africa

Archean and Paleoproterozoic strata are exceptionally well-preserved and exposed in the Kaapvaal Craton (1,200,000 km2) of southern Africa (Fig. A-1). The Kaapvaal Craton covers the eastern part of Botswana, the north-eastern part of the Republic of South Africa, and the Kingdom of Swaziland. The samples were mainly taken from (1) the Swaziland Supergroup in the Barberton Greenstone Belt (Fig. A-1, A2), (2) the Witwatersrand, Ventersdorp, and Transvaal Supergroups in the Witwatersrand Basin (Fig. A-1, A-3), (3) the Sabie-Pilgrim's Rest region (Fig. A-1), and (4) the Griqualand West region (Fig. A-1). 384

A-2-1. Swaziland Supergroup The Swaziland Supergroup is exposed in the Barberton Greenstone Belt that is located in the eastern part of the Kaapvaal Craton (Fig. A-1, A-2). The Barberton Greenstone Belt (3.4 to 3.1 Ga; de Wit et al., 1992) is made up by the predominantly volcanic supracrustal sequence, and is surrounded by tonalitic, trondhjemitic, and granodioritic (TTG) plutons representing several generations of igneous activity. Surrounding granites represent, in part, an older basement termed the Ancient Gneiss Complex (Fig. A-2; Hunter, 1974). As a representative Archean greenstone belt, numerous studies have been done partially because of the remarkable state of preservation of lithologies within the belt. Since two decades before, geological, geochemical, petrologic, geochronological, paleobiological, and sedimentological studies have provided important information about the development of the Barberton Greenstone Belt and its relevance to the evolution of the early Earth. de Ronde and de Wit (1994) provided a summary of the tectonic history of the Barberton Greenstone Belt. The Swaziland Supergroup consists of three groups: the Onverwacht, Fig Tree, and Moodies Groups in order of deposition (Fig. A-4). The Swaziland Supergroup is intensely deformed, and the principal large-scale structures are NE-SW trending folds with steeply dipping to vertical axes (Vijoen and Viljoen, 1969). In many cases, supracrustal rocks have undergone extensive early metasomatic alteration and greenschist facies metamorphism. More highly metamorphosed facies are developed near plutons (Viljoen and Viljoen, 1969). Sediments in the Swaziland Supergroup show an upward increase in maturity of the mineralogy as demonstrated by a transition from volcaniclastic to quartzofelsphathic (Condie et al., 1970). These changes may reflect progressive unroofing of a sialic provenance (Condie et al., 1970). 385

A-2-1-1. Onverwacht Group The Onverwacht Group, 1~12 km in thickness, is predominantly comprised of mafic and ultramafic rocks with minor sedimentary and felsic volcanic units (Fig. A-4; Viljoen and Viljoen, 1969; Lowe and Byerly, 1999). Carbonaceous cherts in the Onverwacht Group have yielded microfossils and stromatolites that are among the oldest on Earth (Knoll and Barghoorn, 1977; Walsh and Lowe, 1985; Byerly et al., 1986; Walsh, 1992). Deposition of the Onverwacht Group occurred during the mid Archean. The oldest dated supracrustal rocks in the Barberton Greenstone Belt are felsic metatuffs from the Theespruit Formation of the lowermost Onverwacht Group, dated at 3,544 ± 3 to 3,547 ± 3 Ma (Kröner et al., 1996). Precise single zircon evaporation ages at 3,445 ± 4 Ma have been determined for the felsic volcanics of the uppermost Hooggenoeg Group which is a middle part of the Onverwacht Group (Kröner et al., 1991). Chert in the lowermost Kromberg Formation overlying the Hooggenoeg Group has been dated at 3,416 ± 4 Ma (Kröner et al., 1991; Byerly et al., 1996), and chert in the upper Kromberg Formation has also been dated at 3,344 ± 3 Ma (Byerly et al., 1996). The age of the Onverwacht Group is approximately 3.55~3.34 Ga (Armstrong et al., 1990, Kröner et al., 1991, 1996).

A-2-1-2. Fig Tree Group The Fig Tree Group, ~3 km thickness, is composed of a succession of graywacke, shale, chert, conglomerate, dacitic flow, fragmented volcanic rocks, and banded iron- formations (Fig. A-4; Lowe and Byerly, 1999). The Fig Tree Group is subdivided into, from lower to upper, (1) the Sheba Formation, composed mainly of coarse, immature turbiditic sandstone and thin interbedded units of siltstone and shale (Lowe and Byerly, 1999), (2) the Belvue Road Formation, made up largely of shale, turbiditic siltstone and sandstone, chert, and regional coarse 386

volcaniclastic rocks, and (3) the Schoongezicht Formation, composed of coarse felsic volcaniclastic sandstone, conglomerate, breccia, and interbedded mudstone and shale. The oldest age yet measured for the Fig Tree Group is 3,258 ± 3 Ma on a dacitic tuff in the lowermost part of the Group (Byerly et al., 1996). Zircons from fresh clasts of felsic volcanic rocks in conglomerates near the top of the Fig Tree Group have yielded ages of 3,226 ± 6 Ma (Kröner et al., 1991). Similar ages of 3,226 ± 6 Ma and 3,222 +10/-4 Ma have been reported from an ignimbrite in the Schoongezicht Formation and felsic porphyritic intrusive, respectively, by Kamo and Davis (1994). Therefore, the age of the Fig Tree Group is bracketed between 3.26~3.23 Ga (Kröner et al., 1991; Kamo and Davis, 1994; Byerly et al., 1996). Deposition of the Fig Tree Group took place in synorogenic basins containing alluvial, fan-delta, and shallow- to deep-water environments (Nocita, 1989; Nocita and Lowe, 1990; Lowe and Nocita, 1999). Gold mineralization occurs in the Fig Tree Group (Anhaeusser, 1986; de Ronde et al., 1992a; Ward, 1995). A U-Pb analysis of zircon and hydrothermal rutile from a Au mine in the Barberton Greenstone Belt shows that the age of hydrothermal Au mineralization is bracketed between 3,126±21 and 3,084±18 Ma, which followed the intrusion of the nearby Kaap Valley tonalite at 3,227±1 Ma (de Ronde et al., 1992a).

A-2-1-3. Moodies Group The Moodies Group is composed of more than 3.5 km thick of feldspathic and quartzose sandstone, subgraywacke, chert-clast, siltstone, and minor banded iron-formations (Fig. A-4; Eriksson, 1979, 1980). It is the oldest known well-preserved quartz-rich sandstone sequence (Eriksson, 1979, 1980). The age of the Moodies Group has been reviewed by Heubeck and Lowe (1994a, b). Deposition of the Moodies Group began between about 3,225 and 3,222 Ma (Heubeck 387

and Lowe, 1994a). Single zircon dating of an intrusion that crosscut and metamorphosed the Moodies Group has yielded an age of about 3,109 +10/-8 Ma (Kamo and Davis, 1994). An age of 3,214 ± 4 Ma by Ar-Ar method has been reported for the Kaap Valley pluton that magnetically overprinted the Moodies Group (Layer et all., 1992). Therefore, the depositional age of the Moodies Group is bracketed between 3.22~3.11 Ga. According to Jackson et al. (1987), de Wit et al. (1992) and Heubeck and Lowe (1994), the sedimentary rocks of the Fig Tree and Moodies Groups deposited in a foreland or foredeep basin. However, according to Windley (1984), they developed in evolving back- arc or passive continental margins from early rifts to continental shelf and shelf-rise environments. Eriksson (1980) interpreted that the sedimentary strata of the Moodies Group was deposited in subaerial and shallow subaqueous settings, including alluvial-fan, braided-stream, tide-dominated delta, and open-shelf systems.

A-2-2. Witwatersrand Supergroup The Witwatersrand basin extends about 200 km south-west from the Barberton area with an areal extent of ~80,000 km2 (Fig. A-3). The Dominion Group, a lowermost succession of the Witwatersrand Basin, is composed of mafic to felsic volcanic rocks. Its age ranges from 2,830 Ma, based on the U-Pb zircon data of the felsic volcanics (Fig. A-4; Armstrong et al., 1986; Phillps, 1986), to 3,074 ± 60 Ma from the recent ion-microprobe zircon data on the volcanic rocks (Armstrong et al., 1991). The tectonic setting of the Dominion Group has been attributed to an Andean-type arc (Burke et al., 1986) or an extensional failed rift basin (Bickle and Eriksson, 1982; Tankard et al, 1982). The Witwatersrand Supergroup, the basin infill, conformably overlies volcanic rocks of the Dominion Group and uncomformably overlies the Archean basement rock (greenstone- granitoid complexes) of the Kaapvaal Carton (Fig. A-4). The Witwatersrand Supergroup 388 was deposited within the period 2,970 to 2,714 Ma (Robb et al., 1991). The geology of the Witwatersrand Basin has been extensively described (e.g., Robb et al., 1991; Coward et al., 1995; Robb and Meyer, 1995). The Witwatersrand Supergroup is dominated by siliclastic deposits that largely divided into a mixed argillaceous-arenaceous lower phase and a predominantly arenaceous upper phase. The arenaceous rocks include several horizons of predominantly quartz-pebble conglomerate units that host Au and U mineralization ('reefs'). Goldfields occurring around the margins of the Witwatersrand Basin are believed to be located in large-scale fan-delta deposits that prograded across finer-grained lacustrine or marine deposits (Minter, 1978). The Witwatersrand Supergroup is divided into two groups; the lower West Rand Group and the upper Central Rand Group (Fig. A-4). The maximum thickness of the West Rand Group is 5 km and that of the Central Rand Group is 3 km (Phillips et al., 1989). The West Rand Group is bimodaly composed of shale and quartzite. A mafic lava from the upper West Rand Group has been dated to be 2,914 ± 8 Ma (Armstrong et al., 1991). The West Rand Group is subdivided into, in order of deposition, the Hospital Hill, Government, and Jeppestown Subgroups. Depositional age of the basal Hospital Hill Subgroup can be approximated to be 2.96 Ga. Shallow marine or tidal marine with minor alluvial depositional environments have been suggested for the West Rand Group (Tankard et al., 1982). The Central Rand Group is composed mainly of subgraywacke, quartzite, and conglomerate with relatively minor shale (Pretorius, 1976). The Central Rand Group is divided into the Johannesburg and the overlying Turffontein Subgroups. Deposition of the Central Rand Group took place between 2,840 ~ 2,714 Ma (Robb et al., 1991). Suggested depositional environments for the Central Rand Group are marine-lacustrine-alluvial- subaerial, frequently switching among them (Minter, 1978). 389

A-2-3. Ventersdorp Supergroup The Ventersdorp Supergroup, attaining a maximum thickness of ~8 km, conformably overlies the Witwatersrand Supergroup and the Dominion Group of the Kaapvaal Craton (Fig. A-4). The Ventersdorp Supergroup is mainly composed of continental flood basalt with minor sediments (Winter, 1976), and divided into, from lower to upper, the Klipriviersberg (~2 km think), Platberg (~5 km thick) and Pniel (~1 km thick) Groups. The Klipriviersberg and Platberg Groups are largely composed of volcanics, and the Pniel Group is composed of clastic sediments and volcanics. Felsic volcanics (Makwassie Quartz Porphyry) in the Klipriviersberg Group yield an age of 2,709 ± 4 Ma, and mafic volcanics in the overlying Platberg Group yield an age of 2,714 ± 8 Ma, respectively, from the precise zircon U-Pb method (Armstrong et al., 1991). The lithology and sedimentation of both the Witwatersrand and Ventersdorp Supergroups were strongly influenced by tectonic processes, such as subduction, formation of island arcs, and rifting (Burke et al., 1985, 1986). An alluvial plain depositional setting has been suggested for the sediments of the Pniel Group (Tankard et al., 1982).

A-2-4. Transvaal Supergroup The Transvaal Supergroup, mainly composed of clastic and chemical sedimentary rocks, uncomformably overlies the Ventersdorp Supergroup (Fig. A-4). The Transvaal Group is divided into, in order of deposition, the Wolkberg, Chuniespoort and Pretoria Groups (Fig. A-4).

A-2-4-1. Wolkberg Group The Wolkberg Group is composed of fluvial and shallow marine feldspathic quartzite, arkoses, siltstones, pelites, carbonate, and conglomerates, as well as subaerially to subaqueously extruded basalt (Button, 1973). Sediments of the Wolkberg Group were 390

probably deposited in a rift basin, the rapidly subsiding Selati trough, to yield a sediment pile up to ~2 km in thickness (Tyler, 1978; Söhnge, 1986; Button, 1986). The Black Reef Formation of the Wolkberg Group shows a basal upward-fining succession of conglomerates, cross-bedded sandstones and uppermost mudrocks, which developed locally in paleotopographic lows (Eriksson et al., 1993). The top of the Black Reef Formation is marked by a transgressive mudrock bed which grades up into the succeeding dolomites of the Chuniespoort Group. The depositional age of the Wolkberg Group is bracketed between 2,640 and 2,560 Ma (Jahn et al., 1990; Armstrong et al., 1991; Eriksson et al., 1995).

A-2-4-2. Chuniespoort Group The Chuniespoort Group is mostly composed of stromatolitic carbonates and iron- formations with an average thickness of ~3 km (Button, 1986). It conformably overlies the Wolkberg Group (Fig. A-4). The Chuniespoort Group is subdivided into, in order of deposition, the Oak Tree, Monte Christo, Lyttelton, Eccles, Frisco, Penge, and Duitschland Formations. The former four Formations belong to the Malmani Subgroup. Direct dating by the Pb-Pb isochron method of stromatolitic carbonates in the Schmidtsdrif (Reivilo) Formation of the Campbellrand Group in the Griqualand West region, which is though to be equivalent to the Chuniespoort Group, indicates a depositional age of 2,557 ± 49 Ma (Jahn et al., 1990). U-Pb zircon dating of the same Formation yield an age of 2,552 ± 11 Ma (Barton et al., 1994). The Oak Tree Formation, lowermost of the Chuniespoort Group, has been dated to be 2,550 ± 3 Ma by Pb-Pb single zircon method (Walraven and Martini, 1995). The depositional setting for the Chuniespoort Group has been interpreted to be shallow marine or intertidal environments from the presence of stromatolitic carbonates 391

(Truswell and Eriksson, 1975; Eriksson, 1983; Button, 1986; Söhnge, 1986; Cledenin et al., 1988).

