The Pennsylvania State University
The Graduate School
College of Earth and Mineral Sciences
MOLECULAR AND ISOTOPIC PERSPECTIVES ON EARLY HUMAN HABITATS AT
OLDUVAI GORGE, TANZANIA
A Dissertation in
Geosciences and Biogeochemistry
by
Clayton R. Magill
© 2013 Clayton R. Magill
Submitted in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
May 2013
The dissertation of Clayton R. Magill was reviewed and approved* by the following:
Gail M. Ashley Professor of Earth and Planetary Sciences, Rutgers University Special Signatory
Katherine H. Freeman Professor of Geosciences Dissertation Advisor Co-Chair of Committee
Nina G. Jablonski Distinguished Professor of Anthropology
Lee R. Kump Professor of Geosciences Head of the Department of Geosciences Co-Chair of Committee
Mark E. Patzkowsky Associate Professor of Geosciences
*Signatures are on file in the Graduate School
ii
ABSTRACT
Early human (hominin) evolution was linked to dramatic changes in regional hydrology and ecosystem composition. Emerging hypotheses linking hominin evolution and the environment present specific, testable predictions for hominin evolutionary responses to different modes of environmental change. Yet, discontinuous terrestrial sediment sequences coupled with indirect proxy indicators for plants and water obscure environmental perspectives on the hominin fossil record. Opportunity to overcome these challenges is presented by lake and soil sediments from a preeminent hominin archaeological locality – Olduvai Gorge, Tanzania. These extensive, well- constrained sediments have abundant organic matter, allowing for unparalleled resolution of biomarker data through time and space for an individual locality.
To link the hominin archaeological record to local environmental conditions at Olduvai Gorge, I characterized distributions and stable isotope compositions (δ13C and δD) for a diverse suite of modern and sedimentary biomarkers. Towards this end, I established a quantitative framework for reconstructing plant functional type (PFT) relative abundances in tropical ecosystems based on biomarker δ13C values from living plants by compiling previously published data relating (a)
PFT to tropical ecosystem structure, (b) PFT to soil organic matter δ13C values, and (c) soil organic matter δ13C values to biomarker δ13C values. Separation of previously published δD values for biomarkers from living plants according to PFT can elucidate differences in the biochemical and physiological influences on apparent fractionation between source-waters and biomarkers (εlipid/water). By extension, ‘landscape’ apparent fractionation factors are calculated
13 from isotopic mass-balance of εlipid/water values via biomarker δ C estimates for relative PFT
iii abundances. Using this novel approach to establish fractionation factors in dynamic environments, I then reconstructed δD values for precipitation and lake waters at Olduvai Gorge during the past.
The early Pleistocene is associated with a key adaptive juncture in hominin evolution – the emergence of our direct ancestor Homo erectus/ergaster. To constrain local environmental conditions during this juncture, I measured δ13C and δD values for sedimentary biomarkers from a continuous sequence of lake sediments deposited between about 2.0 and 1.8 million years ago
(i.e., the early Pleistocene) at Olduvai Gorge. Biomarker δ13C values correlate strongly with changes in orbital geometry and tropical sea-surface temperatures during this interval, and reveal cyclic catchments-scale ecosystem shifts between closed woodlands and open grasslands. After correcting measured biomarker δD values for ‘landscape’ apparent fractionation factors, reconstructed source-water δD values indicate lower amounts of precipitation fell on open grasslands (about 250 mm yr-1) as compared to closed woodlands (about 700 mm yr-1). The scale and pace of environmental changes at Olduvai Gorge contrast with long-held views of directional or step-wise aridification in eastern Africa during the early Pleistocene.
Faunal evolutionary responses to environmental change are strongly influenced by microhabitat
(<100 m2) features. To reconstruct fine-scale spatial heterogeneity in plants and water associated with hominin habitation, we measured the distributions and δ13C values for plant biomarkers
(leaf-waxes, lignin and 5-n-alkyresorinols) preserved in time-equivalent soil sediments across the iconic FLK Zinjanthropus archaeological Level 22 (FLK Zinj) locality at Olduvai Gorge.
Central trenches of FLK Zinj are associated with low δ13C values for leaf-waxes and lignin
iv monomers indicative of closed woodland habitat. About 200 m to the north of the central trenches, abundant aquatic-plant and sedge biomarkers occur in conjunction with mound-like tufa deposits, suggesting wetland habitat near freshwater springs. In contrast, southern trenches contain high δ13C values for leaf-waxes and lignin monomers indicative of open grassland habitat. Taken together, these data delimit a heterogeneous microhabitat mosaic that is obscured in catchment-scale records of environmental change.
v
TABLE OF CONTENTS
LIST OF TABLES ...... xi
LIST OF TABLES ...... xiv
ACKNOWLEDGMENTS ...... xv
CHAPTER 1: Introduction
1.1. Introduction ...... 1 1.2. Environmental hypotheses of hominin evolution ...... 2 1.3. Records of environmental change for eastern Africa ...... 3 1.4. The rise of arid ecosystems in Africa ...... 3 1.5. Organic matter signals for ecosystem ...... 4 1.6. Organic geochemical proxies for hydroclimate ...... 5 1.7. Isotopic notation ...... 6 1.8. Research objectives and dissertation outline ...... 7 1.9. Communications and publications from this dissertation ...... 9 1.10. References ...... 10
CHAPTER 2: Ecosystem variability and early human habitats in eastern Africa
2.1. Abstract ...... 12 2.2. Introduction ...... 13 2.3. Background ...... 14 2.3.1. Site descriptions ...... 14 2.3.2. Sedimentary organic matter ...... 16 2.3.3. Carbon isotopes in leaves, biomarkers and soil organic matter ...... 16 2.3.4. Structural classification for ancient ecosystems ...... 18 2.4. Results and discussion ...... 19 2.4.1. Ecosystem change and woody cover ...... 19 2.4.2. Biogeochemical variability at the ecosystem scale ...... 21 2.4.3. Mechanisms of ecosystem change ...... 22 2.4.4. Mechanisms of hydroclimate change ...... 23 2.4.5. Ecosystems and hominin evolution ...... 25 2.5. Conclusions ...... 25 2.6. Figures ...... 27 2.7. References ...... 33
vi
CHAPTER 3: Water, plants and early human habitats in eastern Africa
3.1. Abstract ...... 39 3.2. Introduction ...... 40 3.3. Background ...... 41 3.3.1. Sample locality ...... 41 3.3.2. Precipitation patterns in eastern Africa ...... 41 3.3.3. Leaf-lipid apparent fractionation factors ...... 42 3.3.4. Algal-lipid apparent fractionation factors ...... 45 3.4. Results ...... 47 3.5. Interpretations and discussion ...... 48 3.5.1. Precipitation in eastern Africa ...... 48 3.5.2. Lake-water evaporation in eastern Africa ...... 50 3.5.3. Tracing isotopic hydrology at Olduvai Gorge ...... 51 3.5.4. Water availability and ecosystem dynamics ...... 52 3.5.5. Water and early human evolution ...... 53 3.6. Conclusions ...... 54 3.7. Figures ...... 55 3.8. References ...... 61
CHAPTER 4: Plant biomarker patterns at Olduvai Gorge establish social meat consumption by pre-modern humans
4.1. Abstract ...... 67 4.2. Introduction ...... 68 4.3. Background ...... 69 4.3.1. Soil organic matter ...... 69 4.3.2. Plant biomarker signatures ...... 70 4.4. Results ...... 71 4.5. Discussion ...... 72 4.5.1. Microhabitat distributions ...... 72 4.5.2. Hominin dietary resources ...... 73 4.5.3. Hominin behavioral implications ...... 73 4.6. Figures ...... 75 4.7. References ...... 78
CHAPTER 5: Differentiation of wholegrain biomarkers based on compound-specific 13C signatures of 5-n-alkylresorcinols
5.1. Introduction ...... 80 5.2. Experimental ...... 81 5.2.1. Reagents, standards and samples ...... 81
vii
5.2.2. Extraction and separation of lipids ...... 82 5.2.3. Instrumental analysis ...... 82 5.3. Results and discussion ...... 84 5.3.1. Saturated homologue abundances ...... 84 5.3.2. Compound-specific δ13C values ...... 85 5.3.3. Statistical analyses ...... 85 5.4. Conclusions ...... 86 5.5. Figures ...... 87 5.6. Tables ...... 89 5.7. References ...... 91
CHAPTER 6: Efficient in-cell separation of lipid biomarkers using sequential accelerated solvent extraction
6.1. Abstract ...... 93 6.2. Introduction ...... 93 6.3. Experimental ...... 95 6.3.1. Wetland soil samples ...... 96 6.3.2. Reagents and standards ...... 96 6.3.3. Sample extraction ...... 97 6.3.4. Sequential ASE method ...... 97 6.3.4.1. Column construction ...... 97 6.3.4.2. Elution scheme ...... 98 6.3.5. Reference procedures ...... 98 6.3.6. Instrumental analysis ...... 99 6.4. Theory ...... 100 6.4.1. Method development ...... 100 6.4.1.1. Column construction and elution scheme ...... 100 6.4.1.2. Extraction temperature ...... 101 6.4.1.3. Extraction cycles, static time and flush volume ...... 102 6.5. Results and discussion ...... 102 6.5.1. Recovery and separation efficiency ...... 102 6.5.1.1. Standard mixture ...... 102 6.5.1.2. Wetland soil samples ...... 103 6.5.2. Comparison between methodologies ...... 104 6.6. Conclusions ...... 105 6.7. Figures ...... 106 6.8. Tables ...... 109 6.9. References ...... 110
viii
CHAPTER 7: Research summary and future considerations
7.1. Research summary ...... 113 7.2. Future directions ...... 116 7.3. References ...... 118
APPENDIX A: Supporting information appendix for ‘Ecosystem variability and early human environments in eastern Africa’
A.1. Dataset description ...... 119 A.2. Age estimates and model uncertainties ...... 120 A.3. Correlation and regression analyses ...... 121 A.4. Tables ...... 123 A.5. Figures ...... 124 A.6. References ...... 127
APPENDIX B: Supporting information appendix for ‘Water, plants and early human habitats in eastern Africa’
B.1. Dataset description ...... 132 B.2. Salinity reconstructions ...... 132 B.3. Monthly and regional amount effects ...... 135 B.4. Evaporative balance reconstructions ...... 136 B.5. Uncertainty in ε31/model values ...... 138 B.6. Determination of environmental water δ18O values ...... 138 B.7. Figures ...... 140 B.8. References ...... 146
APPENDIX C: Supporting information appendix for ‘Plant biomarker patterns at Olduvai Gorge establish social meat consumption by pre-modern humans’
C.1. Biomarker extractions, separation and characterization ...... 151 C.2. 5-n-Alkylresorcinols in sedges and sediments ...... 152 C.3. Figures ...... 154 C.4. References ...... 155
ix
APPENDIX D: Data for δ13C values of soil organic matter vs. leaf tissues
D.1. Dataset description ...... 156 D.2. Tables ...... 157 D.3. References ...... 169
APPENDIX E: Data for δ13C values of leaf lipids vs. leaf tissues
E.1. Dataset description ...... 172 E.2. Tables ...... 173 E.3. References ...... 179
APPENDIX F: Review of published leaf-lipid δD data
F.1. Dataset description ...... 180 F.2. Tables ...... 181 F.3. References ...... 191
APPENDIX G: Measured δ13C and δD values for leaf lipids from modern soils near Olduvai Gorge
G.1. Dataset description ...... 193 G.2. Tables ...... 194
APPENDIX H: Measured δ13C and δD values for leaf lipids from FLK Zinjanthropus archaeological Level 22
H.1. Dataset description ...... 196 H.2. Tables ...... 197
APPENDIX I: Measured δ13C values for n-alkanes and 5-n-alkylresorcinols in selected human dietary resources
I.1. Dataset description ...... 199 I.2. Tables ...... 200
x
LIST OF FIGURES
Figure 2.1...... 27 Present-day precipitation patterns in eastern Africa
Figure 2.2...... 28 Histograms and box-and-whisker plots for apparent fractionation factors between soil organic matter and plant biomarkers
Figure 2.3...... 29 Stable carbon-isotope compositions for soil organic matter and plant biomarkers relative to woody plant cover, ecosystem and plant community composition
Figure 2.4...... 30 Biomarker and isotopic signatures for organic matter preserved in lake sediments at Olduvai Gorge
Figure 2.5...... 31 Time series for global, regional and local proxy indicators during the early Pleistocene
Figure 3.1...... 55 Modern precipitation patterns in eastern Africa during the ‘long rains’ and the ‘short rains’ with respect to average monthly precipitation in northern Tanzania
Figure 3.2...... 56 Schematic depiction of apparent fractionation factors between modeled precipitation and plant biomarkers as a function of reconstructed ecosystem
Figure 3.3...... 57 Plant biomarker and algal-lipid hydrogen-isotope compositions in relation to ancillary proxy data for Olduvai Gorge
Figure 3.4...... 58 Plant biomarker and algal-lipid hydrogen-isotope compositions and reconstructed source waters with respect to sedimentary total organic carbon
Figure 3.5...... 59 Modern isotopic relationships for precipitation and lake waters in eastern Africa
xi
Figure 4.1...... 75 Depositional environments at Olduvai Gorge during the early Pleistocene
Figure 4.2...... 76 Plant biomarker carbon-isotope compositions and 5-n-alkyresorcinol distributions overlying the ‘junction area’ at Olduvai Gorge
Figure 4.3...... 77 Comparison of plant biomarker carbon-isotope compositions and lignin phenol characteristics
Figure 5.1...... 87 Relative saturated 5-n-alkylresorinol homologue abundances for rye and wheat grains
Figure 5.2...... 88 Compound-specific carbon isotope compositions for saturated 5-n-alkylresorcinol homologues for rye and wheat grains
Figure 6.1...... 106 Column construction and elution scheme for sequential in-cell separation of total lipid extracts using accelerated solvent extraction
Figure 6.2...... 107 Combined selected ion mass chromatograms for wetland soil total lipid extracts and lipid fractions
Figure A.1...... 124 Age model for lake sediments from Olduvai Gorge using dated tie points
Figure A.2...... 125 Cross-correlation functions for calculated orbital precession in relation to tropical sea-surface temperatures, reconstructed polar ice volumes and total organic carbon carbon-isotope compositions
Figure A.3...... 126 Conceptual diagram describing carbon-isotopic relationships between carbon dioxide, leaf tissues, plant biomarkers and soil organic matter
Figure B.1...... 140 Published plant biomarker hydrogen-isotope compositions cross-plotted by photosynthetic pathway and growth habit
Figure B.2...... 141 Bathymethric contours and reconstructed salinity for paleolake Olduvai
xii
Figure B.3...... 142 Modeled monthly precipitation hydrogen-isotope compositions versus measured monthly precipitation for rainy seasons in central eastern Africa
Figure B.4...... 143 Measured versus modeled monthly precipitation hydrogen-isotope compositions values for eastern African
Figure B.5...... 144 Geographic locations of stations used to calculate amount effects for eastern Africa
Figure B.6...... 145 Alternative ‘landscape’ apparent fractionation factors based on modified proportions of trees versus shrubs
Figure C.1...... 154 Partial mass chromatogram for polar fractions of a representative northern FLK Zinj trench and the sedge Cyperus papyrus
xiii
LIST OF TABLES
Table 5.1...... 89 Total 5-n-alkylresorcinol content, relative saturated homologue composition and ratio of saturated homologues in rye and wheat grains
Table 5.2...... 90 Compound-specific carbon-isotope compositions and between-homologue apparent fractionation factors for saturated 5-n-alkylresorcinols in rye and wheat grains
Table 6.1...... 109 Separation of target compounds from a standard mixture using accelerated solvent extraction under optimized conditions
Table A.1...... 123 Single-factor cross-correlation after rank-transformation between carbon- isotope compositions of total organic carbon, calculated orbital precession, tropical sea-surface temperatures and reconstructed polar ice volumes
Table D.1...... 157 Compliation of published carbon-isotope compositions for plant leaves and soil organic matter, and plant–soil apparent fractionation factors
Table E.1...... 173 Compliation of published carbon-isotope compositions for plant leaves and lipids, and leaf–lipid apparent fractionation factors
Table F.1...... 181 Compilation of published hydrogen-isotope compositions for plant lipids and modeled precipitation, and precipitation–lipid apparent fractionation factors
Table G.1...... 194 Carbon and hydrogen-isotope compositions for plant lipids from modern soils near Olduvai Gorge
Table H.1...... 197 Carbon and hydrogen-isotope compositions for plant lipids from paleosols across FLK Zinjanthropus archaeological Level 22
Table I.1...... 200 Carbon and hydrogen-isotope compositions for plant lipids in selected human dietary resources
xiv
ACKNOWLEDGMENTS
Foremost I would like to thank my family and friends for tireless support during the last five years. Each of you is an inspiration and reminder to have heart.
When I make something I don’t like, I put a gold frame around it!
Pedro Friedeberg
Sincere thanks to my advisor and mentor – Katherine Freeman – for her guidance through both scientific and personal pursuits. You have provided me with unparalleled expertise and direction. I also sincerely thank Gail Ashley for her scientific acumen, selfless collaboration and unwavering advocacy on my behalf.
Finally, I would like to thank my labmates for countless hours dedicated to helping me find my way through the labyrinthine realm of biogeochemistry.
xv
Chapter 1. Introduction
1.1. Introduction
Since Darwin cast our species as part of an evolutionary framework over a century ago (Darwin
1871), marked interest and controversy have focused on the role of climate during the course of early human (hominin) evolution. Strong parallels among hominin speciation, cultural innovation and dramatic climate change in eastern Africa have given rise to numerous environmental hypotheses of hominin – and other vertebrates – evolution (c.f., Potts 1998).
Central to these hypotheses is that key junctures in eastern African faunal evolution over the last ca. 2.6 million years were directly related to climate and environmental change.
Establishing causal perspectives on physical and behavioral variability documented in the hominin fossil record requires comprehensive, high-resolution datasets for local and regional climate. However, sparse and discontinuous terrestrial sediment sequences present major challenges to reconstructing climate both through time and space, limiting the ability to link evolutionary and environmental changes. Well-constrained lake and soil sediments from hominin archaeological localities provide an opportunity to document temporal and spatial changes in local conditions in eastern Africa. This introductory chapter reviews previous research and hypotheses for hominin evolution in relation to paleoenvironmental evidence for terrestrial environmental change in eastern Africa.
1
1.2. Environmental hypotheses of hominin evolution
Environmental change and hominin evolution have been linked since Darwin’s seminal construction (Darwin 1871), which identified hominin evolution as a consequence of adaptive change on the African savanna. For instance, the ‘Savannah Hypothesis’ states that life on the
African plains mediated many of the most ‘human’ adaptations – including bipedality and technology – without a reliance on environmental change (Vrba 1980). In sharp contrast, more recent hypotheses posit that hominin evolution reflects the attendant pressures of environmental change (Potts 1998).
Recent environmental hypotheses of hominin evolution invoke environmental instability as a primary driver of faunal evolution. For instance, the ‘Turnover Pulse Hypothesis’ states that speciation and extension events among diverse African fauna resulted from abrupt changes in regional climate (Vrba 1995). This view accounts for apparent synchronization among fossil evidence for African fauna and flora, but does not explain differences in the magnitudes and frequencies among ‘turnover’ events (Bobe & Behrensmeyer 2004). More recent views, such as the ‘Variability Selection Hypothesis’ (Potts 1996) emphasize environmental variability, rather than step-like or directional environmental changes, as key to adaptive evolutionary responses.
However, a problem with evaluating any of these hypotheses is the lack of high-resolution, terrestrial records of paleoenvironmental conditions (c.f., deMenocal 2004).
2
1.3. Records of environmental change for eastern Africa
Terrestrial and marine sediments present complimentary perspectives on environmental change for eastern Africa during the last 2.6 million years. Coarse pollen spectra from the Ethiopian highlands and the Turkana basin indicate a regional shift towards arid-adapted plants (e.g., shrubs and grasses) between about 2.6 and 2.4 Ma (Bonnefille 1983). Stable carbon-isotope compositions of soil carbonates from the Turkana basin and Olduvai Gorge indicate protracted grassland expansion between about 2.6 and 1.2 Ma (Cerling et al. 2011), with a sharp increase in arid-adapted plants near 1.8 Ma (c.f., Plummer et al. 2009). Marine dust-flux records also document gradual regional aridification beginning about 2.6 Ma, but distinct precession (21 kyr) and obliquity (41 kyr) cycles are juxtaposed on this trend (deMenocal 2004). How specifically these cycles were manifested at local scales across eastern Africa remains unclear.
1.4. The rise of arid ecosystems in Africa
Precipitation variability forms a cornerstone of ecosystem dynamics in eastern Africa via water availability (Scholes & Archer 1997). Surface-water availability directly influences plant functional type (PFT) distribution and abundances across the landscape (Coughenour & Ellis
1993), particularly for woody plants and grasses. Today, woody plant cover shows a strong positive correlation with mean annual precipitation in arid and semi-arid environments, whereas grasses have an inverse relationship (Sankaran et al. 2004). Soil carbonate records indicate the distribution of open, arid-adapted (C4) grasslands have varied since their emergence in the
Miocene (Cerling et al. 2011). The mechanisms of C4 expansion likely differ depending on
3 scale, but recent work suggests an amalgamation of factors that include precipitation amounts and seasonality, fire regime and ecosystem fragmentation (Osborne & Beerling 2006).
1.5. Organic matter signatures for ecosystems
Buried in place, soils incorporate organic matter from standing biomass. Soil organic matter
(SOM) primarily derives from vascular plant litter (Kögel-Knabner 2002), and stable carbon-
13 isotope compositions of SOM (δ CSOM) are related to parent plant materials. In eastern Africa, most woody plants use the C3 photosynthetic pathway, whereas arid-dapted grasses use the C4
13 photosynthetic pathway. Because these two pathways differ in discrimination against CO2
13 during photosynthesis, δ CSOM values can serve as an indicator of plant community composition. Recently, Cerling et al. (2011) established a method for quantifying the fraction of
13 woody plant cover in tropical ecosystems based on δ CSOM values. This approach presents a major advance for descriptive reconstructions of ancient ecosystems, but is limited by uncertainties about the contribution and effects of microbial processes to refractory SOM components (Simpson et al. 2007).
Biochemical remains of ancient plants – or ‘biomarkers’ – present an opportunity to circumvent challenges associated with SOM provenance and preservation. Biomarkers carry ecologically specific molecular and isotopic signals that can illustrate plant functional type (PFT) characteristics. For instance, long-chained n-alkanes such as nonacosane (nC29) and hentriacontane (nC31) are leaf-waxes from terrestrial plants. Carbon-isotope compositions of
4
13 13 leaf-waxes (δ Cwax), like δ CSOM values, can reveal plant community composition, though the
13 13 relationships between δ CSOM values and δ Cwax values remains unclear.
1.6. Organic geochemical proxies for hydroclimate
Hydrogen-isotope compositions (δD) of biomarkers reflect biochemical, physiological and environmental influences on the isotopic fractionation of source-waters (Sachse et al. 2012).
Plant-water δD values reflect soil-water δD values (Ehlringer & Dawson 1992), which, in turn, generally reflect precipitation δD values in arid and semi-arid environments (Gibson et al. 2008).
Since soil water is not fractionated significantly during uptake, stem-water δD values are often similar to soil-water δD values (Ehleringer & Dawson 1992). In contrast, leaf-water δD values can vary markedly from soil-water δD values because of transpiration (Farquhar et al. 2007).
The relative importance of stem-water versus leaf-water during biosynthesis remains unclear
(McInerney et al. 2011) and could account for differences in apparent fractionation between source-waters and lipids (εlipid/water) among living plants when grouped according to PFT (Sachse et al 2012).
Source-water δD values are recorded in biomarker δD values from plants, but relationships between these two values vary among PFTs and different biomarkers. Previous studies establish
εlipid/water values between source-waters and nC29 in relation to photosynthetic pathway and growth form (Sachse et al. 2012). However, nC29 is less abundant than nC31 in leaves of tropical plants (Rommerskirchen et al. 2006, Vogts et al. 2009) and in sediments. Further complications
5 arise because quantitative methods for constraining relative PFT abundances from biomarkers prevent application of εlipid/water values in many paleoenvironmental (i.e., sedimentary) contexts.
1.7. Isotopic notation
Carbon-isotope compositions are reported in delta (δ) notation in units of per mil (‰; i.e.,
1000×) relative to the standard Vienna Pee Dee Belemnite (VPDB):
13 13 13 13 13 12 δ C = (( Rsample / RVPDB) – 1) where R = ( C / C)
Hydrogen-isotope compositions are expressed relative to the standard Vienna Standard Mean
Ocean Water (VSMOW):
2 2 2 2 1 δD = (( Rsample / RVSMOW) – 1) where R = ( H / H)
Apparent isotopic fractionation factors between two reservoirs is defined as:
αA–B = (RA / RB) where R = (N* / N) of a heavy (N*) to light (N) isotope
Typical αA–B are inconvenient for comparison because values vary in the fifth decimal place (i.e.,
1000×). Thus, isotopic fractionation is often expressed instead in Delta (Δ) and epsilon (ε) notation:
ΔA–B = ((δB – δA) / (1 + (δA / 1000))) where δ = ((RA / RB) – 1) and R = (N* / N)
3 εA–B = (10 (δA + 1000) / (δB + 1000)) where δ = ((RA / RB) – 1) and R = (N* / N)
This notation mixes disciplinary nomenclature, but values can be compared direction to larger bodies of ecological (ΔA–B) and geological (εA–B) literature. In this dissertation, I do not mix notations and processes – that is, ΔA–B describes isotopic fractionation during photosynthesis, but
εA–B describes fractionation during biosynthesis. Although ΔA–B and εA–B are algebraically equivalent, respective sources and products are reversed. For example, ΔA–B measures
6 fractionation of δB relative to δA, but εA–B measures fractionation of δA relative to δB. Thus, ΔA–B values are approximately equivalent to negative εA–B values, assuming differences between δA and δB less than about 30‰.
1.8. Research objectives and dissertation outline
The following dissertation chapters aim to refine the link between hominin evolution and environmental change through characterization of biochemical, physiological and environmental influences on molecular and isotopic signatures in modern and ancient sedimentary organic matter. This dissertation focuses on the following insights and outstanding questions:
Chapter 2: Carbon-isotope compositions of plant biomarkers vary with relative PFT abundances,
but quantitative relationships between isotopic signatures and plant communities
have not been described. How do plant biomarker δ13C values scale with the relative
abundances of woody plants, herbaceous plants and grasses? How variable were
hominin habitats, particularly vegetation, in eastern Africa during the early
Pleistocene?
Chapter 3: Hydrogen-isotope compositions of plant and algal biomarkers record source-water
signatures, but vary with PFT. Can plant biomarker δD values be paired with δ13C
values to account for biochemical and physiological influences on fractionation?
How variable was hydroclimate in eastern Africa during the early Pleistocene?
7
What is the link between tropical hydroclimate and hominin habitats, particularly
vegetation?
Chapter 4: Plant biomarker signatures in lake sediments reflect integrated, catchment-scale
environmental conditions, but hominin evolutionary responses were a function of
microhabitat features. How heterogeneous were hominin habitats across space?
What are the spatial associations between vegetation and hominin artifacts?
Chapter 5: Plant-derived 5-n-alkylresorcinols are a potentially useful tool for identifying wheat
and rye products in the human diet, but the natural abundance of these compounds in
grains remains unclear. Can distributions and isotope compositions of 5-n-
alkylresorcinol homologues distinguish between closely related dietary resources?
Chapter 6: Extraction and chromatographic isolation of plant biomarkers for molecular and
isotopic characterization currently demands extensive manual labor and time,
limiting data resolution. Accelerated solvent extraction techniques significantly
reduce extraction times, but have not adapted for column chromatography. Can
techniques be developed (and optimized) allow for more rapid and efficient isolation
of plant biomarkers?
8
1.9. Communications and publications from this dissertation
Each chapter of this dissertation is intended to constitute a publishable (stand-alone) unit of research that is also part of a cohesive dissertation. Current and anticipated publications arising from this dissertation include:
Chapter 2: Magill C, Ashley G & Freeman K (2013) Ecosystem variability and early human habitats in eastern Africa. Proceedings of the National Academy of Sciences 110:1167–1174.
Chapter 3: Magill C, Ashley G & Freeman K (2013) Water, plants and early human habitats in eastern Africa. Proceedings of the National Academy of Sciences 110:1175–1180.
Chapter 4: Magill C, Ashley G, Dominguez-Rodrigo M & Freeman K (Submitted) Plant biomarker patterns at Olduvai Gorge establish social meat consumption by pre-modern humans. Nature Communications.
Chapter 5: Magill C & Freeman K (To Be Submitted) Differentiation of wholegrain biomarkers based on compound-specific 13C signatures of 5-n-alkylresorcinols. Food Chemistry.
Chapter 6: Magill C & Freeman K (To Be Submitted) Efficient in-cell separation of lipid biomarkers using sequential accelerated solvent extraction. Organic Geochemistry.
9
1.10. References
Bobe R & Behrensmeyer A (2004) The expansion of grassland ecosystems in Africa in relation to mammalian evolution and the origin of the genus Homo. Palaeogeography, Palaeoclimatology, Palaeoecology 207:399.
Bonnefille R (1983) Evidence for a cooler and drier climate in the Ethiopian Uplands towards 2.5 Myr ago. Nature 303:487.
Cerling T, et al. (2011) Woody cover and hominin environments in the past 6 million years. Nature 476:51.
Coughenour M & Ellis J (1993) Landscape and climatic control of woody vegetation in a dry tropical ecosystem: Turkana District, Kenya. Journal of Biogeography 20:383.
Darwin C (1871) The Descent of Man and Selection in Relation to Sex (Murray Press). deMenocal P (2004) African climate change and faunal evolution during the Pliocene- Pleistocene. Earth and Planetary Science Letters 220:3.
Ehleringer J & Dawson T (1992) Water uptake by plants: perspectives from stable isotope composition. Plant, Cell & Environment 15:1073.
Farquhar G, Cernusak L & Barnes B (2007) Heavy water fractionation during transpiration. Plant Physiology 143:11.
2 18 Gibson J, Birks S & Edwards T (2008) Global prediction of δA and δ H-δ O evaporation slopes for lakes and soil water accounting for seasonality. Global Biogeochemical Cycles 22:GB2031.
Kögel-Knabner I (2002) The macromolecular organic composition of plant and microbial residues as inputs to soil organic matter. Soil and Biological Biochemistry 34:139.
McInerney F, Helliker B & Freeman K (2011) Hydrogen isotope ratios of leaf wax n-alkanes in grasses are insensitive to transpiration. Geochimica et Cosmochimica Acta 75:541.
Osborne C & Beerling D (2006) Nature’s green revolution: the remarkable evolutionary rise of C4 plants. Philosophical Transactions of the Royal Society B 361:173.
Plummer T, et al. (2009) Oldest evidence for toolmaking hominins in a grassland-dominated ecosystem. PLOS One 4:e7199.
Potts R (1996) Evolution and climate variability. Science 273:922.
10
Potts R (1998) Environmental hypotheses of hominin evolution. Yearbook of Physical Anthropology 41:93.
Rommerskirchen F, et al. (2006) Chemotaxonomic significance of distribution and stable carbon isotopic composition of long-chain alkanes and alkan-1-ols in C4 grass waxes. Organic Geochemistry 37:1303.
Sachse D, et al. (2012) Molecular paleohydrology, interpreting the hydrogen-isotopic composition of lipid biomarkers from photosynthetic organisms. Annual Review of Earth and Planetary Sciences 40:221.
Sankaran M, et al. (2005) Determinants of woody cover in African savannas. Nature 438:846.
Scholes R & Archer S (1997) Tree-grass interactions in savannas. Annual Reviews of Ecological Systems 28:517.
Simpson A, et al. (2007) Microbially derived inputs to soil organic matter: are current estimates too low? Environtal Science and Technology 41:8070.
Vogts A, et al. (2009) Distribution patterns and stable carbon isotopic composition of alkanes and alkan-1-ols from plant waxes of African rain forest and savanna C3 species. Organic Geochemistry 40:1037.
Vrba E (1980) Evolution, species and fossil: how does life evolve? South African Journal of Science 76:61.
Vrba E (1995) The fossil record of African antelopes (Mammalia, Bovidae) in relation to human evolution and paleoclimate. Paleoclimate and Evolution with Emphasis on Human Origins, eds Vrba E, et al. (Yale University Press).
11
Chapter 2. Ecosystem variability and early human habitats in eastern Africa
Reprinted with permission from the Proceedings of the National Academy of Sciences (2013) 110:1167–1174.
2.1. Abstract
The role of savannas during the course of early human evolution has been debated for nearly a century, in part, because of difficulties in characterizing local ecosystems from fossil and sediment records. Here, we present high-resolution lipid biomarker and isotopic signatures for organic matter preserved in lake sediments at Olduvai Gorge during a key juncture in human evolution about 2.0 million years ago – the emergence and dispersal of Homo erectus (sensu lato). Using published data for modern plants and soils, we construct a framework for ecological interpretations of stable carbon-isotope compositions (expressed as δ13C values) of lipid biomarkers from ancient plants. Within this framework, δ13C values for sedimentary leaf lipids and total organic carbon from Olduvai Gorge indicate recurrent ecosystem variations whereby open C4 grasslands abruptly transitioned to closed C3 forests within several hundreds to thousands of years. Carbon-isotopic signatures correlate most strongly with Earth’s orbital geometry (precession), with tropical sea-surface temperatures as significant secondary predictors in partial regression analyses. The scale and pace of repeated ecosystem variations at Olduvai
Gorge contrast with long-held views of directional or stepwise aridification and grassland expansion in eastern Africa during the early Pleistocene, and provides a local perspective on environmental hypotheses of human evolution.
12
2.2. Introduction
Climate-dependent ecosystem characteristics, such as habitat and water availability, likely influenced natural selection during human evolution (Sabeti et al. 2006). For example, woody plants might have influenced thermoregulatory and dietary adaptations in hominins and other terrestrial mammals since the Pleistocene (Teaford & Ungar 2000, McNabb 2005, Pickering &
Bunn 2007), about 2.6 million years ago (Ma). Unfortunately, in many cases, reconstructions of ecosystem characteristics and climate at hominin archaeological sites are limited by poor preservation and coarse temporal resolution. Moreover, discontinuities are common in terrestrial sediment sequences. As a result, much of the environmental context of human evolution has been interpreted based on regional and global conditions reconstructed from marine records
(deMenocal 2004, Kingston 2007).
The role of savannas in human evolution remains a subject of active debate (deMenocal 2004,
Kingston 2007, Cerling et al. 2011, Wood & Harrison 2011). This stems in part from historically imprecise definition of ‘savanna’ for modern and ancient ecosystems and from the difficulties of estimating plant community compositions – particularly woody cover – from sediments. Recently, Cerling et al. (2011) estimated plant community compositions based on present-day relationships between woody cover and carbon-isotope compositions for soil carbonates and soil organic matter. This approach offers insights into ecosystem structures at hominin archaeological sites (e.g., Omo-Turkana Basin), but is limited to environments supporting ancient soils (paleosols).
13
Here, we extend the approach of Cerling et al. (2011) to include lipid biomarkers archived in lake sediments deposited between about 2.0 and 1.8 Ma at an important hominin archaeological site – Olduvai Gorge. In addition to including key junctures in human evolution (Wood &
Harrison 2011), this time interval is associated with important changes in tropical climate, including strengthening of east-west (Walker) atmospheric circulation across the Indian and
Pacific Oceans (Ravelo et al. 2004). Weakened Walker circulation prior to about 2.0 Ma was similar to conditions projected to accompany the continued rise in greenhouse gas concentrations during the coming century (Haywood et al. 2009). In order to examine connections among ocean and atmospheric circulation, regional climate, and plant community composition, we also compare our organic carbon signatures to reconstructions of polar ice volume and sea-surface temperatures (SSTs) in the Atlantic and Indian Oceans.
2.3. Background
2.3.1. Site descriptions
Olduvai Gorge lies just south of the equator in northern Tanzania (2° 48′S, 35° 06′E), where it cuts across a 50 km rift-platform basin to expose a 2.0 million year sequence of lake and river sediments (Figure 2.1). The basin formed on the western margin of the East African Rift System
(EARS) in response to extension tectonics and the growth of a large volcanic complex (Hay
1976, Ashley & Hay 2002). During the early Pleistocene, the basin area covered an estimated
3500 km2 and included a saline-alkaline lake near its center (Ashley & Hay 2002). Sediments from this central lake are composed primarily of reworked volcanic material and air-fall tuffs
14
(Hay 1976). For our study, we use samples recovered from outcrop exposures near the center of the paleolake that preserve a stratigraphic record of continuous deposition (Hay 1976).
