UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Sedimentary biogeochemistry and palaeoceanography of the South China Sea during the late Pleistocene Kienast, Markus 2002

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2002-750329.pdf [ 17.16MB ]
Metadata
JSON: 831-1.0052772.json
JSON-LD: 831-1.0052772-ld.json
RDF/XML (Pretty): 831-1.0052772-rdf.xml
RDF/JSON: 831-1.0052772-rdf.json
Turtle: 831-1.0052772-turtle.txt
N-Triples: 831-1.0052772-rdf-ntriples.txt
Original Record: 831-1.0052772-source.json
Full Text
831-1.0052772-fulltext.txt
Citation
831-1.0052772.ris

Full Text

Sedimentary Biogeochemistry and Palaeoceanography of the South China Sea during the late Pleistocene by Markus Kienast Diplom, Kiel University, 1997 A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy In The Faculty of Graduate Studies (Department Earth and Ocean Sciences) We accept this thesis as conforming to the required standard The University of British Columbia September 2002 © Markus Kienast, 2002 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2788) Abstract The South China Sea (SCS) is the largest marginal basin off the Asian continent and its surface hydrography as well as sedimentation within the basin are strongly influenced by the SE Asian monsoon and eustatic sea-level changes. The rapidly accumulating sediments in this basin are therefore a unique and sensitive monitor of past variations in monsoonal climate, ocean-continent linkages, deglacial sea-level, and marine biogeochemical processes and their relationship to climate change. This thesis contributes to our understanding of these key aspects of glacial-interglacial palaeoceanography by presenting multi-proxy organic and inorganic sediment geochemical records from a large number of gravity cores and surface sediment samples from throughout the SCS. Records of deglacial sea surface temperature (SST) and summer monsoon variability point to a close coupling with the climate of the circum-North Atlantic realm, whereas sedimentological changes associated with variations in winter monsoonal intensity suggest a concordant deglacial development with SST changes in the open equatorial Pacific. Together, these records demonstrate a complex interaction of Northern and Southern Hemisphere influences on the climate of this region. The biogeochemical cycles of nitrogen and carbon in the open SCS as reflected in their time-varying isotopic composition are not significantly affected by monsoonal climate or the specific sedimentological and geographical setting of this marginal basin but appear instead to reflect the regional marine biogeochemistry. On the other hand, variations in basin configuration related to eustatic sea level changes leave a distinct inprint in the rate and geochemical composition of terrigenous sediment supply to the SCS, which are used to constrain the deglacial history of sea level rise and its impact on near-shore sedimentation. A comparison of various geochemical and micropalaeontological methods to estimate palaeo-sea surface temperatures (SSTs) demonstrates a quantitative agreement between alkenone (UK ' 3 7), foraminiferal Mg/Ca, and foraminiferal tranfer function FP-12E SST estimates in recording an annual average cooling of 2-2.5 °C of the tropical southern SCS during the last glacial period. In contrast, the foraminiferal transfer functions R A M and SIMMAX show an annual average glacial cooling of only <1 °C or no cooling at all, respectively. Both the U K 3 7 and the FP-12E SST estimates, as well as the planktonic foraminiferal 5 l sO values, indicate an abrupt warming (ca. 1°C in <200 years) at the end of the last glaciation, which occurs synchronously (within dating uncertainties) with the Boiling transition as recorded in the Greenland Ice Sheet Project 2 (GISP2) ice core. i i The nitrogen isotopic composition (515N) of surface and down-core sediments spanning the last glacial-interglacial cycle from the entire South China Sea (SCS) has a narrow range (~3.0 to ~6.5%o) with no correlation with discernible palaeoclimatic/oceanographic changes. The absence of any correlation with reconstructed (glacial-interglacial) changes in primary production, terrigenous input, and/or sea level related basin configuration is attributed to the complete consumption of nitrate during primary production in this marginal basin during at least the last 140,000 years. This, in turn, implies that the 5 1 5N of the nitrate used during primary production remained approximately constant during the last climatic cycle. The proposed scenario infers an unchanged nitrogen isotopic composition of the subsurface nitrate in the western Pacific between glacial and interglacial stages as well as during terminations and thus constrains proposed changes in the oceanic N inventory. The carbon isotopic composition of organic matter (5 1 3C o r g) in sediment cores from throughout the open SCS covering the last 220 kyr shows higher values (around -19.5 to -20.5%o) during glacial stages, while lower values (around -21 to -22.5%o) are characteristic of interglacials. Following well established procedures, the S 1 3 C o r g records are converted to local pC02 estimates. Together with other low-latitude 6 1 3 C o r g -pC0 2 estimates from the literature, the results show that 5 1 3C of bulk sedimentary organic matter cannot be used to hindcast past changes in local C0 2( a q). Three crucial pitfalls are identified, namely unreasonable absolute pC0 2 estimates, a lack of correlation between 6 1 3 C o r g -pC0 2 estimates and sedimentary proxies of upwelling intensities, and unexplicable discrepancies in the temporal evolution of atmospheric pC0 2 as recorded in ice cores and marine sedimentary 6 1 3 C o r g -pC0 2 estimates, which caution the use of 5 1 3 C o r g as an unambiguous tracer of dissolved molecular C 0 2 in the surface ocean. This calls for a re-evaluation of the role of the low-latitude ocean on temporal changes in atmospheric C 0 2 . Based on the sedimentological and geochemical variability at core sites along a transect across the outer Sunda Shelf and the continental slope covering the last 20 thousand years, four intervals of significant depositional changes are identified, which closely correlate with environmental shifts on the central shelf. Thus, sedimentation on the southern SCS margin and slope appears to be mainly controlled by the interrelationship between sea level, shelf palaeo-physiography and sediment supply. Because of these complex interactions, the interpretation of nearshore sedimentary records as unequivocal recorders of local climate change (e.g., SE Asian monsoon) is not i i i straightforward. Variations of shelf physiography and sea level need to be included in future palaeoceanographic studies. Major element variations in a core from the northern SCS are used to infer downcore changes in bulk sediment composition, which are interpreted in terms of deglacial changes in monsoonal climate. Thus, S i / A l reveals a two-step deglacial weakening of winter monsoonal winds, possibly linked to the temporal development of equatorial Pacific SSTs. In contrast, the reconstruction of changes in summer monsoonal precipitation, and thus river runoff, based on major element/Al ratios (and grain size indices) is compromised by the antagonistic effects of the deglacial retrogression of the river mouths and changes in fine-grained fluvial sediment delivery due to increasing summer monsoonal precipitation. Nevertheless, an early Holocene peak in summer monsoon intensity can be inferred from a marked clay maximum immediately following the Younger Dryas. Finally, the rapid drop in the supply of terrigenous organic matter to the open SCS also corresponds with a rapid increase in sea-surface temperature during the last deglaciation, corresponding with the Boiling warming at 14.7 kyr B.P. This reflects a rapid retrogression of local rivers due to rapid sea-level rise, strongly implying that the Boiling warming and the onset of melt-water pulse (MWP) l a are synchronous. This phase relation contrasts with the widely cited onset of this melt-water pulse l a at ca. 14 kyr B.P. iv Table of contents Abstract i i List of tables v i i i List of figures ix Preface xi i Acknowledgements xiv 1. General introduction 1.1 Opening remarks 1 1.2 The oceanography of the South China Sea 2 1.3 Sedimentation in the South China Sea 4 1.4 Previous palaeoceanographic studies in the South China Sea 5 1.5 Objectives of this study 6 1.6 References 9 2. Sea surface temperature reconstructions 2.1 Introduction 14 2.2 The amplitude of glacial cooling in the SCS: A comparison of different methods to estimate palaeo-SSTs 17 2.2.1 Alkenone SST estimates of core 18287-3 17 2.2.2 Planktonic foraminiferal SST estimates of core 18287-3 19 2.2.3. Foraminiferal M g / Ca SST estimates from core 18287-3 21 2.2.4 A comparison 23 2.3 The timing of the Boiling Warming in the southern South China Sea 25 2.4 The deglacial pattern of SST change in the SCS 29 2.5 Conclusions 30 2.6 References 33 v 3 . Sedimentary biogeochemistry 3.1 Introductory remarks 38 3.2 Nitrogen isotope records 3.2.1 Introduction 39 3.2.2 Materials 40 3.2.3 Results and discussion The 5 1 5 N surface sediment distribution 42 Down-core records 44 Nitrate utilization 44 Diagenesis 46 Denitrification and upwelling 50 Nitrogen fixation 50 Terrigenous input 50 Cancellation of effects 51 Global implications 53 3.2.4 Summary 55 3.3 Carbon isotope records 3.3.1 Introduction 56 3.3.2 Results and interpretation Marine versus terrigenous organic matter 57 Temperature effect 65 Marine production, upwelling and biological factors 65 Diagenesis 67 3.3.3 Estimating pC0 2 from sedimentary 6 1 3 C o r g 69 3.3.4 Discussion Absolute 6 1 3 C o r g -pC0 2 values 71 The pC0 2 estimates and upwelling intensity 72 Marine 5 1 3 C o r g -pC0 2 estimates in comparison to the ice core C 0 2 record 76 3.3.5 Conclusions 79 3.4 References 81 v i 4. Ocean-continent linkages in the South China Sea 4.1 On the significance of sea-level variations and shelf palaeo-morphology in governing sedimentation in the southern South China Sea during the last deglaciation 4.1.1 Introduction 92 4.1.2 Regional setting of the Sunda Shelf and the continental slope 93 4.1.3 Core stratigraphy 96 4.1.4 Sedimentary proxies used in this analysis 99 4.1.5 Sedimentary records The outer shelf (core 18282-2,151 m water depth) 100 The shelf margin (core 18284-3, 226 m water depth) 101 The upper continental slope (core 18287-3, 598 m water depth) 101 The lower continental slope (core 17964-3,1556 m water depth) 104 A topographic high on the lower slope (core 18294-4, 849 m water depth) 104 4.1.6 Interpretation and discussion 4.1.6.1 The sequence of sedimentation on the shelf margin and continental slope during the last deglaciation 106 4.1.6.2 Climatic controls on the depositional environment? 115 4.1.7 Summary 118 4.2 Reconstructing palaeo-monsoon variability: Downcore variations in the major element composition at site 17940 in the northern SCS 4.2.1 Introduction 120 4.2.2 Results and interpretation 122 4.2.3 Discussion 125 4.2.4 Conclusions 127 4.3 Constraints on the timing of MWP la from the SCS 4.3.1 introduction 129 4.3.2 Results and discussion 130 4.3.3 Conclusions 136 4.4 References 137 5. Closing remarks 144 Appendix I, Sample Material 147 Appendix II, Methods 147 vii List of tables Table 1.1. List of geographic location, water depth, core length and Approximate stratigraphic range of all cores used in this study, and proxies analyzed 8 Table 3.1. Sampling locations, recovery and sedimentation rates of the six sediment cores from the South China Sea 49 Table 4.1. Age fix points (AMS U C dates and analogue ages) of cores 18282-2,18284-1,18284-3,18287-3, and 18294-4 98 Table 4.2. Major changes in the depositional environment off the Sunda Shelf during the last deglaciation 110 Table A l . Latitude, longitude and water depth of all sampling stations 151 Table A2. Si / A l , T i / A l , K / A l , Fe/ A l and M g / A l data of core 17940-2 152 viii List of figures Fig. 1.1. Map of the South China Sea showing the distribution of land surface at present and during the Last Glacial Maximum. 3 Fig. 1.2. Map of the South China Sea showing the locations of sampling sites for this thesis. 7 Fig. 2.1. Map of the South China Sea and adjacent seas showing the location of cores 18252-3 and 18287-3 and the distribution of modern annual average sea surface temperatures. 16 Fig. 2.2. Alkenone (U K 3 7 ) SST estimates of core 18287-3 versus age. 18 Fig. 2.3. Cumulative percentages of major species groups of planktonic foraminifera. 18 Fig. 2.4. Estimated summer and winter sea-surface temperatures of core 18287-3 using different foraminiferal transfer functions. 20 Fig. 2.5. Comparison of alkenone, foraminiferal Mg/Ca and foraminiferal transfer function SST estimates. 22 Fig. 2.6. Planktonic 5 l sO (G.ruber) and alkenone (U K 3 7 ) SST estimates of cores 18252-3 and 18287-3 from the southern South China Sea versus cal. age, compared to the GISP2 8 l sO record. 