A-2-4-3. Pretoria Group The Pretoria Group, with a maximum thickness of ~7.5 km (Eriksson et al., 1993), is a -sedimentary succession. It is mainly composed of quartzites and shales with subordinate volcanics and carbonates. The Pretoria Group overlies the Chuniespoort Group with an angular (Fig. A-4). The Rb-Sr whole rock date of 2,224 ± 21 Ma is obtained on the Hekpoort Formation of the Pretoria Group (Burger and Coertze, 1973). The Rb-Sr whole rock date of 2208 ± 63 Ma is obtained on the underlying Timeball Hill Formation (Hunter and Hamilton, 1978). The Pb-Pb whole rock age of 2,222 ± 13 Ma is obtained on the Ongeluk Formation (Cornell et al., 1996), which is equivalent to the Hekpoort Formation. The whole rock Rb-Sr dating method is thought to be largely unreliable because of its open-system behavior (Walvaren and Martini, 1995), and age data obtained by this method may represent later diagenetic age. Mafic and felsic volcanics overlying the Pretoria Group yield ages of 2,061 ± 27 Ma (Walvaren et al., 1990) and 2,054 ± 3 Ma (Pb-Pb zircon age; Walvaren and Hattingh, 1993), respectively. Therefore, the depositional age for the Pretoria Group can be bracketed between 2.06 and 2.22 Ga. These age data suggest that there is a significant hiatus of ~300 Ma, longer then a Wilson Cycle, between the Chuniespoort Group and the Pretoria Group. Depositional settings of the Pretoria Group have been suggested to be either shallow marine cratonic environments (Button, 1973, 1981b; Tankard et al., 1982) or an intracratonic basin with short-lived marine incursions (e.g., Crockett, 1972; Eriksson et al., 1993). Eriksson et al. (1991) propose a continental rift tectonic setting with half-graben development. It has been suggested that the Kaapvaal Craton was tectonically stable during 392

the depositions of the Chuniespoort Group (Söhnge, 1986) and the lower section of the Pretoria Group (Schreiber et al., 1992).

Eastern Transvaal In the Pilgrim's Rest-Sabie region of the eastern Transvaal (Fig. A-1), covering an area of 600 km2, there are exposures of rocks which belong to the Pretoria and Chuniespoort Groups. There are also some occurrences of mesothermal epigenetic Au mineralization of economic grade in the Malmani Dolomite Subgroup of the Chuniespoort Group (e.g., Boer et al., 1995; Harley and Charlesworth, 1996; Tyler and Tyler, 1996). Tyler (1989) suggested a link between the mineralization and shale unit deposited in supratidal environments. In contrast, Cledenin et al. (1988, 1991) provides substantial evidence of marine transgression of the shale units and a record of basin drowning and cessation of carbonate production.

A-2-5. Metamorphisms and deformation of the Witwatersrand Basin Many previous investigators (e.g., Pretorius, 1981; Phillps et al., 1989; Wronkiewicz and Condie, 1990) have suggested that the Witwatersrand Supergroup, the Ventersdorp Supergroup and the Wolkberg Group of the Transvaal Supergroup were subjected to the greenschist facies metamorphism accompanying the formation of the Vredefort Dome (~2.6 Ga). Such estimate is based on the metamorphic mineral assemblages in the Witwatersrand metapelites, and metamorphic temperature of 350 ± 50 ˚C and pressures of 1~2 kbar have been suggested (Phillps, 1986). The metamorphic grade for the upper Pretoria Group has been interpreted as a zeolite facies from the mineral assemblages (Wronkiewicz and Condie, 1990). It has also been suggested by Miyano and Beukes (1984) that a maximum metamorphic-diagenetic overprint occurred at temperatures between 110 and 170 ˚C and pressures less than 2 kbar. However, locally higher grade 393

metamorphism occurred adjacent to the intrusion of the Bushveld Complex at 2,050 Ma (Hammerbeck, 1986). The Witwatersrand Basin has been affected by several periods of deformation (e.g., Coward et al., 1995); (1) pre-Witwatersrand extension leading to deposition of the underlying Dominion Group in a rift basin at ~3,080 Ma, (2) Witwatersrand-age compression and block faulting, (3) Ventersdorp-age (post-2,714 Ma) and Transvaal-age (post-2,550 Ma) extension and subsidence, and (4) post-Transvaal deformation.

A-3. Australia

Archean and Paleoproterozoic strata are also exceptionally well-preserved and exposed on the Pilbara Craton in the Pilbara-Hamersley regions of NW Australia (Fig. A- 5). The Pilbara Craton is composed of Archean basement (granitic rocks and greenstones) of 3.6-3.0 Ga in age. In the southern part of the Pilbara Craton, the Neoarchean Mt. Bruce Supergroup uncomformably overlies the cratonic basement (Fig. A-5). The Pilbara Craton is separated from the Yilgarn Craton by the Proterozoic Capricorn Orogen and related sedimentary basins (Barley et al., 1992). General review of the geology of the Pilbara Craton is found in Hickman (1983) and Geological Survey of Western Australia (1990).

A-3-1. Mt. Bruce Supergroup The Mt. Bruce Supergroup is a sequence of supracrustal rocks with a maximum thickness of ~10 km (Fig. A-6). Blake and Barley (1992) favors the term "Mt. Bruce Megasequence Set" instead of Mt. Bruce Supergroup based on their sequence stratigraphic research. The Mt. Bruce Supergroup is the infill of the Hamersley Basin that extends over an area of 100,000 km2 (Fig. A-5). 394

The Mt. Bruce Supergroup is traditionally divided into three groups on the basis of lithostratigraphic consideration (Fig. A-6). These are, from lower to upper, the Fortescue, Hamersley, and Turee Creek Groups (Fig. A-6; Trendall, 1979). The deposition of the Mt. Bruce Supergroup took place from 2.77 Ga to 2.44 Ga (Trendall et al., 1990; Arndt et al., 1991; Pidgeon and Horwitz, 1991; Barley et al., 1997). Tectonic evolution of the Mt. Bruce Supergroup has been interpreted as forming during a period of intra-cratonic extension, followed by rifting and the development of a divergent craton margin followed by a period of convergence, culminating in a continent- continent collision (Barley et al., 1992). More detailed discussion are found in Blake (1992), Blake and Barley (1992), and Barley et al. (1992). Metamorphic grade of the Mt. Bruce Supergroup is generally lower than greenschist facies (Trendall, 1979; Barley et al., 1992).

A-3-1-1. Fortescue Group The Fortescue Group, sometimes reaching 6 km in thickness, is mainly composed of mafic volcanics (basalt and komatiite) and volcaniclastic rocks, with subordinate acid volcanics (rhyolite), conglomerate, sandstone, shale, tuffs, and carbonates. The Fortescue Group is divided into, in order of deposition, the Mt. Roe Basalt, Hardey Sandstone, Kylena Basalt, Tumbiana, Nymerina Basalt, Kuruna Siltstone, Maddina Basalt, and Jeerinah Formations (Fig. A-6). The age of the basal Fortescue Group has been suggested to be 2,765 Ma (Arndt et al., 1991). The underlying basement (Munni Munni Complex) has been dated to be 2,925 ± 16 Ma (Arndt et al., 1991). The discrepancy between these two data represent an age gap of ~200 Ma created by the unconformity between the basement and the basal Fortescue Group. The Jeerinah Formation in the uppermost Fortescue Group has been dated to be 2,687 Ma (Arndt et al., 1991), and the Pillingini Tuff Formation that is stratigraphically 395

equivalent to the Tumbiana Formation in the middle Fortescue Group has been dated to be 2,715 ± 6 Ma (Arndt et al., 1991). From the occurrence of extensive flood basalt successions in the Fortescue Group, a rift-based model is favored to explain its deposition (Blake, 1984, 1992; Blake and Groves, 1987; Tyler and Throne, 1990; Thorne, 1990; Blake and Barley, 1992).

A-3-1-2. Hamersley Group The ~2.5 km thick Hamersley Group comformably deposits on the Fortescue Group (Fig. A-6). The Hamersley Group is well known to geoscientists because of the extensive development of BIF (banded iron-formations) in the Group (e.g., Trendall, 1983). Other than BIF, the Hamersley Group contains dolomite, carbonaceous shale, and rhyolite (Fig. A-6). The Hamersley Group is divided into, from lower to upper, the Marra Mamba Iron, Wittenoom Dolomite, Mt. Sylvia, Mt. McRae, Brockman Iron, Weeli Wolli, Woongarra Volcanics, and Boolgeeda Iron Formations (Fig. A-6). A SHRIMP (sensitive high resolution ion microprobe) U-Pb zircon age of 2,470 ± 4 Ma has been reported for a shale unit in the Dales Gorge Member of the Brockman Iron Formation (Trendall et al., 1990). A U-Pb age of 2,603 ± 7 Ma has been reported for a tuff later in the Wittenoom Dolomite Formation (Hassler, 1993). An age of 2,449 ± 3 Ma has been reported for the Woongarra Volcanics (Barley et al., 1997). The deposition of the Hamersley Group can be bracketed between approximately 2.60 and 2.45 Ga. A divergent continental margin as a consequence of earlier continental rifting and ocean formation has been favored to explain the deposition of the Marra Mamba Iron Formation and the Wittenoom Formation in the lower Hamersley Group (the Marra Mamba Supersequence Package: e.g., Barley et al., 1992). Such a depositional environment includes near-shore to shallow shelf and deeper shelf (e.g., Morris and Horwitz, 1983; Morris, 1993). The upper Hamersley Group (the Hamersley Range Megasequence) represents a 396

rock record of a neutral to compressive back-arc cratonic basin with some clastic sediments (Horwitz, 1982, 1987; Barley et al., 1992). BIFs within the Hamersley Group have been partially upgraded to iron ore during Proterozoic orogeny and Tertiary weathering (Morris, 1980) or during synsedimentary orogeny by pre 2.2 Ga oxidizing fluids (Powell et al., 1999; Martin et al., 1998).

A-3-1-3. Turee Creek Group The Turee Creek Group comformably sits on top of the Hamersley Group (Fig. A- 6). The Turee Creek Group is largely composed of graywacke, sandstone, siltstone, quartzite, and dolomite. Diamictite of probably glacial origin occurs in the Meteorite Bore Member of the Kungarra Formation of the Turee Creek Group (Trendall, 1976; Martin, 1999). Deposition of the Turee Creek Group is bracketed between the age of 2,449 ± 3 Ma for underlying Woongarra Volcanics in the upper Hamersley Group and the 2,209 ± 15 Ma for overlying Cheela Springs Basalt in the Wyloo Group (Martin et al., 1998; Barley et al., 1997). Diamictite in the Meteorite Bore Member of the Kungarra Formation in the lower Turee Creek Group has been suggested to be equivalent to the glaciogenic deposits in the Transvaal Supergroup in South Africa and the Huronian Supergroup in Canada (Martin , 1999). The Turee Creek Group was deposited in a foredeeps setting (Blake and Barley, 1992). A continent-continent collision followed the deposition of the Turee Creek Group (Barley et al., 1992), although there is no rock record preserved. The metamorphic grade of the Turee Creek Group does not exceed prehnite- pumpellyite-epidote facies, i.e., subgreenschist facies (Smith et al., 1982) 397

A-3-1-4. Lewin Shale and Carawine Dolomite The Hamersley Group outcrops in two geographically distinct parts of the Hamersley Basin; "the main outcrop area" and "the Oakover river area" (Fig. A-5; Simonson and Hassler, 1997). The Lewin Shale Formation and the Carawine Dolomite Formation occur in the Oakover river area, the northeastern region of the Hamersley Basin (Fig. A-5). The Lewin Shale Formation and the Carawine Dolomite Formation have been thought to be stratigraphically shallower-equivalent to the deeper Jeerinah and Wittenoom Dolomite / Marra Mamba Iron Formation, respectively, in the central-western region of the Hamersley Basin (Fig. A-6; Simonson et al., 1993, 1998; Simonson and Hassler, 1997; Woodhead et al., 1998). The exact stratigraphic correlation between the Lewin Shale / Carawine Dolomite Formations and the Jeerinah / Marra Mamba Iron / Wittenoom Dolomite Formations has not been established. Based on the position and isotopic age of the SMB (spherule marker bed; Simonson et al., 1993, 1998; Simonson and Hassler, 1997; Woodhead et al., 1998) in the Bee Gorge Member of the upper Wittenoom Dolomite and the lower Carawine Dolomite, it is suggested that they are correlative (Woodhead et al., 1998). Simonson and Hassler (1997) suggested the possibility that the Carawine Dolomite was deposited at the same time as part of the major banded iron-formations of the Hamersley Group. The Carawine Dolomite is at least 500 m in thickness and consists mostly of platformal dolomite, whose depositional features are indicative of deposition in shallow- water environments (Simonson et al., 1993). Thinly laminated black shale are interbedded with the dolomite units. The Lewin Shale is mostly composed of black, carbonaceous shales with abundant pyrite. Pb-Pb isotopic dating was performed on carbonates from the Carawine Dolomite, and yielded an age of 2,548 +26/-29 Ma (Woodhead et al., 1998). A diagenetic age of 398

2,541 ± 32 Ma has been reported by Jahn and Simonson (1995), based on a Pb-Pb isochron method for the Carawine Dolomite. 399

References

Anhaeusser, C.R. (1986) Archean gold Mineralization in the Barberton Mountain Land. In Mineral deposits of southern Africa (C.R. Anhaeusser and S. Maske, eds.), Geol. Surv. South Afr., 113-154.

Armstrong, R.A., Compston, W., Retief, E.A., and Wilke, N.J. (1986) Ages and isotopic evolution of the Ventersdorp volcanics. Extended Abstract, Geocongress 1986, Geol. Soc. S. Afr , 89-92.

Armstrong, R.A., Compston, W., de Wit, M.J., and Williams, I.S. (1990) The stratigraphy of the 3.5-3.2 Ga Barberton Greenstone Belt revisited: a single zircon ion microprobe study. Earth Planet. Sci. Lett. 101, 90-106.

Armstrong, R.A., Compston, W., Retief, E.A., Williams, I.S. and Welke, H.J. (1991) Zircon ion microprobe studies bearing on the age and evolution of the Witwatersrand triad. Precam. Res. 53, 243-266.

Arndt, N.T., Nelson, D.R., Compston, W., Trendall, A.F. and Thorne, A.M. (1991) The age of the Fortescue Group, Hamersley Basin, Western Australia, from iron microprobe zircon U-Pb results. Aust. J. Earth Sci. 38, 261-281.

Awramik, S.M., Schopf, J.W., and Walter, M.R. (1983) Filamentous fossil bacteria from the Archean of western Australia. Precam. Res. 20, 357-374.

Awramik, S.M., Schopf, J.W. and Walter, M.R. (1988) Carbonaceous filaments from North Pole, Western Australia: Are they fossil bacteria in Archean stromatolites? A discussion. Precam. Res. 39, 303-309.