Today, annual precipitation patterns at Olduvai Gorge and surrounding regions of eastern Africa are defined by monsoon circulation via two major convergence zones (Nicholson 2000) – the
Intertropical Convergence Zone (ITCZ) and the Interoceanic Confluence (IOC). The ITCZ marks convergence of regional trade winds, whereas the IOC marks zonal confluence of water vapor derived from the Atlantic and Indian Oceans (Figure 2.1). Seasonal migrations of the
ITCZ and the IOC result in an annual cycle consisting of two ‘rainy seasons’ separated by arid conditions that last from May to September. ‘Long rains’ (March to May) produce the largest proportion of total annual precipitation; ‘short rains’ (October to December) are more variable but also correlate with total annual precipitation (Moron et al. 2007). Today, Olduvai Gorge experiences mean annual precipitation (MAP) of about 550 mm; several independent proxy archives suggest that MAP ranged between about 400 and 900 mm during the early Pleistocene
(Bonnefille et al. 1995, Bergner et al. 2003, Sikes & Ashley 2007).
We compare our data with coeval records for alkenone-derived SSTs from the eastern Atlantic and western Indian Oceans (Clemens et al. 1996, Lisiecki & Raymo 2005, Cleaveland & Herbert
2007)(Figure 2.1). Ocean Drilling Program (ODP) site 662 (1° 23′S, 11° 44′W, 3824 m water depth) lies in the eastern Atlantic Ocean in the Gulf of Guinea. ODP site 722 (16° 37′N, 59°
48′E, 2028 m water depth) lies in the northwestern Indian Ocean. Today, SSTs at both sites are sensitive to monsoon-driven seasonal upwelling (McIntyre et al. 1989, Sonzogni et al. 1997,
15
Müller et al. 1998), but surface-sediment calibrations indicate that alkenone signals reflect mean annual SSTs (Sonzogni et al. 1997, Müller et al. 1998).
2.3.2. Sedimentary organic matter
Organic matter in lake sediments derives from bacteria, algae and plants (Mead et al 2005) and integrates contributions from the surrounding watershed (Oline & Grant 2002, Wynn & Bird
2008, Aleman et al. 2012, Teisserenc et al. 2012). In contrast, organic matter in soils depends more strongly on local sources and preservation (Levin et al. 2011, Bai et al. 2012), and can vary spatially over scales of 10s m2 or less (Wynn & Bird 2008). Lipid biomarkers are molecular fossils that have structures with biological specificity (Peters et al. 2005). Those from plants and algae carry isotopic signals derived from terrestrial and aquatic sources, respectively. For instance, long straight-chained hydrocarbons such as nonacosane (nC29) and nC31 occur abundantly in leaves (Eglinton & Hamilton 1967) and indicate sedimentary organic matter inputs from terrestrial plants (Freeman & Colarusso 2001). Although somewhat less specific, certain shorter-chained hydrocarbons (e.g., nC17 and nC19) indicate contributions from algae and cyanobacteria (Peters et al. 2005).
2.3.3. Carbon isotopes in leaves, biomarkers and soil organic matter
Photosynthetic carbon-isotopic fractionation is defined between atmospheric carbon dioxide
13 13 (δ CCO2) and leaf tissues (δ Cleaf), and its magnitude varies with plant functional type and water availability (Bowling et al. 2008, Diefendorf et al. 2010):
16
13 13 εCO2/leaf = ((δ CCO2 + 1000) / (δ Cleaf + 1000)) – 1
We note that ε values are expressed in permil (‰), which is a unit of parts per thousand.
Variability in εCO2/leaf, which is approximately equivalent to negative Δleaf values (widely used in ecological literature; c.f., Diefendorf et al. 2010), is well constrained for different photosynthetic pathways (Bowling et al. 2008), but these modern relationships are not easily applied to ancient plants because of limited preservation of leaf tissues through time. Relatively recalcitrant leaf lipids, such as nC31, afford an opportunity to circumvent this challenge provided fractionation
13 13 between δ Cleaf and nC31 (δ C31) during biosynthesis can be documented for subtropical and tropical plants (Diefendorf et al. 2011).
13 We evaluated carbon-isotopic relationships between soil organic matter (δ CSOM), leaf tissues and leaf lipids using published data for 64 plants species and 288 soils from nearly 300 tropical
13 and subtropical localities (Figure 2.2). For C3 plant-soil systems, we find that nC31 is C-
13 depleted by about 7‰ (n = 45) with respect to leaf tissue, while SOM is C-enriched by about
2‰ (n = 184) relative to leaves. Thus, in C3 ecosystems, there is an isotopic difference (εSOM/31) between nC31 and SOM equal to about 9‰:
13 13 εSOM/31 = ((δ CSOM + 1000) / (δ C31 + 1000)) – 1
13 In C4 plant-soil systems, nC31 and SOM are both C-depleted with respect to leaf tissue by about
10‰ (n = 19) and 1‰ (n = 104), respectively. Accordingly, εSOM/31 in C4 ecosystems also equals
13 13 about 9‰. Therefore, we apply εSOM/31 of 9‰ to estimate δ CSOM from δ C31 values, thus
13 extending to plant waxes the predictive capabilities of δ CSOM for estimating woody cover
(Cerling et al. 2011)(Figure 2.3). We caution that the value of 9‰ for εSOM/31 derived here is
17 based on data from xeric woodlands and scrublands, tropical deciduous forests and C4 grasslands, and may not be representative for other ecosystems (c.f., Diefendorf et al. 2011).
The fractional abundance of C3 plants is inversely related to C4 plants in tropical ecosystems
(Lloyd et al. 2008, Wynn & Bird 2008). However, the relationship between woody cover (fwoody) and C4 abundance is nonlinear (Cerling et al. 2011) because herbaceous C3 plants can occur in
13 both open and wooded ecosystems (Pearcy & Ehleringer 1984). As a result, for δ CSOM values between near-total C3 composition (–30‰) and negligible C3 cover (–14‰), we follow the
13 approach of Cerling et al. (2011) to estimate fwoody (where –13 ≤ δ CSOM ≥ –31‰):
13 2 fwoody = (sin (–1.06688 – 0.08538 δ CSOM))
2.3.4. Structural classification for ancient ecosystems
The broad functional definition for ‘savannas’ as a continuous herbaceous understory with irregular distributions of trees or bushes does not account for differences in fwoody (Ratnam et al.
2011). Therefore, we adopt definitions for African plant communities defined by the United
Nations Educational, Scientific and Cultural Organization (UNESCO)(White 1983). According to this classification scheme, (a) forests display continuous tree cover (>10 m tall) with interlocking crowns and poorly developed understory, (b) woodlands – including bush/shrublands – display open or closed stands of shrubs or trees (up to 8 m tall) with at least
40% woody plant cover and understory with grasses and other herbs, (c) wooded grasslands display 10 to 40% woody plant cover and well developed groundcover with grasses and other herbs, (d) grasslands display less than 10% woody plant cover and well developed groundcover
18 with grasses and other herbs, and (e) deserts display sparse groundcover and sandy, stony or rocky substrate. UNESCO does not distinguish between forests and woodlands in terms of woody plant cover, but here we consider forests to display greater than 80% woody plant cover.
The evolutionary implications of orbital forcing on environmental change were recognized over a century ago (Wallace 1870) but remain controversial today (Pickering & Bunn 2007, Kingston
2007). Marine sediments just off African shores document orbital rhythms in terrestrial inputs during the Pleistocene (deMenocal 2004) and a growing number of terrestrial sequences hint at a similar pacing for environmental changes in eastern Africa (Deino et al. 2006, Ashley 2007).
Marine sequences are often indirectly or too poorly constrained in time to infer relationships between key junctures in human evolution and terrestrial conditions or change (Kingston 2007).
Lithostratigraphic patterns in lake-margin sediments at Olduvai Gorge reveal five episodes of lake expansion between ca. 1.85 and 1.74 Ma, suggesting lake level changes may have tracked orbital precession (Ashley 2007). Here, we use sedimentary organic matter signatures in high- resolution and temporally well-constrained lake sediments to evaluate the magnitude and timing of ecosystem changes associated with lake level changes.
2.4. Results and Discussion
2.4.1. Ecosystem change and woody cover
13 We observe repeated shifts in δ C31 values between about –36 and –20‰ (Figure 2.4), which track closely with orbitally paced lake-margin lithostratigraphic patterns, suggesting ecosystem
19 changes at Olduvai Gorge were also influenced by orbital cycles (Figure 2.5). Total organic
13 13 carbon δ C values (δ CTOC) show a smaller isotopic range of about 9‰ and correlate tightly
13 13 with δ C31 values (r = 0.93); δ CTOC values follow terrestrial-plant inputs but are attenuated,
13 13 likely by algal or macrophytic inputs. Taken together, δ C31 and δ CTOC values suggest rapid local ecosystem shifts between closed C3 woodlands and open C4 grasslands. These changes were comparable to extreme events such as the ‘greening of the Sahara’ about 120 thousand years ago that accompanied the dispersal of modern humans out of Africa (Osborne et al. 2008).
Carbon-isotope evidence for pronounced ecosystem shifts at Olduvai Gorge contrasts with previous reconstructions for eastern Africa that proposed ecosystems were stable at local to regional scales in the early Pleistocene (Reed & Rector 2007, Bamford et al. 2008) and generally lacked closed woodlands near hominin archaeological sites since 6 Ma (Cerling et al. 2011).
13 Dramatic and rapid changes in δ C31 values highlight ecosystem instability in this region;
13 further, δ C31 values indicate that closed woodlands dominated local landscapes for up to 20% of the time.
Differences in the interpretation of ancient ecosystems, in addition to problems related to
‘savanna’ heterogeneity, can stem from inherent proxy biases (Kingston 2007). For instance,
13 13 δ CSOM values can over represent C4 inputs as a result of C-enrichment during organic matter
13 decomposition (Wynn 2007). Values for δ C31 can also over represent C4 inputs as a result of
13 inorganic carbon (e.g., bicarbonate) assimilation by macrophytes, although δ C31 values in arid environments are more likely biased towards wet conditions when plants synthesize most leaf lipids (Post-Beittenmiller 1996). Although specific mechanisms responsible for differences
20 among ecosystem proxies cannot necessarily be reconciled here, we suggest carbon-isotopic
13 signals for leaf lipids can complement δ CSOM values, which otherwise can skew towards C4 – and thus arid – signals on seasonal and longer timescales.
2.4.2. Biogeochemical variability at the ecosystem scale
Organic matter in lake sediments incorporates inputs from both aquatic and terrestrial photosynthetic organisms and can vary in proportion with productivity and deposition in a lake
13 and the surrounding watershed. At Olduvai Gorge, δ CTOC values correlate strongly with total organic carbon (%TOC; r = 0.92), but show no clear relationship with algal versus terrestrial-plant
inputs (r = 0.07)(Figure 2.4):
Palg = (nC17 + nC19) / (nC17 + nC19 + nC29 + nC31)
The abundance ratio of pristane versus phytane is constantly near 2.5, suggesting organic carbon deposition was dominated by terrestrial-plant input (Peters et al. 2005). These observations
13 13 13 corroborate strong correlation between δ CTOC and δ C31 values and suggest that δ CTOC values primarily follow terrestrial-plant inputs.
Macrophytes contain abundant intermediate chain-length n-alkanes (e.g., nC23 and nC25) but limited long chain-length homologues as compared to most terrestrial plants (Ficken et al. 2000,
Lamb et al. 2004). Thus, relative abundances of nC23 and nC25 versus nC29 and nC31 – that is,
Paq – can provide estimates for macrophytic versus terrestrial-plant inputs (Ficken et al. 2000):
Paq = (nC23 + nC25) / (nC23 + nC25 + nC29 + nC31)
21
13 In lake sediments from Olduvai Gorge, δ C31 values correlate weakly both with Paq (r = –0.41)
13 and nC25 δ C values (r = –0.44), suggesting macrophytes did not significantly contribute to nC31 inputs. Note that Paq, as well as Palg, reflect relative homologue abundances and do not necessarily indicate relative organism abundances; instead, these ratios are useful for identifying organic facies or correlations with other biogeochemical signals.
2.4.3. Mechanisms of ecosystem change
Atmospheric CO2 concentrations (pCO2), temperature, seasonality and water availability are potential determinants of C3 versus C4 plant abundance (Pearcy & Ehleringer 1984). Since the middle Pleistocene, records of pCO2 correlate strongly with polar ice volume changes (Petit et al.
1999, Lüthi et al. 2008), which were obliquity paced prior to ~1 Ma (deMenocal 2004). Lake
13 sediment δ CTOC values for Olduvai Gorge correlate weakly with reconstructed polar ice volumes (r = –0.40) based on marine oxygen-isotopic records (Ravel et al. 2004, Lisiecki &
Raymo 2005), suggesting ecosystem changes in this region did not track 41-kyr glacial cycles. If polar ice volume is a representative proxy for pCO2 during the early Pleistocene, then local ecosystem changes were not exclusively tied to pCO2 changes. This conclusion contrasts with suggestions of a dominant role for pCO2 in southern African ecosystems during the early
Pleistocene based on speleothem carbonate δ13C values (Hopley et al. 2007), but is in agreement with marine oxygen-isotopic evidence for eastern African climate sensitivity to polar ice volume
13 only after 1.0 Ma (Trauth et al. 2009). Values for δ CTOC correlate strongly with precession
(ωp) and, thus, do not support temperature as primary determinant of ecosystem change because
ωp negligibly influences mean annual temperatures (Kingston 2005). Similarly, paleosol
22 carbonates indicate mean annual temperatures varied by less than about ±5°C at and around
Olduvai Gorge during the early Pleistocene (Hay & Kyser 2001, Passey et al. 2010). Strong
13 correlation between δ CTOC and ωp further suggests that biotic (e.g., herbivory) or abiotic disturbances such as fire were not primary determinants of ecosystem change, although they may have served as feedback mechanisms that accelerated changes. In agreement with a variety of other studies (Huang et al. 2001, Bobe 2006), we suggest changing C3 and C4 plant abundances at Olduvai Gorge varied with orbital precession in response to water availability.
2.4.4. Mechanisms of hydroclimate change
Cycles of about 21-kyr are common in a variety of hydroclimate proxy records in eastern Africa
13 since the Pliocene (c.f., deMenocal 2004), and δ CTOC values correlate strongly with ωp (r =
0.78) in single factor regression (Appendix A). Although specific mechanisms responsible for these cycles remain unclear, the timing and magnitude of local and regional hydroclimate changes are consistent with theoretical effects of ωp on monsoon strength (Ruddiman 2001) – that is, higher summer insolation would enhance land-ocean temperatures contrasts, resulting in stronger monsoons and increased precipitation.
Insolation alone cannot account for the magnitude of hydroclimate change in eastern Africa
(Clement et al. 2004). Previous reconstructions based on pollen and oxygen-isotope compositions of soil carbonates suggest MAP fluctuated between approximately 400 and 800 mm at Olduvai Gorge and surrounding regions during the early Pleistocene (Bonnefille et al.
1995, Bergner et al. 2003, Ashley 2007, Sikes & Ashley 2007). Yet, in climate simulations,
23 insolation variability accounts for MAP fluctuations of less than 200 mm and mostly affects
‘long rains’ (Clement et al. 2004). Thus, precipitation amounts in the past must have been impacted by multiple factors, as they are today (Nicholson 2000).
Today, precipitation in eastern Africa responds sensitively to SSTs in the Indian Ocean and
Atlantic Ocean (Chiang 2009). In particular, intensifications of ‘short rains’ (up to 200 mm) accompany coordinated warm and cold SST anomalies in the western Indian Ocean and eastern
Atlantic Ocean (Vuille et al. 2005, Balas et al. 2007, Farnsworth et al. 2011), respectively, as a result of transcontinental surface pressure gradients across Africa and monsoon displacement of the IOC from west-to-east (Eltahir & Gong 1996). Partial regressions reveal SSTs for ODP site
662 (SST662) and site 722 (SST722) as significant (p < 0.001) secondary predictors that are statistically independent of covariance with ωp, and the combination of ωp, SST662 and SST722
13 explains 73% of the variability in δ CTOC values in a multiple regression model. During the early Pleistocene, both SST662 and SST722 show strong ωp and 41-kyr (obliquity) periodicity
(Cleaveland & Herbert 2007), but only SST662 shows a consistent relationship with monsoon- driven upwelling (Cleaveland & Herbert 2007). Since upwelling in the eastern Atlantic correlates positively with monsoon strength during late boreal summer (McIntyre et al. 1989,
Müller et al. 1998, Cleaveland & Herbert 2007), we suggest that obliquity-paced cooling in the eastern Atlantic Ocean and monsoon strengthening (and therefore stronger westerly winds) resulted in more frequent eastward displacements of the IOC and intensification of ‘short rains’ in eastern Africa.
24
2.4.5. Ecosystems and hominin evolution
Fossil evidence for pronounced aridification and faunal turnover in eastern Africa between about
2.0 and 1.8 Ma has sparked hypotheses linking the emergence and dispersal of the genus Homo to climate-driven ecosystem change (deMenocal 2004, McNabb 2005, Kingston 2007). Fossil evidence for cranial expansion in pre-modern Homo (e.g., H. erectus sensu lato) has been linked to irregular resource distributions (Snodgrass et al. 2009), and our carbon-isotopic data are consistent with enhanced ecosystem variability as a context for encephalization (Figure 2.5).
During the early Pleistocene, strong ecosystem preferences are not apparent between transitional
(e.g., H. habilis) and archaic (e.g., P. boisei) hominins (Wood & Strait 2004); however, isotopic and fossil data suggest that transitional species accessed a broad spectrum of dietary resources as compared to archaic species (Wood & Strait 2004, van der Merwe et al. 2008, Cerling et al.
2011a). Among primates, quality (i.e., energy density) of dietary resources correlates strongly with brain size (Snodgrass et al. 2009). Assuming that dietary resources were primarily unrelated to technological innovations by transitional species (Wood & Strait 2004), we hypothesize ecosystem variability favored hominin species with large brains that allowed for versatile foraging strategies and dietary diversity.
2.5. Conclusions
This study presents high-resolution biomarker and δ13C records of ecosystem variability from lake sediments at Olduvai Gorge that were deposited during an interval of pronounced shifts in
13 vertebrate community and global climate reorganization, about 2.0 to 1.8 Ma. Values of δ C31
25 indicate rapid and repeated ecosystem restructuring between closed C3 woodlands and open, C4-
13 dominated grasslands. Our δ C records reveal coupled fluctuations between ecosystem and precession with additional variability explained by SST contrasts between the Atlantic and
Indian Oceans, suggesting aridity as opposed to carbon dioxide or temperature controlled woody plant cover in eastern Africa during the early Pleistocene. We conclude that highly variable ecosystems accompanied the emergence and dispersal of the genus Homo. Our study also builds on soil data to construct a new interpretive framework for ecosystem reconstruction based on leaf lipids.
26
2.6. Figures
,0 2,03 ,0 2,03 /,0. 2,, $1 ' *456%7%8&8%9) :--'-9)8;''< AHCI AHCI ,0 => AJC '
!"#$%&#$!"' ()*$+,-$./) /,01 ,
, !"#$"%& !"#'"% D9E54)'F94G5 A)854-%885)8'1845&-B ("#$%&"'*"&%) C9)89#4B':+,,'-< 16&47
!"#$%&'()*"+ +,,,'- ?96'@, !"#$%&'& ()%*+$,*'& ++,,'-
'#""%(
. !""#$%&"'(&)
/ 0 -.
Figure 2.1: Present-day precipitation patterns during ‘long rains’ (a; March to May) and ‘short rains’ (b; October to December)(Mitchell & Jones 2005). Bold horizontal lines mark the ITCZ.
The IOC (b; dashed line) is prominent during ‘short rains’ and its eastward displacement correlates with higher seasonal precipitation (Eltahir & Gong 1996). Targets show Olduvai
Gorge’s location; closed or open circles show the locations of ODP sites 662 and 722, respectively. Panel c; Depositional environments at Olduvai Gorge during the early Pleistocene, about 2.0 to 1.8 Ma. Expanded and contracted lake levels are reconstructed based on detailed stratigraphic correlations (Hay 1976, Ashley & Hay 2002).
27
!'(!'! !( ! ( '!
!$%&' )*+ !4()*+$%&%
#!
"!
!"#$%& !
!$%&' )*+
!'()*+$%&% #!
"!
!"#$%& ! !'(!'! !( ! ( '! ,-&.*"$(/01&#&(!2' !% *0+3
13 13 Figure 2.2: Histograms and box-and-whisker plots for εSOM/31 from published δ CSOM and δ C31 values from subtropical and tropical regions (Appendix A). Published data are binned at 0.5‰
13 13 intervals (counts) based on site averages. Values for δ CSOM (dark gray) and δ C31 (light gray)
13 are plotted relative to δ Cleaf. Values of εSOM/31 equal about 9‰ in both C3 and C4 ecosystems.
Individual ‘boxes’ contain interquartile ranges (IQR). Bold vertical lines within boxes mark median values. Open circles denote mean values. Horizontal ‘whiskers’ mark minimum and maximum values, except values outside of 1.5 IQR (black circles). Notch half-width values indicate confidence in differentiating median values.
28
A; >*#?&>@)@.&! 5;A +9:;-.<=<:>?@A2 !%& !%' !(" !(( !() !(* !"" !"#$%&'()* +,-$.#$/* &!"#$%2 5 &<1$=11.(&!"#$%2&411.(&!"#$%2 5 : ; 5 ; / " +,-./-012 411.*. 87#22"#$. 411."#$. 917*2% 87#22"#$. 0/12(2%*3 +3456782 ,
!"" 411.(&!"#$% 516*7 # +,-./-012 " !!# !!$ !%! !%& !%' !("
A; B1@"&C7D#$@/&E#%%*7&! 5BCE +9:;-.<=<:>?@A2
13 Figure 2.3: Stable carbon-isotope compositions for soil organic matter (δ CSOM, bottom x-axis)
13 and nC31 (δ C31, top x-axis) relative to woody plant cover (a), ecosystem (b) and plant
13 community composition (c). The published relationship between woody plant cover and δ CSOM
13 is described by the function (where –13 ≤ δ CSOM ≥ –31‰)(Cerling et al. 2011):
13 2 fwoody = (sin (–1.06688 – 0.08538 δ CSOM))
13 13 Here, we relate δ CSOM to δ C31 using a value of 9‰ for εSOM/31. The relationship between
13 13 woody plant cover and δ C31 is thus described by the function (where –22 ≤ δ CSOM ≥ –40‰):
13 2 fwoody = (sin (–1.83530 – 0.08538 δ C31))
Ecosystem definitions adhere to African plant communities according to UNESCO terminology
(White 1983).
29
! ! 5$6+#'7,.+%"8' "$% "# 23 /*+('/"0"1' 4 !;<=>AB ! 32 4+,9$% !; !"#$%&'('#)*+,- !.%/012#3%&4%52- !;<=>@?>A< @A>@B>@?>A< !!! ! !"##"$%&'$(')*+,&'-.$
! ! ! 678%#4952&742/95 ! ! 5$6+#'7,.+%"8'4+,9$%' /+:*;!+,."% >,"&6+%*'6$ !+8,$0<=6*' 23 /"6<$&6,+6".,+0<= ><=6+%* 23 ! 4574 ! 4?@ !"#$%&'('#)*+,- !*&:*1- !"#$%&'('#)*+,-
Figure 2.4: Biomarker and isotopic signatures for organic matter preserved in lake sediments at
13 13 13 Olduvai Gorge. a, Leaf lipid δ C values for nC31 (δ C31). b, Total organic carbon δ C values
13 (δ CTOC). c, Total organic carbon percentages (%TOC). d, Diagrammatic depiction of relative lake level changes at Olduvai Gorge based on lake-margin lithostratigraphy (Ashley 2007). e,
Ratios of algal lipids (nC17 + nC19) relative to algal and terrestrial-plant lipids (nC17 + nC19 + nC29
+ nC31) – termed Palg. Higher values reflect relatively increased algal inputs (Peters et al. 2005). f, Ratios of pristane (Pr) to phytane (Ph). In lake sediments, values above 2 generally reflect a dominance of terrestrial-plant inputs (Peters et al. 2005). g, Ratios of macrophytic lipids (nC23 + nC25) relative to macrophytic and terrestrial-plant lipids (nC23 + nC25 + nC29 + nC31) – that is, Paq.
Higher values reflect relatively increased macrophytic inputs (Ficken et al. 2000). h, Values of
13 13 δ C for the macrophytic lipid nC25 (δ C25). Horizontal gray bands highlight periods of time
13 characterized by δ C31 values higher than about –28‰ (i.e., open-grassland and wooded- grassland ecosystems).
30
("% ;HIJ
)"% /*+B/C,(+8*'0*6<*,+4C,*
''@''''' + >G ?"% ! (# //0 ,-./ ,0123456517;<=>/ E#$A+#'F*%49"8 D&4"6+4*& 122 A (' ''''
(* //0 ,-./ 332 (% ''''
8 () +$"$' : $ " !"#"$ 7
:,*6$7*,%'!"#" %&''&()*+'+),-.-/+(-01-
@,A"4+#':,*8*&&"$% 0,+%&"4"$%+#'5$6"%"%& -,89+"8'5$6"%"%& !*.+7$%4'-,89+"8'5$6"%"%&
!!' !($
@ ,012345651789:/ ;*+(';"<"7 ! 0@= >? !(! >@ !(# ''= ''= >? ?> ! > ,012345651789:/ !(' * A !)' 0$4+#'@,.+%"8'=+,A$% !"#$$ !"#%$ !"&$$ !"##"$%&'$(')*+,&'-.$
Figure 2.5: Time series for global, regional and local proxy indicators during the early
Pleistocene. a, Global benthic oxygen-isotopic composite (Lisiecki & Raymo 2005). Gray lines connect original data and the bolded black line shows a 5-point smoothing. b and c, Alkenone- derived sea-surface temperature estimates for the western Indian Ocean (SST722) and eastern
Atlantic Ocean (SST662), respectively (Clemens et al. 1996, Lisiecki & Raymo 2005, Cleaveland
& Herbert 2007). Gray lines connect original data and bolded black lines show 5-point smoothing. d, Calculated orbital precession (ωp). Values for ωp equal the product of calculated eccentricity (e) and the sine function of the longitude of the perihelion (ω). e, Leaf lipid δ13C
13 values for nC31 (δ C31, white circles connected by bolded black lines) and total organic carbon
13 13 δ C values (δ CTOC, gray lines). G, grassland; WG, wooded grassland; W, bush/shrubland and
31 woodland; F, forest. Ecosystem definitions adhere to African plant communities according to
13 UNESCO terminology (White 1983) based on δ C31 values (see Figure 2.4). Vertical gray
13 bands highlight periods of time characterized by δ C31 values higher than about –28‰ (i.e., open-grassland and wooded-grassland ecosystems). Hominin taxonomic grades and fossil occurrences are illustrated after Wood (2010); skulls denote taxonomic grades that have been identified at Olduvai Gorge.
32
2.7. References
Aleman J, et al. (2012) Reconstructing savanna tree cover from pollen, phytoliths and stable carbon isotopes. Journal of Vegetation Science 23:187.
Ashley G (2007) Orbital rhythms, monsoons, and playa lake response, Olduvai Basin, equatorial East Africa (ca. 1.85-1.74 Ma). Geology 35:1091.
Ashley G & Hay R (2002) Sedimentation patterns in a Plio-Pleistocene volcaniclastic rift- platform basin, Olduvai Gorge, Tanzania. SEPM Special Publication 73:107.
Bai E, et al. (2012) Spatial patterns of soil δ13C reveal grassland-to-woodland successional processes. Organic Geochemistry 42:1512.
Balas N, et al. (2007) The relationship of rainfall variability in West Central Africa to sea- surface temperature fluctuations. International Journal of Climatology 27:1335.
Bamford M, et al. (2008) Late Pliocene grassland from Olduvai Gorge, Tanzania. Palaeogeography, Palaeoclimatology, Palaeoecology 257:280.
Bergner A, Trauth M, & Bookhagen B (2003) Paleoprecipitation estimates for the Lake Naivasha basin (Kenya) during the last 175 k.y. using a lake-balance model. Global and Planetary Change 36:117.
Bintanja R & van de Wal R (2008) North American ice-sheet dynamics and the onset of 100,000-year glacial cycles. Nature 454:869.
Bobe R (2006) The evolution of arid ecosystems in eastern Africa. Journal of Arid Environments 66:564.
Bonnefille R, et al. (1995) Glacial/interglacial record from intertropical Africa, high resolution pollen and carbon data at Rusaka, Burundi. Quaternary Science Reviews 14:917.
Bowling D, Pataki D, & Randerson J (2008) Carbon isotopes in terrestrial ecosystem pools and CO2 fluxes. New Phytologist 178:24.
Brodie et al. (2011) Evidence for bias in C and N concentrations and δ13C composition of terrestrial and aquatic organic materials due to pre-analysis acid preparation methods. Chemical Geology 282:67.
Cerling T, et al. (2011) Diet of Paranthropus boisei in the early Pleistocene of East Africa. Proceedings of the National Academy of Sciences 108:9337.
33
Cerling T, et al. (2011) Woody cover and hominin environments in the past 6 million years. Nature 476:51.
Chatfield C (2004) The Analysis of Time Series: An Introduction (CRC Press).
Chiang J (2009) The tropics in paleoclimate. Annual Review of Earth and Planetary Sciences 37:263.
Cleaveland L & Herbert T (2007) Coherent obliquity band and heterogeneous precession band responses in early Pleistocene tropical sea surface temperatures. Paleoceanography 22:PA2216.
Clemens S, Murray D, & Prell W (1996) Nonstationary phase of the Plio-Pleistocene Asian monsoon. Science 274:943.
Clement A, Hall A, & Broccoli A (2004) The importance of precessional signals in the tropical climate. Climate Dynamics 22:327.
Deino A, et al. (2006) Precessional forcing of lacustrine sedimentation in the late Cenozoic Chemeron Basin, Central Kenya Rift, and calibration of the Gauss/Matuyama boundary. Earth and Planetary Science Letters 247:41. deMenocal P (2004) African climate change and faunal evolution during the Pliocene- Pleistocene. Earth and Planetary Science Letters 220:3.
Diefendorf A, et al. (2010) Global patterns in leaf 13C discrimination and implications for studies of past and future climate. Proceedings of the National Academy of Sciences 107:5738.
Diefendorf A, et al. (2011) Production of n-alkyl lipids in living plants and implications for the geologic past. Geochimica et Cosmochimica Acta 75:7472.
Eglinton G & Hamilton R (1967) Leaf epicuticular waxes. Science 156:1322.
Eltahir E & Gong C (1996) Dynamics of wet and dry years in West Africa. Journal of Climate 9:1030.
Farnsworth A, et al. (2011) Understanding the large scale driving mechanisms of rainfall variability over Central Africa. African Climate and Climate Change 43:101.
Ficken K, et al. (2000) An n-alkane proxy for the sedimentary input of submerged/floating freshwater aquatic macrophytes. Organic Geochemistry 31:745.
Freeman K & Colarusso L (2001) Molecular and isotopic records of C4 grassland expansion in the late Miocene. Geochimica et Cosmochimica Acta 65:1439.
Hay R (1976) Geology of the Olduvai Gorge (University of California Press).
34
Hay R & Kyser T (2001) Chemical sedimentology and paleoenvironmental history of Lake Olduvai, a Pliocene lake in northern Tanzania. Geological Society of America Bulletin 113:1505.
Haywood A, et al. (2009) Pliocene climate, processes and problems. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 367:3.
Hopley P, et al. (2007) High-and low-latitude orbital forcing of early hominin habitats in South Africa. Earth and Planetary Science Letters 256:419.
Huang Y, et al. (2001) Climate change as the dominant control on glacial-interglacial variations in C3 and C4 plant abundance. Science 293:1647.
Kingston J (2005) Orbital controls on seasonality. Seasonality in Primates: Studies of Living and Extinct Human and Non-Human Primates, eds Brockman D & van Schaik C (Cambridge University Press).
Kingston J (2007) Shifting adaptive landscapes: progress and challenges in reconstructing early hominid environments. American Journal of Physical Anthropology 134:20.
Lamb A, et al. (2004) Holocene climate and vegetation change in the Main Ethiopian Rift Valley, inferred from the composition (C/N and δ13C) of lacustrine organic matter. Quaternary Science Reviews 23:881.
Levin N, et al. (2011) Paleosol carbonates from the Omo Group: isotopic records of local and regional environmental change in East Africa. Palaeogeography, Palaeoclimatology, Palaeoecology 307:75.
Lisiecki L & Raymo M (2005) A Plio-Pleistocene stack of 57 globally distributed benthic 18O records. Paleoceanography 20:522.
Lloyd J, et al. (2008) Contributions of woody and herbaceous vegetation to tropical savanna ecosystem productivity: a quasi-global estimate. Tree Physiology 28:451.
Lüthi D, et al. (2008) High-resolution carbon dioxide concentration record 650,000-800,000 years before present. Nature 453:379.
McIntyre A, et al. (1989) Surface water response of the equatorial Atlantic Ocean to orbital forcing. Paleoceanography 4:19.
McNabb J (2005) Hominins and the Early-Middle Pleistocene transition: evolution, culture and climate in Africa and Europe. Geological Society, London, Special Publications 247:287.
35
Mead R, et al. (2005) Sediment and soil organic matter source assessment as revealed by the molecular distribution and carbon isotopic composition of n-alkanes. Organic Geochemistry 36:363.
Mitchell T & Jones P (2005) An improved method of constructing a database of monthly climate observations and associated high resolution grids. International Journal of Climatology 25:693.
Moron V, et al. (2007) Spatial coherence of tropical rainfall at the regional scale. Journal of Climate 20:5244.
K' Müller P, et al. (1998) Calibration of the alkenone paleotemperature index U37 based on core- tops from the eastern South Atlantic and the global ocean (60°N-60°S). Geochimica et Cosmochimica Acta 62:1757.
Nicholson S (2000) The nature of rainfall variability over Africa on time scales of decades to millenia. Global and Planetary Change 26:137.
Oline D & Grant M (2002) Scaling patterns of biomass and soil properties: an empirical analysis. Landscape Ecology 17:13.
Osborne C, et al. (2008) A humid corridor across the Sahara for the migration of early hodern humans out of Africa 120,000 years ago. Proceedings of the National Academy of Sciences 105:16444.
Passey B, et al. (2010) High-temperature environments of human evolution in East Africa based on bond ordering in paleosol carbonates. Proceedings of the National Academy of Sciences 107:11245.
Pearcy R & Ehleringer J (1984) Comparative ecophysiology of C3 and C4 plants. Plant, Cell & Environment 7:1.
Peters K, Walters C, & Moldowan J (2005) The Biomarker Guide (Cambridge University Press).
Petit J, et al. (1999) Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399:429.
Pickering T & Bunn H (2007) The endurance running hypothesis and hunting and scavenging in savanna woodlands. Journal of Human Evolution 53:434.
Post-Beittenmiller D (1996) Biochemistry and molecular biology of wax production in plants. Annual Review of Plant Biology 47:405.
Ratnam J, et al. (2011) When is a ‘forest’ a savanna, and why does it matter? Global Ecology and Biogeography 20:653.
36
Ravelo A, et al. (2004) Regional climate shifts caused by gradual global cooling in the Pliocene epoch. Nature 429:263.
Reed K & Rector A (2007) African Pliocene Paleoecology (Oxford University Press).
Ruddiman W (2001) Earth’s Climate: Past and Future (WH Freeman and Company).
Sabeti P, et al. (2006) Positive natural selection in the human lineage. Science 312:1614.
Scargle J (1982) Studies in astronomical time series analysis. Statistical aspects of spectral analysis of unevenly spaced data. The Astrophysical Journal 263:835.
Sikes N & Ashley G (2007) Stable isotopes of pedogenic carbonates as indicators of paleoecology in the Plio-Pleistocene (upper Bed I), western margin of the Olduvai Basin, Tanzania. Journal of Human Evolution 53:574.
Snodgrass J, Leonard W, & Robertson M (2009) The energetics of encephalization in early hominids. The Evolution of Hominin Diets: Integrating Approaches to the Study of Palaeolithic Subsistence, eds Hublin J & Richards M (Springer Science).
Sonzogni C, et al. (1997) Temperature and salinity effects on alkenone ratios measured in surface sediments from the Indian Ocean. Quaternary Research 47:344.
Teaford M & Ungar P (2000) Diet and the evolution of the earliest human ancestors. Proceedings of the National Academy of Sciences 97:13506.
Teisserenc R, et al. (2010) Integrated transfers of terrigenous organic matter to lakes at their watershed level: a combined biomarker and GIS analysis. Geochimica et Cosmochimica Acta 74:6375.
Trauth M, Larrasoaña J, & Mudelsee M (2009) Trends, rhythms and events in Plio-Pleistocene African climate. Quaternary Science Reviews 28:399. van der Merwe N, Masao F, & Bamford M (2008) Isotopic evidence for contrasting diets of early hominins Homo habilis and Australopithecus boisei of Tanzania. South African Journal of Science 104:153.
Vuille M, et al. (2005) Stable isotopes in East African precipitation record Indian Ocean zonal mode. Geophysical Research Letters 32:L21705.
Wallace A (1870) Man and natural selection. Nature 3:8.
White F (1983) The Vegetation of Africa (United Nations Scientific and Cultural Organization).
Wood B (2010) Reconstructing human evolution: achievements, challenges, and opportunities. Proceedings of the National Academy of Sciences 107:8902.