26 Fig. 2.7. Comparison of the interpolated cal. ages of the midpoint of the Boiling-warming in three cores from within the SCS. 28 Fig. 2.8. Comparison of deglacial U K 37-SST records from tropical and mid/high northern latitude sites. 31 Fig. 3.1. Map of the South China Sea and 5 1 5N surface sediment distribution. 41 Fig. 3.2. Water column profiles of nitrate concentration and its isotopic composition in the southern South China Sea. 43 Fig. 3.3. Down-core records of bulk sedimentary 6 1 5 N and &lsOG.mber w h i t e versus age in kyr. 45 Fig. 3.4. Bulk sedimentary 5 1 5 N and percent total nitrogen (% TN) in box core profiles, including a surficial fluff sample. 47 Fig. 3.5. Downcore records of bulk sedimentary 5 1 5 N and percent TN of cores 17954 and 17964. 49 Fig. 3.6. 8 1 5 N and 5 1 3 C o r g of core 18284 versus age. 52 Fig. 3.7. Map of global sea-air pC0 2 difference and location of cores discussed in this chapter. 58 Fig. 3.8. Bulk sedimentary 5 1 3C o r g , concentration of w-nonacosane in ng/ g, percent clay of the siliciclastic fraction, U K 3 7 SST estimates, concentration of C37-alkenones in ng/g, bulk sedimentary 5 1 5N, and 5 1 8 0 G r „ t e r w h i t e of core 17940-2 in the northern SCS versus core depth. 59 Fig. 3.9. Bulk sedimentary 5 1 3C o r g , concentration of n-nonacosane in ng/g, U K 3 7 SST estimates, concentration of C37-alkenones in ng/ g, bulk sedimentary 5 1 5N, and bwOGmber of core 17954-2 in the western central SCS versus core depth. 60 ix Fig. 3.10. Bulk sedimentary 5 1 3C o r g , concentration of n-nonacosane in ng/ g, percent clay of the siliciclastic fraction, U K 3 7 SST estimates, concentration of C37-alkenones in ng/g, bulk sedimentary 5 1 5N, and 5 1 8 0 G n , t e r w h i t e of core 17961-2 in the southern SCS versus core depth. 61 Fig. 3.11. Bulk sedimentary 6 1 3 C o r g and bulk sedimentary 6 , 5 N of core 17924-3 62 in the northeastern SCS versus core depth Fig. 3.12. Map of the 5 1 3 C o r g surface sediment distribution in the SCS 64 Fig. 3.13. Bulk sedimentary 6 1 3 C o r g of surface sediments versus water depth. 68 Fig. 3.14. Bulk sedimentary 5 1 3 C o r g of fluff samples versus surface sediments. 68 Fig. 3.15. The pC02 estimates of cores 17940-2,17954-2, and 17961-2 from the northern, western central, and southern SCS versus age. 70 Fig. 3.16. Bulk sedimentary 6 1 5 N data from and pC0 2 reconstruction based on bulk 6 1 3 C o r g and on S^Qu^ne of core GeoB 1016-3 from the Angola Basin. 73 Fig. 3.17. (a) Bulk sedimentary 5 1 5N, and (b) bulk sedimentary 5 1 3 C o r g surface sediment distribution along two transects across the eastern equatorial Pacifc up welling zone. 75 Fig. 3.18. Bulk sedimentary 8 1 3 C o r g (corrected for 6 1 3 C D I C using 513CG.„,i,(,,) versus estimated export production (in gCm"2 yr"1) of core 16772 in the eastern equatorial Atlantic. 75 Fig. 3.19. Comparison of the SCS and Angola Basin pC0 2 estimates with the Vostok CO z record. 78 Fig. 4.1: Map of the South China Sea and adjacent seas showing. 94 Fig. 4.2. (a) Planktonic (G. ruber) foraminiferal 5 l sO depth profiles of cores 18284-3,18287-3,18294-4 and 17964-3, and (b) age-depth relation (in calendar years) of the investigated cores. 97 Fig. 4.3. Downcore variations in (a) %> 5 1 3C o r g , (b) CaC0 3 (wt.%), (c) Si / A l , (d) clay content (% <2 um), and (e) oxygen isotope data (6180) of G. ruber and G. sacculifer of core 18284-3. (f) Estimated sea-level curve for the Sunda. 102 Fig. 4.4. Downcore variations in (a) %0 8 1 3C o r g , (b) CaC0 3 (wt.%), (c) S i / A l , (d) clay content (% < 2jum), (e) U k ' 3 7 SST, and (f) oxygen isotope data (5180) of G. ruber of core 18287-3. 103 Fig. 4.5. Downcore variations in (a) %o 5 1 3C o r g , (b) n-nonacosane concentration (ng/g), (c) CaC0 3 (wt.%), (d) clay content (% <2 urn), (e) U k ' 3 7 SST, and (f) oxygen isotope data (6lsO) of G. ruber of core 17964-3. 105 Fig. 4.6. Downcore variations in (a) %o 5 1 3C o r g , (b) CaC0 3 (wt.%), (c) S i / A l , and (d) oxygen isotope data (5 lsO) of G. ruber of core 18294-4. 107 Fig. 4.7. Intervals of depositional changes on the outer Sunda Shelf, the shelf margin and continental slope in response to the flooding of the shelf, and the interrelation between shelf palaeo-physiography and changes in sea-level. 108 Fig. 4.8. Down core records of silt median and total mode diameters (both in um), percent clay, and alkenone sea surface temperature (SST) estimates of core 17940-2 versus depth. 121 x Fig. 4.9. Down core distributions of S i / A l , t i / A l , K/ A l , Fe /Al and M g / A l of core 17940-2 versus depth. 123 Fig. 4.10. Alkenone (UK 3 7) sea surface temperature estimates and n-nonacosane (C 2 9 M-alkane) concentrations of core 17940-2 versus depth. 131 Fig. 4.11. Alkenone sea surface temperature and benthic (C. wuellerstorfi) oxygen isotope records of cores 17964-3 (a) and 17940-2 (b) from the southern (06°09'N, 112°13'E, 1556 m water depth) and northern (20°07'N, 117°23'E, 1727 m water depth) SCS, respectively, versus depth, (c) 5 1 8 0 C a , records of cores 17964-3 and 17940-2 versus age. 133 Fig. 4.12. Estimated timing of MWP la based on the Sunda Shelf tidal-organic and Barbados coral-reef records as the only data sets bounding the pulse. 134 Fig. Al. Relationship between nitrate concentrations in SCS sea water and Hg N recovered on the filter after ammonia diffusion. 149 xi Preface Some of the material presented in this thesis has been published or is in press. The very nature of multi-proxy palaeoceanographic studies implies that most of these publications are the result of collaboration and are thus joint papers of a number of authors /co-workers. In the following I detail my own contribution to the co-authored publications adopted in this thesis to justify their incorporation into this thesis; the first-authored papers are largely based on my ideas, data and writing, with some of the data and discussion contributed by my co-authors. Their contribution to the data sets is clearly referenced throughout this thesis, and summarized in Table 1.1. on page 8. 1) The downcore records of sea surface temperature (SST) estimates published by Kienast et al. (2001a) in Science and by Steinke et al. (2001) in QR have been amalgamated in chapter 2, modified to be suitable for this dissertation, and complemented by a record of foraminiferal M g / C a SST estimates kindly provided by Dr. D. Hastings. The verbatim adoption of parts of the paper by Steinke et al. in this thesis is justified by my significant contribution to the interpretation of the data, as well as to the writing of this paper. As clearly stated in chapter 2, all foraminiferal data have been generated by Stephan Steinke as part of his Ph.D. studies, whereas the alkenone SST estimates have been analyzed at UBC, partly in collaboration with Stephan during his visit in late fall 1999. Furthermore, I discuss the revised foraminiferal transfer function SST estimates based on the R A M method published in the 'comment' by Waelbroeck and Steinke (2002) rather then ones originally presented in Steinke et al. (2001), which are erroneous. 2) A n updated and extended version of a paper published in Paleoceanography (Kienast 2000) constitutes chapter 3.2. 3) Chapter 3.3 is largely based on a paper by Kienast et al. (2001b) in Global Biogeochemical Cycles. 4) Chapter 4.1 is based on a manuscript (Steinke, Kienast & Hanebuth, 2002) provisionally accepted for publication in a special volume of Marine Geology on 'Asian Monsoons and Global Linkages on Milankovitch and Sub-Milankovitch Timescales'. This paper is the joint presentation and discussion of parts of three separate studies (theses) on the temporal and spatial variability of sedimentation on and off the Sunda xii Shelf. Early on in the 'Sundaflood' (SONNE 115) project Till Hanebuth, Stephan Steinke and myself agreed to split the work along this transect according to water depths, with Till specializing on the history of the flooding of the shelf itself, Stephan focusing on the cores recovered from intemediate water depths and myself contributing the 'deep sea' story. As shown in table 1.1 (page 8), I contributed all 6 1 3 C o r g and S i / A l data to this work; the other data (8 1 8 O G m b e r , %clay, %CaC0 3) have been generated by Stephan Steinke or are cited from previous studies by Pelejero et al. (1999) and Wang et al. (1999). As specified in the manuscript, all authors contributed equally to the interpretation as well as the writing of the manuscript. 5) Chapter 4.3 is adopted from a manuscript accepted for publication in Geology (Kienast et al. 2002). References Kienast, M . , Constant nitrogen isotopic composition of organic matter in the South China Sea during the last climatic cycle: Global implications, Paleoceanography, 15, 244-253, 2000. Kienast, M. , S. Steinke, K. Stattegger, and S. E. Calvert, Synchronous tropical South China Sea SST change and Greenland warming during deglaciation., Science, 291, 2132-2134, 2001a. Kienast, M . , S. E. Calvert, C. Pelejero, and J. O. Grimalt, A critical review of marine sedimentary 8 1 3C o r g-/?C0 2 estimates: New paleo-records from the South China Sea and a revisit of other low-latitude 8 1 3 C o r g -pC0 2 records, Global Biogeochemical Cycles, 15,113-127, 2001b. Kienast, M . , T.J.J. Hanebuth, C. Pelejero, S. Steinke, Synchroneity of Meltwater Pulse l a and the TMling warming: New evidence from the South China Sea. Geology, in press 2002. Pelejero, C , M . Kienast, L . Wang, and J. O. Grimalt, The flooding of Sundaland during the last deglaciation: imprints in hemipelagic sediments from the southern South China Sea, Earth and Planetary Science Letters, 171, 661-671, 1999. Steinke, S., M . Kienast, U . Pflaumann, M . Weinelt, and K. Stattegger, A high resolution sea surface temperature record from the tropical South China Sea (16,500 - 3000 yr B.P.), Quaternary Research, 55, 352-362, 2001. Steinke, S., M . Kienast, T.J.J. Hanebuth, The importance of sea-level variations and shelf palaeo-morphology in governing shelf margin and slope sedimentation: Examples from the southern South China Sea during the last deglaciation. Marine Geology, in press, 2002. Waelbroeck, C. and S. Steinke, Comment on "A high resolution sea surface temperature record from the tropical South China Sea (16,500 - 3000 yr B.P.)" by Steinke et al., Quaternary Research, 57, 432-433, 2002. Wang, L., M . Sarnthein, H . Erlenkeuser, J. Grimalt, P. Grootes, S. Heilig, E. Ivanova, M . Kienast, C. Pelejero, and U. Pflaumann, East Asian monsoon climate during the late Quaternary: High-resolution sediment records from the South China Sea, Marine Geology, 156, 245-284,1999. xii i Acknowledgements My graditute towards and admiration of my supervisor Steve Calvert is beyond words. I thank him for being a true Doktoruarer, an outstanding advisor. He provided me with roots in the form of sound knowledge and wings in the form of freedom to pursue my own ideas. I am greatly indebted to him for his selfless sharing of ideas, inspiration, data, and not least of all his research money. I am very grateful to Drs. Tom Pedersen, Paul Harrison, Marc Bustin, and Les Lavkulich for co-supervising my thesis. Their stimulating input and enthusiastic support throughout my stay at UBC greatly furthered my work. It is great to have such fabulous friends and colleagues as Stephanie Kienast, Till Hanebuth, Thorsten Kiefer, Joe Needoba, Carles Pelejero, Stephan Steinke and the late Luejiang Wang. I am indebted to all of them for continent-encompassing collaboration, endless discussions, reviving laughter, and continuous inspiration. I am equally indebted to the 'GeochemGroup' at UBC for stimulating discussions, invaluable feedback, as well as well for kind hospitality and friendship. My thanks are due to Bente Nielsen and Kathy Gordon, Drs. Pieter Grootes and Helmut Erlenkeuser, Joe Needoba, Christine Elliot and particularly and foremost to Maureen Soon for invaluable help with many aspects of everyday laboratory life, as well as for careful sample analyses. I am especially grateful to Maureen for her friendship and imperturbable cheerfulness. Furthermore, I am grateful to Alex Allen and Carol Leven for administrative support throughout my stay at UBC. The work presented here benefited greatly from many thorough and thoughtful reviews of the manuscripts which form parts of this thesis. My thanks are due to Drs. Mark Altabet, Helge Arz, Jay Brandes, Roger Francois, Thorsten Kiefer, Hermann Kudrass, David Lea, John Milliman, Masao Minagawa, Joseph Ortiz, George Postma, Yair Rosenthal, Michael Sarnthein, Danny Sigman, Eric Steig, as well as anonymous colleagues. For highly stimulating discussions throughout my time as a Ph.D. student I would like to vicariously thank Drs. Roger Francois, David Lea, Andrew Bush, Peter Clark, Min-Te Chen, Shawn Marshall, David Hastings and Gerald Haug. xiv Special thanks are due to Drs. Karl Stattegger, Michael Sarnthein and Ein-Fen Yu for generously providing sample material. Generous funding through