Barley, M.E., Blake, T.S. and Groves, D. (1992) The Mount Bruce Megasequence Set and eastern Yilgarn Craton: examples of Late Archaean to Early Proterozoic divergent and convergent craton margins and controls on mineralization. Precam. Res. 58, 55- 70.

Barley, M.E., Pickard, A.L., and Sylvester, P.J. (1997) Emplacement of a large igneous province as a possible cause of banded iron formation 2.45 billion year ago. Nature 385, 55-58.

Barnicoat, A.C., Henderson, I.H.C., Knipe, R.J., Yardley, B.W., Napier, R.W., Fox, N.P.C., Kenyon, A.K., Muntigh, D.J., Strydom, D., Winkler, K.S., Lawrrence, S.R., and 400 Conford, C. (1997) Hydrothermal gold mineralization in the Witwatersrand basin. Nature 386, 820-824.

Barton, E.S., Altermann, W., Williams, I.S., and Smith, C.B. (1994) U-Pb zircon age for a tuff in the Campbellrand Group, Griqualand West Sequence, South Africa: implications for Early Proterozoic rocks accumulation rates. Geology 22, 343-346.

Bickle, M. and Eriksson, K.W. (1982) Evolution and subsidence of early Precambrian sedimentary basins. Phil. Trans. Roy. Soc. London A305, 225-269.

Blake, T.S. (1984) The lower Fortescue Group of the northern Pilbara Craton: stratigraphy and paleogeography. In Arcchaean and Proterozoic Basins of the Pilbara, Western Australia: Evolution and Mineralization Potential (eds., Muhling, J.R., Groves, D.I., and Blake, T.S.). Univ. W. Aus., Geol. Dep. and Univ. Extension Pub. 9. 123- 143.

Blake, T.S. (1992) Late Archean crustal extension, sedimentary basin formation, flood basalt volcanism and continental rifting: The Nullagine and Mount Jope Supersequences, Western Australia. Precam Res. 60, 185-242.

Blake, T.S. and Groves, D.I. (1987) Continental rifting and the Archean-Proterozoic transition. Geology 15, 229-232.

Blake, T.S. and Barley, M.E. (1992) Tectonic evolution of the late Archean to early Proterozoic Mount Bruce Megasequence set, western Australia. Tectonics 11, 1415- 1425.

Boer, R.H., Meyer, F.M., Robb, L.J., Graney, J.R., Vennemann, T.W., and Kesler, S.E. (1995) Mesothermal-type mineralization in the Sabie-Pilgrim's Rest gold field, South Africa. Econ. Geol. 90, 860-876.

Brocks, J.J., Logan, G.A., Buick, R., and Summons, R.E. (1999) Archean Molecular Fossils and the Early Rise of Eukaryotes. Science 285, 1033-1036.

Burger, A.J. and Coertze, F.J. (1973) Age-determination - April 1972 to March 1974. Ann. Geol. Surv. S. Afr. 10, 135-141.

Burke, K., Kidd, W.S.F., and Kushy, T.M. (1985) Is the Ventersdorp rift system of southern Africa related to a continental collision between the Kaapvaal and Zimbabwe Craton at 2.64 Ga ago? Tectonophys. 115, 1-24.

Burke, K., Kidd, W.S.F., and Kusky, T.M. (1986) Archean foreland basin tectonics of the Witwatersrand, South Africa. Tectonics 5, 436-456. 401 Button, A. (1973) A regional study of the stratigraphy and development of the in the eastern and northeastern Transvaal. Ph.D. thesis, Univ. of Witwatersrand, Johannesburg, S. Afr., 352 pp.

Button, A. (1981a) The Ventersdorp Supergroup. In Precambrian of the Southern Hemisphere (ed. D. R. Hunter), Elsevier, Amsterdam, Netherlands, 520-527.

Button, A. (1981b) The Transvaal Supergroup. In Precambrian of the Southern Hemisphere (ed. D. R. Hunter), Elsevier, Amsterdam, Netherlands, 527-536.

Button, A. (1986) The Transvaal sub-basin of the Bushveld floor in the eastern Transvaal. Trans. Geol. Soc. S. Afr. 79, 3-12.

Byerly, G.R., Lowe, D.R., and Walsh, M.M. (1986) Stromatolites from the 3,300-3,500 Myr Swaziland Supergroup, Barberton Mountain Land, South Africa. Nature 319, 489-491.

Byerly, G.R., Kröner, A., Lowe, D.R., Todt, W., and Walsh, M.M. (1996) Prolonged magmatism and time constraints for sediment deposition in the early Archean Barberton Greenstone Belt: Evidence from the Upper Onverwacht and Fig Tree Groups. Precam. Res. 78, 125-138.

Cledenin, C.W., Charlesworth, E.G., and Maske, S. (1988) Tectonic style and mechanism of early Proterozoic successor basin development, southern Africa. Tectonophys. 156, 275-291.

Clendenin, C.W., Henry, G., and Charlesworth, E.G. (1991) Characteristics of and influences on the Black Reef depositional sequence in the eastern Transvaal. S. Afr. J. Geol. 94, 321-327.

Condie, K.C., Macke, J.E., and Reimer, T.O. (1970) and geochemistry of early Precambrian graywackes from the Fig Tree Group, South Africa. Geol. Soc. Am. Bull. 81, 2759-2776.

Cornell, D.H., Schütte, S.S., and Eglinton, B.L. (1996) The Ongeluk formation in Griqualand West, South Africa: submarine alteration in a 2222 Ma Proterozoic Sea. Precam Res. 79, 101-123.

Coward, M.P., Spencer, R.M., and Spooner, C.E. (1995) Development of the Witwatersrand Basin, South Africa. In Early Precambrian Processes (eds., M.P. Coward and A.C. Reis), Geol. Soc. Spec. Pub. 95, 243-269.

Crockett, R.N. (1972) The Transvaal System n Botswana; its geotectonic an depositional environment and special problems. Trans. Geol. Soc. S. Afr. 75, 275-292. 402 de Ronde, C. E. J., Kamo, S., Davis, D. W., de Wit, M. J., and Spooner, E. T. E. (1991b) Field, geochemical and U-Pb isotopic constraints from hyperbyssal felsic intrusions within the Barberton greenstone belt, South Africa: Implications for tectonic and the timing of gold mineralization. Precam. Res. 49, 261-280. de Ronde, C.E.J., de Wit, M.J. and Spooner, E.T.C. (1993) Early Archean (>3.2 Ga) Fe- oxide-rich, hydrothermal discharge vents in the Barberton greenstone belt, South Africa. Geol. Soc. Am. Bull. 106, 86-104. de Ronde, C.E.J., Spooner, E.T.C., de Wit, M.J. and Bray, C.J. (1992) Shear zone-related, Au Quartz vein deposits in the Barberton greenstone belt, South Africa: Field and petrographic characteristics, fluid properties, and light stable isotope geochemistry. Econ. Geol. 87, 366-402. de Wit, M.J., Roering, C., Hart, R.J., Armstrong, R.A., de Ronde, C.E.J., Green, R.W.E., Tredoux, M., Peberdy, E. and Hart, R.A. (1992) Formation of an Archean continent. Nature 357, 553-562.

Eriksson, K.A. (1979) Marginal marine depositional processes from the Archean Moodies Group, Barberton Mountain Land, South Africa: Evidence and significance. Precam. Res. 8, 153-182.

Eriksson, K.A. (1980) Transitional sedimentation styles in the Moodies and Fig Tree Groups, Barberton Mountain Land, South Africa: Evidence favoring an Archean continental margin. Precam. Res. 12, 141-160.

Eriksson, K.A. (1983a) A paleohydrologic model for early Proterozoic dolomitization and silification. Precam. Res. 21, 299-321.

Eriksson, K.A. (1983b) Siliciclastic-hosted iron-formations in the early Archaean Barberton and Pilbara sequences. J. Geol. Soc. Am. 30, 473-482.

Eriksson, P.G., Hattingh, P.J., and Altermann, W. (1995) An overview of the geology of the Transvaal Sequence and Bushveld Complex, South Africa. Mineral. Deposita 30, 98-111.

Eriksson, P.G., Schreiber, U.M., Van der Neut, M. (1991) A review of the sedimentology of the early Proterozoic Pretoria Group, Transvaal Sequence, South Africa: implications for tectonic setting. J. Afr. Earth Sci. 10, 107-119.

Eriksson, K.A. (1980) Hydrodynamic and paleogeographic interpretation of turbidite deposits from the Archean Fig Tree Group of the Barberton Mountain Land, South Africa. Geol. Soc. Amer. Bull. 91, 21-26. 403 Eriksson, P.G., Schweitzer, J.K., Bosch, P.J.A., Schreiber, U.M., Van Deventer, J.L., and Hatton, C.J. (1993) The Transvaal Sequence: an overview. J. Afr. Earth Sci. 16, 25- 51.

Geological Survey of Western Australia (1990) Geology and mineral resources of Wesern Australia. W. Aus. Geol. Surv., Memoir 3, 827p.

Hammerbeck, E.C.I. (1986) Gold outside the Witwatersrand triad. In Mineral resources of the Republic of South Africa (ed. C.B. Coetzee), Geol. Surv. S. Afr., Pretoria, 75-92 pp.

Harley, M. and Charlesworth, E.G. (1996) The role of fluid pressure in the formation of bedding-parallel, thrust-hosted gold deposits, Sabie-Pilgrim's Rest goldfield, eastern Transvaal. Precam. Res. 79, 125-140.

Hassler, S.W. (1993) Depositional history of the Main Tuff interval of the Wittenoom Formation, late Archean-early Proterozoic Hamersley Group, Western Australia. Precam. Res. 60, 337-359.

Heubeck, C. and Lowe, D.R.(1994a) Depositional and tectonic setting of the Archean Moodies Group, Barberton Greenstone Belt, South Africa. Precam. Res. 68, 257- 290.

Heubeck, C. and Lowe, D.R.(1994b) Late syndepositional deformation and detachment tectonics in the Barberton Greenstone Belt, South Africa. Tectonics 13, 1514-1536.

Hickman, A.H. (1983) Geology of the Pilbara Block and Its Environs. Geol. Surv. W. Aus. Bull. 127. 268p.

Horwitz, R.C. (1982) Geological history of the Early Proterozoic Paraburdoo hinge zone, Western Australia. Precam. Res. 19, 191-200.

Horwitz, R.C. (1987) Structural trends of the Archean to Lower Proterozoic Hamersley Province, Western Australian Shield. CSIRO Division of Minerals and Geochemistry, Rep. MG 31.

Horwitz, R.C. (1990) Palaeogeographic and tectonic evolution of the Pilbara Craton, Northwestern Australia. Precam. Res. 48, 327-340.

Hunter, D.R. (1974) Crustal development in the Kaapvaal Craton, I. The Archaean. Precam. Res. 1, 259-294.

Hunter, D.R. and Hamilton, P.J. (1978) The Bushveld Complex. In Evolution of the Earth's Crust (ed. D.H. Tarling), Academic Press, London, pp. 107-173. 404 Jackson, M.P.A., Eriksson, K.A., and Harris, C.W. (1987) Early Archean foredeep sedimentation related to crustal shortening: a reinterpretation of the Barberton Sequence, southern Africa. Tectonophys. 136, 360-366.

Jahn, B.M., Bertrand-Sarfati, J., and Macé, N.M.J. (1990) Direct dating of stromatolitic carbonates from the Schmidtsdrif Formation (Transvaal Dolomite), South Africa, with implications on the age of the Ventersdorp Supergroup. Geology 18, 1211- 1214.

Kamo, S.L. and Davies, D.W. (1994) Reassessment of Archean crustal development in the Barberton Mountain Land, South Africa, based on U-Pb dating. Tectonics 13, 167- 192.

Knoll, A.H. and Barghoorn, E.S. (1977) Archean microfossils showing cell division from the Swaziland System of South Africa. Science 198, 396-398.

Kröner, A., Byerly, G.R., and Lowe, D.R. (1991) Chronology of Early Archean granite- greenstone evolution in the Barberton Mountain Land, South Africa, based on precise dating by single zircon evaporation. Earth Planet. Sci. Lett. 103, 41-54.

Kröner, A., Hegner, E., Wendt, J.I., and Byerly, G.R. (1996) The oldest part of the Barberton granitoid-greenstone terrain, South Africa: Evidence for crustal formation between 3.5 and 3.7 Ga. Precam. Res. 78, 105-124.

Layer, P.W., Kröner, A., and York, D. (1992) Pre-3000 Ma thermal history of the Archean Kaap Valley pluton, South Africa. Geology 20, 717-720.

Lowe, D.R. and Byerly, G.R. (1999) Geologic Evolution of the Barberton Greenstone Belt, South Africa. Geol. Soc. Amer. Spec. Paper 329, 319p.

Lowe, D.R. and Nocita, B.W. (1999) Foreland basin sedimentation in the Mapepe Formation, southern-facies Fig Tree Group. In Geologic Evolution of the Barberton Greenstone Belt, South Africa (eds. D.R. Lowe and G.R. Byerly). Geol. Soc. Amer. Spec. Paper 329, 233-258.

Martin, D. McB., Li, Z.X., Nemchin, A.A., and Powell, C.M. (1998) A pre-2.2 Ga age for giant hematite ores of the Hamersley Province, Australia? Econ. Geol. 93, 1084- 1090.

Martin, D. McB. (1999) Depositional setting and implications of Paleoproterozoic glaciomarine sedimentation in the Hamersley Province, Western Australia. Geol. Soc. Amer. Bull. 111, 189-203. 405 McCourt, S. (1995) The crustal architecture of the Kaapvaal crustal block South Africa, between 3.5 and 2.0 Ga. Mineral. Deposita 30, 89-97.

Minter, W.E.L. (1978) A sedimentological synthesis of placer gold, uranium, and pyrite concentrations in Proterozoic Witwatersrand sediments. Can. Soc. Petrol. Geol. Memoir 5, 801-829.

Miyano, T. and Beukes, N.J. (1987) Physicochemical environments for the formation of quartz-free manganese oxide ores from the Early Proterozoic Hotazel Formation, Kalahari Manganese Field, South Africa. Econ. Geol. 87, 706-718.

Morris, R.C. (1980) A textural and mineralogical study of the relationship of iron ore to banded iron-formation in the Hamersley Iron Province of Western Australia. Econ. Geol. 75, 184-209.

Morris, R.C. (1983) Supergene alteration of banded iron-formation. In Iron-Formations: Facts and Problems (eds, Trendall, A.F. and R.C. Morris), Elsevier Sci. Pub., Amsterdam, 513-534.