37
Wood B & Harrison T (2011) The evolutionary context of the first hominins. Nature 470:347.
Wood B & Strait D (2004) Patterns of resource use in early Homo and Paranthropus. Journal of Human Evolution 46:119.
Wynn J (2007) Carbon isotope fractionation during decomposition of organic matter in soils and paleosols: implications for paleoecological interpretations of paleosols. Palaeogeography, Palaeoclimatology, Palaeoecology 251:437.
Wynn J & Bird M (2008) Environmental controls on the stable carbon isotopic composition of soil organic carbon: implications for modelling the distribution of C3 and C4 plants, Australia. Tellus B 60:604.
38
Chapter 3. Water, plants and early human habitats in eastern Africa
Reprinted with permission from the Proceedings of the National Academy of Sciences (2013) 110:1175–1180.
3.1. Abstract
Water and its influence on plants likely exerted strong adaptive pressures in human evolution.
Understanding relationships among water, plants, and early humans is limited both by incomplete terrestrial records of environmental change and by indirect proxy data for water availability. Here, we present a continuous record of stable hydrogen-isotope compositions
(expressed as δD values) for lipid biomarkers preserved in lake sediments from an early
Pleistocene archaeological site in eastern Africa – Olduvai Gorge. We convert sedimentary leaf and algal-lipid δD values into estimates for ancient source-water δD values by accounting for biochemical, physiological, and environmental influences on isotopic fractionation via published water–lipid enrichment factors for living plants, algae, and recent sediments. Reconstructed precipitation and lake-water δD values, respectively, are consistent with modern isotopic hydrology, and reveal that dramatic fluctuations in water availability accompanied ecosystem changes. Drier conditions, indicated by less negative δD values, occur in association with stable carbon-isotopic evidence for open, C4-dominated grassland ecosystems. Wetter conditions, indicated by lower δD values, are associated with expanded woody cover across the ancient landscape. Estimates for ancient precipitation amounts, based on reconstructed precipitation δD values, range between about 250 and 700 mm yr–1, and are consistent with modern precipitation data for eastern Africa. We conclude that freshwater availability exerted a substantial influence
39 on eastern African ecosystems and, by extension, was central to early human proliferation during periods of rapid climate change.
3.2. Introduction
The role of water and ecosystem change in human evolution remains a subject of active debate
(deMenocal 2004, Potts 2007, Cerling et al. 2011), but experts widely acknowledge these factors likely shaped early human (hominin) foraging strategies and diet (Teaford & Ungar 2000) about
2.0 to 1.8 million years ago (Ma). According to marine records, this juncture occurred during an interval of protracted grassland expansion across eastern Africa (deMenocal 2004). In contrast, coeval terrestrial records from hominin archaeological sites such as Olduvai Gorge indicate recurrent fluctuations between open-grassland and closed-woodland ecosystems (Magill et al.
2013). Although such ecosystem fluctuations are commonly interpreted in relation to aridity, grassland expansion is sensitive to multiple factors (Sankaran et al. 2005) and proxy signals more closely linked to meteoric waters could strengthen interpretations.
Here, we use δD values for lipid biomarkers preserved in lake sediments to reconstruct source- water δD values at Olduvai Gorge during the early Pleistocene. In modern lake basins, precipitation and lake-water δD values correlate strongly with δD values for leaf and algal-lipids, respectively, after accounting for isotopic fractionation during lipid biosynthesis (Sachse et al.
2012). Present and past source-water δD values reflect the combined influences of vapor-source, transport history, and aridity, ultimately linking local hydrologic patterns to atmospheric and oceanic-circulation dynamics (Gibson et al. 2008).
40
3.3. Background
3.3.1. Sample locality
Olduvai Gorge is incised into the eastern margin of the Serengeti Plain in northern Tanzania
(Figure 3.1). Gorge walls expose a thick sequence of volcaniclastic sediments and tuff accumulated in lake and lake-margin environments (Hay 1976, Ashley & Hay 2002). Between about 2.0 and 1.8 Ma, a perennial saline-alkaline lake (paleolake Olduvai) occupied the center of the closed basin (Hay 1976, Ashley & Hay 2002, Ashley 2007). Lake-margin stratigraphy suggests paleolake Olduvai responded sensitively to local and regional precipitation patterns during this 200-kyr interval (Hay & Kyser 2001, Ashley 2007). Stable carbon-isotope compositions (expressed as δ13C values) of leaf-lipids preserved in the lake sediments vary by over 15‰, suggesting pronounced ecosystem fluctuations accompanied lake-level fluctuations
(Magill et al. 2013). Here, we measure leaf-lipid δD values in a subset of the same lake sediment outcrop. All are from locality 80 (Bed I), which accumulated near the depocenter of paleolake Olduvai (Ashley & Hay 2002) and was exposed by stream incision during the late
Pleistocene (Hay 1976).
3.3.2. Precipitation patterns in eastern Africa
In eastern Africa, precipitation patterns reflect interactions between regional highlands and two convergent boundaries (Figure 3.1). The Intertropical Convergence Zone (ITCZ) and closely associated African rainbelt (Nicholson 2000) mark north-south convergence of monsoon
41 airstreams. The Interoceanic Confluence (IOC; also referred to as the Congo Air Boundary) marks east-west convergence of air masses derived from the Indian and Atlantic Oceans (Pohl &
Camberlin 2006). Both the ITCZ and IOC migrate in response to insolation-driven surface heating patterns (Nicholson 2000), yielding two rainy seasons. In eastern Africa, the ‘long rains’
(March to May) account for over 50% of annual totals (Prins & Loth 1988). Precipitation during the ‘short rains’ (October to December) is more variable, but correlates strongly with annual totals (Moron et al. 2007).
Precipitation δD values (δDrain) reflect the origins and histories of the related air masses that produce it. In eastern Africa, δDrain values correlate inversely with precipitation amounts due to the influences of vapor-source, transport history and ‘amount’ effects (Araguás-Araguás et al.
1998). Today, the Indian Ocean is the primary vapor source to eastern Africa (Nicholson 2000); the Atlantic Ocean and continental surface-water evaporation are important secondary sources.
Transport history and amount effects result in δDrain values that differ between the rainy seasons
(van der Ent et al. 2010). On average, δDrain values for the long rains are less negative (–20‰) than for the short rains (–28‰); in central eastern Africa, the average annual δDrain value is about
–22‰ (Nkotagu 1996, Bowen & Revenaugh 2003, IAEA 2006, Bohté et al. 2010).
3.3.3. Leaf-lipid apparent fractionation factors
Plant-water δD values reflect soil-water δD values (δDsoil)(Sachse et al. 2012). Isotopic relationships between δDsoil and δDrain values can be influenced by surface evaporation in arid and semiarid regions (Gibson et al. 2008), but plant-water δD values attenuate evaporative
42 signals by accessing deep soil moisture and increasing productivity during rainy seasons
(Ehleringer & Dawson 1992).
Soil water is not fractionated significantly during uptake by plant roots, and stem-water δD values generally reflect δDsoil values (Ehleringer & Dawson 1992). In contrast, leaf-water δD values can vary markedly from δDsoil values as a result of transpiration (Farquhar et al. 2007).
The relative importance of stem-water versus leaf-water during lipid biosynthesis remains unclear (Feakins & Sessions 2010, McInerney et al. 2011), as does biosynthetic phenology.
These factors could account for differences in apparent fractionation between source-waters and lipids (εlipid/water = (δDlipid + 1000) / (δDwater + 1000) – 1, expressed in permil (‰)) among living plants when grouped according to plant functional type (PFT).
We re-evaluate a global compilation of published leaf-lipid δD values for living plants
(Appendix B) to determine representative εlipid/water values for different PFTs relevant to this study. We target subtropical and tropical taxa in clades most representative of plants in eastern
Africa since the early Pleistocene (Chase & Reveal 2009, Bell et al. 2010). Leaves of subtropical and tropical plants commonly contain the leaf-lipid hentriacontane
(nC31)(Rommerkirschen et al. 2006, Vogts et al. 2010), and nC31 is also abundant in lake sediments from Olduvai Gorge (Magill et al. 2013). Therefore, we use published δD values for nC31 (δD31) to determine εlipid/water values applicable to sedimentary δD31 values from Olduvai
Gorge.
43
We define PFTs in terms of photosynthetic pathway and growth habit: C3 woody plants, C3 herbs, and C4 graminoids. We determine representative εlipid/water values – termed ε31/model values
model – from published δD31 values using modeled annual δDrain values (δDrain )(Bowen & Revenaugh
2003, IAEA 2006). Measured annual δDrain values rarely accompany published leaf-lipid δD
model values, but, when available, generally coincide with δDrain values (Bowen & Revenaugh 2003,
Sachse et al. 2012).
Collectively, ε31/model values average –124‰ (εaverage). Among individual PFTs, the median
ε31/model value for C4 graminoids is most negative (–146 ± 8‰, 95% confidence interval; n = 51).
The median ε31/model value for C3 herbs (–124 ± 10‰; n = 24) is more negative than for C3 woody plants (–109 ± 8‰; n = 84).
We calculate ‘landscape’ apparent fractionation factors (εlandscape) using ε31/model values and
13 13 relative PFT abundances (Figure 3.2) estimated from δ C values for nC31 (δ C31) in lake sediments (Magill et al. 2013). Leaf lipids represent relative PFT abundances in basins ranging from small lakes (Polissar & Freeman 2010) to expansive river systems (Galy & Eglinton 2011), despite differences in basin morphology, transport, and burial processes (Carroll & Bohacs
1999). We multiply each ε31/model value by relative PFT abundance (i.e., fwoody, fherb and fgram for woody plants, herbs and gramminoids, respectively) to calculate εlandscape values:
εlandscape = fwoody (–109‰) + fherb (–124‰) + fgram (–146‰)
Finally, we apply εlandscape values to sedimentary δD31 values to reconstruct δDsoil values.
44
The relative influences of biochemical, physiological, and environmental processes on εlipid/water values are difficult to account for in interpretations of δD31 values. For instance, our ε31/model value for C3 woody plants does not distinguish by canopy height, despite observed differences between modern trees and shrubs (Appendix B). Similarly, annual δDrain values do not distinguish seasonality. In eastern Africa, annual and rainy-season δDrain values can differ in excess of 20‰ (Appendix B), though >85% of annual precipitation occurs during rainy seasons
(Nicholson 2000). Thus, to the extent that plant growth takes place in rainy seasons, annual
δDrain values can overestimate δDsoil values during lipid biosynthesis.
Although propagated uncertainty in εlandscape values (Appendix B) represents a substantial portion of the variability in modern δDrain values across central eastern Africa (about 60‰), sedimentary
δD31 values capture space- and time-integrated signals that attenuate variability of individual plants or species (Polissar & Freeman 2010, Sachse et al. 2012). Thus, uncertainty in εlandscape values largely reflects ecosystem-scale differences in transpiration and phenology (Sachse et al.
2012), which are at least partially accounted for by ε31/model values, provided living plants are representative of their ancient counterparts. Despite some shortcomings in the state of the art,
εlandscape values provide a useful interpretational framework to account for biological and physical influences on leaf-lipid δD values – a factor often overlooked in hydrologic reconstructions.
3.3.4. Algal-lipid apparent fractionation factors
Aquatic photosynthetic organisms acquire hydrogen for lipid biosynthesis from ambient waters
(Sachse et al. 2012). Therefore, algal δD values reflect lake-water δD values – which integrate
45 precipitation, groundwater, runoff, and evaporation – as modified by biosynthetic fractionation.
Biosynthetic fractionation, in turn, incorporates biological and physical factors (Sachse et al.
2012). Culture studies indicate algal εlipid/water values vary between species (Zhang & Sachs
2007), but space- and time-integration appears to minimize these effects in sediments (Sachse et al. 2012).
Field studies indicate algal εlipid/water values vary in relation to salinity, and must be accounted for when interpreting algal-lipid δD values (Sachs & Schwab 2010). The response of algal εlipid/water
–1 values to salinity is markedly consistent at 0.9 ± 0.2‰ ppt (Sachse et al. 2012). Heptacosane
(nC17) is a general biomarker for algae (Han & Calvin 1969) and is abundant in lake sediments from Olduvai Gorge. Modern studies establish an apparent fractionation between freshwater and nC17 equal to –172‰ (Sachse et al. 2012), and we use this value to determine algal εlipid/water values at different salinities – termed εlake values: εlake = 0.9 (salinity) – 172‰.
We construct a basic lake-water evaporation model to constrain εlake values in the past (Appendix
B). Briefly, we estimate the total solute load for paleolake Olduvai based on stratigraphic evidence for maximum lake area (about 200 km2)(Ashley & Hay 2002) and lake level (about 5 m)(Hay & Kyser 2001) during the early Pleistocene and fossil evidence for minimum salinity
(about 20 ppt)(Hay 1976). Then, we infer changes in lake level from changes in sedimentary total organic carbon (%TOC) because these values co-vary in many modern lakes in eastern
Africa (Talbot & Livingstone 2009). Next, we assume a conservative solute balance and use estimates for paleolake volume to estimate salinities during lake contraction. Finally, we apply
εlake values to sedimentary δD17 values to reconstruct lake-water δD values (δDlake). Salinity
46 estimates for paleolake Olduvai range from about 20 to 100 ppt, resulting in εlake values that vary by up to 88‰ from freshwater algal εlipid/water values.
Biosynthetic processes responsible for the influence of salinity on δD17 values are unclear
(Sachse et al. 2012), and not all possible mechanisms result in linear relationships. Still, εlake values range between –158 and –84‰ for paleolake Olduvai, highlighting the importance of salinity when interpreting δDlake values from sedimentary δD17 values.
3.4. Results
13 Sedimentary δ C31 values range from –36.3 to –21.4‰, with an average value of –27.8‰
(Figure 3.3). Sedimentary δD31 values range from –148 to –132‰, and correlate weakly with
13 δ C31 values (r = 0.33) and %TOC (r = 0.28). Reconstructed δDsoil values show an increased isotopic range of 54‰, from –38 to +16‰ (Figure 3.4).
Sedimentary δD17 values range from –150 to –30‰ (Figure 3.4). Measured values correlate
13 strongly with δ C31 values (r = 0.91) and %TOC (r = 0.93). Reconstructed δDlake values show a
13 relatively smaller the isotopic range from +3 to +59‰, but still correlate strongly with δ C31 values (r = 0.94). Interestingly, sedimentary δD17 values correlate weakly with sedimentary δD31 values (r = 0.33), but reconstructed δDlake values correlate strongly with reconstructed δDsoil values (r = 0.92).
47
3.5. Interpretations and discussion
3.5.1. Precipitation in eastern Africa
Historical precipitation patterns serve as a framework for interpreting reconstructed hydrologic
1 3 patterns over timescales of 10 to 10 kyr (Laepple & Lohmann 2009), although regional tectonism and the intensification of zonal atmospheric (Walker) circulation during the early
Pleistocene could weaken this interpretational link (Nicholson 2000, deMenocal 2004). Modern
δDrain values reveal a regional meteoric waterline (RMWL) for eastern Africa (Figure 3.5):
18 δDrain = 7.9 δ Orain + 11.3‰
Today, annual δDrain values in eastern Africa range from about –30 and –10‰, whereas monthly
δDrain values range from about –50 and +30‰ (Onodera et al. 1995, Nkotagu 1996, Rozanski et al. 1996, Bowen & Revenaugh 2003, IAEA 2006, Bohté et al. 2010). We pair reconstructed
18 δDsoil values with published δ Orain values derived from lake and soil-carbonate minerals in closely associated lake (Hay & Kyser 2001) and lake-margin (Cerling & Hay 1986) sediments
18 (Appendix B). Reconstructed δDsoil–δ Orain values plot within the range of modern precipitation data for eastern Africa and show strong agreement with the RMWL (Figure 3.5).
A wide range in δDsoil values (and, by inference δDrain values) indicates pronounced changes in vapor source, transport history, amount of precipitation, or a combination thereof. Today, changes in vapor source account for up to 10‰ variability in δDrain values (Rozanski et al. 1996).
–1 Sea-surface temperatures in vapor-source regions influence δDrain values by about 1‰ °C (Steig
2001), accounting for up to 5‰ variability during the early Pleistocene (Cleaveland & Herbert
48
2007). Transport history is difficult to constrain (Worden et al. 2007), but can account for up to
10‰ variability today. Together, changes in vapor source and transport history can account for nearly half of the variability (25‰) in reconstructed δDsoil values for eastern Africa. If so, changes in the amount of precipitation account for the remaining variability (29 of 54‰) in reconstructed δDsoil values.
The relationship between modern δDrain values and amount of precipitation is difficult to evaluate in many tropical regions because of sparse measurements, but available data suggest comparable effects at seasonal, annual, and interannual timescales (Worden et al. 2007, Risi et al. 2008). In eastern Africa, available rainy-season δDrain values correlate inversely with precipitation at a
-1 slope of –0.125‰ mm (Appendix B). If modern sensitivity is representative for the past, and changes in vapor-source and transport history were important, then a 29‰ range in δDsoil values translate to 225 mm precipitation range; the full 54‰ range translates to 415 mm precipitation range.
Historical data for annual δDrain values (–22‰)(Onodera et al. 1995, Bohté et al. 2010) and mean annual precipitation (550 mm)(Hay 1976, Prins & Loth 1988) provide a local reference point from which to project reconstructed δDsoil values. Since rainy seasons account for about 85% of mean annual precipitation (MAP) in northern Tanzania (Prins & Loth 1988), we use an amount effect for rainy-season months to reconstruct MAP in the past:
-1 MAP = (δDsoil + 22‰) / (–0.13‰ mm ) + 550 mm
MAP estimates for the full range of δDsoil values are from ~700 to 250 mm. This range is consistent with reconstructions based on pollen spectra (~750 mm)(Bonnefille et al. 1995) and
49 soil carbonates (<400 mm)(Sikes & Ashley 2007) during wetter and drier intervals, respectively.
Woody cover strongly co-varies with MAP today in eastern Africa (Sankaran et al. 2005):
fwoody = (0.14 (MAP) – 14.2) / 100
We find a similar relationship between our estimates of MAP and woody cover (Magill et al.
2013). For instance, in the modern calibration, MAP of 700 mm yields an fwoody value of 0.84,
13 which is consistent with a δ C31-derived (–36.3‰) fwoody value of 0.90 (Magill et al. 2013).
These observations suggest amount effects influenced δDrain values more than changes in vapor- source or transport history at Olduvai Gorge and highlight the importance of using εlandscape values to reconstruct hydrologic patterns from leaf-lipid δD values.
3.5.2. Lake-water evaporation in eastern Africa
16 Loss of lighter isotopic species (H2 O) during evaporation progressively enriches residual lake
16 18 18 waters in DH O and H2 O. In eastern Africa, lake-water δD values (δDlake) and δ O values
18 (δ Olake) define an isotopic trajectory – called a local evaporation line (LEL) – with a slope that is lower than that of the RMWL (Gibson et al. 2008). LEL slopes are primarily a function of relative humidity (h); in general, very low h values (e.g., 0.25) result in slopes close to 4, whereas higher h values result in slopes closer to 6 (Gibson et al. 2008). Modern δDlake and
18 δ Olake values yield a LEL for eastern Africa:
18 δDlake = 5.6 δ Olake + 1.6‰
Modern δDlake values range from about –30‰ in humid regions of eastern Africa to +80‰ or higher in extremely arid regions (Darling et al. 1996, Odada 2001, Kebede et al. 2004, Delalande et al. 2008).
50
The LEL defines source-water composition at its intersection with the RMWL. For modern waters in eastern Africa, LEL and RMWL intersect at a source-water δD value of –22‰ (Figure
3.5), which closely matches historical data (Nkotagu 1996, Bowen & Revenaugh 2003, IAEA
2006, Bohté et al. 2010). In closed basins, lake waters derive primarily from precipitation
(Yechieli & Wood 2002).
3.5.3. Tracing isotopic hydrology at Olduvai Gorge
To compare modern and ancient lake-waters, we pair reconstructed δDlake values with published
18 δ Olake values that were determined from authigenic clays (Hay & Kyser 2001) in associated
18 sediments (Appendix B). Reconstructed δDlake–δ Olake values show close agreement with the modern LEL (Figure 3.5). Further, reconstructed δDlake and δDsoil values strongly correlate, suggesting lake-water compositions shifted largely due to changes in precipitation. Evaporation rates decrease at high salinity due to the decreased activity of water in high ionic-strength solutions. As a result, potential evaporation can exceed lake-water evaporation by up to 100-fold
(Yechieli & Wood 2002). Thus, while reconstructed δDlake values vary only slightly more than
δDsoil values, changes in source-water and amount of precipitation would have been accompanied by large changes in potential evaporation. Reconstructed lake evaporation relative to meteoric input (E/I) based on our data suggest higher evaporation during intervals of reduced precipitation
(E/I = 2.9) than during increased precipitation (E/I = 1.3) and are consistent with historical and modeled E/I values for eastern Africa (Appendix B).
51
3.5.4. Water availability and ecosystem dynamics
Reconstructed δDrain and δDlake values reveal strong relationships between water and carbon- isotopic data for ecosystem type. Lower δDrain and δDlake values, which reflect increased MAP and decreased evaporation, respectively, correspond with increased woody cover (fwoody = 0.90).
Although the organic-carbon-derived indicators we use to determine εlandscape and εlake values may be codependent (Meyers & Lallier-Vergés 1999), reconstructed values for fwoody and lake level are consistent with independent indicators for ecosystem type and paleolake level (Figure 3.3).
Much like today (Sankaran et al. 2005), aridity was a dominant control on ecosystem change in eastern Africa during the Pleistocene.
Contrasting proxy records have fueled debate about the pace and patterns of environmental change in eastern Africa during the Pleistocene. Pollen and fossil abundance records suggest expansion of arid-adapted species beginning near 2.0 Ma and culminating around 1.8 Ma (Bobe
2006, Bonnefille 2010). Marine dust-flux records and soil-carbonate δ13C values also suggest shifts towards more arid conditions around 1.8 Ma (deMenocal 2004), though geomorphic evidence suggests regionally wetter conditions (c.f., deMenocal 2004).
Lipid biomarkers from Olduvai Gorge point to rapid changes in plants and water between about
2.0 and 1.8 Ma, and we suggest this environmental variability both influenced and can reconcile proxy records. For instance, increased seasonality can lead to C4 graminoid expansions (Gritti et
13 al. 2010), but can also lead to unrepresentatively positive C4-like δ C values in soil carbonates
(Breecker et al. 2009). Similarly, rapid wet-to-dry transitions can simultaneously produce both
52 increased dust and elevated lake-levels (Trauth et al. 2010). Over the past several million years, modulation of marine dust-flux records from the Arabian Sea has been tightly coupled with orbital eccentricity, resulting in distinct intervals of exceptionally high-amplitude variability during orbital-eccentricity maxima (deMenocal 2004). We hypothesize that high-amplitude, orbital precession-paced environmental variability, as opposed to gradual or stepwise aridification, characterized eastern Africa during the early Pleistocene.
3.5.5. Water and early human evolution
In semiarid regions, precipitation primarily determines water availability (Prins & Loth 1988).
Today, water availability shapes primate behaviors through its influence on vegetation and resource distributions (Teaford & Ungar 2000, Preutz & Bertolani 2009). For example, regions with MAP <700 mm do not support chimpanzee populations (Copeland 2009). Water likely shaped behavioral adaptations in the genus Homo (Rose & Marshall 1996). Our evidence for dramatic variability is consistent with water as a strong selective pressure in human evolution
(deMenocal 2004, Potts 2007). However, thermoregulatory and dietary constraints function at microhabitat scales (Hill 2006) and many hominin fossil sites – including Olduvai Gorge (Hay
1976) – are associated with ephemeral or saline water sources (Ashley et al. 2010). Our reconstructions of precipitation and lake-chemistry indicate that, even during maximum lake expansion, lake waters at Olduvai Gorge were likely not potable (Hay & Kyser 2001).
Groundwater-fed freshwater springs could have aided hominin existence and proliferation
(Ashley et al. 2010).
53
3.6. Conclusions
This study presents a continuous record of δD values for lipid biomarkers from lake sediments at
Olduvai Gorge that were deposited during a key juncture in human evolution, ~2.0 to 1.8 Ma.
We pair sedimentary leaf-lipid δD values with corresponding δ13C values to account for physiological and environmental influences on reconstructed precipitation δD values. We use a basic lake-water evaporation model to account for the influence of salinity on algal-lipid δD values and reconstructed lake-water δD values. Sedimentary leaf and algal-lipid δD values show a weak relationship, but ‘corrected’ values correlate strongly. We compare reconstructed precipitation and lake-water δD values with isotopic data for environmental waters in modern eastern African to estimate ancient precipitation amounts and evaporative losses, respectively.
Our results indicate Olduvai Gorge received about 250 mm of mean annual precipitation during arid intervals and ~700 mm during wetter intervals. Given the magnitude and variability in water availability revealed by our reconstructions, we hypothesize freshwater springs were important for hominin subsistence in highly variable environments.
54
3.7. Figures
,? !)#& ,? !)#& +,, >,?3 ( 1 ! !"#$%&%'('%)* "##!#$%&'!!( @ABC @ABC ,? *7, )* @2B !
!"#$%&#$!"' ()*$+,-$./) >,?1 ,
0 *5, $ 2 /$9:;%!<$;=:
*,, )*
6, '#""%( 3 !!"##!#$%&'!!(
5, 7!8# -.#"(/#0!"#$%&%'('%)* -./0/--01234 !"#$"%& !"#'"%& +)*',
Figure 3.1: Modern precipitation patterns in eastern Africa (Mitchell & Jones 2005) during the
‘long rains’ (a; March to May) and the ‘short rains’ (b; October to December) with respect to average monthly precipitation in northern Tanzania (c; data from http://climexp.knmi.nl).
Panels a and b, bold horizontal lines mark the position of the Intertropical Convergence Zone, whereas dashed lines mark the Interoceanic Confluence (Nicholson 2000). Target symbols mark the location of Olduvai Gorge (2° 48′S, 35° 06′E). Panel c, the bold line reflects observed average monthly rainfall (1964–1984); gray envelops variation for average monthly precipitation. The dashed line reflects modeled average monthly rainfall for Tanzania (Kalnay et al. 1996). Panel d shows depositional environments surrounding paleolake Olduvai during the early Pleistocene (Hay 1976). Contracted (dashed line) and expanded (bold line) lake margins are based on correlated stratigraphic sections (Hay 1976, Ashley & Hay 2002, Ashley 2007).