fellowships from the German Academic Exchange Service (DAAD), the University of British Columbia (UGF), as well as from the Izaak Walton Killam Trust (Killam pre-doctoral fellowship) is gratefully acknowledged. A l l analyses in this thesis have been paid for by NSERC and GSA student research grants. I am particularly thankful to the department of Earth and Ocean Sciences for repeatedly supporting my participation in international conferences through EOS student travel funds. I am very grateful to my parents for their boundless support of my dreams, and I owe my deepest thanks to Stephanie for her wonderfully stimulating presence in my life! xv 1 1. General Introduction 1.1 Opening remarks Palaeoceanography is a relatively new discipline in Earth Sciences, dealing with the reconstruction of past changes in the physical, chemical, biological and geological conditions in the oceans. Time-scales range from months, for example in studies based on coraline samples and laminated sediments, to millions of years, i.e., tectonic time-scales. In recent years, one main focus of palaeoceanography has been the reconstruction of the ocean's role in the global carbon cycle as manifested in changing atmospheric concentrations of carbon dioxide (C02), as well as in rapid (decadal to centennial) events of global climate change observed, for example, during, the last glacial-interglacial transition (ca. 22-10 kyrs B.P.) and during the last glacial period (ca. 75-30 kyrs B.P.). This is driven by the need to understand the effects of anthropogenic emissions of C 0 2 to the atmosphere, which by now amount to the difference between glacial and interglacial atmospheric C 0 2 levels, and are likely to result in global climate change. In addition, many of the most rapid and dramatic jumps of Earth's climate system from one state of operation to another during the late Pleistocene occurred on human time-scales (years to decades) rather than on orbital time-scales as previously thought. By reconstructing the temporal development and interaction of various players of the global climate system, such as the hydrosphere, atmosphere, cryosphere and biosphere from sedimentary records, palaeoceanography thus seeks to provide insight into the mechanisms of global climate change. These reconstructions are based on various organic and inorganic geochemical as well as isotopic and palaeontological proxies, i.e., measurable descriptors of a desired variable [Wefer et al, 1999], such as for example alkenone unsaturation ratios as a proxy for sea surface temperatures (SSTs) or sedimentary nitrogen isotope ratios as a proxy for past changes in biological nitrate utilization. By combining various proxy records from one core (multi-proxy approach), temporal and, by inferrence, causal relations between different parts of Earth's climate system can be established. Some of the most spectacular palaeoceanographic records to date have been retrieved from marginal basins, such as the Santa Barbara and the Cariaco Basins. This is not only due to the fact that these basins accumulate sediments under (partly) anoxic conditions, which facilitates the preservation of laminated sedimentary records, but is also due to the high sedimentation rates, which afford the chance to reconstruct palaeoceanographic and palaeoclimatic variations at high temporal resolution. 2 Furthermore, sedimentary indicators of past continental and oceanic changes are preserved in the same sedimentary record, thus enabling the detailed reconstruction of ocean-continent linkages. One such marginal basin is the South China Sea (SCS), the largest basin off the Asian margin. It is connected to the Pacific via the Bashi Strait (sill depth ca. 2000 m) in the northeast and to the tropical Indo-Pacific via the Sunda Shelf (average depth 70-80 m, minimum depth <40 m) in the southwest (Fig. 1.1). The SCS waters are part of the Western Pacific Warm Pool (WPWP), as defined by the 28 °C isotherm. Its surface circulation as well as the sedimentation within the basin are strongly influenced by the Asian monsoon, which is an important climate system transporting significant amounts of moisture and heat between low and high latitudes [Hastenrath and Lamb, 1980; Webster, 1987], and is comparable to the ocean conveyor in its global importance [Webster, 1994]. The monsoonal climate manifests itself in annually reversing winds (NW monsoon during winter, SE during summer), which, in addition to forcing the surface circulation, deliver significant amounts of aeolian sediments during the winter, and cause increased precipitation/riverine sediment input to the SCS during summer. Therefore, the SCS is a unique and sensitive monitor of Asian monsoonal climate variability, ocean-continent-atmosphere linkages and WPWP dynamics. 1.2 The oceanography of the South China Sea The SCS surface oceanography is dominated by the annually reversing monsoonal winds [Wyrtki, 1961; La Violette and Frontenac, 1967], caused by differential heating between ocean and land [Halley, 1686; Hadley, 1735; Webster, 1987]. During winter, the cooling of Central Asia develops a quasi-stable high pressure system that leads to a radial out-flow of cold, dry air towards the western Pacific [Webster, 1987]. During summer the system reverses and central Asia is warm compared to the Pacific Ocean, developing a quasi-stable low-pressure cell. As a result, humid, maritime air flows in from the western Pacific. Accordingly, the winter (NE) monsoon drives cold surface waters (<20 °C) from the adjacent western Pacific southwestward into the SCS [Wyrtki, 1961]. In contrast, warmest sea surface temperatures (SSTs) are reported in August (27-29 °C), when the SW monsoon-induced current pattern drives tropical waters to the northeast. Heavy monsoonal precipitation and high riverine input lead to sea surface salinities (SSSs) of <33 throughout the SCS [Wyrtki, 1961; La Violette and Frontenac, 1967], which is markedly lower than in the adjacent tropical Indo-Pacific and western 3 Fig. 1.1. Map of the South China Sea showing the distribution of land surface at present and during the Last Glacial Maximum, assuming a 120 m sea level lowering. Former drainage systems on the Sunda Shelf (dashed lines) according to Molengraaff'[1921]. 4 Pacific. Even lower SSSs are observed seasonally off the mouths of the Mekong and the Pearl Rivers (ibid.; Fig. 1.1). The high sea surface temperatures in conjunction with low salinities create a stable, low density surface layer with an extremly strong discontinuity layer between 100 and 300 m [Wyrtki, 1961]. Below this discontinuity layer, the deep water hydrography of the SCS is mainly controlled by inflow of the Pacific Intermediate Water through the Bashi Strait in the northeast [Shaw, 1991; Chao et al, 1996]. According to radiocarbon data obtained on both sides of the Bashi Strait by Broecker et al. [1986], the renewal of this deep water in the SCS is rapid (<100 years). The actual circulation pattern of the deep water is, however, controversial. Temperature-salinity characteristics clearly show a seasonal (late summer to late spring) intrusion of Kuroshio water into the SCS, possibly related to seasonal variations in the transport of the Kuroshio [Shaw, 1989, 1991]. Although climatology-driven circulation models predict seasonal upwelling of this deep water along the coast of Vietnam [Chao et al, 1996; Wiesner et al, 1996], on the edge of the Sunda Shelf [Chao et al, 1996], on the west coast of the Philippines [Shaw and Chao, 1994; Shaw et al, 1996] and southwest of Taiwan [Chao et al, 1996], earlier regional oceanographic studies failed to show upwelling-related pertubations of the SST and SSS distribution in the areas of interest [Wyrtki, 1961; La Violette and Frontenac, 1967]. This could be due to the fact that, for example, the upwelling off NW Luzon is associated with decreased sub-surface temperatures rather than sea surface temperatures [Udarbe-Walker and Villanoy, 2001]. However, more recent studies using satellite imagery [Tang et al, 1999; Ho et al, 2000; Liu et al, 2002, in press] seem clearly to depict weak upwelling off northwest Luzon and off the east coast of Vietnam. 1.3 Sedimentation in the South China Sea Sediment supply to the SCS is dominated by terrigenous material [Chen, 1981; Schdnfeld and Kudrass, 1993; Huang and Wang, 1998], mainly supplied by the three major rivers draining into the basin, i.e. the Pearl River in the north, the Red River in the northwest and the Mekong River in the southwest [Milliman and Meade, 1983; Hay, 1998; see Fig. 1.1]. Despite the large fluvial input into the SCS, the lithogenic material delivered to a sediment trap in the central northern SCS is postulated to be of aeolian origin [Wiesner et al, 1996]. This observation is in agreement with earlier studies suggesting that most of the riverine material is trapped on the shelves and upper continental slope regions [Schdnfeld and Kudrass, 1993]. Moreover, studies of global atmospheric dust fluxes [Duce et al, 1991; Gao et al, 1997; Kohfeld and Harrison, 2001] 5 suggest dust deposition rates of up to 10g/m 2/yr in the vicinity of continental China, further emphasizing the significance of aeolian sedimentation within the SCS. 1.4 Previous palaeoceanographic studies in the South China Sea Early marine geological studies in the SCS focused on the reconstruction of the glacial emergence and postglacial flooding of the Sunda Shelf in the southern SCS using detailed bathymetric maps [Molengraaff and Weber, 1920; Molengraaff, 1921; Umbgrove, 1949], shallow seismic profiles [Batchelor, 1979; Emmel and Curray, 1982], reconstructions of shore lines [Tjia, 1970, 1980; Tjia et al, 1972; Gupta et al, 1987] and the distribution of benthic and planktonic foraminifera in shallow push cores [Biswas, 1973, 1976]. Since the 1950's, the composition and sources of the bottom sediments of the SCS have been studied by van Baren and Kiel [1950], Niino and Emery [1961], Emery and Niino [1963], Pimm [1964], Keller and Richards [1967], Chen [1978], Chen [1981], Fung [1982], and Khrustalev et al. [1987]. More recently, attempts have been made to compare and contrast the geochemical composition of Recent sediments with those deposited during the late Quaternary [Astakhov et al, 1990; Calvert et al, 1993; Markov, 1994; Su and Wang, 1994]. Palaeoceanographic studies in the SCS have focused mainly on glacial-interglacial changes in bulk sediment accumulation rates [Broecker et al, 1988; Wang et al, 1992; Schdnfeld and Kudrass, 1993; Huang and Wang, 1998; Wang, 1999], variations in sea surface temperature and, by inferrence, surface circulation [Wang and Wang, 1990; Miao et al, 1994; Huang et al, 1997a,b; Chen and Huang, 1998; Pelejero et al, 1999a,b; Wang et al, 1999], the origin of carbonate cycles [Rottman, 1979; Wang et al, 1995, 1997; Chen et al, 1997; Jian and Wang, 1997] and changes in marine primary production [Thunell et al, 1992; Winn et al, 1992; Miao and Thunell, 1996; Jian et al, 1999, 2001; Kuhnt et al, 1999]. In addition, attempts have been made to obtain records of fluctuations in monsoon intensity, as inferred from continental loess [Kukla et al, 1988; An et al, 1993; Ding et al, 1995; Porter and An, 1995; Xiao et al, 1995; Guo et al, 1996; An and Porter, 1997] and pollen profiles [Liu, 1988; Morley and Heusser, 1989; Heusser and Morley, 1997] from SCS marine sediment sequences [Wang and Sun, 1994; Huang et al, 1997b; Wang et al, 1999]. Pleistocene sea level fluctuations of the order of 120 m [Shackleton, 1987] caused dramatic variations in the palaeoceanography of the SCS [Wang et al, 1999]. Sea level lowering during glacial times resulted in the emergence of vast shelf areas in the northern and especially in the southern SCS, and a concomitant closure of the major connecting seaways, notably to the tropical Indo-Pacific in the southwest (Fig. 1.1). Consequently, the SCS became a semi-enclosed gulf, connected only to the western 6 Pacific through the Bashi Strait in the northeast [e.g., Wang and Wang, 1990; see Fig. 1.1]. With two of the world's largest rivers, i.e. the Mekong in the southwest and the Pearl River in the northwest (Fig. 1.1) draining into the basin [Milliman and Meade, 1983; Hay, 1998] and the development of additional drainage systems on the emergent shelves (the Molengraaff River on the Sunda Shelf, Fig. 1.1; Molengraaff, 1921), terrigenous input into the SCS increased enormously during glacials due to the movement of the sediment depocentres on the shelves into the deep basin [Schdnfeld and Kudrass, 1993]. Moreover, the closure of the connection to the tropical Indo-Pacific during glacials caused changes in sea surface temperature due to the cessation of warm water inflow from the south and a change from the present transbasinal surface circulation pattern to surface gyres [Wang and Wang, 1990; Miao et al, 1994; Wang et al, 1995; Pelejero et al, 1999a]. Finally, it is speculated that increased riverine input, strengthened inflow of nutrient-enriched West Pacific surface and intermediate waters, changes in regional upwelling, and /or changes in the siliceous to calcareous plankton production ratio led to higher glacial primary production [Thunell et al, 1992; Miao and Thunell, 1996; Kuhnt et al, 1999; Wang et al, 1999]. 