Morris, R.C. (1993) Genetic modelling for banded iron-formation of the Hamersley Group, Pilbara Craton, Western Australia. Precam. Res. 60, 243-28.

Morris, R.C. and Horwitz, R.C. (1983) The origin of the iron-formation-rich Hamersley Group of Western Australia - deposition on a platform. Precam. Res. 21, 273-297.

Nocita, B.W. (1989) Sandstone petrology of the Archean Fig Tree Group, Barberton Greenstone Belt, South Africa: Tectonic implications. Geology 17, 953-956.

Nocita, B.W. and Lowe, D.R. (1990) Fan-delta sequence in the Archean Fig Tree Group, Barberton Greenstone Belt, South Africa. Precam. Res. 48, 375-393.

Phillips, G.N. (1986) Metamorphism of shales in the Witwatersrand Goldfields. Univ. Witwatersrand Econ. Geol. Res. Unit, Info. Circ. 192, 25p.

Phillips, G.N., Myers, R.E., Law, J.D.M., Bailey, A.C., Cadle, A.B., Beneke, S.D., and Giusti, L. (1989) The Witwatersrand gold fields: Part I. Postdepositional history, synsedimentationary processes, and gold distribution. Econ. Geol. Monogr. 6, 585- 597.

Pidgeon, R.T. and Horwitz, R.C. (1991) The origin of olistoliths in Proterozoic rocks of the Ashburton Trough, Western Australia, using zircon U-Pb isotopic characteristics. Aust. J. Earth Sci. 38, 55-63. 406 Powell, C.M., Oliver, N.H.S., Li, Z.-X., Martin, D.M., and Ronaszeki, J. (1999) Synorogenic hydrothermal origin for giant Hamersley iron oxide ore bodies. Geology 27, 175-178.

Pretorius, D.A. (1976) The nature of the Witwatersrand gold-uranium deposits. In Handbook of Strat-Bound Stratiform Ore Deposits (ed. K.H. Wolf), Elsevier, Amsterdam, pp 29-88.

Pretorius, D. A. (1981) The Witwatersrand Supergroup. In Precambrian of the Southern Hemisphere (ed. D.R. Hunter), Elsevier, Amsterdam, Netherlands, 511-520.

Robb, L.R., Davies, D.W., and Kamo, S.L. (1990) U-Pb ages on single detrital zircon grains from the Witwatersrand Basin: constraints on the age of sedimentation and on the evolution of granites adjacent to the basin. J. Geol. 98, 311-328.

Robb, L.R., Davies, D.W., and Kamo, S.L. (1991) Chronological framework for the Witwatersrand Basin and environs: towards a time-constrained depositional model. S. Afr. J. Geol. 94, 86-95.

Robb, L.J. and Meyer, F.M. (1995) The Witwatersrand Basin, South Africa: Geological framework and mineralization processes. Ore Geol. Rev. 10, 67-94.

SACS (South African Committee for Stratigraphy) (1980) Stratigraphy of South Africa: Part I: Lithostratigraphy of the Republics of South Africa, South West Africa/Namibia and Republics of Bophuthatswana, Transkei, and Venda. Geol. Soc. S. Afr. Handbook 8, 690p.

Schreiber, U.M. (1991) A paleoenvironmental study of the Pretoria Group, in the eastern Transvaal. Ph.D. thesis, Univ. of Pretoria, Pretoria, S. Afr., 308 pp.

Schreiber, U.M., Eriksson, P.G., Van der Neut, M., and Snyman, C.P. (1992) Sedimentary petrography of the Early Proterozoic Pretoria Group, Transvaal Sequence, South Africa: implications for tectonic setting. Sed. Geol. 80, 89-103.

Simonson, B.M., Schubel, K.A. and Hassler, S.W. (1993) Carbonate sedimentology of the early Precambrian Hamersley Group of Western Australia. Precam. Res. 60, 287- 335.

Smith, R.E., Perdrix, J.L., and Parks, T.C. (1982) Burial metamorphism in the Hamersley Basin, Western Australia. J. Petrol. 23, 75-102.

Simonson, B.M. and Hassler, S.W. (1997) Revised correlations in the Early Precambrian Hamersley Basin based in a horizon of resedimented impact spherules. Aus. J. Earth Sci. 44, 37-48. 407 Simonson, B.M., Davies, D., Wallace, M., Reeves, S., and Hassler, S.W. (1998) Iridium anomaly but no shocked quartz from Late Archean microkrystite layer: Oceanic impact ejecta? Geology 26, 195-198.

Söhnge, A.P.G. (1986) Mineral province of southern Africa. In Mineral Deposits of Southern Africa (eds. C.R. Anhaeusser and S. Maske), Geol. Soc. S. Afr., Johannesburg, pp 1-23.

Tankard, A.J., Jackson, M.P.A., Eriksson, K.A., Hobday, D.K., Hunter, D.R., and Minter, W.E.L. (1982) Crustal evolution of southern Africa, 3.8 billion years of the history. Springer, Berlin, 523 pp.

Thorne, A.M. (1990) Hamersley Basin. In 3rd International Archean Symposium, Perth (1990) Excursion Guidebook (eds. S.E. Ho, J.E. Glover, J.S. Myers, and J.R. Muhling), Univ. of W. Aust, Dept. of Geol. and Univ. Extension Pub. No. 21, 13- 14.

Thorne, A.M. and Blake, T.S. (1990) Fortescue Group. In 3rd International Archean Symposium, Perth (1990) Excursion Guidebook (eds. S.E. Ho, J.E. Glover, J.S. Myers, and J.R. Muhling), Univ. of W. Aust, Dept. of Geol. and Univ. Extension Pub. No. 21, 14-18.

Trendall, A.F. (1976) Striated and faceted boulders from the Turee Creek Formation - Evidence for a possible Huronian glaciation on the Australian continent. Geol. Surv. W. Aus., Ann. Rep. 1975, 88-92.

Trendall, A.F. (1979) A revision of the Mount Bruce Supergroup. Geol. Surv. West. Aust. Ann. Rep., 48-53.

Trendall, A.F. (1983) The Hamersley Basin. In Iron-Formations: Facts and Problems (eds. A.F. Trendall and R.C. Morris), Elsevier, Amsterdam, 69-129.

Trendall, A.F., Compston, W., Williams, I.S., Armstrong, R.A., Arndt, N.T., McNaughton, N.J., Nelson, D.R., Barley, M.E., Beukes, N.J., de Laeter, J.R., Retief, E.A. and Thorne, A.M. (1990). Precise U-Pb chronological comparison of the volcano- sedimentary sequences of the Kaapvaal and Pilbara between about 3.1 and 2.4 Ga. 3rd Int. Archean Symp. Extended Abstracts, 81-83.

Truswell, J.F. and Eriksson, K.A. (1975) Stromatolitic associations and their paleoenvironmental significance: a reappraisal of a lower Proterozoic locality from the northern Cape Province, South Africa. Sed. Geol. 10, 1-23. 408 Tyler, N. (1978) A stratigraphic analysis of the pre-Chuniespoort Group strata around the Makoppa Dome, west-central Transvaal. M.Sc. Thesis, Univ., Witwatersrand, South Africa. 206 pp.

Tyler, I.M. and Thorne, A.M. (1990) The northern margin of the Capricorn Orogen, Western Australia - an example of an early Proterozoic collision zone. J. Struct. Geol. 12, 485-702.

Tyler, R. and Tyler, N. (1996) Stratigraphic and structural controls on gold mineralization in the Pilgrim's Rest goldfield, eastern Transvaal, South Africa. Precam. Res. 79, 141- 169.

Viljoen, M.J. and Viljoen, R.P. (1969) An introduction to the geology of the Barberton granite-greenstone terrain. geol. Soc. South Afr. Spec. Pub. 2, 9-28.

Wagener, J.F.H. (1986) The Agnes gold mine, Barberton greenstone belt. In Mineral Deposits in Southern Africa (C.R.Anhaeusser and S.Maske, eds.), vol. 1, Geol. Soc. South Afr., Johannesburg, 181-185.

Walsh, M.M. (1992) Microfossils and possible microfossils from the Early Archean Onverwacht Group, Barberton Mountain Land, South Africa. Precam. Res. 54, 271- 293.

Walsh, M.M. and Lowe, D.R. (1985) filamentous microfossils from the 3,500-Myr-old Onverwacht Group, Barberton Mountain Land, South Africa. Nature 314, 530-532.

Walvaren, F. and Martini, J. (1995) Zircon Pb-evaporation age determinations of the Oak Tree Formation, Chuniespoort Group, Transvaal Sequence: implications for Transvaal-Griqualand West basin correlations. S. Afr. J. Geol. 98, 58-67.

Walvaren, F., Armstrong, R.A., and Kruger, F.J. (1990) A chronostratigraphic framework for the north-central Kaapvaal Craton, the Bushveld Complex and the Vredefort structure. Tectonophys. 171, 23-48.

Walvaren, F and Hattingh, E. (1993) Geochronology of the Nebo Granite, Bushveld Complex. S. Afr. J. Geol. 96, 31-41.

Ward, J.H.W. (1995) Geology and metallogeny of the Barberton greenstone belt: a survey. J. Afr. Earth Sci. 21, 213-240.

Windley, B.F. (1995) The Evolving Continents. 3rd ed. John Willey and Sons, 526 pp.

Winter, H. de la R. (1976) A lithostratigraphic characteristics of fluvial deposits. Soc. Econ. Paleont. Miner. Spec. Pub. 16, 84-97. 409 Woodhead, J.D., Hergt, J.M., and Simonson, B.M. (1998) Isotopic dating of an Archean bolide impact horizon, Hamersley Basin, Western Australia. Geology 26, 47-50.

Wronkiewicz, D.J. and Condie, K.C. (1990) Geochemistry and mineralogy of sediments from Ventersdorp and Transvaal Supergroup, South Africa: Cratonic evolution during the early Proterozoic. Geochim. Cosmochim. Acta 54, 343-354. 410

20˚E 30˚E N MOÇANBIQUE Sabie - Pilgrim's Rest BOTSWANA region

Barberton 25˚S Greenstone Belt Pretoria NAMIBIA

Witwatersrand Basin SWAZILAND LESOTHO 30˚S Atlantic Ocean Durban

SOUTH AFRICA Transvaal Supergroup

Ventersdorp Supergroup

Witwatersrand Supergroup Cape Town Barberton Greenstone Belt Indian Ocean 400 km

Fig. A-1. Geologic map of South Africa. Only surface / near-surface exposures of Archean and Paleoproterozoic rocks from the Swaziland, Witwatersrand, Ventersdorp, and Transvaal Supergroups are shown. Made from SACS (1980). 411 30' ˚ SOUTH AFRICA

31 SWAZILAND Fig Tree Group Syanite, granodiorite Trondjemite Intrusion Moodies Group Tonalite Intrusion (Kaap Valley Pluton) Onverwacht Group

20 km

Barberton Greenstone Belt STENTOR PLUTON STENTOR 10 ULUNDI SYNCLINE 0 EUREKA SYNCLINE ˚ 31 NELSPRUIT BATHOLITH BARBERTON S S ˚ ˚ 25 30 KAAP VALLEY PLUTON 400 km

MOÇANBIQUE Durban E

SWAZILAND ˚

30

a

ri

o

t

e

r P Barberton Greenstone Belt Ocean Indian

LESOTHO Pretoria E ˚ NELSHOOGTE PLUTON

BOTSWANA 20 30' ˚ N ˚

Cape Town

TRANSVAAL GROUP SEDIMENTS GROUP TRANSVAAL SOUTH AFRICA

26

25 30' ˚ 30 NAMIBIA Atlantic Ocean Fig. A-2. Geologic map of the Barberton Greenstone Belt. Modified after Anhaeusser (1986). 412

26˚00' 27˚00' 28˚00'Pretoria 29˚00'

26˚00'

Johannesburg

MSF6

DRH13 Potchefstroom

Klerksdorp Orkney 27˚00' Vredefort

50 km

28˚00'

THE WITWATERSRAND BASIN

Witwatersrand Supergroup Central Rand Group West Rand Group

29˚00'

Fig. A-3. Geologic map of the Witwatersrand Basin. Drilling sites of drillcores MSF6 and DRH13 are indicated with open circles in the north-western margin of the Basin. Modified after Barnicoat et al. (1997). age data. are notproportionaltothoseshowninthefigure. Seetextforreferencesof where thesamplesweretakenforthisstudy.Actual thicknessoftheGroups Rand, Platberg,Wolkberg,Chuniespoort,andPretoria Groups)indicate Ventersdorp, andTransvaalSupergroups.Shaded Groups(FigTree,West Fig. A-4.LithostratigraphiccolumnoftheSwaziland, Witwatersrand,

Swaziland Supergroup Dominion Witwatersrand Ventersdorp Transvaal Supergroup Supergroup Supergroup Supergroup Onverwacht Fig Tree Moodies Platberg Pniel Pretoria Rand West Rand Central Wolkberg poort Chunies berg Kliprivers Group v v v + v v + v v v v v + v v v v v v v + + 3,074 Ma 2,914 Ma 2,714 Ma 2,557 Ma 2,224 Ma 3,44 3 ,230 Ma 5 Ma Legend ++ v v granite/granitoid mafic lava dolomite shale/siltstone graywacke conglomerate quartzite 413 414

20 ˚S N 117 ˚E 120 ˚E Port Hedland

Karratha

Ripon Hill

Wittenoom

23 ˚S

Turee Creek Group Mt. Bruce Hamersley Group Supergroup Fortescue Group Perth Granite-granitoid Archean Mafic basement Basement

Fig. A-5. Geologic map of the Pilbara Craton, Western Australia. Modified after Geological Survey of Western Australia (1990). 415

Central-western Hamersley Ages Formation

Woongarra v v Volcanics v v Weeli Wooli Iron

Brockman 2,470 Ma Iron

Mt.McRae, Mt.Sylvia Shale

Witteonoom Eastern Hamersley

Hamersley Group Dolomite 2,603 Ma Formation Age

Marra Carawine Mamba Dolomite Iron 2,548 Ma 2,687 Ma Jeerinah Lewin Mt. Bruce Supergroup v v Shale v Maddina v v Basalt v v v v v

Fortescue Group Legend iron-formation

500 m Pilingini shale/siltstone Tuff v v mafic lava dolomite

Fig. A-6. Lithostratigraphic column of the middle Mt. Bruce Supergroup. Only the lower Hamersley Group and the upper Fortescue Group are shown. Shaded Formations of Wittenoom Dolomite, Marra Mamba Iron, Jeerinah, and Pillingini Tuff indicate where the samples were collected for this study. Pilingini Tuff Formation is thought to be equivalent to the Tumbiana Formation. Made from Geological Survey of Western Australia (1990) and CRA Exploration unpublished report 13140 and 15041. Appendix B

Samples

B-1. Introduction

While geologic samples from outcrops are generally altered by recent weathering and not particularly suitable for geochemical analysis, the drillcore samples recovered from deep in the ground are relatively modern weathering-free; therefore the drillcore samples are much more suitable for geochemical analyses than outcrop samples. In this study, drillcore samples of Archean and Paleoproterozoic black shales were used. These samples were collected in South Africa and Australia (see Appendix A for their geological settings). In Appendix B, information on the samples is provided including (1) sampling, (2) drillcore locality, (3) age, (3) Group / Formation, (4) depth of the drillcores where the samples were taken, and (5) dominant lithology. Lithostratigraphic columns and simplified descriptions of individual samples are also provided.