55
!"" !"#$%&'"()*+
(78$6&/3&04 (="8-4 (93305()#$/%4 ,-./0$/1" : ; : < : < $ " ()*+,*-./ 9330"0 78$44#$/0 9330#$/0 *38"4% 78$44#$/0 213454%"6 (012345/ - !!"" ! ?8*+?@* B< >"$?@>&A&0(! :/ Figure 3.2: Schematic depiction of ε31/model values as a function of reconstructed ecosystem. 13 Panel a, we use δ C31 values to estimate relative abundances for three different plant functional types (PFTs): C4 graminoids, C3 herbs, and C3 woody plants. Panel b, we relate relative PFT abundances to ecosystem (Magill et al. 2013) based on UNESCO terminology (White 1983). Panel c, we calculate ‘landscape’ apparent fractionation factors (εlandscape) for deuterium by mass balance. 56 !"#$%!&'&( !"#$%!&'&( 456#/-789#1&3 ;/9#/%!&'&( )* !0*) +#8:51 ! +*) 0=26->/=? !0)< 9< 9? !G%124/#3%&5%7/- !"#$%&'('#)*+,- !2#5K #LM& - !"#$%&'('#).DEF- !"#$%&'('#).DEF- 9:; 9<= 9<> > ? < 9?;> 9?<> 9=> ? >@A > 9?;> 9?<> 9=> 9B> # ?@=A> @&//&512-5$-C"#82-;95 ?@C>> # : 3 ( "$ ?>> A> > 9?;> 9?<> 9=> H124%& IJG%& 9?;> 9?<> 9=> 9B> ! @5&26-A#:&6#6 !/#1(23#'" !#,"-!"."/ /#," B5.&(2 !"#$'#).DEF- !./&0/12&03415#678%&%75%-# !"#$'#).DEF- !3%&5%7/#!"#/J/0N-# Figure 3.3: Relationships between leaf- and algal-lipid δD values and ancillary proxy data for 13 Olduvai Gorge. Panel a plots sedimentary δ C31 values (Magill et al. 2013). Panel b shows dust fluxes into the Arabian Sea (gray line) and closed- or moist-habitat bovids by percent of total bovids (black circles)(Kappelman 1984, deMenocal 2004). Panel c plots sedimentary δD31 values (black circles) alongside εecosystem values (gray line). Panel d shows stratigraphic evidence for lake levels (Hay 1976, Ashley & Hay 2002, Ashley 2007). Horizons that contain faunal evidence for low (down arrows) or high (up arrows) lake levels are also shown (Hay & Kyser 2001). Panel e plots sedimentary total organic carbon. Panel f shows sedimentary δD17 values (black circles) alongside εlake values (gray line). 57 (! ,+71 % 0 ! ! !0 &(! ,+71 )*+,-./.*012345 3 &'!! $%&'$()*+&,-./*-)'$ 9" ! !0 ! &'(! 9" (! 1%2) 8 0 ! ! !01%2) &(! )*+,-./.*012345 3 &'!! $%&'$()*+&,-./*-)'$ "# ! !0 ! &'(! "# ! !"# !"$ !"% !"? '"! 4+-%1$5.6%&7*$!%.8+& 67,89:;*<,-=,>;5 Figure 3.4: Leaf and algal-lipid δD values and reconstructed source waters with respect to sedimentary total organic carbon (%TOC). Panel a plots measured δD31 (hollow circles) and reconstructed δDsoil values (black circles) with respect to %TOC. For reference, a gray line marks the linear regression for values based on the average ε31/model value of –124‰. Panel b plots measured δD17 (hollow circles) and reconstructed δDlake values (black circles) with respect to %TOC. A gray line marks the linear regression for values based on a freshwater ε17/water value of –172‰. 58 !"#$ $%$ %" 1 ! + &" !,5*&6-)75-,80!/.9, " + !,5*&6-)75-,80!).(& =>-4,/5:5-?'@ABC #&" '()*+,-./,0(.(121(3)-,4,)15 %&'()*&+,&-./0#.-,)0 64,/2*,-73)18+9-./,0(.(121(3) ':/;20,-<21,/ #%" #!" #$ " $!" !$ %&'()*&+,&-./0#.-,)0!34 2 =>-4,/5:5-?'@ABC Figure 3.5: Modern isotopic relationships for precipitation and lake waters in eastern Africa. We plot a regional meteoric waterline (RMWL, bold line) based on modern monthly δDrain and 18 δ Orain values and single precipitation events (Onodera et al. 1995, Nkotagu et al. 1996, Rozanski et al. 1996, Bowen & Revenaugh 2003, IAEA 2006): 18 δDrain = 7.9 δ Orain + 11.3‰ For reference, we also plot the global meteoric waterline (GMWL, dashed line). A large black 18 circle marks the point defined by a single reconstructed δDsoil and published δ Orain value reconstructed from minerals in closely associated lake-margin sediments (Appendix B). Propagated uncertainties are shown (Appendix B). For reference, a cross marks the δD value based on the average ε31/model value of –124‰. 59 18 Modern δDlake and δ Olake values yield a local evaporation line (LEL) for eastern Africa (Darling et al. 1996, Odada 2001, Kebede et al. 2004, Delalande et al. 2008a, Kebede & Travi 2012, Delalande et al. 2008b): 18 δDlake = 5.6 δ Olake + 1.6‰ A large black square marks the point defined by a single reconstructed δDlake and published 18 δ Olake value reconstructed from authigenic clays in associated lake sediments (Appendix B) while the overlying cross marks the δD value based on a freshwater ε17/water value of –172‰. 60 3.8. References Araguás-Araguás L, Froehlich K & Rozanski K (1998) Stable isotope composition of precipitation over southeast Asia. Journal of Geophysical Research 103:721. Ashley G (2007) Orbital rhythms, monsoons, and playa lake response, Olduvai basin, equatorial East Africa (ca. 1.85-1.74 Ma). Geology 35:1091. Ashley G & Hay R (2002) Sedimentation patterns in a Plio-Pleistocene volcaniclastic rift- platform basin, Olduvai Gorge, Tanzania. SEPM Special Publication 73:107. Ashley G et al. (2010) Paleoenvironmental and paleoecological reconstruction of a freshwater oasis in savannah grassland at FLK North, Olduvai Gorge, Tanzania. Quaternary Research 3:333. Bell C, Soltis D & Soltis P (2010) The age and diversification of the angiosperms re-revisited. American Journal of Botany 97:1296. Bobe R (2006) The evolution of arid ecosystems in eastern Africa. Journal of Arid Environments 66:564. Bohté R, et al. (2010) Hydrograph separation and scale dependency of natural tracers in a semi- arid catchment. Hydrology and Earth System Sciences Discussions 7:1343. Bonnefille R (2010) Cenozoic vegetation, climate changes and hominid evolution in tropical Africa. Global and Planetary Change 72:390. Bonnefille R, et al. (1995) Glacial-interglacial record from intertropical Africa, high resolution pollen and carbon data at Rusaka, Burundi. Quaternary Science Reviews 14:917. Bowen G & Revenaugh J (2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39:1299. Breecker D, Sharp Z & McFadden L (2009) Seasonal bias in the formation and stable isotopic composition of pedogenic carbonate in modern soils from central New Mexico, USA. Geological Society of America Bulletin 121:630. Carroll A & Bohacs K (1999) Stratigraphic classification of ancient lakes: Balancing tectonic and climatic controls. Geology 27:99. Cerling T & Hay R (1986) An isotopic study of paleosol carbonates from Olduvai Gorge. Quaternary Research 25:63. 61 Cerling T, et al. (2011) Woody cover and hominin environments in the past 6 million years. Nature 476:51. Chase M & Reveal J (2009) A phylogenetic classification of the land plants to accompany APG III. Botanical Journal of the Linnean Society 161:122. Cleaveland L & Herbert T (2007) Coherent obliquity band and heterogeneous precession band responses in early Pleistocene tropical sea surface temperatures. Paleoceanography 22:PA2216. Copeland S (2009) Potential hominin plant foods in northern Tanzania: semi-arid savannas versus savanna chimpanzee sites. Journal of Human Evolution 57:365. Darling W, Gizaw B & Arusei M (1996) Lake-groundwater relationships and fluid-rock interaction in the East African Rift Valley: isotopic evidence. Journal of African Earth Sciences 22:423. Delalande M, Bergonzini L & Massault M (2008a) Mbaka lakes isotopic (18O and 2H) and water balances: discussion on the used atmospheric moisture compositions. Isotopes in Environmental and Health Studies 44:71. Delalande M, et al. (2008b) Hydroclimatic and geothermal controls on the salinity of Mbaka Lakes (SW Tanzania): Limnological and paleolimnological implications. Journal of Hydrology 359:274. deMenocal P (2004) African climate change and faunal evolution during the Pliocene- Pleistocene. Earth and Planetary Science Letters 220:3. Ehrlinger J & Dawson T (1992) Water uptake by plants: perspectives from stable isotope composition. Plant, Cell and Environment 15:1073. Farquhar G, Cernusak L & Barnes B (2007) Heavy water fractionation during transpiration. Plant Physiology 143:11. Feakins S & Sessions A (2010) Controls on the D/H ratios of plant leaf waxes in an arid ecosystem. Geochimica et Cosmochimica Acta 74:2128. Galy V & Eglinton T (2011) Protracted storage of biospheric carbon in the Ganges-Brahmaputra basin. Nature Geoscience 4:843. 2 18 Gibson J, Birks S & Edwards T (2008) Global prediction of δA and δ H-δ O evaporation slopes for lakes and soil water accounting for seasonality. Global Biogeochemical Cycles 22:GB2031. Gritti E, et al. (2010) Simulated effects of a seasonal precipitation change on the vegetation in tropical Africa. Climate of the Past 6:169. 62 Han J & Calvin M (1969) Hydrocarbon distribution of algae and bacteria, and microbiological activity in sediments. Proceedings of the National Academy of Sciences 64:436. Hay R (1976) Geology of the Olduvai Gorge (University of California Press). Hay R & Kyser T (2001) Chemical sedimentology and paleoenvironmental history of Lake Olduvai, a Pliocene lake in northern Tanzania. Geological Society of America Bulletin 113:1505. Hill R (2006) Thermal constraints on activity scheduling and habitat choice in baboons. American Journal of Physical Anthropology 129:242. IAEA (2006) Isotope Hydrology Information System. Accessible at: http://www.iaea.org/water. Kalnay E, et al. (1996) The NCEP/NCAR 40-year reanalysis project. Bulletin of the American Meteorological Society 77:437. Kappelman J (1984) Plio-Pleistocene environments of Bed I and lower Bed II, Olduvai Gorge, Tanzania. Palaeogeography, Palaeoclimatology, Palaeoecology 48:171. Kebede S & Travi Y (2012) Origin of the δ18O and δ2H composition of meteoric waters in Ethiopia. Quaternary International 257:4. Kebede S, et al. (2004) Lake-groundwater relationships, oxygen isotope balance and climate sensitivity of the Bishoftu crater lakes, Ethiopia. The East African Great Lakes: Limnology, Palaeolimnology and Biodiversity (CRC Press). Laepple T & Lohmann G (2009) Seasonal cycle as template for climate variability on astronomical timescales. Paleoceanography 24:PA4201. Magill C, Ashley G & Freeman K (2013) Ecosystem variability and early human habitats in Africa. Proceedings of the National Academy of Sciences 110:1167. McGill R, Tukey J & Larsen W (1978) Variations of box plots. American Statistician 32:12. McInerney F, Helliker B & Freeman K (2011) Hydrogen isotope ratios of leaf wax n-alkanes in grasses are insensitive to transpiration. Geochimica et Cosmochimica Acta 75:541. Meyers P & Lallier-Vergés E (1999) Lacustrine sedimentary organic matter records of Late Quaternary paleoclimates. Journal of Paleolimnology 21:345. Mitchell T & Jones P (2005) An improved method of constructing a database of monthly climate observations and associated high-resolution grids. International Journal of Climatology 25:693. 63 Moron V, et al. (2007) Spatial coherence of tropical rainfall at the regional scale. Journal of Climate 20:5244. Nicholson S (2000) The nature of rainfall variability over Africa on time scales of decades to millenia. Global and Planetary Change 26:137. Nkotagu H (1996) Application of environmental isotopes to groundwater recharge studies in a semi-arid fractured crystalline basement area of Dodoma, Tanzania. Journal of African Earth Sciences 22:443. Odada E (2001) Stable isotopic composition of East African lake waters. Use of Isotope Techniques in Lake Dynamics Investigations (IAEA). Onodera S, Kitaoka K & Shindo S (1995) Stable isotopic compositions of deep groundwater caused by partial infiltration into the restricted recharge area of a semiarid basin in Tanzania. IAHS Publications 227:75. Pohl B & Camberlin P (2006) Influence of the Madden-Julian Oscillation on East African rainfall: intraseasonal variability and regional dependency. Quarterly Journal of the Royal Meteorological Society 132:2521. Polissar P & Freeman K (2010) Effects of aridity and vegetation on plant-wax δD in modern lake sediments. Geochimica et Cosmochimica Acta 74:5785. Potts R (2007) Environmental hypotheses of Pliocene human evolution. Hominin Environments in the East African Pliocene: An Assessment of the Faunal Evidence, Vertebrate Paleobiology and Paleoanthropology, eds Bobe R, Alemseged Z & Behrensmeyer A (Springer Netherlands). Preutz J & Bertolani P (2009) Chimpanzee (Pan trglodytes verus) behavioral responses to stresses associated with living in a savanna-mosaic environment, implications for hominin adaptations to open habitats. PaleoAnthropology 2009:252. Prins H & Loth P (1988) Rainfall patterns as background to plant phenology in northern Tanzania. Journal of Biogeography 15:451. Risi C, Bony S & Vimeux F (2008) Influence of convective processes on the isotopic composition (δ18O and δD) of precipitation and water vapor in the tropics: physical interpretation of the amount effect. Journal of Geophysical Research 113:D19306. Rommerskirchen F, et al. (2006) Chemotaxonomic significance of distribution and stable carbon isotopic composition of long-chain alkanes and alkan-1-ols in C4 grass waxes. Organic Geochemistry 37:1303. Rose L & Marshall F (1996) Meat eating, hominid sociality, and home bases revisited. Current Anthropology 37:307. 64 Rozanski K, et al. (1996) Isotope patterns of precipitation in the East African region. The Limnology, Climatology and Paleoclimatology of the East African Lakes, eds Johnson T & Odada E (CRC Press). Sachs J & Schwab V (2010) Hydrogen isotopes in dinosterol from the Chesapeake Bay estuary. Geochimica et Cosmochimica Acta 75:444. Sachse D, et al. (2012) Molecular paleohydrology, interpreting the hydrogen-isotopic composition of lipid biomarkers from photosynthetic organisms. Annual Review of Earth and Planetary Sciences 40:221. Sankaran M, et al. (2005) Determinants of woody cover in African savannas. Nature 438:846. Sikes N & Ashley G (2007) Stable isotopes of pedogenic carbonates as indicators of paleoecology in the Plio-Pleistocene (upper Bed I), western margin of the Olduvai Basin, Tanzania. Journal of Human Evolution 53:574. Steig E (2001) No two latitudes alike. Science 293:2015. Talbot M & Livingstone D (1989) Hydrogen index and carbon isotopes of lacustrine organic matter and lake level indicators. Palaeogeography, Palaeoclimatology, Palaeoecology 70:121. Teaford M & Ungar P (2000) Diet and the evolution of the earliest human ancestors. Proceedings of the National Academy of Sciences 97:13506. Trauth M, et al. (2010) Human evolution in a variable environment: the amplifier lakes of Eastern Africa. Quaternary Science Reviews 29:2981. van der Ent R, et al. (2010) Origin and fate of atmospheric moisture over continents. Water Resources Research 46:W09525. Vogts A, et al. (2009) Distribution patterns and stable carbon isotopic composition of alkanes and alkan-1-ols from plant waxes of African rain forest and savanna C3 species. Organic Geochemistry 40:1037. White F (1983) The Vegetation of Africa (United Nations Scientific and Cultural Organization). Worden J, et al. (2007) Importance of rain evaporation and continental convection in the tropical water cycle. Nature 445:528. Yechieli Y & Wood W (2002) Hydrogeologic processes in saline systems: playas, sabkhas, and saline lakes. Earth-Science Reviews 58:343. 65 Zhang Z & Sachs J (2007) Hydrogen isotope fractionation in freshwater algae: variations among lipids and species. Organic Geochemistry 38:582. 66 Chapter 4. Plant biomarker patterns at Olduvai Gorge establish social meat consumption by pre-modern humans 4.1. Abstract Over a half century ago, Mary Leakey began excavating the iconic FLK Zinjanthropus archaeological Level 22 (FLK Zinj) at Olduvai Gorge with the goal of identifying evolutionary changes in ‘early man’. Since then, rich co-occurrences of stone artifacts, cut-marked bones and hominin fossils across the site have fueled interest and controversy about the origins of behavioral sophistication and the importance of meat to early members of the genus Homo. Spatial heterogeneity in archaeological materials suggests pre-modern humans (hominins) concentrated their activities within the landscape, but specific behavioral implications of discriminate land-use remain obscure because of sparse or indirect metrics for microhabitat (< 0.1 km2) structure and function. Here, we use plant biomarkers preserved in soils from FLK Zinj and the surrounding landscape, which were instantaneously sealed in place under volcanic ash, to directly reconstruct microhabitat patterning. Molecular and carbon-isotopic signatures resolve sharp ecological transitions across the site, delimiting an isolated spring-fed wetland and forest patch within expansive grassland. Taken together, these data provide a unique glimpse into the daily lives of our ancestors, in which hominins evidently used FLK Zinj as a ‘central-foraging place’ for cooperative butchery and social meat consumption because the site afforded freshwater near arboreal refuge in an otherwise arid, open landscape. 67 4.2. Introduction Excavations of FLK Zinj have uncovered a dense assemblage of artifacts and fossils (Leakey 1971), which contrasts with sparse archaeological scatters across the surrounding landscape (Domínguez-Rodrigo et al. 2010). This site has been variously interpreted as a ‘carnivore (nonhuman predator) kill’ site (Binford 1981), a hominin ‘stone cache’ for resource processing (Potts 1988), ‘refuge’ (Blumenschine et al. 1994) or ‘central-foraging place’ (Isaac 1983, Rose & Marshall 1996), among others with less empirical support. In turn, land-use models implicate hominin behavioral ecology, which is directly contingent upon spatial patterns of microhabitat structure and function (Sikes 1994, Whiten & Erdal 2012). ‘Carnivore kill’ and ‘stone caching’ models imply hominins passively scavenged from carcasses to avoid direct competition with other predators in homogenous or open habitats. Although ‘refuge’ and ‘central-foraging place’ models both imply selective use of referential locations by hominins within a heterogeneous landscape, these two models invoke very different cognition and behaviors (Domínguez-Rodrigo et al. 2007). In ‘refuge’ models, hominins forage like nonhuman primates, in which referential locations are used temporarily for reclusive consumption of scavenged remains near predator- competitive habitats. In ‘central-foraging place’ models, hominins possess unique behavioral sophistication, in which resources were transported from the surrounding landscape to a central, referential location for cooperative processing and social consumption (Isaac 1983). Previous studies suggest FLK Zinj was part of a generic lake-margin ‘savanna’ ecosystem (Hay 1976, Sikes 1994, Ashley et al. 2010, Domínguez-Rodrigo et al. 2010), but do not provide the spatial constraints on microhabitat patterns needed to explicitly evaluate hominin land-use models (Domínguez-Rodrigo et al. 2007, Hopley & Maslin 2010). 68 To resolve microhabitat patterns, we measured stable carbon-isotope compositions (expressed as δ13C values) of diverse plant biomarkers preserved in an ancient soil (paleosol) horizon that spans FLK Zinj and 15 surrounding trenches of stratigraphic equivalency across the ca. 2000 m2 ‘junction area’ of Olduvai Gorge in northern Tanzania (Figures 4.1 and 4.2). Results include analyses of 53 samples from recent excavations of a distinctive ca. 15-cm thick horizon associated with in situ accumulation of artifacts and fossils. The horizon is composed of waxy, green to olive-brown clays initially deposited on a lake-margin floodplain. Blocky ped structures, weak stratification and irregular iron-staining suggest the horizon was modified by soil forming processes during lake contractions (Ashley et al. 2010, Domínguez-Rodrigo et al. 2010). Artifacts and fossils occur primarily on the upper surface of the horizon and occasionally protrude into overlying airfall tuff (Tuff IC)(Leakey 1971), dated to 1.848 ± 0.003 million years ago (Deino 2012), indicating instantaneous burial of exposed archaeological materials under volcanic ash. 4.3. Background 4.3.1. Soil organic matter Buried in place, paleosols incorporate organic matter from standing biomass over spatial scales of less than ca. 10 m2 (Kögel-Knabner 2002). Soil organic matter (SOM) derives primarily from vascular plant litter, but δ13C values for SOM can differ by over 2‰ versus standing biomass (Santruckova et al. 2000, Magill et al. 2013a) due to microbial degradation of plant litter. In 69 order to circumvent challenges related to SOM provenance and preservation, we turn to plant biomarkers. 4.3.2. Plant biomarker signatures Plant biomarkers reveal structural and functional characteristics of microhabitats. Long-chained n-alkanes (e.g., hentriacontane, nC31) are leaf-waxes from terrestrial plants (Diefendorf et al. 2011). Medium-chained n-alkanes (e.g., tricosane, nC23) are primarily from non-emergent aquatic plants (Ficken et al. 2000). Thus, abundance ratios of long-chained to medium-chained n-alkanes (expressed as Paq values) track relative organic matter inputs from aquatic versus terrestrial plants and equals ((nC23 + nC25) / (nC23 + nC25 + nC29 + nC31)). In modern plants, Paq is less than 0.1 for terrestrial plants, less than 0.4 for emergent aquatic plants, and higher than 0.4 for floating aquatic plants (Ficken et al. 2000). Resorcinolic lipids occur in many photosynthetic organisms (Kozubek & Tyman 1999), but 5-n-alkylresorcinols with medium-chained n-alkyl moieties are dominant in sedges (Appendix C). Vanillyl (V), syringyl (S) and cinnamyl (C) phenols are degradation products of lignin, an aromatic biopolymer that surrounds plant cell walls. The phenols V, S and C are characteristic of vascular, angiosperm and herbaceous tissue inputs, respectively (Hedges & Mann 1979). Consequently, abundance ratios of cinnamyl to vanillyl phenols (C/V) track relative organic matter inputs from herbaceous plants. Plant biomarker δ13C values are a function of photosynthetic pathway, community composition and water availability (Hobbie & Werner 2004, Diefendorf et al. 2011). In eastern Africa, most woody and herbaceous plants use C3 photosynthesis, whereas arid-adapted grasses use C4 70 photosynthesis (Magill et al. 2013a). Sedges include both C3 and C4 lineages, but species common in alkaline wetlands often use C3 photosynthesis (Ficken et al. 2000). In tropical regions, plant biomarker δ13C values track ecosystem structure (Magill et al. 2013). 13 In forests with woody cover in excess of ~80%, most plants use C3 photosynthesis and δ C values for nC31 (δ31) are low (ca. –36.0‰). Conversely, C4 plants dominate open grasslands and δ31 values are relatively higher (ca. –20.0‰). Herbaceous C3 plants occur in both closed and open ecosystems, and the fraction of woody cover (fwoody) can be estimated from a non-linear function of δ31 values (between –40.0 ≤ x ≤ –20.0‰)(Magill et al. 2013a): 2 fwoody = (sin (–1.8353 – 0.08538 δ31)) 13 Syringyl phenol δ C values (δS) provide additional evidence for woody cover, and cinnamyl 13 phenol abundances and δ C values (δC) track herbaceous inputs (Huang et al. 1999). Lignin 13 degradation phenols from modern C3 and C4 plants have δ C values of ca. –34.0 and –14.0‰, respectively (Huang et al. 1999). 4.4. Results Our molecular and isotopic data for plant biomarkers preserved in paleosols from FLK Zinj and the surrounding landscape resolve microhabitat patterns (Figures 4.2 and 4.3). Northern trenches 13 have intermediate but variable δ31 (–22.6 ≤ x ≤ –29.0‰) and Paq (0.2 ≤ x ≤ 0.6) values and C- depleted 5-n-alkylresorcinols (ca. –32.0‰)(Appendix C), indicating the area harbored emergent and floating aquatic plants, including C3 sedges (e.g., Cyperus pectinatus). In contrast, central trenches show low values for δ31 (–36.1‰), δS (–33.4‰) and C/V, indicating the area was a 71 forest patch less than 100 m in radius (Lejuene et al. 2002), likely composed of Acacia trees (Leakey 1971, Domínguez-Rodrigo et al. 2010) and palms (Ashley et al. 2010). Negligible cinnamyl phenol abundances indicate understory plants were sparse or absent. Several hundred meters to the south of FLK Zinj, high values for δ31 (–19.4‰), δS (–15.2‰) and δC (–14.9‰) indicate a sharp transition to open grassland. Thus, the dense archeological assemblage of stone artifacts, cut-marked bones and hominin fossils at FLK Zinj accumulated within the confines of an isolated forest patch just 200 m from a spring-fed wetland. 4.5. Discussion 4.5.1. Microhabitat distributions Freshwater is critical to the survival of animals in tropical environments – including humans – and its availability likely shaped the behavioral ecology of hominins during the Pleistocene. Today, in eastern Africa, landscape area versus freshwater shows a strong, inverse relationship with annual precipitation (Chamaillé-Jammes et al. 2007). Regions receiving less than 250 mm yr-1 provide spatially diffuse and ephemeral standing water, with less than 4% of the landscape within 2 km of freshwater (Chamaillé-Jammes et al. 2007). Stable hydrogen- and carbon- isotopic abundances for plant biomarkers in lake sediments indicate Olduvai Gorge received ca. –1 250 mm yr precipitation about 1.848 million years ago (Magill et al. 2013b), and C4 grasses dominated the open landscape (Magill et al. 2013a). Under arid conditions, extant primates selectively use sites near freshwater and, at finer spatial scales, prefer forested microhabitats 72 (Pruetz & Bertolani 2009). By analogy, we suggest recurrent hominin activities at FLK Zinj reflected the unique proximity of perennial freshwater and forest. 4.5.2. Hominin dietary resources Like freshwater, diet tangibly links an organism to the environment (Figure 4.2). Thus, our microhabitat reconstructions constrain dietary resource availabilities. Recent isotopic and microscopic (i.e., dental wear) data suggest Paranthropus boisei from eastern Africa consumed a diet of abrasive C4 plants (Cerling et al. 2011). In eastern Africa, C4 plants are predominantly grasses, although C4 sedges can be abundant near wetlands (Ficken et al. 2000). A growing number of hypotheses identify C4 sedges as a key dietary resource for P. boisei (see Cerling et al. 2011). However, low δ13C values for 5-n-alkylresorcinols indicate sedges nearby FLK Zinj used C3 photosynthesis (Appendix C). This finding indicates P. boisei was not a wetland sedge specialist. Low occlusal relief of molars in this species is counter to expectations for that of a folivore (Cerling et al. 2011), suggesting grass blades were also not a major dietary resource. Thus, P. boisei likely augmented a diet of seasonally available grass seeds, meristems and underground storage organs (USOs) with significant quantities of 13C-enriched animal resources, such as insects, shellfish and meat. 4.5.3. Hominin behavioral implications Hominin diets were a function of foraging behavior and land-use strategies (Rose & Marshall 1996, Cerling et al. 2011). Behaviors adopted by early Homo to secure meat were potentially 73 key to the evolutionary success of our direct ancestors (Binford 1981, Potts 1988, Blumenschine et al. 1994, Domínguez-Rodrigo et al. 2010). Skeletal and mortality profiles for FLK Zinj are statistically distinct from those associated with extant predators (e.g., Panthera pardus, leopard; Crocuta crocuta, spotted hyena)(Bunn & Pickering 2010, Domínguez-Rodrigo et al. 2010), indicating hominins did not passively scavenge from carcasses. Rather, cut-marks on mid-shaft long bones (Domínguez-Rodrigo et al. 2010) from prime adult medium-sized animals (Bunn & Pickering 2010) indicate hominins acquired meat via selective ambush hunting. Plant biomarker signatures establish that portions of at least 48 large-animal carcasses (>100 kg)(Domínguez- Rodrigo et al. 2010) were transported to an isolated forest patch, indicating multiple hominins used FLK Zinj for butchery. Taken together, these data reveal FLK Zinj was a ‘central-foraging place’ used recurrently by hominins as a focal point for cooperative gathering-hunting (Domínguez-Rodrigo et al. 2010) and correlative socialization (Bunn & Pickering 2010, Whiten & Erdal 2012), which are at the heart of what made us human. 74 4.6. Figures +)** ' 9'8*:0'6;45$#)++7<)'45 &+536'/3*7$)'8* !,%-%&./01#2%/.%0/3%2%3' 1)*2'34+5 =9>$!"#$ ()** ? !"##$%& (')*+,+) !"#' (')*+,+) !"#$%&' -.$/0 Figure 4.1: Depositional environments at Olduvai Gorge (3° 00′S, 35° 16′E) during the early Pleistocene. Panel a: Incipient soils at FLK Zinj formed during lake contractions from clays deposited during episodic lake-margin flooding (Hay 1976, Ashley et al. 2010). Lake levels are reconstructed based on detailed stratigraphic correlations (Hay 1976). Panel b: Representative stratigraphic sequence for northern FLK Zinj trenches include airfall tuff (Tuff IC), freshwater spring carbonate (tufa) and paleosol horizons. All samples for this study are from the upper paleosol horizon. 75 !"# . 902,C42+ . . ()*+,-./01*2342- . . !"& . . !"#$%&'('#)*+,- . D4+,2*) %#$")'* !"' "#$%"!&'( ! !$" $5 E0:,C42+ #1 "6# !"#$%&'( !$% % 5 7 9:1;42-0<-=*1>)4= ?02@4-2/1 A:<*-B4>0=/, !8 13 Figure 4.2: Plant biomarker δ C values for leaf-waxes (δ31) alongside 5-n-alkylresorcinol distributions (aR) overlying a pseudocolor image of the present-day ‘junction area’ at Olduvai Gorge (source: http://wms.jpl.nasa.gov/). Collectively, δ31 values span from –19.4 to –36.1‰. Central trenches average low δ31 values (–33.9‰); FLK Zinj in particular has a δ31 value of – 36.1‰. Southern trenches average high δ31 values (–20.1‰). Northern trenches average intermediate δ31 values (–25.8‰). Mound-like tufa deposits (Ashley et al. 2010) and 5-n- alkylresorcinols are limited to northern trenches. 76 !"# A94;0+9+!B99=.0C/,+7D &" 39?7<;4+ ! 0 !$% @947<;4+ A94;0) !$# 3.4*+5./ )*++,-./ > 0C/,+7D !"#$%&'('#)*+,- !&% );+74,/ 6/,+708*9-,4:;4 , E !&# %%'$ %'> %'( !&#!&% !$# !$% !"# !"% )*++,-./0102,+*//./ 6<;+9/0!30,+=0!) !"#$%&'('#)*+,- 13 Figure 4.3: Comparison of plant biomarker δ C values for leaf-waxes (nC31, δ31) and lignin phenols (syringyl, δS; cinnamyl, δC) in relation to excavation trench (northern, gray symbols; central, black symbols; southern, white symbols). Panel a: Cross-plot between δ31 and C/V values. Central trenches show low values for δ31 and C/V. In contrast, southern trenches show high δ31 and C/V values. Panel b: Phenol δS values (triangles) show a nearly linear relationship 13 with δ31 values. Comparatively C-enriched cinnamyl phenols (squares) show a more parabolic relationship with δ31 values. Central trenches have little to no cinnamyl phenol, indicating understory plants were sparse or absent. 77 4.7. References Ashley G, et al. (2010) A spring and wooded habitat at FLK Zinj and their relevance to origins of human behavior. Quaternary Research 74:304. Binford L (1981) Bones: Ancient Men and Modern Myths. (Academic Press). Blumenschine R, Cavallo J & Capaldo S. (1994) Competition for carcasses and early hominid behavioral ecology: a case study and conceptual framework. Journal of Human Evolution 27:197. Bunn H & Pickering T (2010) Bovid mortality profiles in paleoecological context falsify hypotheses of endurance running-hunting and passive scavenging by early Pleistocene hominins. Quaternary Research 74:395. Cerling T, et al. (2011) Diet of Paranthropus boisei in the early Pleistocene of East Africa. Proceedings of the National Academy of Sciences 108:9337. Chamaillé-Jammes S, Fritz H & Murindagomo F (2007) Climate-driven fluctuations in surface- water availability and the buffering role of artificial pumping in an African savanna: potential implication for herbivore dynamics. Austral Ecology 32:740. Deino A (2012) 40Ar/39Ar dating of Bed I, Olduvai Gorge, Tanzania, and the chronology of early Pleistocene climate change. Journal of Human Evolution 63:251. Diefendorf A, et al. (2011) Production of n-alkyl lipids in living plants and implications for the geologic past. Geochimica et Cosmochimica Acta 75:7472. Domínguez-Rodrigo M, et al. (2010) New excavations at the FLK Zinjanthropus site and its surrounding landscape and their behavioral implications. Quaternary Research 74:315. Domínguez-Rodrigo M, Egido R & Egeland C (2007) Decomstructing Olduvai:A Taphonomic Study of the Bed I Sites (Springer). Ficken K, et al. (2000) An n-alkane proxy for the sedimentary input of submerged/floating freshwater aquatic macrophytes. Organic Geochemistry 31:745. Hay R (1976) Geology of the Olduvai Gorge. (University of California Press). Hedges J & Mann D (1979) The characterization of plant tissues by their lignin oxidation products. Geochimica et Cosmochimica Acta 43:1803. Hobbie E & Werner R (2004) Intramolecular, compound-specific, and bulk carbon isotope patterns in C3 and C4 plants: a review and synthesis. New Phytologist 161:371. 78 Hopley P & Maslin M (2010) Climate-averaging of terrestrial faunas: an example from the Plio- Pleistocene of South Africa. Paleobiology 36:32. Huang Y, et al. (1999) δ13C of individual lignin phenols in Quaternary lake sediments: a novel proxy for deciphering past terrestrial vegetation changes. Geology 27:471. Isaac G (1983) Bones in contention: competing explanations for the juxtaposition of Early Pleistocene artifacts and faunal remains. Animals and Archaeology I: Hunters and Their Prey, eds Clutton-Brock J & Grigson C (Archaeopress). Kögel-Knabner I (2002) The macromolecular organic composition of plant and microbial residues as inputs to soil organic matter. Soil Biology and Biochemistry 34:139. Kozubek A & Tyman J (1999) Resorcinolic lipids, the natural non-isoprenoid phenolic amphiphiles and their biological activity. Chemical Reviews 99:1. Leakey M (1971) Olduvai Gorge: Volume 3 (Cambridge at the University Press). Lejeune O, Tlidi M & Couteron P (2002) Localized vegetation patches: a self-organized response to resource scarcity. Physical Reviews E 66:10901. Magill C, Ashley G & Freeman K (2013a) Ecosystem variability and early human habitats in eastern Africa. Proceedings of the National Academy of Sciences 110:1167. Magill C, Ashley G & Freeman K (2013b) Water, plants and early human environments in eastern Africa. Proceedings of the National Academy of Sciences 110:1175. Potts R (1988) Early Hominid Activities at Olduvai (Aldine de Gruyter). Pruetz J & Bertolani P (2009) Chimpanzee (Pan troglodytes verus) behavioral responses to stresses associated with living in a savanna-mosaic environment: implications for hominin adaptations to open habitats. PaleoAnthropology 252. Rose L & Marshall F (1996) Meat eating, hominid sociality, and home bases revisited. Current Anthropology 37:307. Sikes N (1994) Early hominid habitat preferences in East Africa: paleosol carbon isotopic evidence. Journal of Human Evolution 27:25. Santruckova H, Bird M & Lloyd J (2000) Microbial processes and carbon-isotope fractionation in tropical and temperate grassland soils. Functional Ecology 14:108. Whiten A & Erdal D (2012) The human socio-cognitive niche and its evolutionary origins. Philosophical Transactions of the Royal Society B 367:2119-2129. 79 Chapter 5. Differentiation of wholegrain biomarkers based on compound-specific 13C signatures of 5-n-alkylresorcinols 5.1. Introduction A growing number of epidemiological studies link increased dietary whole-grains to decreased risk for a spectrum of diet-related diseases, including coronary heart disease, diabetes and some types of cancer (Slavin et al. 1997, Truswell 2002). Assessment of the relationships between diet and heath traditionally are based on food frequency questionnaires, which are hindered by inaccuracies and memory bias in respondents (Kristal et al. 1997). Biomarkers present an independent and objective approach to dietary assessment. Emerging consensus in nutritional literature indicates 5-n-alkylresorcinols (1,3-dihydroxyl-5-n-alkylbenzene derivatives) are promising biomarkers for dietary whole-grains (Zamora-Ros et al. 2012). 