1.5 Objectives of this study In this thesis, multi-proxy organic and inorganic sediment geochemical records from a large number of gravity cores and surface sediment samples (Fig. 1.2, Tab. 1.1.; see also Appendix I) are presented in an attempt to resolve in detail the timing and amplitude of deglacial sea surface temperature changes in the southern SCS (Chapter 2), to assess glacial-interglacial variations in the biogeochemical cycles of nitrogen and organic carbon as reflected in their time-varying isotopic composition (Chapter 3), and, finally, to differentiate between the effects of variable sea level/shelf physiography and/ or continental climate on bulk sediment geochemistry (Chapter 4). The general introduction (see above) of this thesis is kept rather short because each of the chapters has its own introductory sub-chapter, which summarizes the current state of our understanding, and highlights the topics that will be delt with in the chapter. This structure of the thesis is also a reflection of the fact that most sub-chapters have been published or submitted as separate publications (see Preface). To further minimize redundancy, all analytical procedures are given in the Appendix. 7 105 110 115 120 125°E Fig. 1.2. Map of the South China Sea showing the locations of sampling sites for this thesis. Plain black numbers indicate sites where surface sediment samples only have been taken, bold red numbers indicate sites where both surface sediment samples and gravity cores have been taken. Sites 17920 to 17965 were sampled during the R / V SONNE 95 cruise (April-June 1994; Sarnthein et al, 1994), sites 18260 to 18321 in the southern and southwestern SCS during the R / V SONNE 115 cruise (December 1996 - January 1997; Stattegger et al, 1997), and sites 18376 to 18425 off Vietnam during the R / V SONNE 140 cruise (April-May 1999; Wiesner et al, 1999). Sites GGC1 to GGC19 (black squares) in the southeastern SCS were sampled during a cruise in May 1988 (e.g., Calvert et al, 1993; Miao et al, 1994; Thunell et al, 1992). 8 ON 00 ao 00 00 00 in N 00 NC ON I D o >* ON CN ON I * * w o bb c o 6 6 o T3 •5 SO 5 o 60 C CU <u 60 C re £ 1 x o X o ex P 53 P 53 H • c a-(0 U 60 2 •5 CO 2 3 2 33 2 •5 3 y •5 CO 2 2 c o c -a u C o u 6 O c O U IB u =1 01 in .. S3* 2 ON C ON co A ; -a2 co O OI .ii -5? 10 M ~ C 10 o T 3 ai • in 3 in cu CO ON ON ON a, QJ a c c « cu . . O- o ? <" "5T S 60 in 5; UN J " PH CO O u co 0) 60 2 U *-l ^N D- co 2 — oo"33 CO CN : 00 i S . CN a B _ (0 S3 co _ .§ § S .2 OH CO c co co o in -in ai ^3 co .—1 _ CO 60 C c « __ o S in 3 OH 0 : 1 ai 'in 7 3 fu ai *^  > in - v (0 X> CO •S .2 7 3 o • a> > O J u ^N C in — O co •43 C co cu OH _ 3 •• too o 00T3 C § CO - C — °13 i i cu ai cu a, ^ 9 1.6 References An, Z., and S. C. Porter, Millenial-scale climatic oscillations during the last interglaciation in central China, Geology, 25, 603-606,1997. An, Z., S. C. Porter, W. Zhou, Y. Lu, D. J. Donahue, M . J. Head, X. Wu, J. Ren, and H . Zheng, Episode of strengthened summer monsoon climate of Younger Dryas age on the Loess Plateau of central China, Quaternary Research, 39, 45-54,1993. Astakhov, A . S., Y. D. Markov, and C. T. Khieu, Effect of the Mekong River on late Quaternary sedimentation in the South China Sea, Translated from Litologiya i Poleznye Iskopaemye, 3,112-127,1990. Batchelor, B. C , Discontinously rising Late Cainozoic eustatic sea-levels, with special reference to Sundaland, Southeast Asia, Geologie en Mijnbouw, 58,1-20, 1979. Biswas, B., Quaternary changes in sea-level in the South China Sea, Geol. Soc. Mallaysia Bulletin, 6, 229-256,1973. Biswas, B., Bathymetry of Holocene foraminifera and Quaternary sea-level changes on the Sunda Shelf, Journal of Foraminiferal Research, 6,107-133,1976. Broecker, W. S., W. C. Patzert, J. R. Toggweiler, and M . Stuiver, Hydrography, chemistry, and radioisotopes in the Southeast Asian basins, Journal of Geophysical Research, 91,14345-14354,1986. Broecker, W. S., M . Andree, M . Klas, G. Bonani, W. Wolfli, and H . Oeschger, New evidence from the South China Sea for an abrupt termination of the last glacial period, Nature, 333,156-158,1988. Calvert, S. E., T. F. Pedersen, and R. C. Thunell, Geochemistry of surface sediments of the Sulu and South China Seas, Marine Geology, 114, 207-231,1993. Chao, S.-Y., P.-T. Shaw, and S. Y. Wu, Deep water ventilation in the South China Sea, Deep Sea Research, 43, 445-466,1996. Chen, M.-P., Grain size distribution, mineral composition and carbonate content in bottom sediments of the South China Sea, Proc. Natl. Sci. Counc. ROC(A), 5, 92-105, 1981. Chen, M.-T., and C.-Y. Huang, Ice-volume forcing of winter monsoon climate in the South China Sea, Paleoceanography, 13, 622-633,1998. Chen, M.-T., C.-Y. Huang, and K.-Y. Wei, 25,000-year late Quaternary records of carbonate preservation in the South China Sea, Palaeogeography, Palaeoclimatology, Palaeoecology, 129,155-169,1997. Chen, P.-Y., Minerals in bottom sediments of the South China Sea, Geological Society of America Bulletin, 89, 211-222,1978. Ding, Z., T. Liu, N . W. Rutter, Z. Yu, Z. Guo, and R. Zhu, Ice-volume forcing of East Asian winter monsoon variations in the past 800,000 years, Quaternary Research, 44, 149-159,1995. Duce, R. A., P. S. Liss, and J. T. Merrill, The atmospheric input of trace species to the world ocean, Global Biogeochemical Cycles, 5,193-259,1991. Emery, K. O., and H . Niino, Sediments of the Gulf of Thailand and adjacent continental shelf, Geological Society of America Bulletin, 74, 541-554,1963. Emmel, F. J., and J. R. Curray, A submerged late Pleistocene delta and other features related to sea level changes in the Malacca Strait, Marine Geology, 47,197-216,1982. Fung, C , Specific features of the relief and bottom sediments on the shelf in the tropical zone, Oceanology, 22,188-191,1982. Gao, Y., R. Arimoto, R. A. Duce, X. Y. Zhang, G. Y. Zhang, Z. S. An, L. Q. Chen, M . Y. Zhou, and D. Y. Gu, Temporal and spatial distributions of dust and its deposition to the China Sea, Tellus, 49B, 172-189,1997. Guo, Z., T. Liu, J. Guiot, N . Wu, H . Lii , J. Han, and Z. Gu, High frequency pulses of East Asian monsoon climate in the last two glaciations: link with the North Atlantic, Climate Dynamics, 12, 701-709,1996. 10 Gupta, A., A. Rahman, P. P. Wong, and J. Pitts, The Old Alluvium of Singapore and the extinct drainage system to the South China Sea, Earth Surface Processes and Landforms, 12, 259-275,1987. Hadley, G., Concering the cause of the general trade winds, Phil. Trans., Roy. Soc. London, 39, 58-62,1735. Halley, E., An historical account of the trade winds and the monsoons, observable in the seas between and near the tropicks, wih an attempt to assign the phisical cause of the said winds, Phil. Trans., Roy. Soc. London, 16,153-168,1686. Hastenrath, S., and P. J. Lamb, On the heat budget of hydrosphere and atmosphere in the Indian Ocean, Journal of Physical Oceanography, 10, 694-708,1980. Hastings D, Whitko A, Kienast M , Steinke S, Comparison of three independent paleotemperature estimates from a high resolution record of deglacial SST records in the tropical South China Sea, 2002, in prep. Hay, W. W., Detrital sediment fluxes from continents to oceans, Chemical Geology, 145, 287-323,1998. Heusser, L., and J. Morley, Monsoon fluctuations over the past 350 kyr: high-resolution evidence from northeast Asia/northwest Pacific climate proxies (marine pollen and radiolarians), Quaternary Science Reviews, 16, 565-581,1997. Ho, C.-R., Q. Zheng, Y. S. Soong, N.-J. Kuo, and J.-H. Hu, Seasonal variability of sea surface height in the South China Sea observed with TOPEX/Poseidon altimeter data, Journal of Geophysical Research, 105,13,981-913,990, 2000. Huang, C.-Y., P.-M. Liew, M . Zhao, T.-C. Chang, C.-M. Kuo, M.-T. Chen, C.-H. Wang, and L.-F. Zheng, Deep sea and lake records of the Southeast Asian paleomonsoons for the last 25 thousand years, Earth and Planetary Science Letters, 146, 59-72,1997a. Huang, C.-Y., S.-F. Wu, M . Zhao, M.-T. Chen, C.-H. Wang, X. Tu, and P. B. Yuan, Surface ocean and monsoon climate variability in the South China Sea since the last glaciation, Marine Micropaleontology, 32, 71-94,1997b. Huang, W., and P. Wang, A quantitative approach to deep-water sedimentation in the South China Sea: changes since the last glaciation, Science in China (Series D), 41, 195-201, 1998. Jian, Z., and L. Wang, Late Quaternary benthic foraminifera and deepwater paleoceanography in the South China Sea, Marine Micropaleontology, 32, 127-154, 1997. Jian, Z., L. Wang, M . Kienast, M . Sarnthein, W. Kuhnt, H . Lin, and P. Wang, Benthic foraminiferal paleoceanogrphy of the South China Sea over the last 40,000 years, Marine Geology, 156,159-186,1999. Jian, Z., B. Huang, W. Kuhnt, and H.-L. Lin, Late Quaternary upwelling intensity and East Asian Monsoon forcing in the South China Sea, Quaternary Research, 55, 363-370, 2001. Kienast, M . , S. Steinke, K. Stattegger, and S. E. Calvert, Synchronous tropical South China Sea SST change and Greenland warming during deglaciation., Science, 291, 2132-2134, 2001a. Kienast, M . , S. E. Calvert, C. Pelejero, and J. O. Grimalt, A critical review of marine sedimentary 5 1 3 C o r g -pC0 2 estimates: New paleo-records from the South China Sea and a revisit of other low-latitude 5 1 3 C o r g -pC0 2 records, Global Biogeochemical Cycles, 15,113-127, 2001b. Keller, G. H. , and A. F. Richards, Sediments of the Malacca Strait, Southeast Asia, Journal of Sedimentary Petrology, 37,102-127,1967. Khrustalev, Y. P., L. Y. Khrustalev, D. S. Turovskii, and L. T. Khop, Mineralogy of bottom sediments of the western shelf and continental slope of the South China Sea, Translated from Litologiya i Poleznye Iskopaemye, 2, 3-11,1987. Kohfeld, K. E., and S. P. Harrison, DIRTMAP: the geological record of dust, Earth Science Reviews, 54, 81-114, 2001. 11 Kuhnt, W., S. Hess, and Z. Jian, Quantitative composition of benthic foraminiferal assemblages as a proxy indicator for organic carbon flux rates in the South China Sea, Marine Geology, 156,123-157,1999. Kukla, G., F. Heller, L. X. Ming, X. T. Chun, L. T. Sheng, and A. Z. Sheng, Pleistocene climates in China dated by magnetic susceptibility, Geology, 16, 811-814,1988. La Violette, P. E., and T. R. Frontenac, Temperature, salinity, and density of the World's seas: South China Sea and adjacent gulfs, 1967. Liu, K., Quaternary history of the temperate forests of China, Quaternary Science Reviews, 7,1-20,1988. Liu, K.-K., S.-Y. Chao, P.-T. Shaw, G.-C. Gong, C.-C. Chen, and T. Y. Tang, Monsoon-forced chlorophyll distribution and primary production in the South China Sea: observations and a numerical study, Deep Sea Research, 2002, in press. Markov, Y. D., Recent sedimentogenesis on the Ille des Cendres volcanoes and adjacent shelf, South China Sea, Vulkanologiya i Seismologiya English, 15, 565-578,1994. Miao, Q., and R. C. Thunell, Late Pleistocene-Holocene distribution of deep-sea benthic foraminifera in the South China Sea and Sulu Sea: paleoceanographic implications, journal of Foraminiferal Research, 9-23,1996. Miao, Q., R. C. Thunell, and D. M . Anderson, Glacial-Holocene carbonate distribution and sea surface temperatures in the South China and Sulu seas, Paleoceanography, 9, 269-290,1994. Milliman, J. D., and R. H . Meade, World-wide delivery of river sediment to the oceans, Journal of Geology, 91,1-21,1983. Molengraaff, G. A . F., Modern deep-sea research in the East Indian Archipelago, The Geographical Journal, 57, 95-121,1921. Molengraaff, G. A. F., and M . Weber, On the relation between the Pleistocene glacial period and the origin of the Sunda Sea (Java- and South China- Sea), and its influence on the distribution of coral reefs and on the land- and freshwater fauna, Proceedings Royal Acad. Amsterdam, 23, 394-428,1920. Morley, J. J., and L. E. Heusser, Late Quaternary atmospheric and oceanographic variations in the western pacific inferred from pollen and radiolarian analyses, Quaternary Science Reviews, 8, 263-276,1989. Niino, H. , and K. O. Emery, Sediments of the shallow portions of East China Sea and South China Sea, Geological Society of America Bulletin, 72, 731-762,1961. Pelejero, C , J. O. Grimalt, S. Heilig, M . Kienast, and L. Wang, High resolution U K 3 7 -temperature reconstructions in the South China Sea over the past 220 kyr, Paleoceanography, 14, 224-231,1999a. Pelejero, C , M . Kienast, L. Wang, and J. O. Grimalt, The flooding of Sundaland during the last deglaciation: imprints in hemipelagic sediments from the southern South China Sea, Earth and Planetary Science Letters, 171, 661-671,1999b. Pimm, A. C , Seabed sediments of the South China Sea, Borneo Reg. Malaysia Geol. Survey Ann. Rept., 122-146,1964. Porter, S. C , and Z. An, Correlation between climate events in the North Atlantic and China during the last glaciation, Nature, 375, 305-308,1995. Rottman, M . L., Dissolution of planktonic foraminifera and pteropods in South China Sea sediments, Journal of Foraminiferal Research, 9, 41-49,1979. Sarnthein, M . , U . Pflaumann, P. X. Wang, and H . K. Wong, Preliminary report of the Sonne-95 cruise "Monitor Monsoon" to the South China Sea, Rep. 68, Berichte -Reports, Geol.