B-2. Sampling method

The drillcore samples were collected by the author during fieldwork conducted in South Africa and Australia in 1996, 1998, and 1999. The samples are recovered from 417 drilling by mining companies in 1970's and 1980's, and had been stored in their coreyards until the collection. More than 100 black shale samples from 12 Formations have been analyzed by the author for geochemical investigation. South African samples include the ~3.2 Ga Fig Tree (drillcore PU1308 and MRE10), the ~2.9 Ga West Rand (drillcore DRH13), the ~2.7 Ga Platberg (drillcore MSF6), the ~2.6 Ga Wolkberg (drillcore JPBR), the ~2.5 Ga Chuniespoort (drillcore MSF6), and the ~2.2 Ga Pretoria Groups (drillcore MSF6, PTB3, and SA1677). Australian samples include the ~2.7 Ga Fortescue (drillcore WRL1 and RHDH2A) and the ~2.5 Ga Hamersley Groups (drillcore WRL1 and RHDH2A). Drillcore samples were examined at the coreyards in the Barberton, Transvaal, Witwatersrand, and Griqualand West regions of South Africa during fieldwork in 1996 and 1998, and in the Pilbara region of Australia during fieldwork in 1999. After examination, the samples were collected by the author and used for this study. Table B-1 summarizes the localities of drillholes, ages, Formation / Groups, and dominant lithology of the samples. During drillcore sampling, efforts were made to select representative samples from each stratigraphically characteristic unit of the drillcores, to make even spacing between samples where possible, and to avoid fragmented or apparently modern-weathered samples. Diameter of the drillcore samples is typically between 3 and 7 cm. Long drillcore samples were cut by a hammer into a small specimens of about 5~10 cm length. Samples were named and the sample IDs were marked on their surface by a marker with arrows indicating stratigraphical upward direction, stored in sealed plastic bags on which sample IDs were labeled, and shipped to The Pennsylvania State University. 418

B-3. Drillcore PU1308

Mesoarchean black shale samples were collected from drillcore PU1308 at the Agnes gold mine in the Barberton Greenstone Belt, South Africa (Wagener, 1986; Fig. A-1, A-2). The Agnes gold mine is located 11 km south-west of Barberton and is situated on the north-western flank of the Barberton Greenstone Belt (Fig. A-2). The samples belong to the lowermost part of the ~3.2 Ga Sheba Formation, which is the lowermost succession of the Fig Tree Group, the Swaziland Supergroup (Fig. B-1). Although the beds of the Barberton Greenstone Belt at the Agnes mine (and elsewhere) have been tilted to a near vertical position (Wagener, 1986), sedimentary textures are often well preserved in the rocks. The shale samples have undergone only a minor metamorphism less than or equal to the lower greenschist facies (Viljoen and Viljoen, 1969). After examination of the collected rock specimens, 17 samples were selected from the upper ~45 m section of drillcore PU1308, whose total length is 89.32 m. They were used in further geochemical and isotopic analyses based on the criteria that the samples contained neither vein-rich, silicified, nor significantly deformed parts. The samples examined in this study were free of Au mineralization. The samples are finely laminated,

typically showing an alternation of Corg-rich and Corg-poor parts. Some samples contain disseminated and/or layered pyrite. Effort was made not to collect samples containing such mineralized layers of pyrite. Some samples contain a substantial amount of iron carbonate (siderite), which has been confirmed by XRD and petrographic microscope. One sample is more enriched in silica compared to the other samples. The uppermost part of drillcore PU1308 (not used in this study) exhibits textures of soft sediment deformation and fuchsite, a mineral of Cr-rich clay minerals. 419

B-4. Drillcore MRE10

Mesoarchean graywacke samples were collected from drillcore MRE10 at the Sheba gold mine in the Barberton Greenstone Belt, South Africa (Wagener and Wiegand, 1986; Fig. A-1, A-2). The Sheba gold mine is located 12 km north-east of Barberton (Fig. A-2). The graywacke samples belong to the ~3.2 Ga Sheba Formation of the Fig Tree Group, the Swaziland Supergroup (Fig. B-1). Sedimentary textures are well preserved in the samples. The samples have undergone only a minor metamorphism less than or equal to the lower greenschist facies (Viljoen and Viljoen, 1969). The exact stratigraphic position of the samples from drillcore MRE10 relative to those from drillcore PU1308 (see the above section B-2) is unclear. About 90 samples were collected from the ~300 m long drillcore MRE10. They are free of Au mineralization. The samples are generally enriched in lithic, angular fragments supported by a fine-grained matrix. Graywackes in drillcore MRE10 irregularly alternate with black shales at various thickness scales (cm), with the former mostly dominant. Transitions between graywackes and black shales show both gradational and sharp contacts. Angular fragments of black shales are found in the graywackes. Graywackes and black shales in drillcore MRE10 are generally pyrite-lean; however, in rare cases pyrite is found in some black shale units. After examination of the collected rock specimens, 12 (shale part-free) graywacke samples were selected from the upper 150 m section of the core for further geochemical and isotopic analyses. 420

B-5. Drillcore DRH13

Neoarchean black shale samples were collected from drillcore DRH13 which was recovered from the farm "Rhnosterhoek 299-IP" near Klerksdorp in the north-western margin of the Witwatersrand Basin, South Africa (Fig. A-1, A-3). The drilling was conducted by Anglo American Prospecting Service, LTD. The black shale samples belong to the 2.96 Ga Parktown Formation of the Hospital Hill Subgroup, West Rand Group, Witwatersrand Supergroup (Fig. A-4, B-2). The samples have undergone a minimal grade of metamorphism and well preserved sedimentary textures such as fine-scale laminations. About 130 shale samples were collected from the ~800 m long drillcore DRH13 (Fig. B-2). The majority of the samples are finely-laminated black shales, with minor massive layers of black claystone / siltstone without clear lamination. The uppermost part of the drillcore (top 100 m) is composed of conglomerate overlain by lava. Below this level, there are several shale units which are separated by lava flows, some of which occupy depth ranges of ~400 - ~415 m and ~455 - ~480 m. In this study, ~5 lithologically representative samples were chosen from each of three shale units (15 samples in total) in the Parktown Formation and used for further geochemical and isotopic analyses. The upper shale unit ranges in depth from ~370 to ~400 m, the middle shale unit is between ~415 to ~455 m, and the lower shale unit range from ~485 to 495 m (Fig. B-2). While pyrite are rare and an alternation of laminated and non- laminated (massive) black shales are observed throughout the upper and the middle shale units, the lower unit contains disseminated pyrite and is mostly composed of massive black shales (rare laminated ones). It should be noted that the samples from drillcore DRH13 are generally rich in (ferrous) iron, most likely due to the abundance of clays (chlorite). 421

B-6. Drillcore MSF6

The long drillcore MSF6 (reaching 4,390 m) was recovered near Klerksdorp in the north-western margin of the Witwatersrand Basin (Fig. A-3). The drillcore covers part of the Paleoproterozoic Transvaal Supergroup and the Neoarchean Ventersdorp Supergroup (Fig. A-4). In the Transvaal Supergroup, drillcore MSF6 contains the Timeball Hill (~500 to ~900 m depth) and Rooihoogte (~900 to ~1,000 m depth) Formations of the Pretoria Group, and the Eccles (~1,000 to ~1,400 m depth), Lyttleton (~1,400 to ~1,570 m depth), Monte Christo (~1,570 to 2,113 m depth), and Oak Tree (2,113 to 2,317 m depth) Formations of the Chuniespoort Group (Fig. B-3). In the Ventersdorp Supergroup, the drillcore contains the Black Reef Formation (2,317 m to 2,352 m depth) of the Wolkberg Group, the Rietgat Formation (2,352 to 3,141 m depth) of the Platberg Group, and Goedgenoeg (3,141 to > 3,560 m depth), Kameelsdoorm (< 3,625 to > 3,640 m depth), Elsburg Quartzite (< 3,800 to > 3,900 m depth), and Kimberley Conglomerate (> 3,900 m depth) Formations of the Klipriviersberg Group (Fig. B-3). From drillcore MSF6, nearly 200 samples in total (including black shales, carbonate, conglomerate, and lava) were collected at the Orkney coreyard of the Anglo American Prospecting Service, Ltd. in Klerksdorp. Out of 200 samples, 4 samples from the 2.22 Ga Timeball Hill Formation, 3 samples from the 2.56 Ga Oak Tree Formation, and 7 samples from the 2.71 Ga Rietgat Formation were selected and chemically analyzed for this study. All of these samples are black shales; carbonate, conglomerate, and lava were avoided. The samples examined in this study were not subjected to deformation and only suffered from relatively low-grade regional metamorphism (greenschist facies). 422

B-7. Drillcore JPBR

Drillcore JPBR was recovered in the north-western margin of the Witwatersrand Basin (information of exact location unavailable). Drillcore JPBR covers the 2.64 Ga Black Reef Formation of the Wolkberg Group in the uppermost Ventersdorp Supergroup (Fig. B- 3). The Black Reef Formation is mostly composed of quartzite and conglomerate, with minor abundance of shale and carbonate rocks. The samples used in this study came from the upper Black Reef Formation, with a drillcore depth ranging from 2,344 to 2,452 m (~100 m thick unit). In total, 15 samples were collected from the drillcore, and 4 samples were analyzed for geochemical and isotopic investigation in this study. The upper two samples are carbonate-rich black shale without clear lamination, and the lower two samples show lamination of black and dark gray layers. The lowermost sample is not enriched in carbonate and contains pyrite grains ~1 mm in diameter.

B-8. Drillcore PTB3

Drillcore PTB3 was recovered by Transvaal Gold Mine Estate (TGME), Ltd. in the Sabie - Pilgrim's Rest region of eastern Transvaal (Fig. A-1). The drillcore contains part of the Timeball Hill Formation of the Pretoria Group, Transvaal Supergroup (Fig. B-3). In this region, Au mineralization occurs in the underlying Chuniespoort Group. The examined section of the drillcore is the upper ~600 m, and the black shales of the 2.22 Ga Timeball Hill Formation are found in the depth range from ~150 to ~280 m. In total, 40 samples, including 25 samples from the black shale unit, were collected from drillcore MSF6. The 12 samples were selected from the black shale unit and used for further geochemical and isotopic investigation in this study. The used samples are mostly 423 finely laminated black shales, some of which contain disseminated pyrite. The samples examined in this study were not subjected to Au mineralization and deformation, and only suffered from relatively low-grade regional metamorphism (greenschist facies).

B-9. Drillcore SA1677

This study focuses on the geochemistry of black shales. However, some drillcore samples of red shales were also used in this study to compare their elemental and isotopic geochemical characteristics with those of black shales. Drillcore SA1677 was recovered from the Sishen iron mine, in the Griqualand West region / Northern Cape Province of South Africa (Fig. A-1). The drillcore covers the Pauling Member of the Gamagara Formation, or, depending on the stratigraphy of local geology, the Mapedi Shale Formation of the Olifantshoek Group (Fig. B-4, B-5). The examined section of the ~250 m long drillcore SA1677 ranges in depth from 200 to 245 m. It contains fine-grained red shales and minor red sandstones, with conglomerate and breccia in the lowermost part of the examined section. Some of the red shales alternate with minor white (bleached) layers of variable thickness (mm - cm). In total, 23 samples were collected from drillcore SA1677, and 7 samples were used in this study for geochemical and isotopic investigation. Only red shales were used from drillcore SA1677. The red shale samples examined in this study were not subjected to deformation and only suffered from relatively low-grade regional metamorphism (greenschist facies). 424

B-10. Drillcore WRL1

Drillcore WRL1 was recovered by CRA Exploration, LTD. near Wittenoom in the Pilbara Craton of north-western Australia (Fig. A-5). The drillcore is 1,963 long, and covers part of the Neoarchean-Paleoproterozoic Mt. Bruce Supergroup (Fig. A-6, B-6). The drillcore contains the Wittenoom Dolomite (0 to 525 m in depth) and Marra Mamba Iron (524 to 679 m) Formations of the Hamersley Group, and Jeerinah (679 to 779 m), Maddina Basalt (779 to 1,274 m), Pillingini Tuff (1,274 to 1563 m), Kylena Basalt (1,563 to 1,812 m), and Hardey Sandstone (1,812 to 1,911 m) Formations of the underlying Fortescue Group (Fig. B-6). The drillcore was previously used to successfully extract Archean cyanobacterial and eukaryotic biomarkers from the Jeerinah Formation (Brocks et al., 1999). The finely-laminated black shale samples used in this study were collected from the 2.72 Ga Pillingini Tuff Formation, the 2.69 Ga Jeerinah Formation, the >2.60 Ga Marra Mamba Iron Formation, and the 2.60 Ga Wittenoom Dolomite Formation (Fig. A-6, B-6). From the drillcore, 6 samples from the Wittenoom Dolomite Formation, 8 samples from the Marra Mamba Iron Formation, and ~20 samples from the Jeerinah Formation were collected from a 5 m section, a 70 m section, and a 50 m section, respectively, of drillcore WRL1 at the coreyard of the CRA Exploration in Karratha, Western Australia (Fig. B-6). From the collected samples, 3 samples from the Wittenoom Dolomite Formation, 4 samples from the Marra Mamba Iron Formation, 6 samples from the Jeerinah Formation, and 3 samples from the Pillingini Tuff Formation were selected for further geochemical and isotopic investigation in this study (Fig. B-6). The selected samples are mostly black shales. Samples from the Jeerinah Formation contain abundant disseminated, nodular, and layered pyrite. One sample from the Wittenoom Dolomite Formation is rich in carbonate. 425

The black shale samples from drillcore WRL1 examined in this study were not subjected to deformation and only suffered from relatively low-grade regional metamorphism (greenschist facies).