5-n-Alkylresorcinols (ARs) are phenolic lipids and occur in a wide variety of plants as odd- numbered, side-chain homologues (Kozubek & Tyman 1999). Among plants common in the human diet, ARs are primarily limited to rye and wheat grains (Ross et al. 2004). Previous studies differentiate rye and wheat using the abundance ratio of saturated AR homologues with C17 (AR17) and C21 (AR21) side-chains (Chen et al. 2004). Abundance ratios can be sensitive to genetic and environmental (growth conditions) factors, which strongly influence AR biosynthesis in plants (Andersson et al. 2010a, Shewry et al. 2010). Compound-specific 13C signatures of plant lipids are also sensitive to growth conditions (Diefendorf et al. 2011), but relative differences in stable carbon-isotope compositions (expressed as δ13C values) among homologues are more consistent and can be diagnostic of agricultural plant sources (e.g., 80 Wiesenberg et al. 2004). In this study, we report δ13C values for saturated AR homologues from rye and wheat grains to explore the utility of compound-specific 13C signatures for differentiating between whole-grain sources. 5.2. Experimental 5.2.1. Reagents, standards and samples Isomeric hexanes, dichloromethane (DCM), ethyl acetate (EtOAc) and methanol (MeOH) of analytical-grade purity (>98.5%) were purchased from EMD Chemicals, Inc. Silica gel (60 Å, 70-230 mesh) and neutral alumina (60 Å, 50-150 mesh) were purchased from Mallinckrodt Baker, Inc. Silver-impregnated alumina (5% mass/mass) was prepared as described previously (Chakraborty & Raj 2007). Chromatographic adsorbents were heated for 8 hr at 450°C to remove organic contaminants, stored in a dessicator, and activated for 2 hr at 150°C prior to use. Pelletized diatomaceous earth (DE) and Ottawa (quartz) sand were purchased from EMD Chemicals Inc. Glass (GF/A) microfiber filter disks came from Whatman International, Ltd. Benzene-1,2-dicarboxylic acid (commonly called phthalic acid)(Schimmelmann Standards) and 2,4-dihydroxybenzoic acid (commonly called α-resorcylic acid)(Alfa Aesar) were used as internal standards. Alcoholic and acidic functional groups were derivatized using N,O- bis(trimethylsilyl)trifluoracetamide (BSTFA)(Alfa Aesar). 81 Five different samples each of rye (Secale cereal) and wheat (Triticum durum) grains were provided by Wegmans Food Markets, Inc. Grains were freeze-dried and, directly prior to analysis, powdered to pass through a 0.5 mm sieve. 5.2.2. Extraction and separation of ARs from wholegrain cereals To generate total lipid extracts (TLEs), powdered grains were extracted with DCM/methanol (90:10 volume/volume)(v/v) using a Dionex accelerated solvent extraction (ASE) 200 system. Powdered grains were added to ASE cells capped with glass microfiber disks and 4 g sand. Remaining dead volume was filled with DE/sand (50:50 v/v). Then, ASE cells were extracted in 3 cycles of 5 min at 100°C and 1500 psi (10.3 MPa) with a flush volume of 70%. Saturated AR homologues were separated from TLEs using modified ‘flash’ column chromatography (Still et al. 1978). Reconstituted TLEs were sequentially eluted with 4 mL DCM and DCM/EtOAC (85:15 v/v), over 1.0 g activated silica gel. DCM/EtOAC fractions were blown to dryness under nitrogen and reconstituted in 250 µL of DCM. Reconstituted DCM/EtOAC fractions were then eluted with 2 mL of DCM/EtOAC (85:15 v/v) over 0.5 g silver-impregnated alumina. 5.2.3. Instrumental analysis Saturated AR homologues were identified by gas chromatography-mass spectrometry (GC/MS) using a Hewlett-Packard (HP) 6890 GC connected to an HP 5973 quadropole MS with electron- 82 impact ionization and with a split/splitless injector operated in pulsed splitless mode at 320 °C. The HP GC was equipped with a 60 m DB5 fused-silica column (0.32 mm × 0.25 µm). Column flow rate was 2 ml min-1 and helium was used as the carrier gas. Oven temperature was held at 60°C for 1 min, then ramped to 320°C at 6°C min–1, and held at 320°C for 20 min. Injector temperature was 320 °C. Detector temperature was 320°C. The MS was operated with a scanning mass range of m/z 40-600 at 3 scans per second and ionization energy of 70 eV. Saturated AR homologues were detected as trimethylsilyl (TMS) derivatives and identified using characteristic molecular and fragment ions (m/z = 268 and 281) and retention times via published mass spectra and chromatographs (Kozubek & Tyman 1999, Avsejs et al. 2002). Quantification of concentrations and isotopic measurements were performed on a Varian model 3400 GC equipped with a split/splitless injector operated in splitless mode with a 60 m DB5 fused-silica column (0.32 mm × 0.25 µm). Column flow rate was 2 ml min-1 and helium was used as the carrier gas. Oven temperature was held at 60°C for 1 min, then ramped to 320°C at 6°C min–1, and held at 320°C for 20 min. Injector temperature was 320°C. Compounds were combusted over nickel and platinum wire with oxygen in helium (1% v/v) at 1000°C. Carbon dioxide was monitored using a Finnigan MAT 252. To determine concentrations, known aliquots of each saturated AR fraction were spiked with phthalic acid and α-resorcylic acid. Peak areas were normalized relative to internal standards using response curves in concentrations ranging from 1 to 1000 µg mL-1. Accuracy and precision of internal standard concentrations equaled 6.3% and 6.8% (1σ s.d.; n = 36), respectively. 83 Isotopic values were determined relative to reference carbon dioxide gas calibrated to Vienna Pee Dee Belemnite (VPDB) and expressed in per mil (‰) units: 13 13 12 δ C = 1000 × ((Rsample / Rstandard) – 1), R = C / C Within-run precision and accuracy were determined with co-injected internal standards and equaled 0.62‰ and 0.20‰ (n = 36), respectively. 5.3. Results and discussion 5.3.1. Saturated homologue contents In rye grains, total saturated AR concentrations averaged 1087±249 µg/g (n = 5), ranging from 772 to 1401 µg/g dry weight (Table 5.1). Rye grains had peak abundance of AR19 (Figure 5.1) and an average AR17-to-AR21 ratio of 0.94±0.07 (n = 5). In wheat grains, total saturated AR concentrations were lower than in than rye grains (Table 5.1), averaging 518±90 µg/g (n = 5). Wheat grains had peak abundance of AR21 and an average AR17- to-AR21 ratio of 0.11±0.04 (n = 5). Total saturated AR contents, peak abundances, and AR17-to- AR21 ratios for both rye and wheat grains were similar to previous studies (Ross et al. 2003, Chen et al. 2004, Landberg et al. 2006, Andersson et al. 2010b). 84 5.3.2. Compound-specific δ13C values for saturated homologues 13 13 In rye grains, average δ C values for AR17 (δ CAR17) to AR25 ranged from –28.4 to –29.6‰ (Figure 5.2 and Table 5.2). Saturated AR δ13C values decreased with increasing side-chain length. Isotopic fractionation during biosynthesis, expressed as ε values for individual saturated AR homologues (expressed as εARx, where x is hydrocarbon side-chain length) relative to AR17: 13 13 13 13 εARx = 1000 ((δ CARx + 1000) / (δ CAR17 + 1000) – 1) ≈ δ CARx – δ CAR17 also decreased with increasing side-chain length. In rye grains, average εARx values ranged from –0.3 to –1.2‰. These results are consistent with equivocal evidence for hydrocarbon side-chain formation via sequential addition of C2 units derived from acetyl-CoA (i.e., ‘acetogenic’ pathway)(Kozubek & Tyman 1999), which show similar isotopic depletion with increasing chain length in even-numbered n-alkanoic acid homologues from C26 to C34 (up to about 1.5‰)(Conte et al. 2003). Saturated AR homologues in wheat grains, like those in rye grains, generally became more 13C- depleted with increasing side-chain length (Table 5.2). However, average δ13C values for saturated AR homologues were lower than in rye grains, ranging from –27.5 to –28.1‰. Average εARx values in wheat grains ranged from –0.2 to –0.6‰. 5.3.3. Statistical analyses 13 Statistical comparison of average δ C values for saturated AR homologues and εARx values indicated significant differences between rye and wheat grains (Table 5.2). Specifically, 85 13 δ CAR25 (p = 0.039) and εAR23 (p = 0.045) values showed statistical differences at the α=0.05 level. Further, εAR25 (p = 0.006) and summed εARx (p = 0.002) values showed statistically different at the α=0.01 level. 5.4. Conclusions Previous studies indicate saturated AR homologues are rapidly absorbed by humans in the upper intestine (Ross et al. 2003), and occur intact in human plasma and erythrocytes (Linko & Aldercreutz 2005). Lipids show nominal isotopic fractionation when directly incorporated (routed) during digestion (Stott et al. 1997). Thus, isotopic differences among saturated AR homologues can potentially function as biomarkers for rye and wheat grains in the human diet, though future research is needed to constrain isotopic signatures for saturated AR homologues in other whole-grains (e.g., × Triticosecale) and whole-grain mixtures. 86 5.5. Figures !" *" )" !"#$%#&'( (" +,#-' )*# %" $+,-./0+12-.34-.+516787,793+1:78;7 Figure 5.1. Relative saturated alkylresorcinol (AR) composition for AR homologues with C17 hydrocarbon side-chains (AR17) to AR25 in rye and wheat grains. Vertical bars indicate 1σ standard deviation (n = 5). 87 !"#$% ./-01 !"&$% :4:67;6<=/=6> +( ! !"#$%#&'()* !"'$% +,- -./01./2345676869024 !(%$% )*+# )*+' )*"+ )*"( )*", Figure 5.2. Compound-specific δ13C values for saturated alkylresorcinol (AR) homologues with C17 hydrocarbon side-chains (AR17) to AR25 in rye and wheat grains. 88 5.6. Tables !"#! $%#&'! !""#"$% !""#"$% (!)*+!&% '()*"+",), *)-"+"'), (!),+!&% '()*"+".)' '')'"+"()/ (!)-+!&% .,)-"+"'), *-)$"+"()0 (!-."!&% ,,)*"+"')/ ,')-"+"')( (!-/"!&% ..)."+"')$ *)/"+"')' (!-/0'10(!)- ()/*"+"()(- ()''"+"()(* 21'&3+(!" '(0-"+".*/ $*."+"-/ !1232% ! 4567895:";<"=5;>";>:"<9;>:;8:":5?@;9@7>"!'A%) Table 5.1: Total alkylresorcinol (AR) content, relative saturated homologue composition and ratio of the saturated AR homologue with a C17 hydrocarbon side-chain (AR17) to AR21 in samples of wholegrain rye and wheat. 89 "#$ !"#! $%#&'! !()#*' +!,- !%"#"$% !%"#"$% 13 ! CAR25 (‰) &'()*+,+-)- &'.)-+,+/). & 0'()$ *+,)"-".,) *.,/"-".,0 && +!,. 13 ! CAR23 (‰) &'()-+,+-)/ &'1)(+,+/). %2 0'()0 *.,1"-".,) *.,2"-".,) & +!,/ 13 ! CAR21 (‰) &'.).+,+-)3 &'1).+,+/)( %2 0'()+ *.,2"-".,) *.,0"-".,) %2 +!/0 13 ! CAR19 (‰) &'.)1+,+-)' &'1).+,+/)1 %2 0'(+3 *.,0"-".,+ *.,)"-".,+ %2 +!/1 13 ! CAR17 (‰) &'.)4+,+-)' &'1)5+,+/). %2 2"+!# *),/"-".,2 *+,$"-".,2 && ! 456789":8;<:=8>"59"?85@"5@>"9=5@>5:>">8AB5=B<@"!+C%, " D=5=B9=BE56"9BF@BGBE5@E8"5="H#.,.$"!&%"5@>"H#.,.+"!&&%"68A869, $ I<"9=5=B9=BE56">BGG8:8@E8"!%2%"G<:"B@>8;8@>8@="95?;689, 13 Table 5.2: Compound-specific δ C and εARx values for saturated alkylresorcinol (AR) homologues in samples of wholegrain rye and wheat. 90 5.7. References Andersson A, Kamal-Eldin A & Åman P (2010a) Effects of environment and variety on alkylresorcinols in wheat in the HEALTHGRAIN diversity screening. Journal of Agricultural and Food Chemistry 58:9299. Andersson A, et al. (2010b) Alkylresorcinols in wheat and rye flour and bread. Journal of Food Composition and Analysis 8:794. Avsejs L, et al. (2002) 5-n-Alkylresorcinols as biomarkers of sedges in an ombrotrophic peat section. Organic Geochemistry 33:861. Chakraborty K & Raj R (2007) Eicosapentaenoic acid enrichment from sardine oil by argentation chromatography. Journal of Agricultural and Food Chemistry 55:7586. Chen Y, et al. (2004) Alkylresorcinols as markers of whole grain wheat and rye in cereal products. Journal of Agricultural and Food Chemistry 52:8242. Conte M, et al. (2003) Molecular and carbon isotopic composition of leaf wax in vegetation and aerosols in a northern prairie ecosystem. Oecologia 135:67. Diefendorf A, Freeman K & Wing S (2012) Distribution and carbon isotope patterns of diterpenoids and triterpenoids in modern C3 trees and their geochemical significance. Geochimica et Cosmochimica Acta 85:342. Kozubek A & Tyman J (1999) Resorcinolic lipids, the natural non-isoprenoid phenolic amphiphiles and their biological activity. Chemical Reviews 99:1. Kristal A, et al. (1997) Associations of race/ethnicity, education, and dietary intervention with the validity and reliability of a food frequency questionnaire. American Journal of Epidemiology 146:856. Landberg R, et al. (2006) Alkylresorcinol content and homologue composition in durum wheat (Triticum durum) kernels and pasta products. Journal of Agricultural and Food Chemistry 54:3012. Linko A & Aldercruetz H (2005) Whole-grain rye and wheat alkylresorcinols are incorporated into human erythrocyte membranes. Journal of British Nutrition 93:11. Ross A, et al. (2003) Alkylresorcinols in cereals and cereal products. Journal of Agricultural and Food Chemistry 51:4111. 91 Ross A, Kamal-Eldin A & Åman P (2004) Dietary alkylresorcinols: absorption, bioactivities, and possible use as biomarkers of whole-grain wheat- and rye-rich foods. Nutrition Reviews 62:81. Shewry P (2010) Effects of genotype and environment on the contents and composition of phytochemicals and dietary fiber components in rye in the HEALTHGRAIN diversity screening. Journal of Agricultural and Food Chemistry 58:9372. Slavin J, Jacobs D & Marquart L (1997) Whole-grain consumption and chronic disease: protective mechanisms. Nutrition and Cancer 27:14. Still W, Kahn M & Mitra A (1978) Rapid chromatographic technique for preparative separations with moderate resolution. Journal of Organic Chemistry 43:2923. Stott A, Davies E & Evershed R (1997) Monitoring the routing of dietary and biosynthesized lipids through compound-specific stable isotope (δ13C) measurements at natural abundance. Naturwissenschaft 84:82. Truswell A (2002) Cereal grains and coronary heart disease. European Journal of Clinical Nutrition, 56:1. Wiesenberg G, et al. (2004) Source and turnover of organic matter in agricultural soils derived from n-alkane/n-carboxylic acid compositions and C-isotope signatures. Organic Geochemistry 35:1371. Zamora-Ros R, et al. (2012) Application of dietary phenolic biomarkers in epidemiology: past, present, and future. Journal of Agricultural and Food Chemistry 60:6648. 92 Chapter 6. Efficient in-cell separation of lipid biomarkers using sequential accelerated solvent extraction 6.1. Abstract A novel method is described for sequential in-cell separation of sedimentary lipids via accelerated solvent extraction (ASE). Total lipid extracts were loaded into ASE cells packed with silica gel, alumina and silver-impregnated alumina and then eluted sequentially with solvent mixtures of increasing polarity to yield four fractions: saturated hydrocarbons, unsaturated/aromatic hydrocarbons, esters/ketones/alcohols and polar compounds including carboxylic acids. Recoveries for target lipids ranged from 74 to 99%, and were statistically identical to traditional methods (i.e., flash column chromatography) based on gas chromatography–mass spectrometry analysis. Both a standard lipid mixture and more complex sedimentary lipids from wetland soils were separated efficiently, with minimal co-elution among fractions. This study highlights the utility of sequential in-cell separation using ASE for rapid and efficient characterization of sedimentary lipids for environmental and geochemical studies. 6.2. Introduction Biological markers (biomarkers) are a heterogeneous group of organic compounds widely used for environmental and geochemical studies (Peters et al. 2005). Biomarkers derive from the biomass of living organisms and are preserved in ancient soils and sediments (Kögel-Knabner 2002). Geochemical studies often focus on lipid biomarkers, particularly hydrocarbons such as long-chained n-alkanes derived from plant leaves (Castaneda & Schouten 2011), which are both 93 abundant and recalcitrant in sedimentary matrices. However, recent analytical and instrumental advances (e.g., gas chromatography-time-of-flight mass spectrometry) have spurred interest in more functionalized and biologically specific lipid biomarkers, such as dicarboxylic acids from archaea (Castaneda & Schouten 2011) and ω-hydroxyacids from plant roots (Otto & Simpson 2006), for example. Solid-phase extraction (SPE) is a widely used preparative technique that allows for selective analyte isolation and matrix simplification (clean-up) of environmental samples. For over 30 years, SPE has been used to separate target lipids from complex mixtures (c.f., Poole 2003) by way of stepwise elution over columns with a single stationary phase (i.e., silica gel or alumina). Yet, SPE clean-up is subject to distinct analytical challenges, particularly partial loss of target lipids and co-elution among lipid classes. Using secondary stationary phases during SPE affords enhanced selectivity, but also demands significantly higher sample loads (Bodennec et al. 2000), greater solvent volumes and is labor intensive. Recent studies have adapted ASE techniques to improve analytical efficiency versus SPE by integrating lipid extraction and separation. The exhaustive lipid extraction capabilities of ASE are well established for a spectrum of biological and sedimentary matrixes (Carabias-Martinez et al. 2005, Mottaleb & Sarker 2012). Building on this strength, a growing body of work has demonstrated the capability of ASE for selective retention of organic interferences during lipid extraction, resulting in better isolation of target lipids. In particular, these methods add stationary phases (e.g., silica gel) that are homogenized with a sample matrix prior to loading them into ASE cells (Huie 2002, Kim et al. 2003, Ong et al. 2003, Wiberg et al. 2007). Parallel 94 developments use multiple mobile phases (i.e., solvents), generally of increasing polarity, to achieve sequential in-cell separation of target analytes (Lundstedt et al. 2006, Poerschmann & Carlson 2006, Canosa et al. 2007, Hussen et al. 2007). Although most applications add a single stationary phase to ASE cells, environmental and geochemical studies often demand multiple- stationary phase separations for adequate isolation of target lipids (Still et al. 1978, Bastow et al. 2007). Material properties of sample matrices (e.g., leaves vs. soils vs. marine sediments) vary significantly in environmental and geochemical studies and can directly influence extraction and separation efficiencies (Carabias-Martinez et al. 2005, Mottaleb & Sarker 2012). Multiple- stationary phase in-cell clean-up techniques have been proposed for matrix simplification in the analysis of steroid hormones (Hansen et al. 2011) and polycyclic aromatic hydrocarbons (Zhang et al. 2012), but the utility of adding multiple-stationary phases for sequential in-cell separation of sedimentary lipids remains unclear. In this study, we present a sequential in-cell separation method for sedimentary lipids that takes advantage of the efficiency, expediency and reproducibility of ASE and the enhanced selectivity (separation power) of multiple solvents over multiple-stationary phases. We first extract lipid biomarkers from sample matrices using ASE, and then separate the resulting total lipid extract (TLE) using stepwise elution over multiple-stationary phases. In adapting ASE for multi- dimensional chromatographic separation, we use a novel cell inversion step and are able yield four lipid fractions separated cleanly both by polarity and unsaturation. 6.3. Experimental 95 6.3.1. Wetland soil samples Wetland soils (A and Bh horizons) were collected from below a shoreline patch of Carex trisperma (three-seeded sedge) in an area of bog wetland in central Pennsylvania known as Bear Meadows Natural Area (40.815°N, 77.925°W). Sediments were freeze-dried and powdered prior to extraction. 6.3.2. Reagents and standards Analytical-grade hexanes (total hexane isomers, 98.5%,), dichloromethane (DCM, 99.9%) and methanol (99.8%) were purchased from EMD Chemicals Inc. (Gibbstown, NJ) and degassed prior to use. Silica gel (60Å, 70-230 mesh) and neutral alumina (60Å, 50-150 mesh) were purchased from Mallinckrodt Baker Inc. (Phillipsburg, NJ) and baked for 8 hr at 450°C. Silver- impregnated alumina (5% mass/mass)(w/w) was prepared according to Chakraborty & Raj (2007). Adsorbents were activated for 2 hr at 150°C prior to use. Pelletized diatomaceous earth (DE) and Ottawa sand were purchased from EMD Chemicals Inc. and baked for 8 hr at 450°C. Glass microfiber disks served as frits at the top and bottom of ASE cells and came from Whatman International Ltd. (Piscataway, NJ). A standard mixture consisted of the n-alkanes C15 (nC15) through nC22 plus nC29 and nC31 (99.0%; Alfa Aesar, Ward Hill, MA), the saturated hydrocarbons pristane (99.0%; Acros Organics, Geel, Belgium) and 5-α-cholestane (99.0%; Acros Organics), the unsaturated hydrocarbons 1-eicosene (99.0%; Acros Organics) and squalene (98.0%; Alfa Aesar), the polyaromatic hydrocarbon (PAH) pyrene (98.0%; Alfa Aesar), the n-alcohols C12 (nC12OH), nC18OH and nC28OH (98.0%; Alfa Aesar), the n-carboxylic acids 96 C16 (nC16OOH) and nC26OOH (98.0%; Alfa Aesar), the n-dicarboxylic acid C18 (nC18diOOH; 98.0%; Alfa Aesar), benzene-1,2-dicarboxylic acid (commonly called phthalic acid; 98.0%; Alfa Aesar) and, finally, 2,4-dihydroxybenzoic acid (commonly called α-resorcylic acid; 98.0%; Alfa Aesar). 6.3.3. Sample extraction To generate TLEs, powdered sediments were extracted with DCM/methanol (90:10 v/v) using a Dionex ASE 200 system. Either the standard mixture or powdered sediment (about 10 g) was added to cells capped with glass microfiber disks and about 1 g of Ottawa sand. Remaining dead volume was filled with DE/sand (1:1 v/v). Cells were extracted in 3 cycles of 5 min at 100°C and 1500 psi (10.3 MPa) with a flush volume of 70% (EPA 1995, Richter et al. 1996, Wiesenberg et al. 2004). Resulting TLEs were blown to dryness under nitrogen and reconstituted in 250 µL of DCM. 6.3.4. Sequential ASE method 6.3.4.1. Column construction Cells were packed as illustrated in Figure 6.1. Cells were capped with glass microfiber disks and about 1 g of DE/sand (1:1 v/v). Then, cells were packed with 1.5 g of silver-impregnated alumina (Ag+alumina), 1.5 g silica gel, 1 g alumina and another 1 g silica gel, respectively. A second glass microfiber disk was added and topped with either the standard mixture or 97 reconstituted TLE. Lastly, cells were packed with an additional 1.5 g silica gel. Remaining dead volume was filled with DE/sand (1:1 v/v). 6.3.4.2. Elution scheme Ultimately, packed cells were sequentially eluted with hexanes/DCM (95:5 v/v) to separate saturated hydrocarbons. Cells were then inverted and sequentially eluted first with hexanes/DCM (75:25 v/v) to separate unsaturated/aromatic hydrocarbons, then 100% DCM to separate esters/ketones/alcohols and finally with the azeotrope methanol/DCM (75:25 v/v) to separate polar compounds including oxygen-containing and polyfunctionalized compounds (Still et al. 1978, Bastow et al. 2007). Elutions proceeded in 1 cycle of 1 min at 50°C and 500 psi (3.4 MPa) with a total flush volume of 75%, except for the final elution which occurred at 75°C. Lipid fractions were blown to dryness under nitrogen and reconstituted in 250 µL of DCM. 6.3.5. Reference procedures Sequential in-cell separations were validated with reference to Soxhlet extraction and flash column chromatography (Still et al. 1978, Hawthorne et al. 2000, Bastow et al. 2007). The standard mixture or powdered wetland soils (about 10 g) were transferred to cellulose extraction thimbles and Soxhlet extracted with 500 mL of DCM/methanol (90:10 v/v) for 18 hr. Resulting TLEs were blown to dryness under nitrogen and reconstituted in 250 µL of DCM. 98 Reconstituted TLEs were first loaded onto flash columns prepared with 2 g of silica gel (Still et al. 1978, Bastow et al. 2007). Hydrocarbon, ester/ketone/alcohol and polar-compound fractions were eluted with 8 ml of hexanes/DCM (95:5 v/v), DCM and methanol/DCM (75:25 v/v), respectively. Lipid fractions were blown to dryness under nitrogen and reconstituted in 250 µL DCM. Reconstituted hydrocarbon fractions were further separated on a second flash column prepared with 2 g of Ag+alumina (Chakraborty & Raj 2007). Saturated hydrocarbons were eluted with 8 mL of hexanes/DCM (95:5 v/v) while unsaturated and aromatic hydrocarbons were eluted with 8 mL of hexanes/DCM (75:25 v/v). 6.3.6. Quantitative analysis Recoveries and separation efficiencies were evaluated using gas chromatography-mass spectrometry (GC/MS). A Hewlett-Packard 6890 Series GC was equipped with a 60 m DB5 fused-silica column (0.32 mm × 0.25 µm) and connected to a Hewlett-Packard 5973 mass selective detector. Reconstituted lipid fractions were injected in splitless mode via a Hewlett- Packard 7683 series autosampler. The GC oven temperature was programmed to 60°C for 1 min, then ramped to 320°C at 6°C min–1 and held at final temperature for 20 min. Injector temperature was held at 320°C. Detector temperature was held at 320°C. Helium was used as a carrier gas. Functionalized lipids were derivatized using N,O- bis(trimethylsilyl)trifluoracetamide (BSTFA) and detected as trimethylsilyl (TMS) derivatives. Lipids were identified by comparison with retention times and mass spectra for reference compounds, targeting those in the standard mixture. Peaks were quantified using response factors determined for the standard mixture over a range from 1 to 100 ng µL-1. 99 6.4. Theory 6.4.1. Method development 6.4.1.1. Column construction and elution scheme The sequential in-cell separation method developed here builds on established SPE and flash chromatography techniques for separating TLEs with two or more columns (Still et al. 1978, Huie 2002, Kim et al. 2003, Ong et al. 2003, Lundstedt et al. 2006, Poerschmann & Carlson 2006, Bastow et al. 2007, Canosa et al. 2007, Hussen et al. 2007, Wiberg et al. 2007, Hansen et al. 2011, Zhang et al. 2012). Yet, ASE affords sealed cells that can be inverted without compromising column chemistry or integrity, ultimately distilling multiple single-stationary phase columns into a single, multiple-stationary phase column (Figure 6.1). Hexanes/DCM (95:5 v/v) efficiently eluted only saturated hydrocarbons because silica gel retained the esters/ketones/alcohol and more polar compounds (Still et al. 1978, Wiesenberg et al. 2004, Bastow et al. 2007) and Ag+alumina retained unsaturated/aromatic hydrocarbon compounds (Wiberg et al. 2007). The second eluent, hexanes/DCM (75:25 v/v), eluted unsaturated/aromatic hydrocarbon fractions from Ag+alumina (Chakraborty & Raj 2007), even as silica gel continued to retain esters/ketone/alcohol and polar-compound factions (Bastow et al. 2007). During the second elution step, a hexanes/DCM ratio of 75:25 v/v balanced unsaturated/aromatic hydrocarbon recoveries and co-elution versus other hexanes/DCM ratios (i.e., 70:30 and 80:20 v/v). 100 The relatively low polarity of hexanes/DCM (95:5 v/v) and hexanes/DCM (75:25 v/v) prevented mobilization of ester/ketone/alcohol and polar-compound fractions so that they were retained on silica gel and never reached Ag+alumina. Cell inversion following the first elution step redirected flow and prevented polar-compound retention on the transition metal-impregnated adsorbent (Christie 1987, Cagniant 1992). Cell inversion also prevented caustic accumulation of transition metals in methanolic fractions (Christie 1987) because solvated silver is rapidly scavenged by activated alumina (Cagniant 1992). Thus, ‘backflush’ elution over silica gel with DCM and methanol/DCM (75:25 v/v) optimized separations ester/ketone/alcohol and polar- compound fractions, respectively. Caution should be exercised when performing sequential in-cell separations using ASE. Assuming conservative lipid holding capacities of about 10 mg g-1 for silica gel and alumina (Cagniant 1992, Bastow et al. 2007), the columns described in section 6.3.4. can handle less than 50 mg lipids. Although sorbent ratios should remain accurate (Christie 1987), samples of higher lipid contents will require proportionally larger amounts of stationary phases. 6.4.1.2. Extraction temperature Higher extraction temperatures generally increase lipid recoveries during solvent extraction, but also decrease separation efficiencies during chromatography (Haidacher et al. 1996). Lipid separations according to unsaturation are particularly sensitive to extraction temperatures because electron donor–acceptor (EDA) complexes formed between unsaturated/aromatic 101 compounds and Ag+alumina typically dissociate between 50 and 100°C (Cagniant 1992, van Beck & Subrtova 1995). Extraction temperatures above 100°C can also result in thermal degradation of lipids (Hawthorne et al. 2000, Wiesenberg et al. 2004). Therefore, lipid recoveries and separation efficiencies were evaluated at 50 to 100°C in 10°C intervals. Higher temperatures nominally increased target lipid recoveries, but substantially decreased separation efficiencies. Except for polar-compound fractions, extraction temperatures of 50°C balanced target lipid recoveries and separation efficiencies. Raising temperatures to 75°C in the final elution step optimized polar-compound recoveries. 6.4.1.3. Extraction cycles, static time and flush volume Increased extraction cycles, static time and flush volume can increase lipid recoveries (Zhang et al. 2012), but also increase labor and cost. Therefore, target lipid recoveries were evaluated for 1 to 5 cycles of 1 to 10 min static times with 50 to 150% flush volumes. For 1 cycle of 1 min with 75% flush volume, target lipid recoveries were comparable to those for 5 cycles of 10 min with 150% flush volume. 6.5. Results and discussion 6.5.1. Recovery and separation efficiency 6.5.1.1. Standard mixture 102 Under optimized conditions (section 6.3.4.), sequential in-cell separation of the standard mixture was highly efficient, with fractional recoveries for target lipids of 74 to 99% (Table 6.1). Lowest lipid recoveries were associated with polyfunctional (e.g., α-resorcylic acid) and polyunsaturated (e.g., squalene) compounds, although such compounds are often difficult to recover (Bennett & Larter 2000, Wiesenberg et al. 2004, Bastow et al. 2007). Overall, lipid fractions in the standard mixture were efficiently separated (Table 1). Saturated hydrocarbon fractions eluted exclusively in hexane/DCM (95:5 v/v). Trace amounts of 1- eicosene were also eluted with hexane/DCM (95:5 v/v). With the exception of trace amounts of 1-eicosene, and polyunsaturated compounds such as squalene that complex strongly with silver- impregnated stationary phases (Cagniant 1992), unsaturated/aromatic hydrocarbons eluted only in hexanes/DCM (75:25 v/v). Likewise, ester/ketone/alcohol and polar-compound fractions were separated cleanly. 6.5.1.2. Wetland soils The complex nature of soil and other sediments present a challenge when evaluating quantitative lipid recoveries and separation efficiencies. In particular, isomerism can produce unresolved complex mixtures (UCMs) of overlapping peaks in chromatograms. For example, UCMs are apparent in unsaturated/aromatic hydrocarbon and ester/ketone/alcohol fractions of the surface soil samples used for this study (Figure 6.2). Therefore, separation efficiencies were evaluated qualitatively by monitoring mass chromatograms for individual and combined lipid fractions using selected ions (m/z) of 85 (major ion of hydrocarbon fragments), 75 (major ion of TMS 103 derivatives), 74 and 103 (major ions of ester/ketone/alcohol TMS derivatives) and 117 (major ion of carboxylic acid TMS derivatives)(Peters et al. 2005). Like the standard mixture, sequential in-cell separation of lipid fractions from wetland soils was highly efficient (Figure 6.2). Hexanes/DCM (95:5 v/v) fractions show chromatograms dominated by n-alkanes, and n-alkane distributions appear identical to those for the TLEs. Hexanes/DCM (70:30 v/v) fractions do not contain traces of n-alkanes. BSTFA does not result in discernible differences between derivatized and underivatized mass chromatograms in hexanes/DCM (75:25 v/v) fractions, suggesting an absence of functionalized lipids. Both DCM and methanol/DCM (75:25 v/v) fractions show negligible peaks without BSTFA (Figure 6.2), suggesting that lipids in these two fractions are dominantly functionalized. However, derivatized DCM fractions show high m/z ratios between of 75 and 74 and 103 whereas derivatized methanol/DCM (75:25 v/v) fractions show high m/z ratios between 75 and 117. This suggests DCM fractions contain ester/ketone/alcohol compounds but methanol/DCM (75:25 v/v) fractions contain polar compounds such as dicarboxylic acids. 6.5.2. Comparison between methodologies Lipid recoveries and separation efficiencies between sequential in-cell and traditional separations were compared using the standard mixture based on Student’s t-test for unpaired values. Calculated p-values for target lipids in the standard mixture equaled greater than 0.05 for all 104 fractions (Table 6.1). Thus, lipid recoveries and separation efficiencies for the two methods were are not different at the 5% significance level. 6.6. Conclusions An ASE method was developed for sequential in-cell separation of lipid fractions from sedimentary samples. Lipid mixtures were separated cleanly into four fractions based on unsaturation and polarity: saturated hydrocarbons, unsaturated/aromatic hydrocarbons, esters/ketones/alcohols and polar compounds. Lipid recoveries and separation efficiencies for the ASE method compared favorably with traditional methods (i.e., Soxhlet extraction and flash column chromatography), but the ASE method is also automated, reproducible and very efficient. 105 6.7. Figures @ !"! $#%&'() @ !"! 839'(3:%$;<.=>,?, *+,-.&/0/1'.230 @. !"# A+,-.&/0/1'.230 !"! *-.4056/(' *+,-.&/0/1'.230 *+,-.4-74056/(' $#%&'() 839'(3:%$;<.=B,?C,. $;<.'().<3DE'(F0%$;<.=BA?GA. Figure 6.1: Column construction and elution scheme for sequential in-cell separation of lipid mixtures using pressurized liquid extraction. Cells were built from the bottom up using silver- impregnated alumina (Ag+alumina) and activated alumina and silica gel. Lipid mixtures were loaded onto glass microfibers. Dead volume was filled with a mixture of diatomaceous earth (DE) and sand (1:1 w/w). Elution with hexanes/DCM (95:5 v/v) proceeded ‘forward,’ yielding saturated hydrocarbon fractions. After cell inversion, elution with and hexanes/DCM (75:25 v/v), DCM and methanol/DCM (75:25 v/v) proceeded ‘backward,’ yielding unsaturated/aromatic hydrocarbon, ester/ketone/alcohol and polar-compound fractions, respectively. 106 !"" !>!GEEF !> EF !>%! #"#" !D !>$= !>$DEF '0-12-*,/ !> EEF 0&%*'(12 !>$" $G !> >@ $! H) !>!# !>!; ?1 :) ?C AB !"*+#%,"'-$#"$.%#/ " !>%! !>$= !> !> ?1 !# !; !> !> >@ $" $! ;"<+$".=89: '=#<#/ :)?C AB ;"<+$".=89: ';"<%"/ '.234-567-1.+.689:/ !"*+#%,"'-$#"$.%#/ !> EF 0&%*'1%3%4'56+7#%&$. $D !>!DEF AB 89: !>!GEEF !>$GEEF H) :"#;+$&*=89: ';"<%"/ !" $" %" &" #" !"#"$#%&$'(%)" '()*+,-./ 107 Figure 6.2: Combined selected ion mass chromatograms (m/z 85, 74, 75, 103 and 117) for surface soil total lipid extracts (TLEs) and lipid fractions. The complex nature of soil organic matter yielded extensive co-elution during gas chromatography-mass spectrometry analysis (top panel). Lipid fractions were eluted according to polarity and unsaturation via sequential in-cell separation using ASE with hexanes/DCM (95:5 v/v, saturated hydrocarbons), hexanes/DCM (75:25 v/v, unsaturated/aromatic hydrocarbons), DCM (esters/ketones/alcohols) and methanol/DCM (75:25 v/v, polar compounds). In the lower panel, relative signal intensity is scaled versus TLEs. Bold chromatograms show derivatized fractions while underlying chromatograms show underivatized fractions. Note that not all target lipids in the standard mixture occur in TLEs. Pr, pristane; Ei, 1-eicosene; Py, pyrene; Di, nC18diOOH; Ch, 5-α- cholestane; Sq, squalene. 