-Palaont. Inst. Univ. Kiel, 1994. Schonfeld, J., and H.-R. Kudrass, Hemipelagic sediment accumulation rates in the South China Sea related to Late Quaternary sea-level changes, Quaternary Research, 40, 368-379,1993. Shackleton, N . J., Oxygen isotopes, ice volume and sea level, Quaternary Science Reviezos, 6, 183-190,1987. 12 Shaw, P.-T., The intrusion of water masses into the sea southwest of Taiwan, Journal of Geophysical Research, 94,18213-18226,1989. Shaw, P.-T., The seasonal variation of the intrusion of the Philippine sea water into the South China Sea, Journal of Geophysical Research - Part C, 96, 821-827,1991. Shaw, P.-T., and S.-Y. Chao, Surface circulation in the South China Sea, Deep Sea Research, 41,1663-1683,1994. Shaw, P.-T., S.-Y. Chao, K.-K. Liu, S.-C. Pai, and C.-T. Liu, Winter upwelling off Luzon in the northeastern South China Sea, Journal of Geophysical Research, 101, 16,435-416,448, 1996. Stattegger, K., W. Kuhnt, H . Wong, et al., Cruise report SONNE 115: Sundaflut - Sequence stratigraphy, late Pleistocene-Holocene sea level fluctuations and high resolution record of the post-Pleistocene transgression on the Sunda Shelf, 211 pp., 1997. Steinke, S., M . Kienast, U . Pflaumann, M . Weinelt, and K. Stattegger, A high resolution sea surface temperature record from the tropical South China Sea (16,500 - 3000 yr B.P.), Quaternary Research, 55, 352-362, 2001. Steinke, S., M . Kienast, T.J.J. Hanebuth, The importance of sea-level variations and shelf palaeo-morphology in governing shelf margin and slope sedimentation: Examples from the southern South China Sea during the last deglaciation. Marine Geology, in press, 2002. Su, G., and T. Wang, Basic characteristics of modern sedimentation in South China Sea. in Oceanology of China Seas, edited by Zhou, D., Y. Liang and C. Zeng, pp. 407-418, Kluwer Academic Publishers, 1994. Tang, D.-L., I.-H. Ni , D. R. Kester, and F. E. Mtiller-Karger, Remote sensing observations of winter phytoplankton blooms southwest of the Luzon Strait in the South China Sea, Marine Ecology Progress Series, 191, 43-51,1999. Thunell, R. C , Q. Miao, S. E. Calvert, and T. F. Pedersen, Glacial-Holocene biogenic sedimentation patterns in the South China Sea: productivity variations and surface water pCO^ Paleoceanography, 7,143-162,1992. Tjia, H . D., Quaternary shore lines of the Sunda Land Southeast Asia, Geologie en Mijnbouw, 49,135-144,1970. Tjia, H . D., The Sunda Shelf, Southeast Asia, Z. Geomorph. N.F., 24, 405-427,1980. Tjia, H . D., S. Fujii, K. Kigoshi, A . Sugimura, and T. Zakaria, Radiocarbon dates of elevated shorelines, Indonesia and Malaysia. Part I, Quaternary research, 2, 487-495, 1972. Udarbe-Walker, M . J. B., and C. L. Villanoy, Structure of potential upwelling areas in the Philippines, Deep Sea Research, Part I, 48,1499-1518, 2001. Umbgrove, J. H . F., Structural history of the East Indies, Cambridge University Press, 1949. van Baren, F. A., and H . Kiel, Contribution to the sedimentary petrology of the Sunda Shelf, Journal of Sedimentary Petrology, 20,185-213,1950. Wang, L., and P. Wang, Late Quaternary paleoceanography of the South China Sea: glacial-interglacial contrasts in an enclosed basin, Paleoceanography, 5, 77-90,1990. Wang, L., Z. Jian, and J. Chen, Late Quaternary pteropods in the South China Sea: carbonate preservation and paleoenvironmental variation, Marine Micropaleontology, 32,115-126,1997. Wang, L., M . Sarnthein, H . Erlenkeuser, J. Grimalt, P. Grootes, S. Heilig, E. Ivanova, M . Kienast, C. Pelejero, and U. Pflaumann, East Asian monsoon climate during the late Quaternary: High-resolution sediment records from the South China Sea, Marine Geology, 156, 245-284,1999. Wang, P., Response of western Pacific marginal seas to glacial cycles: paleoceanographic and sedimentological features, Marine Geology, 156, 5-39,1999. Wang, P., and X. Sun, Last glacial maximum in China: comparison between land and sea, Catena, 23, 341-353,1994. 13 Wang, P., Z. Jian, and Z. Liu, Late Quaternary sedimentation rate in the South China Sea. in Contributions to late Quaternary paleoceanography of the South China Sea, edited by Ye, Z. and P. Wang, pp. 23-41, Qingdao Ocean University Press, 1992. Wang, P., L. Wang, Y. Bian, and Z. Jian, Late Quaternary paleoceanography of the South China Sea: surface circulation and carbonate cycles, Marine Geology, 127, 145-165, 1995. Webster, P. J., The elementary monsoon, in Monsoons, edited by Fein, J. S. and P. L. Stephens, pp. 3-32, Wiley-Interscience Publication, 1987. Webster, P. J., The role of hydrological processes in ocean-atmosphere interactions, Revieios of Geophysics, 32, 427-476,1994. Wefer, G., W. H . Berger, J. Bijma, and F. Fischer, Clues to ocean history: A brief overview of proxies, in Use of proxies in paleoceanography: Examples from the South Atlantic, edited by Fischer, G. and G. Wefer, pp. 1-68, Springer-Verlag, 1999. Wiesner, M . G., L. Zheng, H . K. Wong, Y. Wang, and W. Chen, Fluxes of particulate matter in the South China Sea. in Particle flux in the ocean, edited by Ittekot, V., P. Schafer, S. Honjo and P. J. Depetris, pp. 293-312, Wiley & Sons, 1996. Wiesner, M . G., K. Stattegger, W. Kuhnt, et al., Cruise Report SONNE 140 - Siidmeer III, Rep. 7, Berichte - Reports, Inst, fur Geowiss., Univ. Kiel, 1999. Winn, K., L. Zheng, H . Erlenkeuser, and P. Stoffers, Oxygen/carbon isotopes and paleoproductivity in the South China Sea during the past 110 000 years, in Marine geology and geophysics of the South China Sea, edited by Jin, X., H . R. Kudrass and G. Pautot, pp. 154-166, China Ocean Press, 1992. Wyrtki, K., Scientific results of marine investigations of the South China Sea and the Gulf of Thailand, N A G A report, 1961. Xiao, J., S. C. Porter, Z. An, H . Kumai, and S. Yoshikawa, Grain size of quartz as an indicator of winter monsoon strength on the Loess Plateau of Central China during the last 130,000 yr, Quaternary Research, 43, 22-29,1995. 14 2. Sea surface temperature reconstructions 2.1 Introduction Reconstructing past tropical sea-surface temperatures (SSTs), particularly the magnitude of glacial cooling and the timing and pattern of the deglacial warming, is required for a complete understanding of the role of the tropics in global climate change [e.g., Cane, 1998; Cane and Clement, 1999]. Discrepancies in the magnitude of cooling of the tropics during the last glacial period as reconstructed by different proxy parameters, e.g., foraminiferal transfer functions [e.g., CLIMAP project members, 1981; Mix et al, 1999a], alkenone [U K ' 3 7 ; Lyle et al, 1992; Sonzogni et al, 1998; Pelejero et al, 1999] and foraminiferal M g / C a [Hastings et al, 1998; Lea et al, 2000; Numberg et al, 2000] palaeothermometry, coral geochemistry [Guilderson et al, 1994, 2001; Beck et al, 1997; Tudhope et al, 2001], the oxygen isotopic composition of planktonic foraminifera [Broecker, 1986] and ice cores [Thompsori et al, 1995; Pierrehumbert, 1999], snow line depressions [see Porter, 2001 for a recent summary], pollen and plant-macrofossils [see Farrera et al, 1999 for a recent summary], and, finally, ground water thermometry [Stute et al, 1995, 1997; Weyhenmeyer et al, 2000], has limited our understanding of the sensitivity of the climate system to changes in radiative forcing [see Crowley, 2000 for a recent summary]. Estimates of the Last Glacial Maximum (LGM)-to-Holocene temperature increase in the tropics of the Pacific region, for example, range from 1-6 °C [CLIMAP project members, 1981; Lyle et al, 1992; Beck et al, 1997; Stute et al, 1997; Mix et al, 1999a; Pelejero et al, 1999; Lea et al, 2000; Rosenthal et al, 2000b]. Previous studies of the phase relation between tropical and high latitude warming during the last deglaciation came to contrasting conclusions: the tropical ocean was either synchronous with [Bard et al, 1997], or led [Ruhlemann et al, 1999; Lea et al, 2000] the northern hemisphere deglacial temperature increase. Antiphasing between changes in tropical Atlantic sea surface temperature (SST) and temperature over Greenland is expected based on the so-called bipolar seesaw mechanism [Broecker, 1998; Stocker, 1998, 2000]. Thus, heat is exported from the tropical Atlantic to the north via the thermohaline circulation; during times when deep water formation in the North Atlantic ceased, the thermohaline circulation comes to a halt, and heat accumulates in the tropical Atlantic at the same time as the North Atlantic cools due to the lack of northward advection of warm surface water. On the other hand, the timing of deglacial SST increases in the Pacific and Indian Oceans relative to high latitude warming is still controversial. Based on a radiocarbon-dated alkenone thermometry (UK 3 7)-SST record from the tropical northwestern Indian Ocean, Bard et al. [1997] inferred an 15 interhemispheric synchrony of deglacial warming in the Arabian Sea and Greenland, specifically during the so-called Boiling-Transition at the end of the last glaciation. This Indian Ocean UK ' 3 7-SST change, however, leads planktonic foraminiferal S 1 8 0 from the same core [Duplessy et al, 1991] during this abrupt event. A similar lead of foraminiferal M g / Ca-derived SST estimates versus S l sO, as well as the correspondence between equatorial Pacific foraminiferal Mg/Ca and Antarctic temperature records, however, led Lea et al. [2000] to postulate a lead of tropical Pacific deglacial SST increase versus ice volume, and a synchroneity with Antarctic warming during the last deglaciation. The last deglaciation was not a smooth transition from one climate state to another, but rather occurred in a series of abrupt warming steps interrupted by sudden shifts towards near-glacial conditions. The most prominent warming phase of the northern hemisphere is the Boiling transition at the end of the last glaciation that initiated the Bolling-Allerod (B/A) warm period [12,600-10,880 1 4 C yrs B.P., equivalent to 14,700-12,800 cal. yr B.P.; ages adopted from Stuiver et al, 1995]. This is followed by a change to near-glacial conditions, the so-called Younger Dryas (YD) period [10,800-10,000 1 4 C yrs B.P., equivalent to 12,800-11,600 cal. yr B.P; ages adopted from Stuiver et al, 1995]. Distinctly different climate conditions during both of these periods are well recorded in Greenland ice cores [Dansgaard et al, 1993; Dansgaard et al, 1989; Grootes et al, 1993], as well as in European [Mangerud et al, 1974; von Grafenstein et al, 1999] and North [Engstrom et al, 1990] and central [Ledru et al, 2002] American terrestrial records, and North Atlantic [Duplessy et al, 1981; Bond et al, 1993] and northeast Pacific marine sedimentary [Mix et al, 1999b; S. Kienast and McKay, 2001] records. Changes in surface ocean conditions in the tropical western Pacific and Indian Oceans during the YD climate event have been inferred from 14C-dated foraminiferal oxygen isotope records from the Sulu Sea [cores ODP769 and S058-69KL/S049-82KL; see Fig. 2.1 for core location; Linsley and Thunell, 1990; Kudrass et al, 1991], the South China Sea [SCS; core 17940-2, see Fig. 2.1 for core location; Wang et al, 1999], and the Arabian Sea [Anderson and Thunell, 1993]. However, considerable debate exists on whether the heavier oxygen isotope ratios during this time interval are indeed a reflection of decreased SSTs or of other environmental processes. Based on foraminiferal abundances in SCS core V35-5 [see Fig. 2.1 for core location; Broecker et al, 1988], as well as on foraminiferal transfer function SST estimates from the Arabian Sea [Anderson and Thunell, 1993] and the eastern SCS [core GGC-9, see Fig. 2.1 for core location; Miao and Thunell, 1996], these authors concluded that the cooling of the higher northern latitudes during the YD did not extend southward to the tropical ocean. However, recent SST estimates based on the U K ' 3 7 method seem to indicate a minor cooling (ca. 0.5 16 105° 110° 115° 120° E Fig. 2.1. Map of the South China Sea and adjacent seas showing the location of cores 18252-3 and 18287-3 (asterisks). The locations of other cores discussed in the text are also indicated (solid circles): GGC-9 [Thunell and Miao, 1996], SCS-12 [Jian et al, 1998], 17940, 17961, and 17964 [Pelejero et al, 1999; Wang et al, 1999], V35-5 [Broecker et al, 1988; Duplessy et al, 1991], ODP 769 [Anderson and Thunell, 1993; Linsley and Thunell, 1990] and S049/S058 [Kudrass et al, 1991]. The 100 m isobath (thick dark line) approximately represents the coastline during the last glacial maximum. The inferred course of the glacial Molengraaff or North Sunda river system is shown according to Molengraaff (1921; stippled lines). Isotherms represent the modern annual sea surface temperatures (in °C) at 0 m water depth [according to Levitus and Boyer, 1994]. 