B-11. Drillcore RHDH2A

The drillcore RHDH2A was recovered by CRA Exploration, LTD. near Ripon Hill in the eastern edge of the Pilbara Craton of north-western Australia (Fig. A-5). The drillcore is 501 m long, and covers part of the Neoarchean-Paleoproterozoic Mt. Bruce Supergroup (Fig. A-6, B-7). The drillcore contains the Carawine Dolomite (70 to 254 m in depth) and Lewin Shale (254 to 501 m in depth) Formations (Fig. A-6, B-7). The Carawine Dolomite Formation has been thought to be stratigraphically (shallow facies) equivalent to the deep facies Wittenoom Dolomite Formation, and the Lewin Shale has also been thought to be stratigraphically equivalent to the Jeerinah Formation (Fig. A-6). Drillcore RHDH2A was previously used for S isotope study (Bottomley et al., 1992). About 10 and 20 samples of the 2.60 Ga Carawine Dolomite and the 2.69 Ga Lewin Shale Formations, respectively, were collected from drillcore RHDH2A at the coreyard of the CRA Exploration in Karratha, Western Australia (Fig. A-5). Out of ~30 collected samples, 6 samples from the Carawine Dolomite and 8 samples from the Lewin Shale Formation were selected for further geochemical and isotopic investigation in this study (Fig. B-7). Samples from the Lewin Formation are finely laminated and contain pyrite. The term "Lewin Shale Formation" is used here instead of the stratigraphically equivalent "Jeerinah Formation", in order to distinguish between the two. 426

The black shale samples from drillcore RHDH2A examined in this study were not subjected to deformation and only suffered from relatively low-grade regional metamorphism (greenschist facies). 427

References

Anhaeusser, C.R. (1986) Archean gold mineralization in the Barberton Mountain Land. In Mineral Deposits of Southern Africa (eds. C.R. Anhaeusser and S. Maske), Geol. Soc. S. Afr., Johannesburg, pp 113-154.

Beukes, N.J. and Smit (1987) New evidence for thrust faulting in Griqualand West, South Africa: implications for stratigraphy and the age of red beds. S. Afr. J. Geol. 90, 378-394.

Bottomley, D.J., Veizer, J., Nielsen, H., and Moczydlowska, M. (1992) Isotopic composition of disseminated sulfur in Precambrian sedimentary rocks. Geochim. Cosmochim. Acta 56, 3311-3322.

Brocks, J.J., Logan, G.A., Buick, R., and Summons, R.E. (1999) Archean Molecular Fossils and the Early Rise of Eukaryotes. Science 285, 1033-1036.

CRA Exploration (1984) Log of the drillcore RHDH2A. Unpu. Report No. 13140.

CRA Exploration (1984) Log of the drillcore WRL1. Unpub. Report No. 15041.

Viljoen, M.J. and Viljoen, R.P. (1969) An introduction to the geology of the Barberton granite-greenstone terrain. Geol. Soc. South Afr. Spec. Pub. 2, 9-28.

Wagener, J.H.F. (1986) The Agnes gold mine, Barberton Greenstone Belt. In Mineral Deposits of Southern Africa (eds. C.R. Anhaeusser and S. Maske), Geol. Soc. S. Afr., Johannesburg, pp 181-185.

Wagener, J.H.F. and Wiegand, J. (1986) The Sheba gold mine, Barberton Greenstone Belt. In Mineral Deposits of Southern Africa (eds. C.R. Anhaeusser and S. Maske), Geol. Soc. S. Afr., Johannesburg, pp 155-161. 428

Moodies conglomerate Belvue Road Quartzite Formation tuff

Fig Tree Greywacke, shale, chert Onverwacht

Mafic, ultramafic, felsic lava with minor shale terrigenous graywacke sediment chert Schoongezicht Formation Swaziland Supergroup chert

graywacke and shale

Sheba Formation banded ferrugenous chert 2 km

Fig. B-1. Lithostratigraphic column of the ~3.2 Ga Fig Tree Group, the Swaziland Supergroup, Barberton Greenstone Belt, South Africa. Only dominant lithologies are shown. Samples of black shales and graywackes were collected from drillcores PU1308 and MRE10, respectively, covering the Sheba Formation. 429

Core depth (m) v v v v Ventersdorp Supergroup 200 ~2.9 Ga Witwatersrand Supergroup West Rand Group Hospital Hill Subgroup

300

Samples Legend

v v lava 400 conglomerate shale quartzite siltstone + + + + + + tuff ++ ++ + + + intrusive 500 + + + + + + + ~3.1 Ga Dominoin Group

Fig. B-2. Lithostratigraphic column of the ~2.9 Ga Parktown Formation in the Hospital Hill Subgroup of the West Rand Group,Witwatersrand Supergroup, South Africa. Samples were collected from three shale units (indicated by bars) in drillcore DRH13. Core depth (in meters) is indicated to the left of the column. 430

Formation v Hekpoort v v v v v Andesite + + +++

Timeball Hill MSF6, PTB3 Pretoria Group

Rooihoogte

Eccles

Lyttelton Transvaal Supergroup 500 m

Monte Cristo Chunispoort Group

Oak Tree MSF6 Black Reef JPBR

MSF6 Rietgat

Legend v v v v v v

Platberg Group Goedgenoeg quartzite v v v v v v dolomite Kameelsdoorn v v vv v Brecciated Central v v v v v v shale Rand Rocks + + intrusive

Ventersdorp Supergroup Elsburg Quartzite Kimberley v v volcanics Conglomerate Livingstone conglomerate Conglonerate fault zone Florida Quartzite

Fig. B-3. Lithostratigraphic column of the Transvaal and Ventersdorp Supergroups, South Africa. Samples used in this study (drillcores MSF6, JPBR, and PTB3) were taken from the shaded (Timeball Hill, Oak Tree, Black Reef, and Rietgat) Formations. Approxomate Group Subgroup Formation Lithology thickness (m)

Volop Hartley Quartzite 3500

Formation Andesitic lava 700

Lucknow Quartzite 450 Olifantshoek Mapedi Shale, quartzite, lava 10-1500

Mooidraai Dolomite, chert Voëlwater 250 Hotazel Iron-formation, manganese lava

Ongeluk Andesitic lava 900

Postmasburg Makganyene Diamictite 50-150

Rooinekke Iron-formation 100

Naragas Quartz wacke, shale Koegas Kwalwas Riebeckite slate 240-600 Doradale Iron-formation Transvaal Supergroup Ghaap Pannetjie Quartz-wacke, shale

Griquatown Clastic-textured iron-formation 200-300 Asbesheuwels Kuruman Micorbanded iron-formation 150-750

Fig. B-4. Lithostratigraphy of a Paleoproterozoic succession in Griqualand West, South Africa. Samples of drillcore SA1677 are from the Mapedi Formation (shaded). From Beukes and Smit (1987). 431 Group Formation Member Lithology ~thickness (m) Volop Quartzite >4000 Amygdaloidal andesitic lavawith intercalated tuff, Hartley Andesite breccia, quartzite and conglomerate 760 Whitish quartzite with flagstone and subordinate layers Lucknow Quartzite of dolomitic limestone 170

Olifantshoek Mapedi Phyllitic shale with lava bands and quartzite layers Mooidraai Dolomite, chert, banded jasper and lava Voëlwater 340 Jasper Banded red jasper, chert, dolomite, banded ironstone, Hotazel manganiferous jasper and lava Ongeluk Andesitic and amygdaloidal lava with occasional bands of 400~900 Andesite red jasper and agglomerate Hartebeeshoek Siltstone Makganyene Diamictite, banded jasper, sandstone, grit, conglomerate, <500 Diamictite Bolham siltstone, mudstone, and dolomite Regional uncomformity Magoloring Banded jasper, diamictite, dolomite, sandstone Grey, greenish, black and reddish shale and ocasional Pauling thin quartize seams Gamagara Marthaspoort Fine-grained whitish and purplish massive quartzite 70~300 Shale Uncomformity Sishen Shale, flagstone, conglomerate Brown jaspilite, flagstone, quartzite, dolomite, and Kwakwaas Koegas banded ironstone Griqualand West Sequence 200~560 Jappilite Mudstone, amphibolite, quartzite,jaspilite with Middelwater crocidolite and conglomerate Brown jaspilite and crocidolite with amphibolite and Asbestos Hills Daniëlskuil clastic sediments at top 350~820

GriquatownBanded Ironstone Cox Minesotanite-riebeckite Kuruman Ironstone with amphibolite bands and crocidolite

Fig. B-5. Lithostratigraphy of a Paleoproterozoic succesion in Griqualand West, South Africa. Samples of

drillcore 1677 are from the Pauling Member of the Gamagara Formation (shaded). Modified after SACS (1980). 432 433

Core depth Formation 0 m

Wittenoom Dolomite

500 Hamersley Group Marra Mamba Iron

Jeerinah v v v v 1000 Maddina Basalt v v v v v Mt. Bruce Supergroup v

Pillingini Tuff Fortescue Group 1500 v v v Kylena Basalt v v v Hardey Sandstone Archean Basement 1963.50 + + ++

Fig. B-6. Lithostratigraphic column of drillcore WRL1 covering the lower Hamersley and upper Fortescue Groups, Mt. Bruce Supergroup, Western Australia. Samples from drillcore WRL1 were taken from the Wittenoom Dolomite, Marra Mamba Iron, Jeerinah, and Pilingini Tuff Formations (indicated by bars). Core depth in meters is indicated to the left of the column. Pilingini Tuff has been thought to be stratigraphically equivalent to the Tumbiana Formation. 434

Core depth (m)

0 Recent Alluvium

Pinkian Chert Breccia

70.0

100

Carawine Dolomite

200

254.46

300

Lewin Shale 400

500

Fig. B-7. Lithostratigraphic column of drillcore RHDH2A covering the Carawine Dolomite and Lewin Shale Formations, in eastern Hamersley, Western Australia. Core depth in meters is indicated to the left of the column. Bars to the right of the column indicate where the samples were collected. 435 Table B-1. Summary of drillcores used in the study.

Drillcore Drillhole locality Age Formation (Fm) / Dominant [Ga] Group (G), Supergroup (SG) lithology

South Africa PU1308 Agnes gold mine, 3.25 Sheba Fm Black shale Barberton Fig Tree G, Swaziland SG (some rich in siderite, silica) MRE10 Sheba gold mine, 3.25 Sheba Fm Graywacke Barberton Fig Tree G, Swaziland SG DRH13 26˚52'S, 26˚23'E, 2.96 Parktown Fm, Black shale near Klerksdorp West Rand G, Witwatersrand Witwatersrand SG MSF6 26˚33'S, 27˚12'E 2.71 Rietgat, Oak Tree, and Black shale near Klerksdorp ~2.22 Timball Hill Fm Witwatersrand Pretoria, Chuniespoort, and Platberg G, Transvaal & Ventersdorp SG JPBR near Klerksdorp 2.64 Black Reef Fm, Carbonate-rich Witwatersrand Wolkberg G, Ventersdorp SG black shale PTB3 24˚55' S, 30˚44'E 2.22 Timeball Hill Fm, Black shale Pilgrim's Rest Pretoria G, Transvaal SG SA1677 Sishen iron mine, ~2.2 Pauling member#, Red shale Griqualand West Gamagara Fm, Griqualand West Seq. Australia WRL1 East of Wittenoom, 2.72 Pillingini Tuff, Jeerinah, Black shale Hamersley ~2.60 Marra Mamba Iron, and Carbonate-rich Wittenoom Dolomite Fm, black shale, Fortescue & Hamersley G, BIF, Chert Mt. Bruce SG RHDH2A Ripon Hill, 2.69 Carawine Dolomite and Carbonate-rich Northeastern ~2.60 Lewin Shale Fm, Black shale Hamersley Hamersley G & Fortescue G Mt. Bruce SG

#: Mapedi Formation of the Olifantshoek Group, according to Beukes and Smit (1987) 436 Table B-1. (continued)

Drillcore Age Number of samples (Fm) Tectonic settings References [Ga]

South Africa PU1308 3.25 17 (Sheba Fm) Cratonic epeiric sea Lowe & Byrly

MRE10 3.25 12 (Sheba Fm) Cratonic epeiric sea 1

DRH13 2.96 15 (Parktown Fm) Deep marine 2

MSF6 2.71 8 (Rietgat Fm) 3 2.56 3 (Oak Tree Fm) 3 2.22 4 (Timeball Hill Fm) Shallow-deep marine 3

JPBR 2.64 4 (Black Reef Fm) 3

PTB3 2.22 12 (Timeball Hill Fm) 3

SA1677 ~2.2 8 (Mapedi Fm) shallow marine 3

Australia WRL1 2.72 3 (Pillingini Tuff Fm) 3 2.69 6 (Jeerinah Fm) Deep marine 3 >2.60 4 (Marra Mamba Iron Fm) Deep marine 3 2.6 3 (Wittenoom Dolomite Fm) Deep marine 3

RHDH2A 2.69 8 (Lewin Shale Fm) Deep marine 3 2.60 6 (Carawine Dolomite Fm) Deep marine 3

Refrencees 1: Lowe and Byrly (1999), 2: Condie et al. (1970), 3. SACS (1980) Appendix C

Analytical methods

C-1. Samples treatment and pulverization

Drillcore samples were rinsed with large volumes of water to remove contaminants such as encrusting dirt. Dried clean drillcore samples were examined for lithological and textural characteristics, and individually placed in clean plastic bags. Samples were selected for chemical and isotopic analyses based on a number of criteria including homogeneity (in , lithology, etc.), veinlet, soft-sediment deformation features, lamination, distribution of sulfide minerals (massive layers, nodules, or disseminated), sizes, and representativeness of a section of interest. Screened samples were crushed into coarse fragments (< 5-10 mm) using a jaw- crusher. To minimize cross-contamination between samples, rock chips less than ~1 mm in diameter produced during crushing were separated using sieves and not used for further preparation. The coarse fraction of the samples were rinsed with water several times to remove attached dust, and washed again with distilled deionized water (18 MΩ) in an ultrasonic bath for ~10 minutes (repeated when necessary). Then they were washed with acetone in an ultrasonic bath for 5 minutes to remove any organic contaminants from the samples. Cleaned samples were dried in a low temperature oven at ~60˚C until dryness. 438

Parts of the cleaned rock chips were pulverized until reaching 150 mesh size (106 µm). An agate ballmill and agate mortar and pestle were used for pulverization at the Ocean Research Institute (ORI) at the University of Tokyo and the Pennsylvania State University. A mild steel mill was also used for samples pulverized at the ACTLAB, Inc. Cleaner sand (ashed at 400 ˚C for overnight) was used to avoid cross-contamination between samples. Unused cleaned rock chips have been stored in clean glass vials with teflon/rubber caps. The powdered samples have also been stored in clean glass vials with teflon/rubber caps. All the glass vials were ultrasonically cleaned with pure water and acetone before use.