108 6.8. Tables 3 4 2 3 3 = 6) s.d. n c 87 81 89 81 83 ref. 0.48 = f d d d d d d d d d d d d d d d d d d -value 88 82 90 84 88 p n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Fraction 4 3 2 4 3 = 6) ( s.d. n c 84 89 13 85 ref. 0.55 = f d d d d d d d d d d d d d d d d d d d -value 88 91 10 84 p n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Fraction 3 3 6 3 4 = 6) ( s.d. n c inal separation which occurred at 75°C. 2 82 70 85 ref. 0.71 = f d d d d d d d d d d d d d d d d d d d -value 2 86 74 84 p n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Fraction 2 3 3 3 2 2 3 3 2 3 3 2 3 3 = 6) ( s.d. n ( c 1 95 80 83 84 92 92 91 85 89 88 85 90 ref. 0.47 = f d d d d d d d d d d -value 2 p 99 82 85 88 90 91 91 84 93 91 89 90 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. recovery (%) recovery (%) recovery (%) recovery (%) Fraction 1 OH OH OH 15 16 17 18 19 20 21 22 29 31 OOH OOH 12 18 28 C C C C C C C C C C di-OOH 16 26 C C C n n n n n n n n n n 16 C C n n n pyrene pristane n n C -cholestane squalene 1-eicosene n ! Compound pthalic Acid pthalic -resorcylic acid 5- ! b b b,e b,e Alcohols Polar Compounds Polar -test for unpaired values. t Saturated Hydrocarbons Unsaturated/Aromatic Hydrocarbons Unsaturated/Aromatic Separations were obtained in 1 cycles of min at 50°C and 500 psi (3.4 MPa) with a total flush volume 75%, except for the f Compounds listed in order of elution. Abbreviations appear in section 6.3.2. Compounds listed in order of elution. Not detected. Detected as TMS derivatives. Detected as Reference procedure (see section 6.3.5.). Based on Student's a b c d e f Table 6.1: Separation of target compounds from a standard mixture using accelerated solvent extraction (ASE) under optimized conditions (see section 6.3.4.). 109 6.9. References EPA (1995) Pressurised fluid extraction: test methods for evaluating solid waste (EPA SW-846). Bastow T, van Aarssen B & Lang D (2007) Rapid small-scale separation of saturate, aromatic and polar components in petroleum. Organic Geochemistry 38:1235. Bennett B & Larter S (2000) Quantitative separation of aliphatic and aromatic hydrocarbons using silver ion-silica solid-phase extraction. Analytical Chemistry 72:1039. Bodennec J, et al. (2000) A procedure for fractionation of shingolipid classes by solid-phase extraction on aminopropyl cartridges. Journal of Lipid Research 41:1524. Cagniant D (1992) Complexation Chromatography (CRC). Canosa P, et al. (2007) Pressurized liquid extraction with in-cell clean-up followed by gas chromatography-tandem mass spectrometry for the selective determination of parabens and triclosan in indoor dust. Journal of Chromatography A 1161:105. Carabias-Martinez R, et al. (2005) Pressurized liquid extraction in the analysis of food and biological samples. Journal of Chromatography A 1089:1. Castaneda C & Schouten S (2011) A review of molecular proxies for examining modern and ancient lacustrine environments. Quaternary Science Reviews 30:2851. Chakraborty K & Raj R (2007) Eicosapentaenoic acid enrichment from sardine oil by argentation chromatography. Journal of Agricultural and Food Chemistry 55:7586. Christie W (1987) High-Performance Liquid Chromatography and Lipids: A Practical Guide (Pergamon Press Oxford). Haidacher D, et al. (1996) Temperature effects in hydrophobic interaction chromatography. Proceedings of the National Academy of Sciences 93:2290. Hansen M, et al. (2011) Determination of ten steroid hormones in animal waste manure and agricultural soil using inverse and integrated clean-up pressurized liquid extraction and gas chromatography-tandem mass spectrometry. Analytical Methods 3:1087. Hawthorne S, et al. (2000) Comparisons of Soxhlet extraction, pressurized liquid extraction, supercritical fluid extraction and subcritical water extraction for environmental solids: recovery, selectivity and effects on sample matrix. Journal of Chromatography A 892:421. Huie C (2002) A review of modern sample-preparation techniques for the extraction and analysis of medicinal plants. Analytical and Bioanalytical Chemistry 373:23. 110 Hussen A, et al. (2007) Selective pressurized liquid extraction for multi-residue analysis of organochlorine pesticides in soil. Journal of Chromatography A 1152:247. Kim J, et al. (2003) One-step pressurized liquid extraction method for the analysis of polycyclic aromatic hydrocarbons. Analytica Chimica Acta 498:55. Kögel-Knabner I (2002) The macromolecular organic composition of plant and microbial residues as inputs to soil organic matter. Soil Biology and Biochemistry 34:139. Lundstedt S, et al. (2006) Simultaneous extraction and fractionation of polycyclic aromatic hydrocarbons and their oxygenated derivatives in soil using selective pressurized liquid extraction. Analytical Chemistry 78:2993. Mottaleb M & Sarker S (2012) Accelerated solvent extraction for natural products isolation. Natural Products Isolation 864:75. Ong R, et al. (2003) Pressurized liquid extraction – comprehensive two-dimensional gas chromatography for fast-screening of polycyclic aromatic hydrocarbons in soil. Journal of Chromatography A 1019:221. Otto A & Simpson M (2006) Sources and composition of hydrolysable aliphatic lipids and phenols in soils from western Canada. Organic Geochemistry 37:385. Peters K, Walters C & Moldowan J (2005) The Biomarker Guide (Cambridge University Press). Poerschmann J & Carlson R (2006) New fractionation scheme for lipid classes based on “in-cell fractionation” using sequential pressurized liquid extraction. Journal of Chromatography A 1127:18. Poole C (2003) New trends in solid-phase extraction. Trends in Analytical Chemistry 22:362. Richter B, et al. (1996) Accelerated solvent extraction: a technique for sample preparation. Analytical Chemistry 68:1033. Still W, et al. (1978) Rapid chromatographic technique for preparative separations with moderate resolution. The Journal of Organic Chemistry 43:2923. van Beek T & Subrtova D (1995) Factors involved in the high pressure liquid chromatographic separation of alkenes by means of argentation chromatography on ion exchangers: overview of theory and new practical developments. Phytochemical Analysis 6:1. Wiberg K, et al. (2007) Selective pressurized liquid extraction of polychlorinated dibenzo-p- dioxins, dibenzofurans and dioxin-like polychlorinated biphenyls from food and feed samples. Journal of Chromatography A 1138:55. 111 Wiesenberg G, et al. (2004) Improved automated extraction and separation procedure for soil lipid analyses. European Journal of Soil Science 55:349. Zhang Q, et al. (2012) Determination of polycyclic aromatic hydrocarbons from soil samples using selective pressurized liquid extraction. Analytical Methods 4:2441. 112 Chapter 7. Research summary 7.1. Research summary The role of climate in hominin evolution has sparked interest and debate for more than a century because of apparent coincidence in changes in global climate and the hominin (and other mammals) fossil record. Biological adaptations are a result of local environmental change, particularly in habitat structure and water availability. However, high-resolution reconstructions of local conditions have been very limited because of (1) sparse or discontinuous terrestrial sediment sequences and (2) indirect proxy signals for vegetation and hydroclimate. To address these limitations, this dissertation develops a quantitative framework for interpreting environmental signals based on stable carbon- and hydrogen-isotope compositions of sedimentary biomarkers preserved in an auspicious suite of well-constrained lake and soil sediments from an individual hominin archaeological locality – Olduvai Gorge. Chapter 2 provides a quantitative framework for interpreting vegetation characteristics (i.e., habitat structure) in tropical regions as a function of plant biomarker δ13C values. Then, I use this framework to interpret changes in habitat structure through time during the early Pleistocene using a continuous sequence of lake sediments from Olduvai Gorge. Plant biomarker δ13C values reveal recurrent, abrupt fluctuations between closed forests and open grasslands that were paced by orbital precession and, to a lesser extent, tropical sea-surface temperatures. The emergence of our direct ancestor – Homo erectus (sensu lato) – is also associated with the early 113 Pleistocene, suggesting climate instability might have mediated hominin morphological and behavioral adaptations. Chapter 3 builds on the previous chapter to account for biochemical and physiological influences on hydrogen-isotope fractionation by plants. Previously published data for plant biomarker δD values and modeled precipitation δD values show clear distinctions in apparent fractionation among plant functional types (PFTs). Then, I use plant biomarker δ13C values from lake sediments at Olduvai Gorge to constrain relative PFT abundances through time and, in turn, calculate ‘landscape’ apparent fractionation factors via isotopic mass balance. Reconstructed precipitation δD values reveal dramatic fluctuations in mean annual rainfall between about 250 and 700 mm yr-1, which were paralleled by fluctuations between open grassland and closed forest ecosystems. Chapter 4 explores the utility of plant biomarker signatures in soil sediments for reconstructing microhabitat (<0.1 km2) distributions in the past. In this study, I use plant biomarkers from a suite of isochronous soil sediments to reconstruct spatial microhabitat variability across an iconic hominin archaeological site at Olduvai Gorge – FLK Zinjanthropus archaeological Level 22 (FLK Zinj). Molecular and isotopic signatures for sedimentary leaf-waxes and lignin-phenols delimit a heterogenous mosaic of microhabitats in which an isolated forest and wetland patch occurs in expansive grassland. Hominin remains occur almost exclusively within the forest patch, suggesting the patch functioned as a ‘central-foraging place’ for butchery and other social activities. 114 In Chapter 5, I investigate molecular and isotopic signatures of 5-n-alkylresorcinols in wheat (Triticum spp.) and rye (Secale cereal). These 5-n-alkylresorcinols are common in both bacteria and plants, but medium-chained homologues (e.g.. 23- and 25-carbon moities) are unique to plants in the order Poales, which are often presumed as dietary staples for hominins. Molecular data show distinct, but highly variable, differences in distributions and concentrations of 5-n- alkylresorcinols in wheat and rye. Likewise, δ13C values for individual homologues are highly variable, ranging up to several permil between interspecific samples. Despite marked variability in absolute values, relative fractionation of 13C among interspecific homologues shows statistical difference. Thus, relative isotopic differences among 5-n-alkylresorcinols might be a useful indicator of hominin dietary resources. Chapter 6 details an automated method for extraction and separation of sedimentary lipids using modified accelerated solvent extraction (ASE) instrumentation. This method consolidates a variety of traditional chromatographic and geochemical techniques to create a single, distilled method. Specifically, sedimentary lipids are loaded into ASE cells packed with silica gel, alumina and silver-impregnated alumina, then eluted with solvent mixtures of increasing polarity to yield four lipid fractions: saturated hydrocarbons, unsaturated/aromatic hydrocarbons, esters/ketones/alcohols and polar compounds. This method produces minimal co-elution between lipid fractions and requires only a fraction of the time and reagents associated with more traditional techniques. 115 7.2. Future directions Today, eastern African ecosystems are highly variable through time and across space, and work from this dissertation reveals similar variability occurred during the region’s past. However, linking ecosystem variability to hominin evolution requires not only descriptive constraints, but also emprical ‘scaling connections’ between biological, environmental and geological processes that operate at vastly different scales (c.f., Muraoka & Koizumi 2009). Scaling connections (e.g., A = kBD where A is a landscape property [e.g., wetland], B is a measure of scale [e.g., patch size], k is a constant, and D is an empirical scaling exponent) are mathematical functions derived from empircal observations of a property of interest, and provide a context for connecting large- scale environmental phenomena to molecular-scale chemical processes. Given scarce high- resolution reconstructions connecting temporal and spatial signals, current assessments of environmental hypotheses of hominin evolution are limited to arbitrary distinctions about the frequency, magnitude and extent of ecosystem variability in the past. This dissertation indicates that isolated wetland and forest patches occurred even in catchments characterized by open grasslands. Future studies should expand the scope and detail of spatial ecosystem reconstructions during the early Pleistocene at Olduvai Gorge to connect temporal and spatial ecosystem variability with hominin archaeological accumulations and, by extension, assess hominin behavioral changes through time. Food resources directly link an organism to its environment. However, functional ecosystem characteristics (e.g., woody cover) do not necessarily scale with availability of specific food resources due, for instance, to dietary preferences. This dissertation identifies 5-n- 116 alkylresorcinol signatures for several dietary resources (e.g., sedges) that could have been important for hominins, but considers only a fraction of the plants available for consumption in modern ‘savanna’ ecosystems. Future studies should explore 5-n-alkylresorcinol signatures for an expanded number of potential dietary resources for hominins (e.g., aquatic plants). 117 7.3. References Muraoka H & Koizumi H (2009) Satellite Ecology (SATECO) – linking ecology, remote sensing and micrometeorology, from plot to regional scale, for the study of ecosystem structure and function. Journal of Plant Research 122:3. 118 Appendix A. Supporting information appendix for ‘Ecosystem variability and early human environments in eastern Africa’ Reprinted with permission from the Proceedings of the National Academy of Sciences (2013) doi:10.1073/pnas.1206276110/-/DCSupplemental. A.1. Dataset description 13 13 We review 38 references to form a compilation of δ C values for soil organic matter (δ CSOM), 13 13 leaf tissues (δ Cleaf) and leaf lipids (δ C31) that includes 64 plant species and 288 soils and from over 289 subtropical and tropical geographic sites (Archer 1990, Ambrose & Sikes 1991, Balesdent et al. 1993, Mariotti & Peterschmidt 1994, Wieden et al. 1995, Desjardins et al. 1996, Schulze et al. 1996, Simoneit 1997, Boutton et al. 1998, Boutton et al. 1999, Buchmann & Schulze 1999, Guillaume et al. 1999, Schweizer et al. 1999, Diels et al. 2001, Guillaume et al. 2001, Biggs et al. 2002, Bowman & Cook 2002, Solomon et al. 2002, Conte et al. 2003, Jessup et al. 2003, Tiessen et al. 2003, Scartazza et al. 2004, Bi et al. 2005, Krull & Bray 2005, Krull et al. 2005, Bayala et al 2006, Krull et al. 2006, Liao et al. 2006, Rommerskirchen et al. 2006, Milliard et al. 2008, Terwilliger et al. 2008, Wang et al. 2008, Vogts et al. 2009, Wang et al. 2009, Aichner et al. 2010, Freier et al. 2010, Milliard et al. 2010, Wang et al. 2010). Data derive from sites with mixed C3/C4 ecosystems and mean annual precipitation of 250 to 5000 mm. Means were calculated for plant species or soil cores (0-10 cm) for each site to remove internal variability. We caution that our εSOM/31 values are not necessarily applicable on a global scale and future additions to our compilation will provide additional refinements for our calculations. 119 A.2. Age estimates and model uncertainties We construct an age model for lake sediments from Olduvai Gorge using dated tie points for Bed I sediments (Figure A.1). Individual ages derive from 40Ar/39Ar dating, tephrocorrelation or magnetic stratigraphy, or a combination of techniques (Hay 1976, Walter et al. 1991, Walter et al. 1992, Tamrat & Thouveny 1995, Hay & Kyser 2001, Blumenschine et al. 2003, McHenry et al. 2008, Stollhofen et al. 2008). Sediments from Bed I at geologic locality 80 do not show evidence for depositional hiatuses or surface erosion (Hay 1976, Hay & Kyser 2001); therefore, we adopt the approach of Hay and Kyser (2001) and apply consistent sedimentation rates for non-tuff sediments between dated tie points. We do not orbitally tune our record because local conditions at Olduvai Gorge may not be regionally or globally correlative (Blaauw 2012). Using this strategy, we also avoid any assumptions about phase relationships between the effects of ωp on seasonal solar heating (insolation) and hydroclimate change (Kingston 2007). Nevertheless, ωp shows tight amplitude 13 modulation with δ CTOC values (Figure A.2). We use the approach of Bintanja and van de Wal (2008) and assume that a global composite record of benthic δ18O values (LR04 Stack) traces polar ice volume during the early Pleistocene (Lisiecki & Raymo 2005). We use previously published age models for Ocean Drilling Program (ODP) sites 662 and 722 as constructed from δ18O values, biological and magnetic stratigraphy (Clemens et al. 1996, 120 Lisiecki & Raymo 2005, Cleaveland & Herbert 2007). Both ODP records are orbitally-tuned with respect to the LR04 Stack (Lisiecki & Raymo 2005). However, absolute differences between orbitally-tuned age estimates and depth-derived age estimates (Huybers & Wunsch 2004) average less than 0.002 Ma. We interpolate SST records at an interval of 2.0 kyr. Interpolation follows original sampling resolutions for site 662 (1.8 kyr) and site 722 (2.0 kyr), corresponding to average sedimentation -1 rates of about 6.7 and 3.1 cm kyr , respectively (Clemens et al. 1996, Lisiecki & Raymo 2005, Cleaveland & Herbert 2007). Models for both sites yield age estimates in close agreement with well-constrained radiometric ages for the top (1.778 ± 0.003 Ma) and bottom (1.945 ± 0.004 Ma) of the Olduvai subchron (Gradstein et al. 2004). A.3. Correlation and regression analyses Proxy records (time series) from terrestrial and marine sediments are difficult to compare, in part, due to methodological differences for estimates of age and age uncertainty. For instance, time series from terrestrial sediments often cannot be orbitally-tuned because of ambiguity regarding phase-relationships or correlations between local, regional and global signals (Huybers & Wunsch 2004, Kingston 2007, Laepple & Lohmann 2009, Blaauw 2012). Time series from marine sediments can be orbitally-tuned, but ambiguity about phase-relationships persists because of ambiguous synchronicity between local, regional and global signals (Hughen et al. 2006, Blaauw 2012) 121 Precise phase-relationships remain unresolved in the absence of a common absolute timescale (Huybers & Wunsch 2004, Blaauw 2012), but we compare time series from terrestrial and marine sediments using rank- and Fourier transformation methods. We use rank-transformations to account for asymmetric and uneven data distributions and to provide robust estimates (Spearman) for monotonic relationships between time series (Conover & Iman 1981). We estimate average time-lags between rank-transformed time series using cross-correlation, which measures how closely two signals resemble each other when shifted forward or backward through time (Figure A.2). A robust approach for calculating cross-correlations between uneven time series is with the Lomb-Scargle algorithm for Fourier transformation (Scargle 1982): -1 * rαβ (τ) = FT (FT{α}FT {β}) The terms FT, FT* and FT-1 denote a Fourier transform, its complex conjugate and inverse, respectively, and rαβ (τ) is the peak cross-correlation between α and β at a time-lag of τ kyr. Then, we evaluate cross-correlations between time series for lags from –10 to 20 kyr to calculate peak cross-correlations (Table A.1): peak cross-correlation = max{rαβ (τ)} We use rank- and Fourier transformed time series at peak lags for our multivariate partial regression models (Bretherton et al. 1992). In all cases, peak lags occur in less than 5 kyr, and 13 we conclude that correlations between δ CTOC and ωP, SST662 and SST722 (Clemens et al. 1996, Lisiecki & Raymo 2005, Cleaveland & Herbert 2007) and reconstructed polar ice volumes from global benthic δ18O values (Lisiecki & Raymo 2005) are robust within depth-derived age uncertainties for dated tie points (Figure A.1). 122 A.4. Tables a Environmental Factor 2 Slope p-value Shift r kyr Precession (!P) 0.61 + <0.0001 -1 SST662 0.44 + <0.001 -2 SST722 0.37 " <0.001 2 Polar Ice Volumeb 0.16 + <0.01 -5 a 13 Positive shifts indicate time-lag in relation to # CTOC values. b 18 Polar ice volumes from global benthic # OLR04 values (12). 13 Table A.1: Single-factor cross-correlation (Fourier) after rank-transformation between δ CTOC values, ωP, SST662 and SST722 (Clemens et al. 1996, Lisiecki & Raymo 2005, Cleaveland & 18 Herbert 2007) and reconstructed polar ice volumes from global benthic δ OLR04 values (Lisiecki & Raymo 2005). 123 A.5. Figures >400/?B &)*$$ >400/?C >400/?A >400/?, &)*%$ &)+$$ >400/?@ 3$00$"&#'"4'56*)#'7! &)+%$ ,-..-"/-0/1234567/849:;<-= $ % &$ &% '$ '% ($ !"#$%$"&'$&'(%)*%$+)*,-$.'/"012& !"# Figure A.1: Age model for Olduvai sediments using dated tie points from Bed I based on 40Ar/39Ar, tephrocorrelation and magnetic stratigraphy (Hay 1976, Walter et al. 1991, Walter et al. 1992, Tamrat & Thouveny 1995, Hay & Kyser 2001, Blumenschine et al. 2003, McHenry et al. 2008, Stollhofen et al. 2008). Bed I sediments from geologic locality 80 show no evidence for depositional hiatuses or surface erosion (Hay 1976, Hay & Kyser 2001). We apply linear sedimentation rates between dated tie points for non-tuff sediments (Hay & Kyser 2001). Gray lines represent stratigraphic age uncertainties based on non-parametric interpolation and stochastic sedimentation variability (Heegard et al. 2005). 124 !$ " - "#&" :<- !? " <6="> 99:&;; "#%" 99:@@; "#$" "#!" , ! ( '!" " !" ;" -+.//'-.++01234.5 6277457 !8 ()*+, '"#&" Figure A.2: Cross-correlation functions calculated by rank-transformation and Fourier methods for calculated orbital precession (ωP) in relation to SST662 and SST722 (Clemens et al. 1996, Lisiecki & Raymo 2005, Cleaveland & Herbert 2007), reconstructed polar ice volumes from 18 13 global benthic δ OLR04 values (Lisiecki & Raymo 2005) and δ CTOC values for lake sediments from Olduvai Gorge. Dashed horizontal lines represent 5% confidence bounds for uncorrelated time series, as calculated from cross-correlation of 1000 data randomizations of original data. Positive shifts indicate time-lag in relation to ωP. 125 0 !*#$%&'() !"#$%&'() .&/ 1-0 !!"#$ !!"#$ ! 3" !"#$ ! %&' " " 1/0 %&'(!)*)+ !)*)+ !.,- 3456"78985:;<=> %&' 1,0 !"#$ +,,-./01&(2# " "!)*)+ %&'(,- !.,- 120 Figure A.3: Conceptual diagram describing stable carbon-isotopic relationships between carbon dioxide, leaf tissues, the nC31 and soil organic matter (SOM) for C3 and C4 plants. 126 A.6. References Aichner B, et al. (2010) Biomarker and compound-specific δ13C evidence for changing environmental conditions and carbon limitation at Lake Koucha, eastern Tibetan Plateau. Journal of Paleolimnology 43:873. Ambrose S & Sikes N (1991) Soil carbon isotope evidence for Holocene habitat change in the Kenya Rift Valley. Science 253:1402. Archer S (1990) Development and stability of grass/woody mosaics in a subtropical savanna parkland, Texas, USA. Journal of Biogeography 17:453. Balesdent J, Girardin C & Mariotti A (1993) Site-related 13C of tree leaves and soil organic matter in a temperate forest. Ecology 74:1713. Bayala J, et al. (2006) Relative contribution of trees and crops to soil carbon content in a parkland system in Burkina Faso using variations in natural 13C abundance. Nutrient Cycling in Agroecosystems 76:193. Bi X, Sheng G, Liu X, Li C & Fu J (2005) Molecular and carbon and hydrogen isotopic composition of n-alkanes in plant leaf waxes. Organic Geochemistry 36:1405. Biggs T, Quade J & Webb R (2002) δ13C values of soil organic matter in semiarid grassland with mesquite (Prosopis) encroachment in southeastern Arizona. Geoderma 110:109. Bintanja R & van de Wal R (2008) North American ice-sheet dynamics and the onset of 100,000-year glacial cycles. Nature 454:869. Blaauw M (2012) Out of tune: the dangers of aligning proxy archives. Quaternary Science Reviews 36:38. Blumenschine R, et al. (2003) Late Pliocene Homo and hominid land use from western Olduvai Gorge, Tanzania. Science 299:1217. Boutton T, Archer S & Midwood A (1999) Stable isotopes in ecosystem science: structure, function and dynamics of a subtropical savanna. Rapid Communications in Mass Spectrometry 13:1263. Boutton T, et al. (1998) δ13C values of soil organic carbon and their use in documenting vegetation change in a subtropical savanna ecosystem. Geoderma 82:5. Bowman D & Cook G (2002) Can stable carbon isotopes (δ13C) in soil carbon be used to describe the dynamics of Eucalyptus savanna-rainforest boundaries in the Australian monsoon tropics? Austral Ecology 27:94. 127 Bretherton C, Smith C & Wallace J (1992) An intercomparison of methods for finding coupled patterns in climate data. Journal of Climate 5:541. Buchmann N & Schulze E (1999) Net CO2 and H2O fluxes of terrestrial ecosystems. Global Biogeochemical Cycles 13:751. Cleaveland L & Herbert T (2007) Coherent obliquity band and heterogeneous precession band responses in early Pleistocene tropical sea surface temperatures. Paleoceanography 22:PA2216. Clemens S, Murray D & Prell W (1996) Nonstationary phase of the Plio-Pleistocene Asian monsoon. Science 274:943. Conover W & Iman R (1981) Rank transformations as a bridge between parametric and nonparametric statistics. American Statistician 35:124. Conte M, et al. (2003) Molecular and carbon isotopic composition of leaf wax in vegetation and aerosols in a northern prairie ecosystem. Oecologia 135:67. Desjardins T, et al. (1996) Changes of the forest-savanna boundary in Brazilian Amazonia during the Holocene revealed by stable isotope ratios of soil organic carbon. Oecologia 108:749. Diels J, et al. (2001) Temporal variations in plant δ13C values and implications for using the 13C technique in long-term soil organic matter studies. Soil Biology and Biochemistry 33:1245. Freier K, Glaser B & Zech W (2010) Mathematical modeling of soil carbon turnover in natural Podocarpus forest and Eucalyptus plantation in Ethiopia using compound specific δ13C analysis. Global Change Biology 16:1487. Gradstein F, Ogg J & Smith A (2004) A Geologic Time Scale (Cambridge University Press). Guillaume K, et al. (1999) Soil organic matter dynamics in tiger bush (Niamey, Niger). Preliminary results. Acta Oecologica 20:185. Guillaume K, et al. (2001) Does the timing of litter inputs determine natural abundance of 13C in soil organic matter? Insights from an African tiger bush ecosystem. Oecologia 127:295. Hay R (1976) Geology of the Olduvai Gorge (University of California Press). Hay R & Kyser T (2001) Chemical sedimentology and paleoenvironmental history of Lake Olduvai, a Pliocene lake in northern Tanzania. Geological Society of America Bulletin 113:1505. 128 Heegaard E, Birks H & Telford R (2005) Relationships between calibrated ages and depth in stratigraphical sequences: an estimation procedure by mixed-effect regression. The Holocene 15:612. Hughen K, et al. (2006) Marine-derived 14C calibration and activity record for the past 50,000 years updated from the Cariaco Basin. Quaternary Science Reviews 25:3216. Huybers P & Wunsch C (2004) A depth-derived Pleistocene age model: uncertainty estimates, sedimentation variability, and nonlinear climate change. Paleoceanography 19:PA1028. Jessup K, Barnes P & Boutton T (2003) Vegetation dynamics in a Quercus-Juniperus savanna: an isotopic assessment. Journal of Vegetation Science 14:841. Kingston J (2007) Shifting adaptive landscapes: progress and challenges in reconstructing early hominid environments. American Journal of Physical Anthropology 134:20. Krull E & Bray S (2005) Assessment of vegetation change and landscape variability by using stable carbon isotopes of soil organic matter. Australian Journal of Botany 53:651. Krull E, et al. (2005) Recent vegetation changes in central Queensland, Australia: evidence from δ13C and 14C analyses of soil organic matter. Geoderma 126:241. Krull E, et al. (2006) Compound-specific δ13C and δ2H analyses of plant and soil organic matter: a preliminary assessment of the effects of vegetation change on ecosystem hydrology. Soil Biology and Biochemistry 38:3211. Laepple T & Lohmann G (2009) Seasonal cycle as template for climate variability on astronomical timescales. Paleoceanography 24:PA4201. Liao J, Boutton T & Jastrow J (2006) Organic matter turnover in soil physical fractions following woody plant invasion of grassland: evidence from natural 13C and 15N. Soil Biology and Biochemistry 38:3197. Lisiecki L & Raymo M (2005) A Plio-Pleistocene stack of 57 globally distributed benthic 18O records. Paleoceanography 20:522. Mariotti A & Peterschmitt E (1994) Forest savanna ecotone dynamics in India as revealed by carbon isotope ratios of soil organic matter. Oecologia 97:475. McHenry L, Mollel G & Swisher III C (2008) Compositional and textural correlations between Olduvai Gorge Bed I tephra and volcanic sources in the Ngorongoro Volcanic Highlands, Tanzania. Quaternary International 178:306. Millard P, et al. (2008) Partitioning soil surface CO2 efflux into autotrophic and heterotrophic components, using natural gradients in soil δ13C in an undisturbed savannah soil. Soil Biology and Biochemistry 40:1575. 129 Millard P, et al. (2010) Quantifying the contribution of soil organic matter turnover to forest soil respiration, using natural abundance δ13C. Soil Biology and Biochemistry 42:935. Rommerskirchen F, et al. (2006) Chemotaxonomic significance of distribution and stable carbon isotopic composition of long-chain alkanes and alkan-1-ols in C4 grass waxes. Organic Geochemistry 37:1303. Scargle J (1982) Studies in astronomical time series analysis. Statistical aspects of spectral analysis of unevenly spaced data. The Astrophysical Journal 263:835. Scartazza A, et al. (2004) Comparisons of δ13C of photosynthetic products and ecosystem respiratory CO2 and their responses to seasonal climate variability. Oecologia 140:340. Schulze E, et al. (1996) Diversity, metabolic types and δ13C carbon isotope ratios in the grass flora of Namibia in relation to growth form, precipitation and habitat conditions. Oecologia 106:352. Schweizer M, Fear J & Cadisch G (1999) Isotopic (13C) fractionation during plant residue decomposition and its implications for soil organic matter studies. Rapid Communications in Mass Spectrometry 13:1284. Simoneit B (1997) Compound-specific carbon isotope analyses of individual long-chain alkanes and alkanoic acids in Harmattan aerosols. Atmospheric Environment 31:2225. Solomon D, et al. (2002) Soil organic matter dynamics in the subhumid agroecosystems of the Ethiopian highlands: evidence from natural 13C abundance and particle size-fractionation. Soil Science Soceity of America Journal 66:969. Stollhofen H, et al. (2008) Fingerprinting facies of the Tuff IF marker, with implications for early hominin palaeoecology, Olduvai Gorge, Tanzania. Palaeogeography, Palaeoclimatology, Palaeoecology 259:382. Tamrat E & Thouveny N (1995) Revised magnetostratigraphy of the Plio-Pleistocene sedimentary sequence of the Olduvai Formation (Tanzania). Palaeogeography, Palaeoclimatology, Palaeoecology 114:273. Terwilliger V, et al. (2008) Reconstructing palaeoenvironment from δ13C and δ15N values of soil organic matter: a calibration from arid and wetter elevation transects in Ethiopia. Geoderma 147:197. Tiessen H, et al. (2003) Organic matter transformations and soil fertility in a treed pasture in semiarid NE Brazil. Plant and Soil 252:195. 130 Vogts A, et al. (2009) Distribution patterns and stable carbon isotopic composition of alkanes and alkan-1-ols from plant waxes of African rain forest and savanna C3 species. Organic Geochemistry 40:1037. Walter R, Manega P & Hay R (1992) Tephrochronology of Bed I, Olduvai Gorge: an application of laser-fusion dating to calibrating biological and climatic change. Quaternary International 13:37. Walter R, et al. (1991) Laser-fusion 40Ar/39Ar dating of Bed I, Olduvai Gorge, Tanzania. Nature 354:145. Wang G, et al. (2008) Paleovegetation reconstruction using 13C of soil organic matter. Biogeosciences Discussions 5:1795. Wang L, et al. (2009) Spatial heterogeneity and sources of soil carbon in southern African savannas. Geoderma 149:402. Wang L, et al. (2010) Patterns and implications of plant-soil δ13C and δ15N values in African savanna ecosystems. Quaternary Research 73:77. Wedin D, et al. (1995) Carbon isotope dynamics during grass decomposition and soil organic matter formation. Ecology 76:1383. 