17 °C) associated with the YD climate interval in the western Indian Ocean [Bard et al, 1997; Sonzogni et al, 1998], as well as in the northern and southern SCS [cores 17940-2 and 17964-3, see Fig. 2.1 for core location; Pelejero et al, 1999]. Cayre and Bard [1999] first attempted to unravel this conflicting evidence by directly comparing the micropalaeontological (planktonic foraminifera) and geochemical (alkenones) palaeo-thermometers in one core (MD77194) from the eastern Arabian Sea, concluding that the YD is indeed associated with a decrease in SSTs of the order of 0.5-1 °C. In the Arabian Sea, however, SST estimates as well as foraminiferal assemblages are significantly affected by changes in monsoon-driven upwelling intensity, potentially masking deglacial changes in atmospherically-conditioned temperature. In this chapter, SST records from the southern SCS covering the late glacial to Holocene transition are presented. In the first part (Chapter 2.2), three different techniques of SST reconstruction from a single core (18287-3, see Fig. 2.1 for core location) are compared in an attempt to resolve the amplitude of deglacial SST changes in the southern SCS. In section 2.3, I evaluate the timing of the Boiling warming in the southern SCS based on two AMS 1 4C-dated UK ' 3 7-SST and foraminiferal 6 l sO records (cores 18252-3 and 18287-3, see Fig. 2.1 for core location). Section 2.4 focuses on the deglacial pattern of SST change in the southern SCS. The key results are summarized in section 2.5. 2.2 The amplitude of glacial cooling in the SCS: A comparison of different methods to estimate palaeo-SSTs 2.2.1 Alkenone SST estimates of core 18287-3 The late glacial part (16.5-14.6 cal. ka, which overlaps with Heinrich Event 1 in the North Atlantic realm) from core 18287-3 shows low alkenone (U K ' 3 7) SST estimates (see Appendix IIA for analytical details) of ca. 26 °C, whereas the Holocene displays warm temperatures of 27.2-28.3 °C, in agreement with modern annual sea surface temperatures (Fig. 2.2). A slight warming during the late glacial is followed by an abrupt warming step (ca. 1 °C in <200 years) at the Belling transition (ca. 14,600 cal yr B.P.; see discussion in Chapter 2.3), whereas the YD period is marked by a temperature decrease of 0.2-0.6 °C compared to the Bolling-Allerod temperatures (see discussion in Chapter 2.4). The second warming step of the last deglaciation, following the YD, occurs rather gradually, and modern SSTs are only reached at ca. 6,500 cal yr B.P. This timing and amplitude of the U K ' 3 7 SST record of core 18287-3 correlates very well with recently 18 28 27 H O £_ H (0 0) CO * 26 25 Y D B A core 18287-3 5°39" N/110°39'E 4 1 - j -10 12 age ka 14 16 18 Fig. 2.2. Alkenone (UK ' 3 7) SST estimates of core 18287-3 versus age (see figure caption 2.6 for details on the age model). Shaded bars indicate the Younger Dryas (YD; blue) and Bolling-Allerod (BA; red) intervals [ages adopted from Stuiver etal, 1995]. 100 n •D C i <0 m 3 E 80 60 40 20 ^cosjoopolltan f YD BA subtropical-tropical " - ^ — / ~ ^ cool-tornp«nrto 8 10 12 cal. age (BP) 14 16 18 Fig. 2.3. (a) Cumulative percentages of major species groups of planktonic foraminifera in core 18287-3 versus age [from Steinke et al. 2001]. Shaded bars indicate the Younger Dryas (YD; blue) and Bolling-Allerod (BA; red) intervals [ages adopted from Stuiver et al, 1995]. 19 published, lower resolution U k 3 7 temperature estimates from two cores from the southern SCS [sites 17961 and 17964, Pelejero et al, 1999]. 2.2.2 Planktonic foraminiferal SST estimates of core 18287-3 The planktonic foraminiferal assemblage in the southern SCS is dominated by Globigerinoides ruber (10-35 %) and Globigerinoides sacculifer (5-30 %), indicating tropical to subtropical conditions throughout the record [Fig. 2.3; all foraminiferal data from Steinke, 2001 and Steinke et al, 2001]. However, higher abundances of cool- to temperate-water species such as Neogloboquadrina pachyderma (dextral) and the pachyderma-dutertrei intergrades (up to 19 %), Globorotalia inflata (up to 4%) and Globigerina bulloides (up to 14%) in the lower section of the core indicate distinctly different environmental conditions during the late glacial period (Fig. 2.3). No distinct faunal change is observed during the YD time interval at site 18287-3 (Fig. 2.3; see discussion below), corroborating observations by Thunell and Miao [1996] in the eastern SCS off Palawan (core GGC-9). Based on principal components analysis, various techniques have been developed to convert planktonic foraminiferal abundances to SST estimates. We have applied three techniques to the foraminiferal census data of core 18287-3, tranfer functions FP-12E [Thompson, 1981] and SIMMAX-28 [Pflaumann and Jian, 1999], and the revised analog method [RAM; Waelbroeck et al, 1998]. SIMMAX SST estimates show very stable summer and winter SSTs of ca. 29 °C and ca. 27 °C during the last 16,500 years (Fig. 2.4a), except for a marked winter SST minimum at ca. 16,000 cal yr B.P. A low similarity coefficient in the lower part of the record indicates that these faunal assemblages are not optimally represented in modern analogs and that therefore the SST estimates are less reliable, specifically the SST minimum noted above. In contrast, RAM-derived estimates reveal nearly unchanged summer estimates of around 29 °C (Fig. 2.4b), but the winter estimates are lower during the late glacial period (ca. 24 °C) compared to the Holocene average of ca. 26.5 °C. The increase in winter SSTs to Holocene levels occurred slightly after the sudden decrease in planktonic S 1 80 (see further discussion below). In contrast to the low variability of the 8 1 80 values of G. ruber in the lower-most part of the core, RAM-derived estimates show markedly higher late glacial winter SSTs before ca. 16,000 cal yr B.P. The average difference between the summer and winter SST estimates is ca. 5 °C during the late glacial period, indicating a significant increase in seasonality compared to the Holocene (ca. 2.5 °C). 20 a « 30 s c 6) E E 3 (0 (0 26 24 22 20 -SIMMAX28 1 1 1 ! 1 1 —r* 0) * -c 1 E E 3 m c 1 E E to (0 10 12 cal age ka 14 8 10 12 cal age ka 14 "i i r 8 10 12 14 cal age ka 16 16 16 at | 0) o.8 m 18 18 18 Fig. 2.4. Estimated summer (red dots) and winter (blue dots) sea-surface temperatures of core 18287-3 using different foraminiferal transfer functions: (a) SIMMAX28 [Pflaumann and Jian, 1999] with similarity indices (black diamonds), (b) R A M [Waelbroeck et al, 1998] with dissimilarity coefficients (black diamonds), (c) FP-12E [Thompson, 1981] with communality (black diamonds). Vertical thin line is drawn at 14.6 ka to highlight the timing /phasing of SST increases during the Belling warming. 21 The FP-12E-derived summer SST estimates are the only foraminiferal SST estimates that show slightly decreased glacial summer SSTs (ca. 28 °C, compared to average Holocene values of ca. 29 °C), whereas winter estimates vary from 21-25 °C during the late glacial period to 25-27 °C during Termination lb and the Holocene, similar to the R A M and SIMMAX estimates (Fig. 2.4c). These FP-12E SSTs of core 18287-3 are in good quantitative agreement with FP-12E SST estimates of Miao et al. [1994] and Jian et al. [1998] from the eastern (core GGC-9) and southwestern (core SCS-12; see Fig. 2.1 for core location) SCS, respectively. The most prominent increase in summer and winter SSTs occurred at ca. 14,600 cal yr B.P., synchronous with the sudden decrease of o 1 80 G . r a t e f . The difference between the winter and summer SSTs is as high as ~7 °C during the late glacial period, indicating that the seasonality was much stronger than today (2.5 °C). However, with all three transfer functions used, the reliability of planktonic foraminiferal SST estimates, as indicated by the dissimilarity, similarity and communality indexes, respectively, is lower during the late glacial period (16,500-14,500 cal yr B.P.) as compared to the estimates for the last 14,000 years (Fig. 2.4a-c). Despite high similarities between modern analogs and YD faunal assemblages, but as expected from the unchanged faunal assemblages during the YD period, neither of the planktonic foraminiferal estimates shows a distinct temperature change associated with this climate interval in the southern SCS. 2.2.3 Foraminiferal Mg/Ca SST estimates from core 18287-3 The observation that the Mg/Ca ratio in inorganic [e.g.,Katz, 1973] and biogenic [e.g., Nurnberg et al, 1996; Rosenthal et al, 1997; Dekens et al, 2002] calcite is temperature dependent has led to the establishment of a new palaeotemperature proxy, foraminiferal Mg/Ca palaeothermometry, which is used on both planktonic and benthic foraminiferal tests to reconstruct surface and deep ocean temperatures, respectively [e.g., Hastings et al, 1998; Mashiotta et al, 1999; Lea et al, 2000; Lear et al, 2000; Nurnberg et al, 2000]. Planktonic foraminiferal (G. ruber) Mg/Ca ratios of core 18287-3 have been determined by Dr. David Hastings at Eckerd College (unpubl. data, 2001), following analytical techniques detailed in Hastings et al. [1998]. Mg/Ca SST estimates were derived using the calibration established by Lea et al. [2000] for the equatorial Pacific. The M g / C a SST estimates of core 18287-3 range between 26.1 and 28 °C during the late glacial, and between 27.4 and 30.3 °C during the Holocene (Fig. 2.5). A pronounced warming step of almost 2 °C occurs at ca. 14.6 ka. Thus, on average, the 22 b o ] i s s 23 M g / C a SST estimates are higher by 1-2 °C than the estimates derived by the alkenones and foraminiferal transfer functions, possibly due to the use of the open Pacific calibration to convert M g / C a ratios to SST estimates instead of a local calibration [Hastings et al, in prep.]. Furthermore, the higher sample-to-sample variability of the M g / C a SST estimates as compared to those derived from alkenones could be due to variable partial dissolution of the planktonic foraminiferal tests, which has been shown to affect the M g / C a ratio significantly [e.g., Rosenthal et al, 2000a]. Measurements of the foraminiferal shell weights will be used to constrain this in the future [Hastings et al, in prep.]. 2.2.4 A comparison In order to compare the foraminiferal transfer function FP-12E SST estimates more reliably with those based on alkenones and foraminiferal M g / C a ratios, we calculated the annual faunal SST by averaging the winter and summer SST estimates (Fig. 2.5). In contrast to the faunal and Mg/Ca SST estimates, the downcore U K 3 7 -SST record is smoother and has less high-amplitude variability. However, despite the larger error associated with the FP-12E SST estimates as compared to the U K ' 3 7 method, and even though lower communality coefficients during the late glacial period urge caution in accepting the foraminiferal SST estimates, both the FP-12E foraminiferal and the U K 3 7 alkenone SST estimates record an identical average late glacial to Holocene (i.e. 16,500 to 14,600 cal yr B.P. vs. 10,000 to 3,000 cal yr B.P.) SST difference of 2.0 °C (Fig. 2.5). Similarly, the foraminiferal M g / C a SST estimates record a late glacial-interglacial SST difference of 2.4 °C (Fig. 2.5). These estimates agree very well with the CLIMAP reconstruction of tropical western Pacific and southern SCS temperature LGM-Holocene anomalies of ca. -2 °C [CLIMAP project members, 1981], and are similar to the L G M -Holocene temperature U K 3 7-SST contrast reported by Pelejero et al. [1999] for the southern SCS (core 17964-3; ca. -2,5 °C), and by Sonzogni et al. [1998] for the tropical Indian Ocean (-1.5 to -2.5 °C). However, the reconstructions of the glacial-interglacial SST difference presented here are slightly lower than foraminiferal Mg/Ca based estimates of 2.8-3.0 °C in the equatorial Pacific [Lea et al, 2000] and the Sulu Sea [Rosenthal et al, 2000b], and significantly lower than temperature reconstructions based on noble gas concentrations in ground water from the Mekong delta [Stute et al, 1997] and on snow-line reconstructions from Borneo [Porter, 2001], which both suggest a glacial cooling of ca. 5 °C. The small difference between the SST estimates of core 18287-3 and the tropical Pacific M g / C a estimates could be due to the fact that core 18287-3 may not record the 24 full glacial-interglacial SST contrasts at this site because the record ends at ca. 16,500 cal yr B.P. (Fig. 2.2 and 2.5). The significant contrast between ground water temperature reconstructions and the U k 37-SST estimates, however, is similar to discrepancies between both methods in Oman and in the northern Arabian Sea [ca. 6.5 vs. <4 °C glacial cooling; Sonzogni et al, 1998; Weyhenmeyer et al, 2000], as well as in lowland Brazil and the adjacent tropical Atlantic [5.4 vs. 3.5 °C; Ruhlemann et al, 1999; Stute et al, 1995]. Clearly, more data are needed to resolve this problem [see Crowley, 2000 for discussion]. Despite