C-2. Preparation of thin sections and microscopic observation

Selected drillcore samples were cut by a rock saw to make a small specimen for thin sections. Thin sections of ~30 µm-thick were made by the author at Penn State and Ocean Research Institute, University of Tokyo by standard technique, and by Petrographic International, Inc. Transmitted light and reflected light were used for observation of the thin sections under the petrographic microscopes (Nikon Optiphot-POL) from at The Pennsylvania State University.

C-3. Decomposition of the powdered samples

Almost all methods of chemical and isotopic analysis require the initial decomposition of the sample. Alkali fusion and mixed-acids digestion are the common methods to dissolve geological samples for the determination of their elemental 439

compositions by ICP-AES (inductively coupled plasma - atomic emission spectrometry) and ICP-MS (inductively coupled plasma - mass spectrometry). Both methods were used in this study. Sample preparation is an essential aspect of all plasma spectrometry and other solution-based analyses. It ultimately limits the range of elements and the precision and accuracy of analytical data which can be obtained from a sample. Clearly, methods which result in the incomplete dissolution of refractory minerals and/or the loss of volatile species will preclude the accurate determination of some elements. No single preparation procedure is suited to the digestion of all materials for the determination of all elements. Combined data from both alkali fusions and mixed-acids digestion enables the accurate and precise determination of a wide range of elements in most samples types. In this section, a brief summary of methods used for the decomposition of samples and associated problems are discussed, followed by the detailed methods used in this study.

C-3-1. Alkali fusion Alkali fusions are commonly used to digest geological samples for the analysis of major elements by ICP-AES. The high-temperature fusion of rock samples with lithium

metaborate (LiBO2) results in the formation of glasses which are readily soluble in dilute nitric acid. This procedure is effective for the dissolution of all major rock-forming silicates as well as many accessory minerals. Silica, which is lost as volatile during acid digestion, is retained in the solution therefore enabling the determination of all major elements from a single preparation. The other advantage in alkali fusion method over acid-digestion method include that hazardous reagents such as HF and HClO4 are not required, and that special apparatus such as teflon-lined digestion bombs is not required. The major disadvantage of all fusion procedures, however, is that they introduce large quantities of TDS (total dissolved solids). High amounts of TDS in sample solutions 440

will necessitate increased dilution to attain TDS levels of <1-2 % for ICP-AES and < 0.1- 0.2 % for ICP-MS, which typically mean ~500-100-fold dilution for ICP-AES and ~5000- fold dilution for ICP-MS. This is to minimize signal drift and maximize precision. These increased dilution factors will invariably push some trace-element concentrations in solution below detection limits of quantitative analysis. Nevertheless, even with the extra dilution required, fusion has been advocated as the preferred dissolution technique for analyzing Ti, Zr, Hf, W, and Cr. These elements are commonly found particularly in refractory minerals. Another disadvantage includes loss of volatile elements such as Cd, Pb, Sb, Sn, Zn from

LiBO2 fusions; there is a possibility of preventing their accurate determination.

In this study, alkali fusion method was performed at the Material Characterization Laboratory (MCL) at Penn State. Detailed procedure of alkali fusion used in this study is as follows (see MCL's web site at http://www.mri.psu.edu/mcl/):

(1) Add accurately weighed ~0.100 or ~0.200 g of 150 mesh powdered samples to a pre-

weighed vials containing 1.000 g of LiBO2. (2) Shake vials by hand, gently but thoroughly mixing the two components. (3) Dump contents of vials into a graphite crucibles and insert crucibles into an oven pre- heated to 1000˚C.

- (4) Pipette 100 ml of a 5% HNO3 into watch-glass-covered Teflon beakers and add a stirring bar to each. (5) After 15 minutes at 1000 ˚C (allow a few minutes for oven to return to 1000 ˚C after inserting samples into the oven), place beakers on magnetic stirrers and start. Then swirl the contents of the crucibles to pick up any uncoalesced beads and dump into beakers. (6) After a few minutes, carefully examine the graphite crucibles for any remaining melt, scrape off if present and add to beakers. 441

(7) Stir at least 15 minutes and transfer to pre-cleaned polyethylene bottles for storage.

C-3-2. Acid digestion Acid digestion with HF is routinely used to digest geological materials for trace element determination. If silica is not to be determined, the digestion of samples in open

teflon beakers using concentrated acid-mixtures of HNO3-HF-HClO4 at ~200 ˚C is

preferred. The addition of HClO4 has advantages including the more efficient attack of refractory minerals (by improving the efficiency of HF, principally due to the increased boiling temperature of the reaction mixture) and organic compounds. This digestion technique is highly effective for the dissolution of many minerals and reduces the amounts

of TDS by removing silica as the volatile silicon tetrafluoride, SiF4. Some other volatile elements such as Se and B will also be lost. Acid-mixture dissolves samples, and then the acids are evaporated to incipient dryness. The residues are digested in a further aliquot of the acid-mixture and a second evaporation is undertaken (this process is repeated when necessary). Finally the samples are dissolved in dilute HNO3 because most elements are

stable in HNO3 and it is the preferred acid matrix for ICP-MS analysis, since it avoids the - addition of excess Cl (if HCl is used instead of HNO3) which can cause severe interferences on some elements. This dissolution procedure is generally favored for trace- element determination by ICP-AES and ICP-MS.

- However, residual Cl from remnant HClO4 may cause interference problems with the determination of As and V, and to a much lesser extent Cr, Fe, Ga, Ge, Se, Ti, Zn (Jarvis and Jarvis, 1992). Even with repeated attacks by HNO3-HF-HClO4 acid-mixtures, refractory minerals such as barite, chromite, zircon, rutile, magnetite, and may remain only partially digested. 442

In this study, the acid digestion method was used to decompose samples. The detailed procedure of acid digestion used in this study is as follows:

(1) Add accurately weighed ~0.1000 g of 150 mesh powdered samples to pre-cleaned Teflon beakers, then moisten with a few droplets of pure water. (2) Add 10 ml of trance-element grade ~29 M HF and 4 ml of trance-element grade ~12 M

HClO4 to the pre-cleaned Teflon beakers.

(3) Overnight drying on a hot plate under a HClO4-tolerant hood to achieve incipient dryness.

(4) Add 4 ml of trance-element grade ~12 M HClO4 . (5) Overnight drying [same as (3)].

(6) Add 10 ml of 5 M HNO3 and dilute with pure water to 50 ml in a volumetric flask. (7) The solution samples are stored in pre-cleaned plastic bottles.

A majority of the samples used in this study were repeatedly treated with the above- mentioned acid-mixture to attain complete digestion. To achieve complete-digestion of the samples containing refractory minerals, the teflon-bomb method (digestion in sealed teflon vessels with increased pressure at elevated temperature) was used. Many samples were successfully digested by these methods and their solutions were very clear. However, there remains a possibility that extremely fine particles of refractory minerals exist in the sample solutions. Samples that were not completely digested were rejected in this study. Powders of such hard-to-dissolve samples were sent to ACTLAB, Inc. for their complete digestion by alkali fusion method and measurement of trace element contents by ICP-MS. 443

C-4. Major and minor element analysis

For major (Si, Al, Ti, ∑Fe, Mn, Mg, Ca, K, Na, and P) and minor (Ba, Nb, Ni, Pb, Rb, Sr, Th, Y, and Zr) element analyses, two methods were used; ICP-AES at Penn State and ACTLAB, Inc., and XRF (X-ray fluorescence spectrometry) at ORI, Univ. of Tokyo. For ICP-AES at Penn State, the model "Leeman Labs PS 3000UV" in the Material Characterization Laboratory was used. Sample preparation for ICP-AES is already described in the previous section. Good precision (± 0.5-1 %) was achieved with ICP-AES for simultaneous multi-elements analyses. Calibration was made with in-house standard materials that were calibrated against international SRM (standard reference material). Procedure blank samples and procedure standard samples (SCo1, MAG1, BCR2, and W2: United States Geological Survey SRM) were also analyzed to track possible contamination during experimental procedures and to assess quality of the analysis. For XRF, the model "Rigaku 3270 X-ray spectrometer" was used with an operating condition of 50mA and 50kV in power and 50 ml/min in gas flow. For major elements analyses, glass beads were made from a 10:1 mixture of Li2B4O7 and powdered rock samples; they were mixed and put in a Pt crucible and subsequently melted by furnace (bead sampler) at 1000 ˚C for 15 minutes and swirled for complete homogenization of the mixture. For minor element analyses, a pressed pellet of powdered rock samples was made with static pressurization (~1000 kg) for 30 seconds. Calibration was made with in-house standard materials that were calibrated against international SRM. Procedure blank samples and procedure standard samples (JB3 and JG1A: Geological Survey of Japan SRM) were also analyzed to track possible contamination during experimental procedures and to assess the quality of the analysis. 444

C-5. Minor, trace, and rare earth elements analysis

For minor and trace (Sc, V, Cr, Co, Ni, Cu, Zn, Ga, As, Rb, Sr, Y, Mo, Cd, Sb, Cs, Ba, Hf, W, Pb, Th, and U) and rare earth (La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu) element analyses, ICP-MS either at Penn State or ACTLAB, Inc. was used. For ICP-MS at Penn State, the model "Finnigan ELEMENT" in the Material Characterization Laboratory was used. Good precision (± 0.5-1 %) was achieved with ICP-MS for simultaneous multi-elements analyses. Calibration was made with in-house standard materials that were calibrated against international SRM. Procedure blank samples and procedure standard samples (SCo1, MAG1, BCR2, and W2: USGS SRM) were also analyzed to track possible contamination during experimental procedures and to assess the quality of the analysis. ICP-MS is the preferred method for quantitative determination of multi-elements and isotopes in a wide variety of sample types at trace and ultra-trace concentration levels. Ultra-high sensitivity and low dark noise result in excellent detection limits at the pg/L and ng/L levels.

C-6. Ferrous iron titration

The Wilson method (Wilson, 1955) was used to measure Fe2+ content after

decomposition of the samples by HF. Boric acid (H3BO3) is added to HF-digested solution, because H3BO3 form complexes with any free HF to make the resultant solution far more stable and less hazardous to handle. Ferric iron content is obtained by difference from the separate determination of the Fe2+ and the total Fe contents of the sample. 445

...... ∑ Fe2O3 [wt.%] - 1.1113 • FeO [wt.%] = Fe2O3 [wt.%] (C-1)

The Wilson method is based on the oxidation of Fe2+ using ammonium

metavanadate (NH4VO3). In order to prevent atmospheric oxidation during the chemical 2+ dissolution of the sample, NH4VO3 is added to the digestion mixture. As soon as Fe is released from the rock matrix, it reacts as follows:

(5+) + 2+ + (4+) 2+ 3+ ...... V O2 + Fe + 2 H ---> V O + H2O + Fe (C-2)

The reduced form of vanadium (V(4+)O2+) is stable to atmospheric oxidation. The reduced vanadium (V4+) content of solution is equivalent to the amount of Fe2+ liberated from the sample. This is determined by adding a known excess of Fe2+ solution (ferrous ammonium

2+ 2- + 5+ sulfate: Fe (SO4 )2(NH4 )2 to each beaker. All unreacted vanadium ions (V ) are then reduced to vanadium (V4+) as in equation above.

(5+) + 2+ 2+ + V O2 (unreacted with Fe ) + Fe (from ferrous ammonium sulfate) + 2 H (4+) 2+ 3+ ...... ---> V O + H2O + Fe (C-3)

2+ 2+ 2- + Each solution then contains excess Fe (from Fe (SO4 )2(NH4 )2), the concentration of which is determined by titration with a standard solution of potassium dichromate

(K2Cr2O7) using barium-diphenylamine p-sulfonic acid as an indicator:

2- 2+ + 3+ 3+ ...... Cr2O7 + 6 Fe + 14 H ---> 2 Cr + 6 Fe + 7 H2O (C-4)

The difference in titration volumes between sample and blank solutions is equivalent to the concentration of ferrous iron released from the rock sample. 446

Associated problems Potential interferences include an atmospheric oxidation of Fe2+ during

3+ crushing/grinding/titrating, reduction of Fe by Corg and S, and incomplete digestion of samples. Aerial oxidation has been generally considered the chief source of error in the determination of Fe2+, particularly during the process of acid decomposition and, to a lesser extent, during the titration. There is a possibility that S2- will reduce some of the Fe3+ present, thus giving a high value for Fe2+ and a correspondingly low value for Fe3+. Carbon may give high results for Fe2+ because it will tend to be oxidized during the titration, or will reduce an added oxidizing agent. Thus, the obtained Fe3+ and Fe2+ contents should be viewed as minimum and maximum values, respectively. However, the Fe2+ content of some SRMs measured by the author agreed very well with their certified values with very good reproducibility (± 0.2 wt.%). The Fe2+ content of many Precambrian samples measured by the author also agreed very well with the values reported from ACTLAB, Inc. Therefore, the Fe2+ contents measured in this study are assumed to be accurate ennough.

C-7. Loss on ignition

Decrease in weight of a sample during high-temperature heating, i.e., LOI (loss of ignition), was measured using a furnace at Penn State (NEY M-525 ) or one at ORI (YAMATO Muffle Furnace FP42). Approximately 1 g of powdered samples was weighed and put into a clean-dry porcelain crucibles. They were dried in a low temperature oven at 110 ˚C for 1~2 hours. Subsequently, they were placed in a desicator to cool in a moisture-free environment, and 447

- weighed. The difference in weight before and after the 110 ˚C drying is assigned to be H2O ; however, it was not significant. Then, 110 ˚C-dried samples in porcelain crucibles were placed in a high temperature oven at 950 ˚C for ~5 hours. After cooling to ~200 ˚C, the samples were placed in a desicator to cool in a moisture-free environment, and weighed. The difference in weight

+ before and after the 950 ˚C drying is assigned to be H2O , or LOI. Since the samples usually contained Corg and Ccarb, the LOI values represent bound H2O content and CO2 content from C, together with negligible contribution from the other volatile species.