131 Appendix B. Supporting information appendix for ‘Water, plants and early human habitats in eastern Africa Reprinted with permission from the Proceedings of the National Academy of Sciences (2013) doi:10.1073/pnas.1209405109/-/DCSupplemental. B.1. Dataset description We compile and review published leaf-lipid δD data to form a compilation (n = 159) of 76 plant species from 31 sites (Chikaraishi & Naraoka 2003, Bi et al. 2005, Chikaraishi & Naraoaka 2006, Krull et al. 2006, Liu et al. 2006, Sachse et al. 2006, Smith & Freeman 2006, Hou et al. 2007, Hou et al. 2008, Liu & Huang 2008, Mügler et al. 2008, Sachse et al. 2009, Feakins & Sessions 2010). Sites are located within 5 biomes and have mean annual precipitations ranging from 183 mm to more than 1200 mm. Leaf-lipid data derive from living subtropical and tropical plants belonging either to the monocot clade or the two largest dicot clades (rosid and asterid) in ecosystems with discontinuous woody cover and herbaceous understory. We calculate species means for individual sites to remove within-species variability (Diefendorf et al. 2010). We use modeled annual δDrain values to calculate representative εlipid/water values because measured δDrain values rarely accompany published leaf-lipid δD values. B.2. Salinity reconstructions Today, saline-alkaline lake waters in eastern Africa show remarkably consistent trends in major solute compositions of water (Talling & Talling 1965). In closed basins, water chemistry is 132 - + - + primarily a function of dissolved Cl , Na , CO3–HCO3 , and K species (Eugster & Jones 1979). The presence of distinctive minerals such as trona (NaHCO3·Na2CO3·2H2O) in Bed I lake sediments suggests a similar water chemistry for paleolake Olduvai (Renaut et al. 1986, Hay & Kyser 2001). Species diversity and composition of aquatic organisms are influenced by the interplay between salinity and habitat availability (Hammer & Heseltine 1988, Verschuren et al. 2000). Fossil remains of two fish taxa (Clarias sp. (catfish) and Orechromis sp. (tilapia)) occur in Bed I sediments from about 1.835 Ma (Stewart 1996, Hay & Kyser 2001). Today, catfish similar to those occurring in Bed I sediments (e.g., C. gariepinus and C. lazera) survive at salinities of up to 20 ppt in shallow benthic habitats (Chervinski 1984, Britz & Hecht 1989). Extant tilapias can survive in open waters at salinities of up to 40 ppt (Kamal & Mair 2005), but most species reproduce in littoral habitats of between 4 and 5 m depth at salinities of 20 ppt or less (Reite et al. 1974, Stickney 1986, Stewart 1996, Kamal & Mair 2005). In shallow saline-alkaline lakes, salinity is controlled primarily by water level fluctuations over long timescales (101 to 103 kyr)(Mason et al. 1994, Yechieli & Wood 2002, Rimmer 2003). Solute balance is largely a function of mineral precipitation and solute diffusion at water- sediment interfaces (Lerman & Jones 1973, Yuretich & Cerling 1983), and we use a conservative solute balance for paleolake Olduvai during deposition of Bed I sediments. Paleolake Olduvai occupied an elliptical conic basin with high surface area-to-volume ratio (Hay 1976, Ashley & Hay 2002). Today, closed basins with similar morphology show strong 133 correlations between water level and surface area (Langbein 1961, Mason et al. 1994). Stratigraphic evidence in lake-margin deposits suggests a maximum water level of about 5 m (Hay 1976, Hay & Kyser 2001), which is consistent with fossil occurrences of tilapia (Stewart 1996). During maximum expansion, paleolake Olduvai extended to 15 km in average diameter (Ashley & Hay 2002). Relative lake levels correlate strongly with sedimentary total organic carbon (%TOC) in many shallow lakes in eastern Africa (Talbot & Livingstone 1989, Verschuren 2001). Although this relationship may not be purely mechanistic, in part, low lake levels result in low %TOC values due to selective removal of unstable organic compounds during bacterial respiration and to sediment dilution (Verschuren 2001). Thus, we interpret %TOC values as a reflection of relative lake levels. Reconstructed levels for paleolake Olduvai agree closely with independent records for lake level (lithological and faunal) during deposition of Bed I sediments (Hay 1976, Frenandez-Jalvo et al. 1998, Hay & Kyser 2001, Ashley 2007). We develop a conservative lake-water evaporation model for paleolake Olduvai based on the strong empirical relationship between observed and modeled salinity in modern saline-alkaline lakes (Langbein 1961, Mason et al. 1994)(Figure B.2): 2/3 S = 55000 (E / D) (0.55 / U) (√AL / D) Here, S is salinity in parts per million, E is annual potential evaporation in feet, D is average lake level in feet, U is the coefficient of variation for lake area change (about 1.75 for shallow lakes in 134 eastern Africa)(Langbein 1961, Hay 1968, Vareschi 1982) and AL is lake area in square miles. Thus, based on stratigraphic evidence for average lake level (about 6.5 feet)(Hay 1976, Hay & Kyser 2001) and area (about 70 square miles)(Hay 1976, Ashley & Hay 2002) during maximum lake expansion, we calculate a salinity of about 20 ppt for paleolake Olduvai during maximum lake expansion, which is consistent with faunal evidence for salinity (Figure B.2). Evidence for wave or current action at several localities in lake sediments from central parts of the paleolake Olduvai basin suggest lake levels ranged between a maximum of 1 to 2 m (average depth of about 1.5 feet)(Hay 1976, Ashley & Hay 2002) during lake contraction with lake areas of about 10 square miles (Hay 1976, Ashley & Hay 2002); thus, we calculate salinity of about 105 ppt during lake contraction, which is consistent with mineralogical evidence for high salinity during low lake levels (e.g., trona and gaylussite)(Hay 1976, Ashley & Hay 2002). Overall, reconstructed salinity primarily fluctuates between about 20 and 80 ppt during and is consistent with the range of modern fluctuations in nearby lakes considered as chemical and sedimentary analogues for paleolake Olduvai (e.g., Natron and Nakuru)(Hay 1968, Hay 1976, Vareschi 1982, Hay & Kyser 2001). B.3. Monthly and regional amount effects Amount effects are strongest in tropical regions and function via re-evaporation and diffusive exchange during precipitation events (Lee & Fung 2008). Thus, amount effects are sensitive to relative humidity and precipitation rate (Risi et al. 2010), resulting in monthly and regional variability (Vimeaux et al. 2005). Modeled monthly δDrain values (Bowen & Revenaugh 2003) and precipitation averages (Oldenborgh 2011) for rainy seasons and the climatologically 135 important (Camberlin & Okoola 2003, Tierney et al. 2011) months that precedes them (February (long rains) and September (short rains)) from 48 stations in central eastern Africa show amount –1 effects of about –0.125‰ mm (Figures B.3 and B.4). Interestingly, the x-intercept for short rains is about 10‰ more negative than for long rains. Thus, amount effects for central eastern –1 Africa are broadly consistent with those for Central America (about –0.125‰ mm )(Lachinet & Patterson 2006) and for tropical and coastal regions receiving less than 750 mm of MAP (about – –1 0.145‰ mm )(Risi et al. 2008). B.4. Evaporative balance reconstructions Isotopic mass balance for evaporative loss in well-mixed lakes with constant volume is equal to (Gibson et al. 1996): I δinput = Q δoutflow + E δevaporation (1) I is input, Q is outflow and E is lake evaporation. Variables δinput, δoutflow and δevaporation represent isotopic compositions of input, outflow and evaporation, respectively. Since δoutflow is similar to the composition of lake water (δlake): E/I = (δinput – δlake) / (δevaporation – δlake) (2) Values for δevaporation cannot be measured directly, but fractionation between δevaporation and δlake depends on temperature, boundary layer and atmospheric conditions. Assuming negligible resistance to mixing (Gat 1996): –3 δevaporation ≈ (α*δlake – h (δinput – ε*) – ε) / (1 – h + 10 εK) (3) 136 Here, α* is equilibrium isotopic fractionation between lake-water and vapor and h is relative humidity. The variable ε equals the sum of equilibrium (ε*) and kinetic (εK) fractionations. We calculate ε* for deuterium using the empirical equation (Horita & Wesolowski): 3 9 2 6 3 9 3 ε* = 1158.8 (T / 10 ) – 1620.1 (T / 10 ) + 794.84 (T/ 10 ) – 161.04 + 2.9992(10 /T ) T is lake surface temperature in Kelvin. We also calculate εK for deuterium (Araguás-Araguás et al. 2000): εK = 12.5 (1 – h) (5) Next, we substitute equation (3) into equation (2): E/I = (δinput – δlake) / (–ε* – εK) = (δlake – δinput) / ε (6) We calculate the ratio of lake evaporation to input for Olduvai Gorge during arid and wetter intervals. We define δlake based on reconstructed δlake values, but must define several other variables based on historical observations: 1. Mean annual δDinput value equals –22‰ (Nkotagu 1996, Araguás-Araguás et al. 2000, Bergonzini et al. 2001, Bowen & Revenaugh 2003, Delalande et al. 2008, Bohté et al. 2010). 2. Mean annual h rose to 75% during wetter intervals but fell to 55% during arid intervals (Williams et al. 1998, Tierney et al. 2011). Mean annual h is currently about 65% (Williams et al. 1998). 3. Mean annual temperature of 23°C with little seasonal variability (Hay & Kyser 2001). Then, we use equation (8) to calculate E/I of 2.9 (δlake = +59‰) during arid intervals and 1.3 (δlake = +16‰) during wetter intervals. Values vary by less than 0.5 if mean annual h is used to calculate E/I. These values are in close agreement with modeled E/I (0.5° × 0.5°)(Willmott et al. 137 1985) near Olduvai Gorge using prescribed MAP values of 250 mm (E/I = 3.2) and 700 mm (E/I = 1.3)(Matsurura et al. 2009). B.5. Uncertainty in ε31/model values We propagate uncertainty in εlandscape values (95% confidence intervals; σlandscape) using a linear combination of 95% confidence interval values for individual C4 graminoids (±8‰; σgram), C3 herbs (±10‰; σherb) and C3 woody plants (±8‰; σwoody) and modeled annual δDrain values (±6‰; 13 σrain)(Matsurura et al. 2009). We account for uncertainty in δ C31-based estimates of relative plant functional type abundances (about 20%)(Cerling et al. 2011) by multiplying respective standard error values by 1.2: 2 2 2 2 2 σ landscape = 1.2σ gram+ 1.2σ herb + 1.2σ woody + 3σ rain Thus, σlandscape is equal to about 20‰. B.6. Determination of environmental water δ18O values In this study, we use a single soil carbonate sample (nodule with sparry calcite) from the eastern lake-margin of Olduvai Gorge that has a δ18O value of –6.2‰ (Cerling & Hay 1986) to 18 determine δ Orain values. We assume a mean annual soil temperature (MAsT) of 25°C in order to calculate apparent fractionation values for oxygen isotopes between environmental water and soil carbonate minerals (εcarb/water = Rcarb / Rwater – 1)(Friedman & O’Neil 1977, Cerling & Hay 18 1986). Uncertainty of ±5°C in MAsT results in about 1‰ uncertainty in δ Orain values. 138 We determine apparent fractionation values for oxygen isotopes between environmental water and clay minerals (εclay/water = Rclay / Rwater – 1) according to bond-type calculations of Savin and 18 Lee (1988) from structural formulas. Our determination of δ Olake values derives from a single sediment sample composed of 97% illite and 3% analcime (w/w). Mean annual temperature at Olduvai Gorge during the deposition of Bed I sediments has been estimated as 14–16°C (Cerling & Hay 1986, Hay & Kyser 2001), as compared to about 22°C in the present, and we use 15°C to 18 calculate εclay/water values. Bulk clay minerals show a δ O value of 27.1‰, to which we apply a εclay/water value of 24.8‰. 139 B.7. Figures *##+,-./01/23#!0#4#56)7 *##?@,A3#!0#4#6=)> *##81129#:;-0<3#!0#4#=7)> !%$$ 0 2% !%'$ / ! !%&$ !"#$%&'() ()*+,-). !""$ !"#$ !"#$ !""$ !%&$ !%'$ !%$$ !"#$ !""$ !%&$ !%'$ !%$$ !"#$ !""$ !%&$ !%'$ !%$$ ()*+,-). !/"10 ()*+,-). !/"10 ()*+,-). !/"10 !"#$%&'() !"#$%&'() !"#$%&'() Figure B.1: Published δD values for the lipids nC31 (δD31) and nC29 (δD29)(Chikaraishi & Naraoka 2003, Bi et al. 2005, Chikaraishi & Naraoaka 2006, Krull et al. 2006, Liu et al. 2006, Sachse et al. 2006, Smith & Freeman 2006, Hou et al. 2007, Hou et al. 2008, Liu & Huang 2008, Mügler et al. 2008, Sachse et al. 2009, Feakins & Sessions 2010), cross-plotted by photosynthetic pathway and growth habit. C4 graminoids: δD31 = 0.94 δD29 – 26‰ (r = 0.94) C3 herbs: δD31 = 1.1 δD29 + 15‰ (r = 0.94) C3 woody plants: δD31 = 1.0 δD29 + 1‰ (r = 0.90) 140 % !"#$%&'()#'"*+,-./%/'0 !"#$ #% *++,- &%% &#% #()$ '%% Figure B.2: Bathymethric contours (lines) and reconstructed salinity (shading) for paleolake Olduvai. Bold outlines represent ancient shorelines during expanded and contracted phases (Pickering 1958, Hay 1976, Hay & Kyser 2001). Fish fossils constrain salinity to 20 ppt or less during maximum lake expansion (Reite et al. 1974, Stickney 1986, Stewart 1996, Kamal & Mair 2005), and we use 20 ppt for our conservative lake water evaporation model. Bathymetric contours occur at approximately 0.5 m intervals. 141 ! *'5. ( # "#! ! 2 ! "$! ) )' " 3-45-67)89 + # * "%! $ %-+&1&+,%-/012, "&! ! '! #!! #'! $!! $'! %&'()*&+,%-./012,3*&4565/'/5-. +,,-,./01"#2 Figure B.3: Modeled monthly δDrain values (Bowen & Revenaugh 2003) versus measured monthly precipitation (P in mm)(Oldenborgh 2011) for rainy seasons from 48 stations in central eastern Africa (Figure B.4). Short rains (n = 6200) and long rains (n = 6335) show amount effects –1 of about –0.125‰ mm : ‘Long rains’: δDrain = –0.132 P – 2‰ (r = –0.93; bold regression) ‘Short rains’: δDrain = –0.138 P – 12‰ (r = –0.95; dashed regression) Bold vertical lines represent uncertainty in modeled monthly δDrain values (95% confidence interval) and dotted horizontal lines represent average monthly precipitation variability. Long rains include the months February (F), March (Ma), April (A) and May (M); short rains include the months September (S), October (O), November (N) and December (D)( Nicholson 2000, Camberlin & Okoola 2003). 142 !" !?, 12)*3(�*$4((6,7#28-#9:-#28;5 15,)&+(�*$4((1,:#3@A#BC+#D,@5 D8-*,.*18E88FG*H8?I8?>8 D8-*,.*18E88FG*H8?I8?>8 " �* J?A,77,G*KL8?M8 J?A,77,G*KL8?M8 / 2/L/L8G*N,?;8 2/L/L8G*N,?;8 ! #!" #$" 6 #%" ()*+,-./.*012345 !)'"-"'(!)*+,-.( #&" #'" #'" #&" #%" #$" #!" " !" #'" #&" #%" #$" #!" " !" !"#$%&"'(!)*+,-.(!/ !"#$%&"'(!)*+,-.(!/�* �* ()*+,-./.*012345 ()*+,-./.*012345 Figure B.4: Measured (IAEA 2006) versus modeled (Bowen & Revenaugh 2003) monthly δDrain values for eastern African sites with at least 2 years of rainy season precipitation data (Dar es Salaam, Tanzania; Entebbe, Uganda; Muguga, Kenya). One ‘short rains’ data point (marked x) has been omitted from linear regression analysis because it is an outlier (jackknife estimate). ‘Long rains’: Modeled δDrain = 0.9621 δDrain – 2‰ (r = 0.95; n = 12) ‘Short rains’: Modeled δDrain = 0.9362 δDrain – 3‰ (r = 0.89; n = 11) 143 (" !"# $"# %"# &"# %%"' %&"' %)"' %*"' Figure B.5: Geographic locations of the 48 stations used to calculate amount effects for eastern Africa. 144 !$# !-./01 !"## "#$!-./01%>4567!,/44 : !""# !;<<8=4 568!,/44%%&"#'!-./014 !"%# ! )*+#-2*34 ,/44 ! ?43@4ABCDE , !"&# !"'# !9/*:;+";# !"(# # #)% #)' #)* #)+ !""#$%&'%()*+,- ,-./0123.405678079.: Figure B.6: Alternative εlandscape values based on modified proportions of trees versus shrubs in C3 woody plants. We use median ε31/model values for C4 graminoids (εgram = –146‰), C3 non- woody plants (εherb = –124‰), C3 shrubs (εshrub = –87‰) and C3 trees (εtree = –121‰)(Chikaraishi & Naraoka 2003, Bi et al. 2005, Chikaraishi & Naraoaka 2006, Krull et al. 2006, Liu et al. 2006, Sachse et al. 2006, Smith & Freeman 2006, Hou et al. 2007, Hou et al. 2008, Liu & Huang 2008, Mügler et al. 2008, Sachse et al. 2009, Feakins & Sessions 2010). εlandscape = fgram εgram + fherb εherb + fshrub εshrub + ftree εtree Taken together, the ε31/model value for combined C3 woody plants is –109‰. Alternative scenarios yield εlandscape values that vary by up to 15‰, although differences are nominal for relative C3 woody plant abundances of less than about 50%. 145 B.8. References Araguás-Araguás L, Froehlich K & Rozanski K (2000) Deuterium and oxygen-18 isotope composition of precipitation and atmospheric moisture. Hydrological Processes 14:1341. Ashley G (2007) Orbital rhythms, monsoons, and playa lake response, Olduvai basin, equatorial East Africa (ca. 1.85-1.74 Ma). Geology 35:1091. Ashley G & Hay R (2002) Sedimentation patterns in a Plio-Pleistocene volcaniclastic rift- platform basin, Olduvai Gorge, Tanzania. SEPM Special Publication 73:107. Bergonzini L, et al. (2001) Water and isotopic (18O and 2H) budget of Lake Massoko, Tanzania. Quantification of exchange between lake and groundwater. Comptes Rendus de l'Academie des Sciences 333:617. Bi X, et al. (2005) Molecular and carbon and hydrogen isotopic composition of n-alkanes in plant leaf waxes. Organic Geochemistry 36:1405. Bohté R, et al. (2010) Hydrograph separation and scale dependency of natural tracers in a semi- arid catchment. Hydrology and Earth System Sciences Discussions 7:1343. Bowen G & Revenaugh J (2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39:1299. Britz P & Hecht T (1989) Effects of salinity on growth and survival of African sharptooth catfish (Clarias gariepinus) larvae. Journal of Applied Ichthyology 5:194. Camberlin P & Okoola R (2003) The onset and cessation of the "long rains" in eastern Africa and their interannual variability. Theoretical and Applied Climatology 75:43. Cerling T & Hay R (1986) An isotopic study of paleosol carbonates from Olduvai Gorge. Quaternary Research 25:63. Cerling T, et al. (2011) Woody cover and hominin environments in the past 6 million years. Nature 476:51. Chervinski J (1984) Salinity tolerance of young catfish, Clarias lazera (Burchell). Journal of Fish Biology 25:147. Chikaraishi Y & Naraoka H (2003) Compound-specific δD-δ13C analyses of n-alkanes extracted from terrestrial and aquatic plants. Phytochemistry 63:361. Chikaraishi Y & Naraoka H (2006) Carbon and hydrogen isotope variation of plant biomarkers in a plant-soil system. Chemical Geology 231:190. 146 Delalande M, et al. (2008) Hydroclimatic and geothermal controls on the salinity of Mbaka Lakes (SW Tanzania): limnological and paleolimnological implications. Journal of Hydrology 359:274. Diefendorf A, et al. (2010) Global patterns in leaf 13C discrimination and implications for studies of past and future climate. Proceedings of the National Academy of Sciences 107:5738. Eugster H & Jones B (1979) Behavior of major solutes during closed-basin brine evolution. American Journal of Science 279:609. Feakins S & Sessions A (2010) Controls on the D/H ratios of plant leaf waxes in an arid ecosystem. Geochimica et Cosmochimica Acta 74:2128. Fernandez-Jalvo Y, et al. (1998) Taphonomy and palaeoecology of Olduvai Bed I (Pleistocene, Tanzania). Journal of Human Evolution 34:137. Friedman I & O’Neil J (1977) Compilation of stable isotope fractionation factors of geochemical interest. U.S. Geological Survey Professional Papers 440:1. Gat J (1996) Oxygen and hydrogen isotopes in the hydrologic cycle. Annual Review of Earth and Planetary Sciences 24:225. Gibson J, Edwards T & Prowse T (1996) Development and validation of an isotopic method for estimating lake evaporation. Hydrological Processes 10:1369. Hammer U & Heseltine J (1988) Aquatic macrophytes in saline lakes of the Canadian prairies. Hydrobiologia 158:101. Hay R (1968) Chert and its sodium-silicate precursors in sodium-carbonate lakes of East Africa. Contributions to Mineralogy and Petrology 17:255. Hay R (1976) Geology of the Olduvai Gorge (University of California Press). Hay R & Kyser T (2001) Chemical sedimentology and paleoenvironmental history of Lake Olduvai, a Pliocene lake in northern Tanzania. Geological Society of America Bulletin 113:1505. Horita J & Wesolowski D (1994) Liquid-vapor fractionation of oxygen and hydrogen isotopes of water from the freezing to the critical temperature. Geochimica et Cosmochimica Acta 58:3425. Hou J, D'Andrea W & Huang Y (2008) Can sedimentary leaf waxes record D/H ratios of continental precipitation? Field, model, and experimental assessments. Geochimica et Cosmochimica Acta 72:3503. 147 Hou J, et al. (2007) Hydrogen isotopic variability in leaf waxes among terrestrial and aquatic plants around Blood Pond, Massachusetts (USA). Organic geochemistry 38:977. IAEA (2006) Isotope Hydrology Information System. Accessible at: http://www.iaea.org/water. Kamal A & Mair G (2005) Salinity tolerance in superior genotypes of tilapia, Oreochromis niloticus, Oreochromis mossambicus and their hybrids. Aquaculture 247:189. Krull E, et al. (2006) Compound-specific δ13C and δ2H analyses of plant and soil organic matter: a preliminary assessment of the effects of vegetation change on ecosystem hydrology. Soil Biology and Biochemistry 38:3211. Lachniet M & Patterson W (2006) Use of correlation and stepwise regression to evaluate physical controls on the stable isotope values of Panamanian rain and surface waters. Journal of Hydrology 324:115. Langbein W (1961) Salinity and hydrology of closed lakes. Geological Survey Professional Paper 412:1. Lee J & Fung I (2008) 'Amount effect' of water isotopes and quantitative analysis of post- condensation processes. Hydrological Processes 22:1. Lerman A & Jones B (1973) Transient and steady-state salt transport between sediments and brine in closed lakes. Limnology and Oceanography 18:72. Liu W, Yang H & Li L (2006) Hydrogen isotopic compositions of n-alkanes from terrestrial plants correlate with their ecological life forms. Oecologia 150:330. Liu Z & Huang Y (2008) Hydrogen isotopic compositions of plant leaf lipids are unaffected by a twofold pCO2 change in growth chambers. Organic geochemistry 39:478. Mason I, et al. (1994) The response of lake levels and areas to climatic change. Climatic change 27:161. Matsuura K, Willmott C & Legates D (2009) Web-Based, Water-Budget, Interactive, Modeling Program (WebWIMP): http://climate.geog.udel.edu/~wimp/ Mügler I, et al. (2008) Effect of lake evaporation on δD values of lacustrine n-alkanes: a comparison of Nam Co (Tibetan Plateau) and Holzmaar (Germany). Organic geochemistry 39:711. Nicholson S (2000) The nature of rainfall variability over Africa on time scales of decades to millenia. Global and Planetary Change 26:137. 148 Nkotagu H (1996) Application of environmental isotopes to groundwater recharge studies in a semi-arid fractured crystalline basement area of Dodoma, Tanzania. Journal of African Earth Sciences 22:443. Oldenborgh G (2011) Koninklijk Nederlands Meteoroligisch Instituut (KNMI) Climate Explorer. Accessible at: http://www.knmi.nl/. Pickering R (1958) A Preliminary Note on the Quaternary Geology of Tanganyika (Proceedings of the CCTA Joint Committee on Regional Geology). Reite O, Maloiy G & Aasehaug B (1974) pH, salinity and temperature tolerance of Lake Magadi Tilapia. Nature 247:315. Renaut R, Tiercelin J & Owen R (1986) Mineral precipitation and diagenesis in the sediments of the Lake Bogoria basin, Kenya Rift Valley. Geological Society, London, Special Publications 25:159. Rimmer A (2003) The mechanism of Lake Kinneret salinization as a linear reservoir. Journal of Hydrology 281:173. Risi C, Bony S & Vimeux F (2008) Influence of convective processes on the isotopic composition (δ18O and δD) of precipitation and water vapor in the tropics: physical interpretation of the amount effect. Journal of Geophysical Research 113:D19306. Risi C, et al. (2010) Evolution of the stable water isotopic composition of the rain sampled along Sahelian squall lines. Quarterly Journal of the Royal Meteorological Society 136:227. Sachse D, Kahmen A & Gleixner G (2009) Significant seasonal variation in the hydrogen isotopic composition of leaf-wax lipids for two deciduous tree ecosystems (Fagus sylvativa and Acer pseudoplatanus). Organic geochemistry 40:732. Sachse D, Radke J & Gleixner G (2006) δD values of individual n-alkanes from terrestrial plants along a climatic gradient: implications for the sedimentary biomarker record. Organic geochemistry 37:469. Savin S & Lee M (1988) Isotopic studies of phyllosilicates. Reviews in Mineralogy and Geochemistry 19:189. Smith F & Freeman K (2006) Influence of physiology and climate on δD of leaf wax n-alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70:1172. Stewart K (1996) A report on the fish remains from Beds I and II sites, Olduvai Gorge, Tanzania. Darmst Beitrag Naturgesch 6:263. Stickney R (1986) Tilapia tolerance of saline waters: a review. The Progressive Fish-Culturist 48:161. 149 Talbot M & Livingstone D (1989) Hydrogen index and carbon isotopes of lacustrine organic matter as lake level indicators. Palaeogeography, Palaeoclimatology, Palaeoecology 70:121. Talling J & Talling I (1965) The chemical composition of African lake waters. Internationale Revue der gesamten Hydrobiologie und Hydrographie 50:421. Tierney J, et al. (2011) Model, proxy and isotopic perspectives on the East African Humid Period. Earth and Planetary Science Letters 307:103. Vareschi E (1982) The ecology of Lake Nakuru (Kenya): abiotic factors and primary production. Oecologia 55:81. Verschuren D (2001) Reconstructing fluctuations of a shallow East African lake during the past 1800 yrs from sediment stratigraphy in a submerged crater basin. Journal of Paleolimnology 25:297. Verschuren D, et al. (2000) Effects of depth, salinity, and substrate on the invertebrate community of a fluctuating tropical lake. Ecology 81:164. Vimeux F, et al. (2005) What are the climate controls on δD in precipitation in the Zongo Valley (Bolivia)? Implications for the Illimani ice core interpretation. Earth and Planetary Science Letters 240:205. Williams K, et al. (1998) Temporally variable rainfall does not limit yields of Serengeti grasses. Oikos 81:463. Willmott C, Rowe C & Mintz Y (1985) Climatology of the terrestrial seasonal water cycle. Journal of Climatology 5:589. Yechieli Y & Wood W (2002) Hydrogeologic processes in saline systems: playas, sabkhas, and saline lakes. Earth-Science Reviews 58:343. Yuretich R & Cerling T (1983) Hydrogeochemistry of Lake Turkana, Kenya: mass balance and mineral reactions in an alkaline lake. Geochimica et Cosmochimica Acta 47:1099. 150 Appendix C. Supporting information appendix for ‘Plant biomarker patterns at Olduvai Gorge establish social meat consumption by pre-modern humans’ C.1. Biomarker extractions, separation and characterization Paleosol samples from FLK Zinj were freeze-dried and powdered prior to accelerated solvent extraction (Dionex ASE 200 system) with dichloromethane (DCM) and methanol (90:10 v/v) in 3 cycles of 5 min at 100°C and 1500 psi (10.3 MPa). Resulting total lipid extracts were blown to dryness under nitrogen. First, total lipid extracts were separated into hydrocarbon, ester/ketone/alcohol, and polar fractions by ‘flash’ column chromatography over activated silica gel with hexanes, DCM and DCM/methanol (50:50 v/v), respectively. All 5-n-alkylresorcinol homologues were separated into polar fractions. Second, hydrocarbon fractions were further separated into saturated and unsaturated fractions over activated silver-impregnated alumina (5% w/w) with hexanes and DCM, respectively. Third, saturated hydrocarbon fractions were separated into unbranched (normal) and branched fraction by zeolitic (5Å) sieving. Extracted sediments were oxidized under alkaline conditions (8% NaOH) with CuO at 155°C for 180 min in stainless steel pressure vessels (Hedges & Mann 1979) and then acidified to pH 1 with concentrated HCl. Lignin- derived oxidation products were recovered by extraction with diethyl ether. Preliminary biomarker characterization was achieved using gas chromatography-mass spectrometry (GC/MS) with a Hewlett-Packard 6890 series GC and Hewlett-Packard 5973 mass selective detector. Biomarker fractions were injected in splitless mode onto a 60 m DB5 fused- 151 silica column (0.32 mm × 0.25 µm) via a Hewlett-Packard 7683 series autosampler. Oven temperature was programmed to 60°C for 1 min, then ramped to 320°C at 6°C min-1 and held at final temperature for 20 min. Injector and detector temperature was held at 320°C. Functionalized lipids were derivatized using N,O-bis(trimethylsilyl)trifluoracetamide and detected as trimethylsilyl (TMS) derivatives. Biomarker δ13C values were measured using gas chromatography-combustion-isotope-ratio monitoring mass spectrometry (GC-C-irMS) with a Varian 3400 model GC connected to a Thermo MAT 252. Biomarker fractions were injected in splitless mode onto a 60 m DB5 fused- silica column (0.32 mm × 0.25 µm) prior to combustion over nickel and platinum wire with oxygen in helium (1% v/v) at 1000°C. Isotopic values were determined relative to reference gas calibrated to Vienna Pee Dee Belemnite (VPDB) and expressed in permil (‰) units: 13 13 12 δ C = 1000 ((Rsample / Rstandard) – 1), R = C / C Within-run precision and accuracy were determined with co-injected internal standards and are equal, respectively, to 0.14‰ and 0.09‰ (n-alkanes; 1σ, n = 106), to 0.49‰ and 0.17‰ (5-n- alkylresorcinols; 1σ, n = 28), and to 0.66‰ and 0.21‰ (lignin monomers; 1σ, n = 54). C.2. 5-n-Alkylresorcinols in sedges and sediments Resorcinolic lipids have been isolated and characterized in leaves, stems, and bacterial cells (Kozubek & Tyman 1999), but odd-numbered, long-chained 5-n-alkylresorcinol homologues occur primarily in the monocotyledonous plant families Poaceae (subfamily Pooideae) and Cyperaceae. In eastern Africa, extant genera of Pooideae are limited to high-altitude 152 environments (>2000 m)(Tieszen et al. 1979), well above the elevation of Olduvai Gorge (1450 m). In contrast, extant genera of Cyperaceae are quite cosmopolitan, occurring in wetlands and seasonal floodplains from sea level to about 2000 m (Stock et al. 2004). Previous studies identify 5-n-alkylresorcinols with saturated C19–C25 sidechains as major homologues in extant Eriophorum (C3) and Tricophorum (C3) sedges (Avsejs et al. 2002). We identify the same major homologues in extant Cyperus (C4) sedges as well as sediments from northern FLK Zinj trenches (Figure C.1). Values for δ13C of 5-n-alkylresorcinol homologues from Cyperus are much less negative than from FLK Zinj. We conclude that sedges fringing FLK Zinj used C3 photosynthesis. 153 C.3. Figures ."" BCD+%&'( *+ @>#A> !$%+ @>.A$ @>.A5 @>#A? 4%#5 !"#+2+#$3 $% $% $% $% " .= #. #> #? ."" )*+,-./0+$+*-./ % %&67')8&+49:(;7(<& ,/0)1 1,/0) ?@'@76EF6G&H*G<)(*6 -+./0+1'2#34' *+ !!,, ) !$%+ @#.A> @#.A? @#.A$ " !"#+2+#$3 !" !# !! !$ %&'&(')*(+,)-& !"#$%&'() Figure C.1: Partial mass chromatogram (m/z = 268) for polar fractions (as TMS derivatives) of a 13 representative northern FLK Zinj trench and the C4 sedge Cyperus papyrus. Also shown are δ C values for 5-n-alkylresorcinol homologues (δaR) with saturated C19–C25 sidechains (aR19–aR25). 154 C.4. References Avsejs L, et al. (2002) 5-n-alkylresorcinols as biomarkers of sedges in an ombrotrophic peat section. Organic Geochemistry 33:861. Hedges J & Mann D (1979) The characterization of plant tissues by their lignin oxidation products. Geochimica et Cosmochimica Acta 43:1803. Kozubek A & Tyman J (1999) Resorcinolic lipids, the natural non-isoprenoid phenolic amphiphiles and their biological activity. Chemical Reviews 99:1. Stock W, Chuba D & Verboom G (2004) Distribution of South African C3 and C4 species of Cyperaceae in relation to climate and phylogeny. Austral Ecology 29:313. Tieszen L, et al. (1979) The distribution of C3 and C4 grasses and carbon isotope discrimination along an altitudinal and moisture gradient in Kenya. Oecologia 37:337. 155 Appendix D. Data for δ13C values of soil organic matter vs. leaf tissues D.1. Dataset description 13 13 13 This is a compilation of δ C values for soil organic matter (δ CSOM) and leaf tissues (δ Cleaf) that includes 288 soils and from subtropical and tropical geographic sites (Archer 1990, Ambrose & Sikes 1991, Balesdent et al. 1993, Mariotti & Peterschmidt 1994, Wieden et al. 1995, Desjardins et al. 1996, Schulze et al. 1996, Simoneit 1997, Boutton et al. 1998, Boutton et al. 1999, Buchmann & Schulze 1999, Guillaume et al. 1999, Schweizer et al. 1999, Diels et al. 2001, Guillaume et al. 2001, Biggs et al. 2002, Bowman & Cook 2002, Solomon et al. 2002, Conte et al. 2003, Jessup et al. 2003, Tiessen et al. 2003, Scartazza et al. 2004, Bi et al. 2005, Krull & Bray 2005, Krull et al. 2005, Bayala et al 2006, Krull et al. 2006, Liao et al. 2006, Rommerskirchen et al. 2006, Milliard et al. 2008, Terwilliger et al. 2008, Wang et al. 2008, Vogts et al. 2009, Wang et al. 2009, Aichner et al. 2010, Freier et al. 2010, Milliard et al. 2010, Wang et al. 2010). Data derive from sites with mixed C3/C4 ecosystems and mean annual precipitation of 250 to 5000 mm. Means were calculated for multiple specimens of the same plant species or soil cores (0-10 cm) for each site to remove internal variability (Diefendorf et al. 2010). 156 D.2. Tables 13 Table D.1. Compliation of published carbon-isotope compositions for plant leaves (δ Cleaf) and 13 soil organic matter (δ CSOM), and plant–soil apparent fractionation factors (εleaf/SOM) according to photosynthetic pathway. 157 1993 1998 et al. et al. Boutton Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -25.1-24.0-24.0-24.0 1.0018-24.0 1.0029-24.0 1.0034-24.0 1.0016 1.8-24.0 1.0033 2.9 -24.0 1.0011 3.4 -24.0 1.0053 1.6 -24.0 1.0048 3.3 -24.0 1.0072 1.1 -24.0 1.0078 5.3 -24.0 1.0038 4.8 -24.0 1.0051 7.2 -24.0 1.0058 Balesdent 7.8 -24.0 1.0023 3.8 -24.0 1.0067 5.1 -24.0 1.0070 5.8 -24.0 1.0059 2.3 -24.0 1.0037 6.7 -25.0 1.0048 7.0 1.0083 5.9 1.0049 3.7 1.0018 4.8 8.3 4.9 1.8 102 soils Ambrose and Sikes 1991 13 ! ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Community leaf C 13 vs. VPDB -27.3-25.6-27.2-25.1 C -29.1 C -28.7 C -31.0 C -31.6 C -27.7 C -29.0 C -29.6 C -26.2 C -30.5 C -30.8 C -29.7 C -27.6 C -28.7 C -32.0 C -28.8 C -26.8 C C C C -26.9 C -26.8 C ! ‰ 158 1996 2002 2003 2005 et al. et al. et al. et al. Krull Tiessen Tiessen Krull & Bray 2005 Desjardins Bowman & Cook 2002 Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -26.3-26.1-26.1-26.1 1.0022-26.1 1.0047-26.1 1.0047-26.1 1.0043 2.2-26.1 1.0042 4.7 -26.1 1.0040 4.7 -28.1 1.0038 4.3 -28.1 1.0033 4.2 -24.8 1.0032 4.0 -24.8 1.0022 3.8 -24.8 1.0015 3.3 -24.7 1.0033 3.2 -24.7 1.0002 Solomon 2.2 -24.7 1.0018 1.5 -24.0 1.0024 3.3 -24.0 1.0020 0.2 -24.0 1.0016 1.8 -24.0 1.0040 2.4 -24.0 1.0042 2.0 1.0035 1.6 1.0036 4.0 1.0041 4.2 3.5 3.6 4.1 13 ! ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Community leaf C 13 vs. VPDB -30.7-30.3-30.2-30.0 C -29.8 C -29.3 C -29.2 C -30.2 C -29.6 C -28.0 C -25.0 C -26.6 C -27.0 C -26.6 C -26.3 C -27.9 C -28.1 C -27.4 C -27.5 C -28.0 C C C C -28.4 C -30.7 C ! ‰ 159 2003 2001 et al. et al. Diels Jessup Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -23.8-23.8-23.8-23.8 1.0039-23.8 1.0044-23.8 1.0049-23.8 1.0030 3.9 -23.8 1.0030 4.4 -23.8 1.0034 4.9 -23.8 1.0035 3.0 -23.8 1.0058 3.0 -23.8 1.0050 3.4 -23.8 1.0050 3.5 -25.2 1.0057 5.8 -25.2 1.0048 5.0 -25.2 1.0049 5.0 -25.2 1.0006 5.7 -25.2 1.0012 4.8 -25.2 1.0015 4.9 -25.2 1.0016 0.6 -25.2 1.0018 1.2 -25.2 1.0023 1.5 -25.2 1.0028 1.6 1.0028 1.8 1.0029 2.3 1.0031 2.8 2.8 2.9 3.1 13 ! ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Community leaf C 13 vs. VPDB -28.1-28.6-26.7-26.7 C -27.1 C -27.2 C -29.4 C -28.7 C -28.7 C -29.3 C -28.5 C -28.6 C -25.8 C -26.4 C -26.7 C -26.8 C -27.0 C -27.4 C -27.9 C -27.9 C -28.0 C -28.2 C C C C -27.6 C ! ‰ 160 1999 1999 2001 2008 2006 2001 2006 et al. et al. et al. et al. et al. et al. et al. Diels Bayala Guillaume Guillaume Buchmann Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -25.2-25.2-25.2-25.2 1.0031-25.2 1.0032-25.2 1.0033-25.2 1.0033 3.1 -25.2 1.0034 3.2 -25.2 1.0034 3.3 -25.2 1.0035 3.3 -25.2 1.0035 3.4 -24.0 1.0036 3.4 -24.2 1.0036 3.5 -24.7 1.0037 3.5 -25.0 1.0021 3.6 -25.0 1.0021 3.6 -25.8 1.0033 3.7 -24.9 1.0022 2.1-24.2 1.0027 2.1 -24.7 1.0015 3.3 -24.5 1.0009 2.2 10 soils-24.5 1.0025 2.7 -24.5 1.0017 1.5 1.0052 0.9 1.0063 2.5 Liao 1.0053 1.7 5.2 6.3 5.3 Millard 13 ! ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Community leaf C 13 vs. VPDB -28.3-28.4-28.4-28.5 C -28.5 C -28.6 C -28.6 C -28.7 C -28.7 C -28.8 C -26.0 C C -27.9 C -27.1 C -27.6-27.3 C -25.8 C -26.6 C -26.4 C C -30.6 C -29.6 C C C -28.2 C -29.5 C -26.2 C ! ‰ 161 2001 2010 2008 et al. et al. et al. Wang Wang Guillaume Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -24.5-24.5-24.5-24.5 1.0037-24.5 1.0079-24.5 1.0050-24.5 1.0036 3.7 -24.5 1.0031 7.9 -24.5 1.0061 5.0 -24.5 1.0035 3.6 -25.0 1.0031 3.1 -25.9 1.0021 6.1 -25.9 1.0044 3.5 -25.9 1.0017 3.1 -25.9 1.0007 2.1 -25.9 0.9994 4.4 -25.9 1.0000 1.7-25.9 1.0013 0.7 -25.9 0.9999 -0.6 -25.9 0.9994 0.0 -25.9 0.9998 1.3 -25.9 1.0001 -0.1 -25.9 1.0007 -0.6 1.0001 -0.2 1.0023 0.1 1.0001 0.7 Wang 0.1 2.3 0.1 13 ! ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Community leaf C 13 vs. VPDB -32.1-29.4-28.0-27.5 C -30.4 C -27.9 C -27.5 C -26.5 C -28.8 C -26.7 C C -25.3 C -25.9 C -27.2-25.8 C -25.3 C -25.7 C -26.0 C -26.6 C -26.0 C -28.1 C -26.0 C C C C -28.1 C -26.6 C ! ‰ 162 2010 . 2009 et al. et al. Wang Wang Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -25.9-25.9-25.9-25.9 1.0006-25.9 1.0014-25.9 1.0019-25.9 1.0011 0.6 -25.9 1.0021 1.4 -25.9 1.0020 1.9 -25.9 1.0012 1.1 -25.9 1.0008 2.1 -25.9 1.0029 2.0 -25.9 1.0019 1.2 -25.9 1.0012 0.8 -25.9 1.0029 2.9 -25.9 1.0010 1.9 -25.9 1.0032 1.2 -25.9 1.0044 2.9 -25.9 1.0017 1.0 -25.9 1.0015 3.2 -25.9 1.0022 4.4 -25.9 1.0025 1.7 -24.9 1.0021 1.5 1.0022 2.2 1.0026 2.5 1.0017 2.1 2.2 2.6 1.7 5 soils Wang 13 ! ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Community leaf C 13 vs. VPDB -27.3-27.7-27.0-27.9 C -27.8 C -27.1 C -26.7 C -28.7 C -27.7 C -27.1 C -28.7 C -26.9 C -29.0 C -30.2 C -27.6 C -27.4 C -28.0 C -28.3 C -27.9 C -28.0 C -28.4 C -26.6 C C C C -26.5 C ! ‰ 163 2008 2004 2010 2009 2009 et al. 1998, 1999 et al. 1998 and 1999 et al. et al. et al. et al. et al. Scartazza Boutton Mariotti & Peterschmidt 1994 Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -25.1-26.1-26.4-27.0 1.0026-25.4 1.0023-25.5 1.0020-25.1 1.0032 2.6-25.5 1.0020 2.3 1.0015 2.0-13.3 1.0024 3.2 10 soils-13.3 1.0017 2.0 -13.3 8 soils 1.5 -13.2 2 soils 0.9981 2.4 -16.2 0.9979 1.7 Millard -16.2 0.9982 Terwillinger -16.2 0.9986 -1.9 -16.2 0.9979 Freier -2.1 -16.2 0.9988 -1.8 -16.2 0.9976 -1.4 Freier -16.2 0.9983 -2.1-16.2 0.9982 -1.2 -16.2 0.9969 -2.4 -12.8 0.9962 -1.7 0.9988 -1.8 0.9978 -3.1 0.9986 -3.8 -1.2 Boutton et al. 1995 Wedin -2.2 -1.4 14 soils Ambrose & Sikes 1991 13 ! ‰ 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Community leaf C 13 vs. VPDB -11.4-11.2-11.5-11.8 C C C C -11.4 C -27.0-27.4-27.2 C C C -14.1-13.8-14.5 C -14.4-13.1 C -12.4 C -15.0 C -14.0 C C C C -27.6 C -28.3-28.3-30.1-27.3 C C C C -15.0 C ! ‰ 164 1996 2002 2005 . 2002 et al. et al. et al et al. Krull Krull & Bray 2005 Bowman & Cook 2002 Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -14.5-12.9-12.9-12.9 0.9979-12.9 0.9992-12.9 0.9992-12.9 0.9990 -2.1-12.9 0.9989 -0.8 -16.3 0.9987 -0.8 -15.5 0.9985 -1.0 -14.3 0.9985 -1.1 -15.1 0.9980 -1.3 -15.1 0.9987 -1.5 -15.1 0.9991 -1.5 -15.1 0.9992 -2.0-15.1 0.9988 Solomon -1.3-15.1 0.9992 -0.9-15.1 0.9979 -0.8 -15.1 0.9988 -1.2 -15.1 0.9984 -0.8 -15.1 0.9989 -2.1 -15.1 0.9992 -1.2 -15.1 0.9972 -1.6 0.9984 -1.1 Biggs Desjardins 0.9977 -0.8 0.9982 -2.8 Archer 1990 -1.6 -2.3 -1.8 13 ! ‰ 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Community leaf C 13 vs. VPDB -11.9-11.8-11.6-11.4 C -11.4 C C C C -13.4 C -12.1 C -14.3-14.3 C -13.9-14.3-13.0 C -13.9 C -13.5 C -14.0 C -14.3 C -12.3 C -13.5 C -12.8 C -13.3 C C C C -12.4 C -14.2 C -12.1 C ! ‰ 165 2001 . 2003 et al et al. Diels Jessup Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -15.1-15.1-15.1-15.1 0.9988-15.1 0.9972-15.1 0.9995-15.1 0.9998 -1.2-15.1 0.9977 -2.8 -15.1 0.9993 -0.5 -15.1 0.9990 -0.2 -15.1 1.0000 -2.3 -15.1 1.0006 -0.7 -15.1 0.9985 -1.0 -15.1 0.9980 0.0 -15.1 0.9985 0.6 -15.1 0.9973 Krull & Bray 2005 -1.5 -15.1 0.9973 -2.0 -15.1 0.9974 -1.5 -15.1 0.9977 -2.7 -15.1 0.9978 -2.7 -15.1 0.9984 -2.6 -15.1 0.9984 -2.3 0.9984 -2.2 0.9986 -1.6 0.9992 -1.6 -1.6 -1.4 -0.8 13 ! ‰ 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Community leaf C 13 vs. VPDB -12.3-14.6-14.9-12.8 C -14.4 C -14.1 C -15.1 C -15.7 C -13.6 C -13.1 C -13.6 C -12.4 C -12.4 C -12.5 C -12.8 C -12.9 C -13.5 C -13.5 C -13.5 C -13.7 C -14.3 C C C C -13.9 C ! ‰ 166 2008 2010 2006 et al. et al. et al. Krull Wang Wang Millard Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C -17 0.9977 -2.3 vs. VPDB -17.8-15.6-15.6-15.6 0.9974-15.6 0.9986-15.6 0.9979-15.6 0.9984 -2.6 -15.6 0.9987 -1.4 -14.2 0.9988 -2.1 -14.2 0.9978 -1.6 -14.2 0.9983 -1.3 -14.2 1.0000 -1.2 -14.2 0.9999 -2.2 -14.2 0.9995 -1.7 -14.2 0.9999 0.0 -14.2 0.9995 -0.1 -14.2 0.9996 -0.5 -14.2 1.0001 -0.1 -14.2 0.9983 -0.5 -14.2 1.0002 -0.4 -14.2 1.0001 0.1 -14.2 0.9987 -1.7 0.9986 0.2 0.9981 0.1 0.9996 -1.3 -1.4 -1.9 -0.4 13 ! ‰ 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Community leaf C 13 vs. VPDB -15.2 C -14.2-13.5-14.0-14.3 C -14.4 C -13.4 C -13.9 C -14.2 C -14.1 C -13.7 C -14.1 C -13.7 C -13.8 C -14.3 C -12.5 C -14.4 C -14.3 C -12.9 C -12.8 C -12.3 C -13.8 C C C C -14.7 C ! ‰ 167 2008 2010 2009 et al. et al. et al. Wang Wang Notes Reference(s) leaf/SOM ! leaf/SOM " SOM C vs. VPDB -14.2-14.2-17.2-15.2 0.9989 1.0003 0.9986 0.9993 -1.1 0.3 -1.4 -0.7 3 soils 3 soils Wang Terwillinger 13 ! ‰ 4 4 4 4 Community leaf C 13 vs. VPDB -14.5-15.8 C C -13.1-14.5 C C ! ‰ 168 D.3. References Ambrose S & Sikes N (1991) Soil carbon isotope evidence for Holocene habitat change in the Kenya Rift Valley. Science 253:1402. Archer S (1990) Development and stability of grass/woody mosaics in a subtropical savanna parkland, Texas, USA. Journal of Biogeography 17:453. Balesdent J, Girardin C & Mariotti A (1993) Site-related 13C of tree leaves and soil organic matter in a temperate forest. Ecology 74:1713. Bayala J, et al. (2006) Relative contribution of trees and crops to soil carbon content in a parkland system in Burkina Faso using variations in natural 13C abundance. Nutrient Cycling in Agroecosystems 76:193. Bi X, et al. (2005) Molecular and carbon and hydrogen isotopic composition of n-alkanes in plant leaf waxes. Organic Geochemistry 36:1405. Biggs T, Quade J & Webb R (2002) δ13C values of soil organic matter in semiarid grassland with mesquite (Prosopis) encroachment in southeastern Arizona. Geoderma 110:109. Bintanja R & van de Wal R (2008) North American ice-sheet dynamics and the onset of 100,000-year glacial cycles. Nature 454:869. Boutton T, Archer S & Midwood A (1999) Stable isotopes in ecosystem science: structure, function and dynamics of a subtropical savanna. Rapid Communications in Mass Spectrometry 13:1263. Boutton T, et al. (1998) δ13C values of soil organic carbon and their use in documenting vegetation change in a subtropical savanna ecosystem. Geoderma 82:5. Bowman D & Cook G (2002) Can stable carbon isotopes (δ13C) in soil carbon be used to describe the dynamics of Eucalyptus savanna-rainforest boundaries in the Australian monsoon tropics? Austral Ecology 27:94. Buchmann N & Schulze E (1999) Net CO2 and H2O fluxes of terrestrial ecosystems. Global Biogeochemical Cycles 13:751. Desjardins T, et al. (1996) Changes of the forest-savanna boundary in Brazilian Amazonia during the Holocene revealed by stable isotope ratios of soil organic carbon. Oecologia 108:749. Diefendorf A, et al. (2010) Global patterns in leaf 13C discrimination and implications for studies of past and future climate. Proceedings of the National Academy of Sciences 107:5738. 169 Diels J, et al. (2001) Temporal variations in plant δ13C values and implications for using the 13C technique in long-term soil organic matter studies. Soil Biology and Biochemistry 33:1245. Freier K, Glaser B & Zech W (2010) Mathematical modeling of soil carbon turnover in natural Podocarpus forest and Eucalyptus plantation in Ethiopia using compound specific δ13C analysis. Global Change Biology 16:1487. Guillaume K, et al. (1999) Soil organic matter dynamics in tiger bush (Niamey, Niger). Preliminary results. Acta Oecologica 20:185. Guillaume K, et al. (2001) Does the timing of litter inputs determine natural abundance of 13C in soil organic matter? Insights from an African tiger bush ecosystem. Oecologia 127:295. Jessup K, Barnes P & Boutton T (2003) Vegetation dynamics in a Quercus-Juniperus savanna: an isotopic assessment. Journal of Vegetation Science 14:841. Krull E & Bray S (2005) Assessment of vegetation change and landscape variability by using stable carbon isotopes of soil organic matter. Australian Journal of Botany 53:651. Krull E, et al. (2005) Recent vegetation changes in central Queensland, Australia: evidence from δ13C and 14C analyses of soil organic matter. Geoderma 126:241. Krull E, et al. (2006) Compound-specific δ13C and δ2H analyses of plant and soil organic matter: a preliminary assessment of the effects of vegetation change on ecosystem hydrology. Soil Biology and Biochemistry 38:3211. Liao J, Boutton T & Jastrow J (2006) Organic matter turnover in soil physical fractions following woody plant invasion of grassland: evidence from natural 13C and 15N. Soil Biology and Biochemistry 38:3197. Millard P, et al. (2010) Quantifying the contribution of soil organic matter turnover to forest soil respiration, using natural abundance δ13C. Soil Biology and Biochemistry 42:935. Millard P, et al. (2008) Partitioning soil surface CO2 efflux into autotrophic and heterotrophic components, using natural gradients in soil δ13C in an undisturbed savannah soil. Soil Biology and Biochemistry 40:1575. Scartazza A, et al. (2004) Comparisons of δ13C of photosynthetic products and ecosystem respiratory CO2 and their responses to seasonal climate variability. Oecologia 140:340. Solomon D, et al. (2002) Soil organic matter dynamics in the subhumid agroecosystems of the Ethiopian highlands: evidence from natural 13C abundance and particle size-fractionation. Soil Science Soceity of America Journal 66:969. 170 Terwilliger V, et al. (2008) Reconstructing palaeoenvironment from δ13C and δ15N values of soil organic matter: a calibration from arid and wetter elevation transects in Ethiopia. Geoderma 147:197. Tiessen H, et al. (2003) Organic matter transformations and soil fertility in a treed pasture in semiarid NE Brazil. Plant and Soil 252:195. Wang G, et al. (2008) Paleovegetation reconstruction using 13C of soil organic matter. Biogeosciences Discussions 5:1795. Wang L, et al. (2009) Spatial heterogeneity and sources of soil carbon in southern African savannas. Geoderma 149:402. Wang L, et al. (2010) Patterns and implications of plant-soil δ13C and δ15N values in African savanna ecosystems. Quaternary Research 73:77. Wedin D, et al. (1995) Carbon isotope dynamics during grass decomposition and soil organic matter formation. Ecology 76:1383. 171 Appendix E. Data for δ13C values of leaf lipids vs. leaf tissues E.1. Dataset description 13 13 13 This is a compilation of δ C values for leaf tissues (δ Cleaf) and leaf lipids (δ C31) that includes 64 plant species from subtropical and tropical geographic sites (Schulze et al. 1996, Krull et al. 2005, Krull et al. 2006, Rommerskirchen et al. 2006, Vogts et al. 2009). Data derive from sites with mixed C3/C4 ecosystems and mean annual precipitation of 250 to 5000 mm. Means were calculated for multiple specimens of the same plant species for each site to remove internal variability (Diefendorf et al. 2010). 172 E.2. Tables 13 Table E.1. Compilation of published carbon-isotope compositions for plant leaves (δ Cleaf) and 13 n-alkanes (hentriacontane, δ C31), and leaf–lipid apparent fractionation factors (εleaf/lipid) according to phylogeny, plant functional type and photosynthetic pathway. 173 2006 1996 et al. 2006 et al. et al. Krull Reference(s) Schulze Rommerskirchen 31/leaf ! 31/leaf " 31 nC lipid C 13 ! 29 -25.1-22.7 -26.5-24.9 -23.8 0.9877-22.9 -25.3 -12.3 0.9895-22.4 -24.3 -10.5 0.9884-21.4 -22.8 -11.6 0.9885-21.7 -21.9 -11.5 0.9899-25.0 -19.8 -10.1 0.9908-22.5 -23.2 0.9936 -9.2 -19.6 -23.4 0.9891 -6.4 -20.6 -20.2 -10.9 0.9881-19.4 -21.7 -11.9 0.9943-18.3 -20.9 0.9928 -5.7 -22.7 -20.1 0.9940 -7.2 -20.8 -24.6 0.9948 -6.0 -21.8 0.9878 -5.2 -12.2 0.9906 -9.4 leaf C 13 vs. VPDB nC -11.6 -14.4 -13.4 -13.9 -12.9 -12.8 -12.8 -13.5 -12.4 -14.6 -14.6 -15.0 -15.0 -12.6 -12.5 ! ‰ Grass Grass Grass Grass Grass Grass Grass Grass Grass Grass Grass Grass 4 4 4 4 4 4 4 4 4 4 4 4 C C C C C C C C C C .C .C p p s s Iseilma Astrebla Astrebla Chloris virgata Chloris gayana Aristida congesta Stipagrostis ciliata Stipagrostis Eragrostis superba Eragrostis Eragrostis nindensis Eragrostis Aristida meridionalis Aristida adscensionisc Genus and/or SpeciesGenus and/or Photosynthesis PFT Stipagrostis hirtigluma Stipagrostis Enneapogon cenchroides Savanna Grasses 174 2006 1996 et al. et al. Reference(s) Schulze Rommerskirchen 31/leaf ! 31/leaf " 31 nC lipid C 13 ! 29 -19.6-22.0 -21.8-21.8 -22.9 0.9920-22.0 -22.4 0.9898 -8.0 -19.2 -22.1 -10.2 0.9907-20.4 -18.7 0.9896 -9.3 -18.9 -20.9 -10.4 0.9930-21.9 -19.5 0.9901 -7.0 -22.0 -21.9 0.9917 -9.9 -20.8 -24.0 0.9900 -8.3 -24.2 -19.4 -10.0 0.9889-21.3 -24.5 -11.1 0.9933 -19.6 0.9881 -6.7 -11.9 0.9919 -8.1 leaf C 13 vs. VPDB nC -11.8 -11.8 -11.1 -11.3 -11.6 -13.9 -12.8 -13.2 -12.0 -13.0 -12.8 -12.8 ! ‰ Grass Grass Grass Grass Grass Grass Grass Grass Grass Grass Grass 4 4 4 4 4 4 4 4 4 4 4 C C C C C C C C C C C Themeda triandra Eragrostis viscosa Eragrostis Panicum maximum Sporobolus ioclados Sporobolus Digitaria milanjiana Panicum arbusculumc Genus and/or SpeciesGenus and/or Photosynthesis PFT Bothriochloa insculpta Brachiaria erucitormis Sporobolus pyramidalis Sporobolus Schmidtia kalahariensis Hyparrhenia filipendula Hyparrhenia 175 2009 2009 et al. et al. Reference(s) Vogts Vogts Vogts 31/leaf ! 31/leaf " 31 nC lipid C 13 ! 29 leaf C 13 vs. VPDB nC -27.2-28.4 -36.1 -35.7 -35.4 -37.0 0.9916 0.9911 -8.4 -8.9 ! ‰ HerbHerb -26.5Herb -26.7 -33.6Herb -27.5 -32.2Herb -34.4 -26.8 -36.0 -34.2 0.9919 -26.4 -34.6 -36.2 0.9923 -8.1 -33.7 -36.7 0.9911 -7.7 -34.8 0.9898 -8.9 -10.2 0.9914 -8.6 ShrubShrub -27.2Shrub -24.9 -32.6Shrub -27.1 -32.3 -32.8Shrub -28.3 -34.0 -33.9 0.9942Shrub -28.5 -33.8 -34.6 0.9908 -5.8 -33.8 -35.1Shrub 0.9923 -9.2 -32.5Shrub 0.9930 -7.7 -25.4 0.9959 -7.0 -27.2 -33.9 -4.1 -36.3 -33.1 -35.5 0.9921 0.9915 -7.9 -8.5 3 3 3 3 3 3 3 3 3 3 3 3 3 C C C C C C C C C C C C C sp. Heliotropium Heliotropium Acacia mellifera Acacia nebrownii Salvadora persica Achyrantes aspera Ipomoea cordofana Abutilon graveolens Dichrostachis cinera Dichrostachis Xanthium brasilicum Grewia flavescens Grewia Genus and/or SpeciesGenus and/or Photosynthesis PFT Gymnosporia buxifolia Catophractes alexandri Crassocephalum mannii Savanna Herbs Savanna Shrubs 176 . 2009 et al Reference(s) Vogts Vogts 31/leaf ! 31/leaf " 31 nC lipid C 13 ! 29 leaf C 13 vs. VPDB nC -24.2-28.3 -30.1 -31.9 -30.6 -32.3 0.9934 0.9959 -6.6 -4.1 ! ‰ TreeTree -29.1Tree -27.7 -34.6Tree -28.0 -32.2Tree -36.3 -27.4 -35.5Tree -33.7 0.9926 -26.2 -31.3Tree -37.1 0.9938 -7.4 -27.1 -33.6Tree -30.1 0.9906 -6.2 -27.0 -32.3Tree -33.0 0.9972 -9.4 -26.3 -36.0Tree -34.9 0.9930 -2.8 -26.3 -31.5 -36.0 0.9920Tree -7.0 -28.3 -34.3 -31.8 0.9908 -8.0 Tree -30.9 0.9944 -9.2 Tree 0.9953 -5.6 -25.2 -36.0Tree -4.7 -26.0 0.9921 -26.8 -32.1 -7.9 -34.7 -30.9 -33.1 -30.1 0.9903 0.9927 -9.7 0.9966 -7.3 -3.4 3 3 3 3 3 3 3 3 3 3 3 3 3 3 C C C C C C C C C C C C C C Acacia kirkii Acacia karroo Boscia foetida Acacia senegal Acacia erioloba Burkea africana Acacia mellifera Acacia luederitzii Adansonia digitata Combretum imberbe Combretum Balanites aegyptiaca Genus and/or SpeciesGenus and/or Photosynthesis PFT Combretum apiculatum Combretum Commiphora glaucescens Colophospermum mopane Savanna Trees Savanna 177 2006 . 2009 et al et al. Krull Reference(s) Vogts Vogts 31/leaf ! 31/leaf " 31 nC lipid C 13 ! 29 -35.8-35.6 -36.0 -31.9 0.9918 0.9946 -8.2 -5.4 leaf C 13 vs. VPDB nC -28.0 -26.6 ! ‰ TreeTree -28.4Tree -24.6 -37.7Tree -27.6Tree -37.7 -28.0 -36.9Tree 0.9904 -28.1 -31.0 -35.5Tree -33.6 -9.6 -28.8 -31.7Tree -37.4 0.9934 0.9938 -24.0 -36.3Tree -31.7 -6.6 0.9903 -6.2 -25.7 -28.1Tree -37.1 0.9963 -9.7 -25.7 -33.0Tree -27.4 0.9915 -3.7 -26.8 -29.9Tree -36.2 0.9965 -8.5 -32.3 -33.8 0.9892 -3.5 -35.0 -10.8 0.9917 0.9916 -8.3 -8.4 3 3 3 3 3 3 3 3 3 3 3 3 C C C C C C C C C C C C sp. sp. Acacia Ataliya Morus alba Ziziphus spina Sesbania sesban Tamarix usneoides Tamarix Ziziphus mucronata Euclea pseudebenus Commiphora merkeri Terminalia prunioides Terminalia Genus and/or SpeciesGenus and/or Photosynthesis PFT Leucaena leucocephala Sclerocarya birrea caffra birrea Sclerocarya 178 E.3. References Diefendorf A, et al. (2010) Global patterns in leaf 13C discrimination and implications for studies of past and future climate. Proceedings of the National Academy of Sciences 107:5738. Krull E, et al. (2005) Recent vegetation changes in central Queensland, Australia: evidence from δ13C and 14C analyses of soil organic matter. Geoderma 126:241. Krull E, et al. (2006) Compound-specific δ13C and δ2H analyses of plant and soil organic matter: a preliminary assessment of the effects of vegetation change on ecosystem hydrology. Soil Biology and Biochemistry 38:3211. Rommerskirchen F, et al. (2006) Chemotaxonomic significance of distribution and stable carbon isotopic composition of long-chain alkanes and alkan-1-ols in C4 grass waxes. Organic Geochemistry 37:1303. Schulze E, et al. (1996) Diversity, metabolic types and δ13C carbon isotope ratios in the grass flora of Namibia in relation to growth form, precipitation and habitat conditions. Oecologia 106:352. Vogts A, et al. (2009) Distribution patterns and stable carbon isotopic composition of alkanes and alkan-1-ols from plant waxes of African rain forest and savanna C3 species. Organic Geochemistry 40:1037. 179 Appendix F. Review of published leaf-lipid δD data F.1. Dataset description We compile and review published leaf-lipid δD data to form a compilation (n = 159) of 76 plant species from 31 sites (Chikaraishi & Naraoka 2003, Bi et al. 2005, Chikaraishi & Naraoaka 2006, Krull et al. 2006, Liu et al. 2006, Sachse et al. 2006, Smith & Freeman 2006, Hou et al. 2007, Hou et al. 2008, Liu & Huang 2008, Mügler et al. 2008, Sachse et al. 2009, Feakins & Sessions 2010). Sites are located within 5 biomes and have mean annual precipitations ranging from 183 mm to more than 1200 mm. Leaf-lipid data derive from living subtropical and tropical plants belonging either to the monocot clade or the two largest dicot clades (rosid and asterid) in ecosystems with discontinuous woody cover and herbaceous understory. We calculate species means for individual sites to remove within-species variability (Diefendorf et al. 2010). We use modeled annual δDrain values to calculate representative εlipid/water values because measured δDrain values rarely accompany published leaf-lipid δD values. 180 F.2. Tables 2 Table F.1. Compilation of published hydrogen-isotope compositions for plant lipids (δ Hlipid) and modeled annual precipitation (Bowen & Revenaugh 2003), and precipitation–lipid apparent fractionation factors (εprecip/lipid) according to phylogeny, plant functional type and photosynthetic pathway. 181 2005 2005 et al. et al. E Bi E Bi ! ! N, 113°20 N, 113°20 ! ! Location Reference 69°06 69°06 OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -95 -38 -59 -199-168-159-174 -38-171 -38-158 -38 -167 -162 -38 -135 -171 -38 -126 -205 -38 -141 -191 -38 -138 -38 -125 -201 -38 -129 -213 -38 -138 -216 -174 -210 -38 -159 -38-157 -38 -169 -185 -38 -182 -143 -185 -130 -38 -179 -207 -38 -38 -124 -38 -153 -38 -109 -96 -176 C vs. VSMOW n ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 Tree C Tree C TreeTree C C Tree C Tree C Herb C Herb C Herb C Shrub C Shrub C Shrub C Shrub C ShrubShrub C C GraminoidGraminoidGraminoid C Graminoid C Graminoid C C C Zea mays Caryota mitis Ficus altissima Zoysia japonica Kigelia africana Syzygium cumini Ficus microcarpa Saccharum sinense Imperata cylindrica Swietenia mahagoni Osmanthus fragrans Alternanthera dentata Codiaeum variegatum Euphorbia pulcherrima Genera and/or SpeciesGenera and/or Habit Photosynthesis Holmskioldia sanguinea Cinnamomum burmanni Araucaria cunninghamii Alternanthera versicolor Bothriochloa ischaemum Alternanthera bettzickiana 182 2006 et al. Sachse Smith & Freeman 2006 E E E E E E E E E " " " " " " " " !! W W W W 49 06 35 46 55 44 22 09 29 ! ! ! ! W ! ! ! ! ! ! ! ! ! ! N, 11°46 N, 10°18 N, 27°10 N, 26°51 N, 24°16 N, 25°05 N, 06°52 N, 10°55 N, 25°08 N, 99°18 ! " " N, 104°47 " " " " " " N, 100°55 N, 101°52 N, 105°16 " ! ! ! ! 46 00 23 20 55 36 10 14 37 Location Reference ! ! ! ! ! ! ! ! ! 47°09 40°42 46°50 43°58 40°00 42°36 61°11 43°50 67°21 67°10 61°50 61°03 50°07 42°45 OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -211 -99 -124 -193-166-186 -104-226 -104-162-168 -93 -99 -161 -91 -69 -148 -91 -103 -136 -60 -149 -142 -42 -78 -42 -115 -234 -39 -124 -224 -46 -111 -232 -101 -117-230 -101 -117-213 -113-209 -133 -113-236 -121 -113-214 -134 -219 -99 -132 -99 -113 -235 -99 -122 -266 -99 -152 -128 -95 -133 -95 -155 -189 C vs. VSMOW n ‰ 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 TreeTree C C Tree C Tree C TreeTreeTree C C Tree C Tree C C GraminoidGraminoid C GraminoidGraminoid C Graminoid C Graminoid C Graminoid C C C Graminoid C Graminoid C Graminoid C Graminoid C Graminoid C sp. Shrub C Myrtus Quercus robur Quercus Quercus cerris Quercus Betula pendula Betula pendula Betula pendula Betula pendula Betula pendula Carpinus betulus Aristida longiseta Quercus variabilis Quercus Bouteloua gracilis Bouteloua gracilis Buchloe dactyloides Andropogon gerardii Andropogon gerardii Andropogon Calamovilfa longifolia Genera and/or SpeciesGenera and/or Habit Photosynthesis Bouteloua curtipendula Schizachyrium scoparium Schizachyrium scoparium Schizachyrium scoparium Schizachyrium scoparium 183 Smith & Freeman 2006 E Chikaraishi & Naraoka 2003 W ! ! W W ! ! N, 99°20 N, 96°35 N, 138°52 ! ! N, 106°53 ! ! Location Reference 38°52 39°12 36°28 34°05 OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -111-111 -49 -49 -65 -65 -116 -49 -70 -204-192-207-213 -70-197 -70-205 -70 -144 -155 -73 -131 -210 -73 -147 -73 -151 -156 -68 -134 -142 -68 -142 -180 -93 -159 -49 -152 -168 -49-185 -49 -113 -128 -49 -98 -168 -49 -138 -49 -116 -49 -125 -176 -49 -143 -183 -83 -127 -125 -49 -49 -49 -134 -141 -82 C vs. VSMOW n ‰ 4 4 4 4 4 4 4 4 4 3 4 3 4 4 3 3 3 3 3 3 3 3 Tree C Tree C Tree C Tree C Tree C Tree C Shrub C Shrub C Shrub C Shrub C GraminoidGraminoidGraminoid C Graminoid C C Graminoid C C GraminoidGraminoid C Graminoid C Graminoid C C Graminoid C Graminoid C Graminoid C Zea mays Acer argutum Acer argutum Cornus kousa Cornus kousa Zoysia japonica Prunus serrulata Albizia julibrissin Acer carpinifolium Acer carpinifolium Panicum virgatum Bouteloua gracilis Camellia sasanqua Quercus acutissima Quercus Miscanthus sinensis Bouteloua eriopoda Sorghastrum nutans Sorghastrum nutans Sorghastrum Andropogon gerardii Andropogon Andropogon gerardii Andropogon Saccharum officinarum Genera and/or SpeciesGenera and/or Habit Photosynthesis Schizachyrium scoparium 184 2007 et al. E Chikaraishi & Naraoka 2003 ! W Hou ! N, 71°58 N, 138°52 ! ! Location Reference 42°05 36°28 OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -155-176-163-179 -49-179 -49-172 -49 -111 -184 -49 -134 -133 -49 -120 -176 -49 -137 -176 -49 -137 -49 -129 -177 -49 -142 -174 -49 -88 -182 -134 -176 -63 -134 -180 -63-180 -63 -122 -198 -63 -118 -213 -63 -127 -210 -63 -121 -178 -63 -125 -172 -63 -125 -175 -63 -144 -63 -160 -63 -157 -63 -123 -116 -120 C vs. VSMOW n ‰ 3 3 3 4 3 3 3 3 4 4 3 3 3 3 3 3 3 3 3 3 3 3 Tree C TreeTreeTree C Tree C Tree C C C Tree C TreeTreeTree C Tree C C C Tree C Herb C HerbHerb C C Shrub C Shrub C Graminoid C GraminoidGraminoid C C sp.sp. Tree Tree C C Carya Carya Betula lenta Acer rubrum Acer rubrum Acer rubrum Acer rubrum Acer palmatum Quercus dentata Quercus Sorghum bicolor Sorghum Plantago asiatica Betula populifolia Betula populifolia Betula populifolia Betula populifolia Betula populifolia Manihot utilissima Artemisia princeps Quercus mongolica Quercus Miscanthus sinensis Taraxacum officinale Taraxacum Saccharum officinarum Genera and/or SpeciesGenera and/or Habit Photosynthesis 185 2007 et al. E Chikaraishi & Naraoka 2006 ! W Hou ! N, 71°58 N, 138°52 ! ! Location Reference 42°05 36°28 OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -111 -49 -65 -116 -49 -70 -185-203-166-150 -63-157 -63-162 -63 -130 -177 -63 -149 -172 -63 -110 -161 -63 -93 -179 -63 -100 -180 -63 -106 -182 -63 -122 -149 -63 -116 -181 -63 -105 -156 -63 -124 -208 -63 -125 -190 -63 -127 -168 -63 -92 -189 -63 -126 -63 -99 -63 -155 -63 -136 -157 -112 -134 -49 -114 C vs. VSMOW n ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 TreeTreeTree C Tree C Tree C C C HerbHerb C Herb C C HerbHerb C Herb C C Shrub C ShrubShrubShrub C Shrub C C C ShrubShrubShrub C C C Shrub C sp. Tree C sp. Tree C Carya Carya Acer argutum Acer argutum Acer argutum Typha latifolia Typha Quercus rubra Quercus rubra Quercus Daucus carota Prunus serotina Lindera benzoin Mimulus ringens Lonicera tatarica Aster Divaricatus Plantago asiatica Trifolium pratense Trifolium Fraxinus americana Fraxinus americana Fraxinus americana Decodon verticillatus Prunus allegheniensis Hamamelis virginiana Genera and/or SpeciesGenera and/or Habit Photosynthesis 186 2006 et al. E Liu E Chikaraishi & Naraoka 2006 ! ! N, 109°29 N, 138°52 ! ! Location Reference 36°36 36°28 OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -111 -49 -65 -113 -48 -68 -127-163-133 -49-186 -49-148 -82 -48 -120 -183 -48-127 -48 -89 -154 -145 -158 -48 -105 -151 -48-143 -48 -142 -188 -48 -83 -200 -48 -111 -173 -48 -116 -206 -48 -108 -174 -48 -100 -134 -48 -147 -131 -48 -160 -153 -48 -131 -48 -156 -48 -132 -48 -90 -87 -110 C vs. VSMOW n ‰ 3 3 3 3 4 3 3 3 4 3 3 4 3 4 4 4 3 4 3 3 3 Tree C Tree C Tree C HerbHerb C C HerbHerb C Herb C C Herb C Shrub C Shrub C Shrub C Graminoid C Graminoid C GraminoidGraminoid C GraminoidGraminoid C C Graminoid C C sp. Shrub C sp. Shrub C Oxytropis Oxytropis Oxytropis Vitex negundo Vitex Sophora velutina Heteropappus less Heteropappus Peganum harmala Lespedeza daurica Acer carpinifolium Acer carpinifolium Acer carpinifolium Artemisia scoparia Asparagus officinalis Pennisetum flaccidum Pennisetum flaccidum Caragana stenophylla Cleistogenes squarrosa Cleistogenes squarrosa Genera and/or SpeciesGenera and/or Habit Photosynthesis Dichanthium ischaemum Dichanthium ischaemum Dichanthium ischaemum Dracocephalum moldavica 187 Feakins & Sessions 2010 E E E E ! ! ! ! N, 118°10 N, 118°35 N, 115°40 N, 116°47 ! ! ! ! Location Reference 34°14 34°05 34°48 33°47 OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -90 -66 -26 -111 -83 -31 -128-150-221-120 -83-153 -83-163 -83 -49 -210 -83 -73 -83 -150 -151 -83 -40 -146 -83 -76 -134 -87 -129 -83 -138 -159 -83-102 -83 -74 -159 -83 -69 -127 -70 -56 -124 -66 -50 -142 -66 -96 -136 -66 -39 -137 -66 -100 -121 -66 -65 -66 -62 -126 -66 -81 -66 -75 -76 -66 -59 -64 C vs. VSMOW n ‰ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Tree C Herb C Herb C Herb C Shrub C Shrub C ShrubShrub C Shrub C C ShrubShrubShrub C C C ShrubShrubShrub C Shrub C C C ShrubShrubShrub C Shrub C C Shrub C C ShrubShrub C C Larrea tridentata Larrea Epilobium canum Pinus lambertiana Phacelia cicutaria Eriogonum wrightii Keckiella cordifolia Quercus chrysolepis Quercus chrysolepis Quercus chrysolepis Quercus chrysolepis Quercus Ceanothus spinosus Artemisia californica Artemisia ludoviciana Genera and/or SpeciesGenera and/or Habit Photosynthesis Arctostaphylos pringlei Arctostaphylos pringlei Arctostaphylos Ceanothus leucodermis Ceanothus megacarpus Ceanothus megacarpus Coleogyne ramosissima Ceanothus integerrimus Adenostoma fasciculatum Arctostaphylos glandulosa Arctostaphylos Malacothamnus fasciculatus 188 2006 et al. E Feakins & Sessions 2010 E Krull ! ! S, 143°57 N, 118°35 ! ! Greenhouse Liu & Huang 2008 Location Reference 23°13 34°05 OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -93 -23 -72 -148-199 -66-154-195 -88 -174 -23-176 -23-169 -23 -180 -151 -23 -134 -23 -176 -140 -23 -155 -159 -23 -157 -127 -149 -137 -38 -131 -158 -38-182 -38 -106 -168 -38 -126 -166 -38 -93 -152 -38 -103 -168 -38 -125 -142 -38 -150 -141 -38 -135 -38 -133 -38 -119 -38 -135 -108 -107 C vs. VSMOW n ‰ 3 3 4 3 4 4 4 4 4 3 3 3 3 3 3 4 3 3 3 3 3 Tree C Tree C Tree C HerbHerb C C HerbHerb C Herb C C HerbHerb C C Shrub C Shrub C Graminoid C sp. Graminoid C sp. Graminoid C sp. Graminoid C sp. Graminoid C sp. Graminoid C sp. Graminoid C sp. Shrub C sp. Tree C Zea mays Acacia Atalya Astrebla Astrebla Iseilema Aristida Panicum Glycine max Glycine max Acer rubrum Acer rubrum Beta vulgaris Sporobolus Sporobolus Enneapogon Trifolium repens Trifolium Helianthus annuus Helianthus annuus Phaseolus vulgaris Platanus racemosa Gossypium hirsutum Gossypium hirsutum Genera and/or SpeciesGenera and/or Habit Photosynthesis 189 Greenhouse Liu & Huang 2008 Location Reference OIPC/31 ! OIPC vs. VSMOW ‰ H 2 ! 31 -170-183-168 -38 -38 -38 -137 -151 -135 C vs. VSMOW n ‰ 3 3 4 Herb C Herb C Graminoid C Zea mays Trifolium repens Trifolium Phaseolus vulgaris Genera and/or SpeciesGenera and/or Habit Photosynthesis 190 F.3. References Bi X, et al. (2005) Molecular and carbon and hydrogen isotopic composition of n-alkanes in plant leaf waxes. Organic Geochemistry 36:1405. Bowen G & Revenaugh J (2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39:1299. Chikaraishi Y & Naraoka H (2003) Compound-specific δD-δ13C analyses of n-alkanes extracted from terrestrial and aquatic plants. Phytochemistry 63:361. Chikaraishi Y & Naraoka H (2006) Carbon and hydrogen isotope variation of plant biomarkers in a plant-soil system. Chemical Geology 231:190. Feakins S & Sessions A (2010) Controls on the D/H ratios of plant leaf waxes in an arid ecosystem. Geochimica et Cosmochimica Acta 74:2128. Hou J, D'Andrea W & Huang Y (2008) Can sedimentary leaf waxes record D/H ratios of continental precipitation? Field, model, and experimental assessments. Geochimica et Cosmochimica Acta 72:3503. Hou J, et al. (2007) Hydrogen isotopic variability in leaf waxes among terrestrial and aquatic plants around Blood Pond, Massachusetts (USA). Organic Geochemistry 38:977. Krull E, et al. (2006) Compound-specific δ13C and δ2H analyses of plant and soil organic matter: a preliminary assessment of the effects of vegetation change on ecosystem hydrology. Soil Biology and Biochemistry 38:3211. Liu W, Yang H & Li L (2006) Hydrogen isotopic compositions of n-alkanes from terrestrial plants correlate with their ecological life forms. Oecologia 150:330. Liu Z & Huang Y (2008) Hydrogen isotopic compositions of plant leaf lipids are unaffected by a twofold pCO2 change in growth chambers. Organic Geochemistry 39:478. Sachse D, Kahmen A & Gleixner G (2009) Significant seasonal variation in the hydrogen isotopic composition of leaf-wax lipids for two deciduous tree ecosystems (Fagus sylvativa and Acer pseudoplatanus). Organic Geochemistry 40:732. Sachse D, Radke J & Gleixner G (2006) δD values of individual n-alkanes from terrestrial plants along a climatic gradient: implications for the sedimentary biomarker record. Organic Geochemistry 37:469. 191 Smith F & Freeman K (2006) Influence of physiology and climate on δD of leaf wax n-alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70:1172. 192 Appendix G. Measured δ13C and δD values for leaf lipids from modern soils near Olduvai Gorge G.1. Dataset description 13 We present δ C and δD values for n-alkanes in modern soils from Olduvai Gorge (2° 59′S, 35° 06′E), Kisima Ngeda (3° 28′S, 35° 21′E), Ukamako River at Hadzabe (3° 52′S, 35° 01′E) and Lake Manyara (3° 22′S, 35° 49′E). Modern soils were collected by Doris Barboni in 2011. These samples serve as modern analogues for depositional environments across FLK Zinj during the early Pleistocene. 193 G.2. Tables 13 Table G.1. Carbon and hydrogen-isotope compositions for hentriacontane (δ C31 and δD31, respectively) from modern soils near Olduvai Gorge. 194 ) ) ) Hyphaene Euphorbia ) and ) ) ) ) ) and Commiphora ) and grasses) ) ) and Acacia Acacia Acacia Acacia Acacia Hyphaene Acacia Acacia Field Notes Cyperus Acacia Cyperus Adansonia Acacia and and Sesbania Typha 31 C 13 vs. VPDB ! ‰ 31 D ! vs. VSMOW ‰ DB11-75 Lake Manyara -131.0 -33.0 Groundwater woodland ( DB11-58 Kisima Ngeda -132.6 -31.6DB11-68DB11-70 Swamp ( DB11-71 Hadzabe Lake Manyara Lake Manyara -125.1 -134.4 -128.6 -35.0 -30.8 -34.1 Groundwater woodland ( Playa of seasonal lake Groundwater woodland ( DB11-55DB11-56 Kisima NgedaDB11-57 Kisima Ngeda -141.0 Kisima Ngeda -134.0 -30.8 -145.1 -31.5 Groundwater woodland ( -30.2 Groundwater woodland ( Groundwater woodland ( DB11-61DB11-62 Ukumako RiverDB11-64 Ukumako River -144.3 -146.1 Hadzabe -30.5 -28.3 Spring -143.2 Spring -29.3 grassland ( Wooded DB11-36DB11-54 Olduvai Kisima Ngeda -150.5 -140.3 -27.9 -30.4 Groundwater woodland ( Bed of seasonal river ( DB11-35 Olduvai -150.9 -28.7 Fringe of seasonal river ( DB11-34 Olduvai -149.4 -27.2 grassland ( Wooded DB11-59A Kisima Ngeda -161.8 -24.5 Swamp ( Sample Identification Locality 195 Appendix H. Measured δ13C and δD values for leaf lipids from FLK Zinjanthropus archaeological Level 22 H.1. Dataset description We present δ13C and δD values for n-alkanes in paleosols from across the FLK Zinjanthropus archaeological Level 22 (FLK Zinj). Paleosols were collected by Gail Ashley. 196 H.2. Tables 13 Table H.1. Carbon and hydrogen-isotope compositions for hentriacontane (δ C31 and δD31, respectively) from paleosols across FLK Zinj. Archaeological localities are also listed for reference. 197 31 C 13 vs. VPDB ! ‰ 31 D ! vs. VSMOW ‰ NW -113 -27.8 NNENNE -133 -136 -23.5 -22.7 NNN -108 28.3 NN-1 -127NN-1 -26.3 -125 -27.1 NN-9NN-9NN-9 -114NN-9 -107NN-9 -102 -27.7 -102 -28.6 -109 28.9 29.2 27.5 NNW -133 -23.9 NN-11NN-11NN-11 -134 -125 -128 -25.8 -26.6 -26.7 NN-12NN-12 -138 -123 -24.5 -27.0 NN-10 -100 -30.4 GA-29-11 GA-28-11 GA-33-11 GA-30-11 GA-32-11 GA-62-11 GA-77-11 GA-81-11 GA-81-11 GA-27-08 GA-58-08 GA-70-09 GA-23-08 GA-92-09 GA-103-11 GA-106-11 GA-102-11 GA-102A-11 Sample Identification Locality 31 C 13 vs. VPDB ! ‰ 31 D ! vs. VSMOW ‰ LCSLCS -20 -19.9 -150 -155 LLK -19.3 -155 FLK-4FLK-4 -116FLK-5 -117 -33.3 -118FLK-3 -33.5 FLK-3 -33.9 -115 -121FLK-3 -33.6 FLK-3 -35.2 -114 -119 -33.9 -34.3 NN-10NN-10 -104NN-10 -120 -29.9 -110 -27.1 -28.5 Corner -118Corner -32.3 -118 -34 Junction -19.4 -150 GA-59-11 GA-60-11 GA-69-11 GA-68-11 GA-65-11 GA-66-11 GA-67-11 GA-61-11 GA-16-08 GA-13-09 GA-12-08 GA-110-11 GA-108-11 GA-114-07 GA-101-11 GA-14B-09 Sample Identification Locality 198 Appendix I. Measured δ13C values for n-alkanes and 5-n-alkylresorcinols in selected human dietary resources I.1. Dataset description We present δ13C values for n-alkanes and 5-n-alkylresorcinols in selected grains, fruits and plants. 199 I.2. Tables 13 13 Table I.1. Carbon-isotope compositions for heptacosane (δ C27), nonacosane (δ C29), 13 13 13 hentriacontane (δ C31), tritriacontane (δ C33), 5-nC19-alkylresorcinol (δ CAR19), 5-nC21- 13 13 alkylresorcinol (δ CAR21), and 5-nC23-alkylresorcinol (δ CAR23) in selected human dietary resources. Sampled plant materials are listed under Type. 200 AR23 C 13 ! AR21 C 13 ! AR19 C 13 ! 33 C 13 vs. VPDB ! ‰ 31 C 13 vs. VPDB ! ‰ 29 C 13 vs. VPDB ! ‰ 27 C 13 vs. VPDB ! ‰ Peel -35.8 -34.9 -35.2 -36.0 -28.3 Stem -22.1 -22.7 -23.6 -23.0 -21.3 -21.5 -21.6 Grain -35.3GrainGrain -35.5 -36.6 -33.8 -36.3 -35.5 -33.9 -35.9 -34.1 -29.4 -34.2 -30.2 -28.7 -27.6 -28.8 -28.0 -30.4 -29.1 -28.0 GrainGrain -21.3 -21.4 -21.5 -34.2 -34.5 -35.3 -27.8 -27.8 -27.9 Kernel -17.8 -19.3 -19.9 -20.1 everta Secale cereale Zea mays Triticum durum Triticum Cyperus papyrus Sorghum bicolor Mangifera indica Horedum vulgare Triticum aestivum Triticum Sample Identification Type 201 CURRICULUM VITAE CLAYTON R MAGILL EDUCATION The Pennsylvania State University University Park, PA 2013 Dual PhD, Geosciences and Biogeochemistry Primary Advisor: Katherine H Freeman Dissertation: Biomarker and isotopic perspectives on early Pleistocene climate in East Africa. University of Cambridge Cambridge, UK 2007 MPhil, Quaternary Science Primary Advisor: Tamsin C O’Connell Thesis: Intraspecific δ18O values as an indication of seasonality in a Mesolithic shell midden. University of Pittsburgh Pittsburgh, PA 2005 BPhil, Chemistry Summa cum Laude Primary Advisor: Michael F Rosenmeier Thesis: Reconstruction of Holocene climate variability within the central Mediterranean using lake sediments from the Akrotiri Peninsula, Crete. RECENT COMMUNICATIONS Ecological insights from plant biomarkers preserved in archaeological horizons. Lamont- Doherty Earth Observatory Seminar Series, March 2013 (invited talk). Plant biomarker and isotopic perspectives on early human habitats at Olduvai Gorge. Society for American Archaeology Annual Meeting, April 2013 (invited talk). Prehistoric weather and the rise of agriculture. Pinnacle Charter School in conjunction with the Pennsylvania Department of Agriculture, October 2012 (invited talk). Plant lipid biomarker evidence for hominin habitat patterning at Olduvai Gorge. Lamont- Doherty Earth Observatory Climate-Evolution Symposium, July 2012 (invited poster). High-resolution habitat reconstruction of the FLK Zinjanthropus site, Olduvai Gorge, using plant biomarker and phytolith evidence. Paleoanthropology Society Annual Meeting, April 2012 (talk). SELECTED HONORS 2013 Cozzarelli Prize, Proceedings of the National Academy of Sciences 2012 ETH/Marie Curie Actions Postdoctoral Fellow, ETH Zürich 2009 American Geophysical Union Outstanding Paper, Paleoclimatology 2006 Winston Churchill Foundation Scholar 202