the clear differences between the late glacial and the Holocene faunal assemblages (Fig. 2.3), the SIMMAX28 method does not record any change in SST associated with the late glacial-Holocene transition in the SCS, and the R A M method records only an increase in winter SSTs, albeit with a similar amplitude (2.1±0.7 °C) as the annual average estimates reported above (Fig. 2.4). The low sensitivity that underestimates the glacial cooling could be due to the lack of suitable analogues from temperate to subpolar regions in the data set used for calibration of these two transfer functions [Pflaumann and Jian, 1999]. In contrast to the SIMMAX28 and R A M calibration data sets, FP-12E is based on an open-ocean data set including mid-latitude samples [Thompson, 1981], and is thus capable of recording the full late glacial-interglacial SST difference. A detailed comparison of the U K ' 3 7 and the foraminiferal SST estimates covering the last 16,500 years in the southern SCS reveals that the foraminiferal transfer function estimates seem incapable of recording SST variability below the error of the method, i.e. below ca. 1 °C. Thus, the minor SST change associated with the YD and Termination lb in the southern SCS, as indicated by the U K 3 7 estimates, remain undetected using the foraminiferal estimates (Fig. 2.4). This could explain why previous studies, based on foraminiferal transfer function estimates, were unable to detect any SST change during the YD in the tropics [Anderson and Thunell, 1993, core ODP 769; Thunell and Miao, 1996, core GGC-9]. Similarly, the Mg/Ca SST estimates do not show any clear cooling during the YD, possibly due to the noisy nature of the record (see disussion above). A close comparison of the different SST estimates reveals a minor, albeit significant, lag of the R A M estimates by 5-15 cm in recording the SST increase associated with Termination la (Fig. 2.4). Given the synchroneity of both the planktonic foraminiferal oxygen isotope record and the FP-12 SST estimates with the U k ' 3 7-SST estimates, this offset in timing cannot be caused by differential bioturbation of foraminifera and coccoliths, as suggested by Sachs et al. [2000]. Alternatively, this offset in timing could be due to different thresholds of indicator species of the different foraminiferal transfer functions. An even larger offset between foraminiferal transfer 25 function and alkenone SST estimates has been observed by Cayre and Bard [1999] in a record from the Arabian Sea. However, given the importance of constraining phasing relationships between different climate parameters for a better understanding of mechanisms governing past global climate change, this discrepancy in timing, even within different foraminiferal SST estimates, highlights the need for a multiproxy approach in palaeoceanographic studies. Given both the lower sensitivity of planktonic foraminiferal transfer function and M g / C a SST estimates and the similarity in patterns of the planktonic S 1 80 record and the U K 3 7 S S T estimates, only the U K 3 7 estimates are assumed to reliably reflect temporal changes in local SST in the following discussion of the deglacial SST variability in the southern SCS. 2.3 The timing of the Belling warming in the southern South China Sea The late glacial parts of two down-core records from the southwestern (core 18252-3) and southern (core 18287-3; see chapter 2.2) SCS, respectively, show low U K 3 7 SSTs of ca. 25.2 °C and ca. 26 °C, respectively, and high planktonic (G. ruber) S 1 80 values (ca. -1.5 %o; Fig. 2.6). In contrast, the Holocene displays warm temperatures of 26.5-28.0 °C (core 18252-3) and 27.2-28.3 °C (core 18287-3), and low 6 l sO values of ca. -3.1 %> (Fig. 2.6). In both records, planktonic 6 l sO and U K 3 7 -SST estimates vary in concert during the transition from the late glacial to the Holocene, including much of the high-resolution, small-scale variability (Fig. 2.6). The most prominent event in both records is the abrupt warming of at least 1 °C at ca. 14.6 cal. ka. The 6 l sO and U K 3 7-SST trends at both SCS sites are interpreted to reflect uniquely changes in local sea surface conditions, unaffected by variations in, for example, riverine input, advection of different water masses, or upwelling. This assertion is based not only on the similarity of both records despite the different local setting of the core sites, but is corroborated by planktonic foraminiferal census counts (core 18287-3; see Chapter 2.2), as well as organic 61 3C and inorganic (major/minor element composition) geochemical data from both cores (see Chapter 4.1.5 for data of core 18287-3) that do not show any significant variability associated with this particular warming event. The abrupt deglacial warming event is the largest amplitude signal in both SCS sedimentary records as well as the Greenland isotope record, and in both locations it is larger than the analytical error of the U K 3 7 -SST estimates (see Appendix IIA) and planktonic foraminiferal 6 1 8 0 determinations [see M . Kienast et al. 2001 for analytical details]. The extreme rapidity of the event in the SCS is analogous to the Boiling-Fig. 2.6. Planktonic 8 1 8 0 (G.ruber) and alkenone (U K' 3 7) SST estimates of cores 18252-3 and 18287-3 from the southern South China Sea (SCS) versus cal. age, compared to the GISP2 8 l s O record of Stuiver and Grootes [2000] on the age scale of Meese et al. [1997]. The chronologies of cores 18252 and 18287 are based on AMS 1 4 C dates, converted to cal. ages using CALIB 4.3 [Stuiver et ai, 1998]. The age of the onset of the Belling warming in Greenland of 14,660±300 compares to an interpolated age of 15,140/14,600/14,400 (15,360-14,340 l a range) in core 18252, and of 14,570 (14,780-14,250 lo range) in core 18287. The subtle lead of 18252-3 is most likely caused by the large lo range of the age fix point below the midpoint of the warming. Note the close parallelism between planktonic 8 l s O and SST estimates. Note also the clearly detectable cooling following the Belling/Allered warm period in the southern SCS. 27 Transition in the Greenland ice cores (Fig. 2.6). Two independent lines of evidence, each of which has been used in previous studies to establish phasing relations between tropical and high latitude climate, suggest synchroneity (referring to synchronous timing within the inevitable uncertainties of absolute chronologies) of this warming in the SCS and at the Boiling-Transition in Greenland. First, the midpoint of the abrupt warming in the southern SCS has interpolated AMS U C ages of 15,140/14,600/14,400 (15,360-14,340 l a range; core 18252-3) and 14,570 (14,780-14,250 l a range; core 18287-3) calibrated years. Thus, within the l a range of the calibrated AMS U C dates, these ages are identical with the age of the midpoint in the Greenland (GISP2) warming of 14,660 ±300 years ago [Stuiver et al, 1995; Meese et al, 1997]. The ages of the midpoint of the warming event in both SCS cores are not interpolated over any major change in sedimentation rates as inferred from unchanged bulk sediment geochemistry between the AMS 1 4 C control points. Moreover, the calibrated AMS 1 4 C ages have been derived assuming a minimal average oceanic resevoir age of -400 years [Bard, 1988]. Adopting a larger reservoir effect [Wang et al, 1999] would result in a lag of SCS versus Greenland warming rather than a lead. Second, based on the revised chronology of the deglacial rise in sea level from the southern SCS [Hanebuth et al, 2000], the first major melt water pulse (MWP) 1A occurred at 14.6-14.3 cal. ka, synchronous (within dating uncertainties) with the Boiling warming in Greenland (see also Chapter 4.3). MWP 1A is associated with a decrease of 6 1 8 O s e a w a t e r of up to ~0.2 %0, which should also be reflected in a synchronous [Anderson and Thunell, 1993] step-like decrease of 6 1 80G n,(, e r. In both SCS cores there is no indication for such a sustained decrease of S 1 8 O G mber either preceding or postdating the abrupt increase in UK 3 7-SSTs (Fig. 2.6). Thus, the 6 l sO decrease related to MWP 1A most likely occurs synchronously with the temperature-related bl80Gn,ber decrease during the Boiling-Transition in the SCS, thus increasing its amplitude slightly. This evidence also corroborates synchronous warming during the Boiling-Transition in the southern SCS and Greenland. Accordingly, neither SCS core records support the idea that tropical SST increases led Greenland warming during the Boiling-Transition. On the contrary, the records support earlier findings Bard et al. [1997] from the northern Indian Ocean of synchronous deglacial warming in the tropics and high northern latitudes. Previous records from the SCS displayed a similar parallelism between U K 3 7-SST and S 1 8 0 G r u b e r [Pelejero et al, 1999]. There is, however, no radiocarbon age control for the abrupt deglacial warming in core 17964-3 from the southern SCS, and the midpoint of the abrupt warming in the northern SCS (core 17940-2) is radiocarbon dated at 15,970 years [Fig. 2.7, Wang et al, 1999]. This large difference in the radiocarbon age between the warming in northern and southern SCS is most likely due to a significantly higher 28 * CO 28 Comparison of cal. ages of the Boiling warming within the SCS 27 A O £L* 26 H (/) 25 A 24 23 core 18287-3 5°39' N, 110°39E ,570 +210/-325 core 17940-2 20°7,N,117,>23E AMS " C dating • i ' • ' i -60 -40 -20 0 20 40 60 std. depth in cm, midpoint of warming = 0 cm Fig. 2.7. Comparison of the interpolated cal. ages of the midpoint of the Belling-warming in three cores from within the SCS [cores 18252-3 and 18287-3, this study; core 17940-2 from Pelejero et al, 1999 and Wang et al, 1999]. The midpoint of the warming has been fixed to 0 cm to show the comparable resolution at all sites. See Chapter 4.3 for further discussion. 29 reservoir age at the northern site (see further discussion in Chapter 4.3), possibly caused by the advection of old Pacific intermediate- to deep-water masses [Wang et al, 1999]. The close parallelism between 5 1 8 0 G r „ b e r and U K ' 3 7 -SSTs during the last deglaciation in the SCS contrasts markedly with the lead of U K ' 3 7 - and Mg/Ca-SST estimates over bwOG„lber in the Arabian Sea [Bard et al, 1997; Cayre and Bard, 1999], the equatorial Pacific [Lea et al, 2000], the Sulu Sea [Rosenthal et al, 2000b], and the tropical Atlantic [Numberg et al, 2000; Ruhlemann et al, 1999]. The variability of SST in the equatorial Pacific upwelling region during the last 250,000 years has previously been interpreted to reflect variable horizontal and/or vertical advection of different water masses [Lyle et al, 1992], processes that are not likely to have affected the SCS given the secluded nature of the basin, particularly during glacial and early deglacial sea-level lowstands (see Fig. 1.1). Thus, equatorial open-ocean SST variability could be significantly influenced by changes in large-scale oceanic circulation patterns, which appear to show an early response to Southern Hemisphere deglaciation [Lyle et al, 1992; Mix and Morey, 1996], and may not affect the surface ocean in semi-enclosed marginal basins such as the SCS. The tropical SST records examined here suggest a diverse pattern of temporal changes in different parts of the tropical ocean during the glacial-interglacial transition. 2.4 The deglacial pattern of SST change in the SCS Whereas climate records from the high northern latitudes indicate a return to near-glacial conditions during the YD following the Boiling/Allerod warm period [Dansgaard et al, 1989, 1993; Grootes et al, 1993], the UK ' 3 7-SST record of core 18287-3 implies only a minor, albeit clearly detectable, cooling of ca. 0.2-0.6 °C (compared to the preceding Bolling-Allerod period) associated with this climate interval in the tropical SCS. Similarly, core 18252-3 displays a SST decrease following the Bolling-Allerod warm period. However, the AMS radiocarbon age model of this core suggests a persistent cooling beyond the end of this climate period in the North Atlantic region at 11.6 ka (Fig. 2.6) Taken together with the U K 3 7 records from the Arabian Sea [Bard et al, 1997] and the northern SCS [core 17940-2; Pelejero et al, 1999], this implies that the YD is indeed associated with a cooling in the tropics. Moreover, a temperature decrease of ca. 0.2-0.6 °C with respect to the Bolling-Allerod period cannot fully explain the global ice effect-corrected 6 1 80 increase of ca. 0.2-0.5 %o associated with the YD interval in the southern SCS. The remainder is potentially associated with a change in surface ocean o 1 80, possibly due to more arid conditions. 