C-8. Mineralogical analysis by X-ray diffraction

X-ray diffraction (XRD) is a basic tool in the mineralogical analysis of rocks. XRD at MCL of Penn State (Rigaku Geigerflex with a Dmax-B controller and a vertical goniometer) was used in this study to identify minerals in the samples. Powdered samples were packed in a cavity mount made of quartz glass. The instrument uses a Cu Kα radiation and is run by continuous mode. Operating conditions were such that the power was 20mA and 40kV, scanning was from 4 to 75˚ (2θ), scanning speed was 5˚ / minute, and sampling interval was 0.02˚. Obtained diffraction patterns were searched for peaks using the automatic peak finding function, and minerals were identified by matching the measured peaks with the known peak patterns of minerals. No effort was made to quantify the mineral abundance with XRD. For more detailed information, visit the MCL's web site at http://www.mri.psu.edu/mcl. 448

C-9. Organic C, carbonate C, H, N, and S elemental analysis

An elemental analyzer (EA), either at Penn State (CE Instruments NA2500) or at Tohoku University (Carlo Erba EA 1108), Japan, was used to measure content of C, H, N,

and S. To measure content of Corg and Ccarb, the bulk samples were decarbonated using 2N

HCl overnight at room temperature. Ccarb content is calculated from a difference in C content between bulk and decarbonated samples. The combustion column in the EA uses

WO3 and reduced Cu short wires. Helium was used as a carrier gas at a rate of 140 ml / minute. Analytical standard samples used for calibration were BBOT (2, 5- Bis- 95-tert.

butyl-benzoxazol-2-yl)-thiopen: C26H26N2O2S) and SAD (Sulfanialammide:

C6H8N2O2S).

C-10. Kerogen extraction

Kerogen was extracted from the bulk rock powder samples using HCl and HF at Penn State (with assistance from Ms. Y. Watanabe) and Tohoku Univ. in Japan (with assistance from Dr. T. Kakegawa). Weighed samples were put into a teflon bottles, along with a small amount of distilled and deionized water (DDW) to moisten the powder samples and then 6 N HCl. The bottle was capped and placed in a shaking bath overnight at ~80 ˚C. After centrifuging of the solution at a speed of 2000 rpm for 15 minutes, the supernatant was discarded. Then, a mixture of concentrated HF and 6 N HCl was added to the bottle and reacted overnight with the samples in the same manner described above. After centrifugation, the supernatant was discarded. Then 6 N HCl was added to the bottle and reacted in the same manner. After centrifuging the solution, the supernatant was discarded and DDW was added to the bottle to dilute and increase the pH. The bottle was gently 449

shaken to disaggregate the compacted powdered samples due to centrifugation. This process of centrifuge-DDW dilution was repeated until the pH reached ~5. Then the sample was oven-dried at ~80 ˚C. After drying, the sample pellet was gently disaggregated with an agate mortar and pestle, and stored in an ashed (at 500 ˚C overnight) glass vial with a teflon inseam. When fluoride formed during the concentrated HF-treatment, supersaturated warm boric acid (H3BO3) was added for decomposition. Samples with elevated ash content (typically C content of kerogen less than 50 %) were considered to be the result of incomplete acid digestion and/or ash formation (fluoride); such poor kerogen samples were not used for further discussion.

C-11. Carbon and oxygen isotope analysis

C-11-1. Off line isotopic analysis of organic C Carbon isotopic compositions of decarbonated and kerogen samples were measured using an EA (CE Instruments NA2500) and a mass spectrometer (Finnigan MAT 252) at

Penn State. Combustion at 1000 ˚C of powdered samples packed in Sn cups generated CO2

gas, which was collected using a liquid N2 trap. The trapped CO2 gas was cryogenically purified and transferred to a glass tube, which was sealed and separated using flame. Then the CO2 sample was measured for its C isotopic compositions with the mass spectrometer. The detailed operational method is described in the manual written by the author (Yamaguchi and Walizer, 1996, 1998a). Repeated analyses of the standard resulted in very good precision, better than ± 0.2‰. 450

C-11-2. On line isotopic analysis of organic C Carbon isotopic compositions of decarbonated and kerogen samples were measured with an EA (Eurovector) and a mass spectrometer at the University of Nevada-Reno. Dr. Simon Poulson performed the measurements. Combustion at 1000 ˚C of powdered samples

packed in Sn cups generated CO2 gas, which was directly transferred to a mass spectrometer. Some of the advantages of the on-line CF-IRMS (continuous-flow isotope ratio mass spectrometry) system are that the C isotopic composition of a sample containing very low amounts of C can be measured with good precision and accuracy and that analytical time is significantly reduced, compared to conventional off-line analyses. SAD

(Sulfanialammide: C6H8N2O2S) standard samples used for elemental analyses with an EA were also used as a standard for C isotopic composition, together with other SRM. Repeated analyses of the SAD standard resulted in very good precision, better than ± 0.2‰.

C-11-3. Carbon and oxygen isotopic analysis of carbonate The "Common acid bath" method was used to measure the C and O isotopic compositions of carbonate fraction of the samples. The bulk powder samples put in Cu-

boats were first roasted in a vacuum at 350 ˚C for ~1 hour to remove Corg. Then the roasted samples were set in a carousel of the "common acid bath" system. Dehydrated phosphoric acid was used to generate CO2 from carbonate fraction in a Corg-free samples. The reaction temperature was maintained between 87.5-87.9 ˚C, and approximately an hour per sample was required for the reaction with acid and measurement of isotope ratios by mass spectrometry (Finnigan MAT252). A lab standard "Biogeochem" (5 samples) and USGS's SRM "NBS19" (4 samples) were used for each batch of 47 samples (including 38 unknown samples). Isotopic compositions from such samples that did not produce enough

CO2 for reliable measurement were discarded and not used in this study. Data obtained data was not used for this study but will be used in future research. 451

C-11-4. Oxygen isotopic analysis of silicate The oxygen isotopic composition of silicate fraction of bulk samples was measured

using the BrF5 method (e.g., Clayton and Mayeda, 1963) at Penn State. Samples were

weighed and put in a Ni-tube for an overnight reaction at 600 ˚C with BrF5. Released

oxygen was reacted with a heated graphite disk to convert it to CO2. The CO2 was cryogenically collected in a glass tube, and sealed and separated from the experimental line.

Isotopic analysis of oxygen of CO2 was performed using a mass spectrometer (Finnigan MAT 252) at Penn State. The detailed operational method is described in the manual written by the author (Yamaguchi and Walizer, 1997). Quartz power with known O isotopic composition was used as a standard. Repeated analyses of the standard resulted in very good precision of better than ± 0.2‰. Obtained data was not used for this study and will be used in future research.

C-12. Sulfur isotopic analysis

The sulfur isotopic compositions of pyrite were measured with an EA (CE Instruments NA2500) and a mass spectrometer (VG Prism II) at Penn State. Mr. Dennis Walizer supervised the operation of the mass spectrometer. Combustion at 1000˚C of powdered samples packed in Sn cups generated SO2 gas, which was collected using a liquid

nitrogen trap. The trapped SO2 gas was cryogenically purified and transferred to a glass

tube, which was flamed off. Then the SO2 sample was measured for its sulfur isotopic composition with the mass spectrometer. The detailed operational method is described in the manual that the author wrote (Yamaguchi and Walizer, 1998b). 452

C-13. Nitrogen isotopic analysis

Nitrogen isotopic compositions of kerogen and decarbonated samples were measured with a CF-IRMS system (Carlo Erba EA1108 Elemental Analyzer combined with a mass spectrometer (Finnigan Delta S) at Tokyo Metropolitan University (TMU). Dr. Hiroshi Naraoka of TMU supervised the analyses with the CF-IRMS. To generate enough

N2 gas (i.e., enough voltage at the detectors) during the mass spectrometer analyses, the amount of samples to be combusted in the EA was adjusted. Nevertheless, some samples containing very low amounts of N could not attain an acceptable voltage of 1V for reliable measurements. Nitrogen isotopic data of such samples were discarded and not used in this study. Repeated analyses of the standard material resulted in very good precision of better than ± 0.2 ‰.

C-14. Degree of pyritization analysis

Degree of pyritization (DOP) is a measure of degree of sulfide formation from Fe pool that had a potential toward forming sulfide. Its definition is given as follows (e.g., Berner, 1970; Raisewell and Berner, 1985):

DOP = [Pyrite-bound Fe] / [Reactive Fe]...... (C-5) [Reactive Fe] = [Pyrite-bound Fe] + [HCl-extractable Fe] ...... (C-6)

In this study, pyrite-bound Fe was determined by calculating the stoichiometric value assuming that total S approximates pyritic S since no sulfate minerals were identified in the 453 samples of this study. To determine HCl-soluble Fe content, there are several methods. Berner (1970) defined reactive Fe in recent marine sediments as the Fe that was extracted from a 100 mg sample after boiling one minute with 5 ml of concentrated (12N) HCl. The solution was rapidly cooled (Raisewell et al., 1988) by adding 245 ml of water to achieve rapid quenching and dilution of acidity. Other methods have been used to determine reactive Fe in ancient sedimentary rocks, including timed reactions with several reagents such as Na- dithionite in a citrate buffer (Canfield, 1988). Leventhal and Taylor (1990) presented the experimental results of DOP determination by various methods and different samples, and proposed the use of a 24 hour room temperature treatment with 1 N HCl. This is because this 24 h over 1 N HCl method produced comparable concentrations of acid-extractable Fe (thus DOP value) to that by other methods such as Berner's, and because laboratory handling is easier and safer. In this study, the 24 h over 1 N HCl method was utilized. Weighed (10-20 mg) powdered samples were placed in test tubes and 50 ml of 1 N HCl was added. After 24 hours, the samples were centrifuged and concentrations of dissolved Fe in the supernatant were measured by the atomic adsorption (AA) (flame emission spectrometry) method using a Perkin Elmer 703 spectrophotometer at the MCL at Penn State. When necessary, the supernatant was diluted with DDW to adjust Fe concentrations suitable for AA analysis. Typically, Fe concentrations were diluted to reach an order of ppm level. Measurements by AA were repeated, and average values of Fe concentration were used to calculate leached Fe content by HCl from the samples. 454

References

Berner, R.A. (1970) Sedimentary pyrite formation. Am. J. Sci. 268, 2-23.

Canfield, D.E. (1988) Sulfate reduction and the diagenesis of iron in anoxic marine sediments. PhD thesis, Yale University, 248pp.

Clayton, R.N. and Mayeda, T. K. (1963) The use of bromine pentafluoride in the extraction of oxygen from oxides and silicates for isotopic analysis. Geochim. Cosmochim. Acta 27, 43-52.

Jarvis, I. and Jarvis, K.E. (1992) Plasma spectrometry in the earth sciences: techniques, applications and future trends. Chem. Geol. 95, 1-33.

Leventhal, J.S. and Taylor, C. (1990) Comparison of methods to determine degree of pyritization. Geochim Cosmochim. Acta 54, 2621-2625.

Raiswell, R. and Berner, R.A. (1985) Pyrite formation in euxinic and semi-euxinic sediments. Am. J. Sci. 285, 710-724.

Raiswell, R., Berner, R.A., and Andersen, T.F. (1988) Degree of pyritization of iron as a paleoenvironmental indicator of bottom-water oxygenation. J. Sed. Petrol. 58, 812- 819.

Taylor, S.R. and McLennan, S.M. (1985) The Continental Crust: its Composition and Evolution. Blackwell Scientific Publications, 311p.

Wilson, A.D. (1955) A new method for the determination of ferrous iron in rocks and minerals. Bull. Geol. Surv. Great Britain 9, 56-58.

Yamaguchi, K.E. and Walizer, D.P. (1996) δ13C and δ18O measurement by Finnigan MAT252 (Program Dixon 2). Unpublished laboratory manual, 3p.

Yamaguchi, K.E. and Walizer, D.P. (1997) Extraction procedure of silicate-oxide oxygen by BrF5 method. Ver 2.2. Unpublished laboratory manual, 17p.

Yamaguchi, K.E. and Walizer, D.P. (1998a) EAGE (Elemental Analyzer-Gas Extraction) method. Unpublished laboratory manual, 8p.

Yamaguchi, K.E. and Walizer, D.P. (1998b) Operating manual of VG Prism II for laser microprobe anbalysis. Ver 2.0. Unpublished laboratory manual, 10p. 455

Rinse

Crushing

Remaining rock chips Thin section Coarse fraction Very fine fraction <~1mm φ Rinse

Microscopic Discarded observation

Preserve Powder

Various type of geochemical analysis

Fig. C-1. Flow diagram of the sample handling toward pulverization. 456

Bulk powder

Analytical Preparative treatment Measurement instruments Fusion at high T

Alkali fusion by LiBO 2 ICP-AES, Major, minor, and HNO3 digestion ICP-MS trace elements

Alkali fusion by Li 2B4O7 XRF Major elements to make glass bead

Roasting At 950 ˚C Loss on ignition

13 18 At Corg at 350 ˚C δ Ccarb, δ Ocarb

Digestion at low T Acid digestion by ICP-MS Trace elements HNO3-HF-HClO4 Kerogen extraction EA, CHNS content 13 34 15 by HF-HCl Mass Spec δ Ckero , δ Spy , δ Norg Leaching Leached Fe content AA by 1 N HCl "Reactive Fe" and DOP Titration by HF digestion Fe2+ content K2Cr2O7 Decarbonation EA, CHNS content, 13 34 by 6N HCl Mass Spec δ Corg, δ Spy

No special treatment Corg, Ccarb content required EA Bulk CHNS content

XRD Minerals abundance

Pressed pellet XRF Minor elements

Fig. C-2. Flow diagram of the sample treatments for the various type of analyses. VITA

Kosei Yamaguchi was born in Tokyo, Japan, on November 23, 1969. He is a son of Masaaki and Sai Yamaguchi, Yokosuka, Kanagawa, Japan. In 1988, he graduated from Yokosuka High School in Yokosuka. In April 1989, he enrolled at the University of Tokyo, where he received the degree of Bachelor of Engineering in Laser Quantum Physics in March 1993. He then studied Precambrian geology at the Ocean Research Institute, The University of Tokyo, from 1993 to 1995. Under the supervision of Prof. Asahiko Taira, he received the degree of Master of Science in Geology in March 1995. In August 1995, he entered the Department of Geosciences, and then NASA Astrobiology Institute - Penn State Astrobiology Research Center since 1998, The Pennsylvania State University, where he received his Ph.D in Geoscience. He is currently a staff scientist at IFREE (Institute for Frontier Research on Earth Evolution), Japan, and a research associate at the Department of Geology and Geophyscis, University of Wisconsin - Madison.

He has been an enthusiastic bicycle traveler, camper, mountaineer, free- and rock- climber, marathon-runner, orienteer, swimmer, oil-painter, juggler, flutist, photographer, etc. etc.... too many to write all -ist and –er to describe myself.

Nothing is harder, yet nothing is more necessary, than to speak of certain things whose existence is neither demonstrable nor probable. --- Hermann Hesse

Top View