30 In contrast to the similarity of SCS and northern hemisphere temperature changes during the Boiling/Allerod and the YD time intervals, the temperature rise at the end of the YD is less pronounced in the southern SCS, and modern SST levels are reached only at ca. 6.5 cal kyr B.P. (Figs. 2.6 and 2.8). This more gradual post-YD warming as compared to the high northern latitude temperature records is again similar to the warming trend observed in the northern Indian Ocean [Bard et al, 1997], the northern SCS [Pelejero et al, 1999], and the tropical Atlantic [Ruhlemann et al, 1999], once more implying a common tropical SST development. Thus, despite the very different temporal records of the onset of deglacial warming in various tropical marine records, there appears to be strong evidence that tropical Holocene SSTs increased steadily from -10 ka to ~6 ka, stabilizing thereafter [Figs. 2.6, 2.8; Bard et al, 1997; Pelejero et al, 1999; Ruhlemann et al, 1999; this study]. On the other hand, high-resolution mid- and high-northern latitude deglacial U K 3 7 -SST records [Doose et al, 1997; Cacho et al, 1999; Bard et al, 2000; S. Kienast and McKay, 2001; Fig. 2.8] show an early Holocene SST optimum at ~10-9 ka. The establishment of maximum SSTs in the tropics after ~6 ka implies a weaker latitudinal SST contrast between ~10 and 6 ka, at the same time as the development of an increased contrast in seasonal insolation evolved between the equator and high latitudes [Berger, 1987]. 2.5 Conclusions There are four important conclusions to be drawn from the SST reconstructions presented in this chapter. 1) Three independent means of SST reconstruction (alkenones, foraminiferal M g / C a ratios, as well as foraminiferal transfer functions FP-12E a n n u a l a v e r a g e and R A M w i n t e r ) indicate a late glacial cooling of the tropical southern SCS of 2-2.5 °C as compared to the Holocene. These estimates are in good agreement with SST reconstructions from the tropical western Pacific but contrast markedly with regional temperature reconstructions based on snow line depressions and groundwater palaeothermometry. The origin of differences in the absolute SST estimates (foraminiferal M g / C a ratios versus U K 3 7 and FP-12E), and minor offsets in the phasing of major SST-transitions (RAM estimates versus 5 l sO and U K 3 7 ) need to be resolved in future studies. 2) The southern SCS experienced an abrupt warming of at least 1 °C at the end of the last glacial period, which, within the recognized uncertainties of absolute chronologies, 31 minimum latitudinal SST gradient 0 5 10 15 20 25 age ka Fig. 2.8. Comparison of deglacial UK' 3 7-SST records from tropical and mid/high northern latitude sites on different SST axes [on left SST axis: red line, core 18287-3, this study; green line, core M35003-4, Ruhlemann et al, 1999; blue line, core MD79257, Bard et al, 1997; on right SST axis: light green line, core MD95-2043, Cacho et al, 1999; grey line, core JT-96-09, S. Kienast and McKay, 2001]. Note the clear early Holocene SST maximum at the mid/high latitude sites, which is absent in the tropical records. 32 is synchronous with the Boiling warming observed in the Greenland (GISP2) ice core record at 14.6 ka. A similar timing of an abrupt warming during the last deglaciation has been postulated for the Cariaco Basin {Lea et al, unpubl.). This synchroneity of the Boiling warming in both these marginal basins with the deglacial warming in Greenland contrasts markedly with the inferred lead of deglacial warming in the equatorial Pacific and Atlantic, as well as in the northern Indian Ocean. This could indicate a different forcing mechanism of temporal SST variability in marginal basins as compared to the open ocean. 3) Following the Boiling/Allerod warm period, U K 3 7 SST estimates indicate a minor, albeit clearly detectable, cooling in the southern SCS of ca. 0.5 °C. In core 18287-3, this colder period is time-equivalent to the YD period in the circum-North Atlantic region, whereas the radiocarbon chronology of core 18252-3 suggests a prolonged cool period compared to the northern hemisphere records. 4) In agreement with other tropical palaeo-SST records, the SST reconstructions from the southern SCS presented here indicate a delay in the establishment of Holocene temperatures as compared to the higher northern latitudes (ca. 6 ka versus ca. 10 ka). This implies a weaker latitudinal SST gradient during the Preboreal, quite possibly in response to intensified summer monsoonal redistribution of incoming solar radiation from low to high latitudes during this time. 33 2.6 References Anderson, D. M. , and R. C. Thunell, The oxygen-isotope composition of tropical ocean surface water during the last deglaciation, Quaternary Science Reviews, 12, 465-473, 1993. Bard, E., Correction of accelerator mass spectrometry 1 4 C ages measured in planktonic foraminifera: paleoceanographic implications, Paleoceanography, 3, 635-645,1988. Bard, E., F. Rostek, and C. Sonzogni, Interhemispheric synchrony of the last deglaciation inferred from alkenone palaeothermometry, Nature, 385, 707-710,1997. Bard, E., F. Rostek, J. L. Turon, and S. Gendreau, Hydrological impact of Heinrich events in the subtropical northeast Atlantic, Science, 289,1321-1324, 2000. Beck, J. W., J. Recy, F. Taylor, R. L. Edwards, and G. Cabioch, Abrupt changes in Holocene tropical sea surface temperature derived from coral records, Nature, 385, 705-707,1997. Berger, A. L., Long-term variations of caloric insolation resulting from the Earth's orbital elements, Quaternary Research, 9,139-167,1987. Bond, G., W. Broecker, S. Johnsen, J. McManus, L. Labeyrie, J. Jouzel, and G. Bonani, Correlations between climate records from North Atlantic sediments and Greenland ice, Nature, 365,143-147,1993. Broecker, W. S., Oxygen isotope constraints on surface ocean temperatures, Quaternary Research, 26,121,1986. Broecker, W. S., Paleocean circulation during the last deglaciation: A bipolar seesaw, Paleoceanography, 13,119-121,1998. Broecker, W. S., M . Andree, M . Klas, G. Bonani, W. Wolfli, and H . Oeschger, New evidence from the South China Sea for an abrupt termination of the last glacial period, Nature, 333,156-158,1988. Cacho, I., J. O. Grimalt, C. Pelejero, M . Canals, F. J. Sierro, J. A . Flores, and N . Shackleton, Dansgaard-Oeschger and Heinrich event imprints in Alboran Sea paleotemperatures, Paleoceanography, 14, 698-705,1999. Cane, M . A., A role for the tropical Pacific, Science, 282, 59-61,1998. Cane, M . A., and A . C. Clement, A role for the tropical Pacific coupled ocean-atmosphere system on Milankovitch and millennial timescales. Part II: Global impacts, in Mechanisms of Global Climate Change at Millennial Time Scales, edited by Clark, P. U., R. S. Webb and L. D. Keigwin, pp. 373-383,1999. Cayre, O., and E. Bard, Planktonic foraminiferal and alkenone records of the last deglaciation from the eastern Arabian Sea, Quaternary Research, 52, 337-342,1999. CLIMAP project members, Seasonal reconstructions of the Earth's surface at the Last Glacial Maximum, Geological Society of America Map and Chart Series, MC-36, 18pp., 1981. Crowley, T. J., CLIMAP SSTs re-revisited, Climate Dynamics, 16, 241-255, 2000. Dansgaard, W., J. W. C. White, and S. J. Johnsen, The abrupt termination of the Youunger Dryas climate event, Nature, 339, 532-534,1989. Dansgaard, W., S. J. Johnsen, H . B. Clausen, D. Dahl-Jensen, N . S. Gundestrup, C. U . Hammer, C. S. Hvidberg, J. P. Steffensen, A. E. Sveinbornsdottir, J. Jouzel, and G. Bond, Evidence for general instability of past climate from a 250-kyr ice-core record, Nature, 364, 218-220,1993. Dekens, P. S., D. W. Lea, D. K. Pak, and H . J. Spero, Core top calibration of M g / C a in tropical foraminifera: Refining paleotemperature estimation, Geochemistry, Geophysics, Geosystems, 3, 2002. Doose, H. , F. G. Prahl, and M . Lyle, Biomarker temperature estimates for modern and last glacial surface waters of the California Current system between 33° and 42°N, Paleoceanography, 12, 615-622,1997. 34 Duplessy, J.-C, G. Delibrias, J. L. Turon, C. Pujol, and J. Duprat, Deglacial warming of the northeastern Atlantic Ocean: Correlation with the paleoclimatic evolution of the European continent, Palaeogeography, Palaeoclimatology, Palaeoecology, 35, 121-144, 1981. Duplessy, J. C , E. Bard, M . Arnold, N . J. Shackleton, J. Duprat, and L. Labeyrie, How fast did the ocean-atmosphere system run during the last deglaciation?, Earth and Planetary Science Letters, 103, 27-40,1991. Engstrom, D. R., B. C. S. Hansen, and H . E. Wright, A possible Younger Dryas record in southeastern Alaska, Science, 250,1383-1385,1990. Farrera, I., S. P. Harrison, I. C. Prentic, and et al., Tropical climates at the Last Glacial Maximum: a new synthesis of terrestrial palaeoclimate data. I. Vegetation, lake-levels and geochemistry, Climate Dynamics, 15, 823-856,1999. Grootes, P. M . , M . Stuiver, J. W. C. White, S. Johnsen, and J. Jouzel, Comparison of oxygen isotope records from the GISP2 and GRIP Greenland ice cores, Nature, 366, 552-554,1993. Guilderson, T. P., R. G. Fairbanks, and J. L. Rubenstone, Tropical temperature variations since 20,000 years ago: modulating interhemispheric climate change, Science, 263, 663-665,1994. Guilderson, T. P., R. G. Fairbanks, and J. L. Rubenstone, Tropical Atlantic coral oxygen isotopes: glacial-interglacial sea surface temperature and climate change, Marine Geology, 172, 75-89, 2001. Hanebuth, T., K. Stattegger, and P. M . Grootes, Rapid flooding of the Sunda Shelf: A late-glacial sea-level record, Science, 288,1033-1035, 2000. Hastings, D. W., A . D. Russell, and S. R. Emerson, Foraminiferal magnesium in Globigerinoides sacculifer as a paleotemperature proxy, Paleoceanography, 13, 161-169, 1998. Hastings D, Whitko A, Kienast M , Steinke S, Comparison of three independent paleotemperature estimates from a high resolution record of deglacial SST records in the tropical South China Sea, 2002, in prep. Jian, Z., M . Chen, H . Lin, and P. Wang, Stepwise paleoceanographic changes during the last deglaciation in the southern South China Sea: records of stable isotope and microfossils, Science in China (Series D), 41,187-194,1998. Katz, A., The interaction of magnesium with calcite during crystal growth at 25-90 °C and one atmosphere, Geochimica et Cosmochimica Acta, 37,1563-1586,1973. Kienast, M . , S. Steinke, K. Stattegger, and S. E. Calvert, Synchronous tropical South China Sea SST change and Greenland warming during deglaciation., Science, 291, 2132-2134, 2001. Kienast, S. S., and J. McKay, Sea surface temperatures in the subarctic northeast Pacific reflect millenial-scale climate oscillations during the last 16 kyrs, Geophysical Research Letters, 28,1563-1566, 2001. Kudrass, H . R., H . Erlenkeuser, R. Vollbrecht, and W. Weiss, Global nature of the Younger Dryas cooling event inferred from oxygen isotope data from the Sulu Sea, Nature, 349, 406-409,1991. Lea, D. W., D. K. Pak, and H . J. Spero, Climate impact of late Quaternary equatorial Pacific sea surface temperature variations, Science, 289,1719-1724, 2000. Lear, C. H., H . Elderfield, and P. A. Wilson, Cenozoic deep-sea temperatures and global ice volumes from Mg/Ca in benthic foraminiferal calcite, Science, 287, 269-272, 2000. Ledru, M.-P., P. Mourguiart, G. Ceccantini, B. Turcq, and A. Sifeddine, Tropical climates in the game of two hemispheres revealed by abrupt climate change, Geology, 30, 275-278, 2002. Levitus, S., and T. Boyer, World Ocean Atlas Volume 4: Temperature, NOAA Atlas NESDIS, U.S. Govt. Printing Office, 1994. 35 Linsley, B. K., and R. C. Thunell, The record of deglaciation in the Sulu Sea: evidence for the Younger Dryas event in the tropical western Pacific, Paleoceanography, 5, 1025-1039, 1990. Lyle, M. , F. G. Prahl, and M . A . Sparrow, Upwelling and productivity changes inferred from a temperature record in the central equatorial Pacific, Nature, 355, 812-815, 1992. Mangerud, J., S. T. Andersen, B. E. Beglund, and J. J. Dormer, Quaternary stratigraphy of Norden, a proposal for terminology and classification, Boreas, 3,109-128,1974. Mashiotta, T. A., D. W. Lea, and H . J. Spero, Glacial-interglacial changes in Subantarctic sea surface temperature and 5lsO-water using foraminiferal Mg, Earth and Planetary Science Letters, 170, 417-432,1999. Meese, D. A., A. J. Gow, R. B. Alley, G. A . Zielinski, P. M . Grootes, M . Ram, K. C. Taylor, P. A. Mayewski, and J. F. Bolzan, The Greenland Ice Sheet Project 2 depth-age scale: Methods and results, Journal of Geophysical Research, 102, 26,411-426,423,1997. Miao, Q., and R. C. Thunell, Late Pleistocene-Holocene distribution of deep-sea benthic foraminifera in the South China Sea and Sulu Sea: paleoceanographic implications, Journal of Foraminiferal Research, 9-23,1996. Miao, Q., R. C. Thunell, and D. M . Anderson, Glacial-Holocene carbonate distribution and sea surface temperatures in the South China and Sulu seas, Paleoceanography, 9, 269-290,1994. Mix, A., and A. E. Morey, Climate feedback and Pleistocene variations in the Atlantic South Equatorial Current, in The South Atlantic: Present and past circulation, edited by Wefer, G., W. H . Berger, G. Siedler and D. J. Webb, pp. 503-525, Springer-Verlag, 1996. Mix, A., A. E. Morey, N . G. Pisias, and S. W. Hostetler, Foraminiferal faunal estimates of paleotemperature: Circumventing the no-analog problem yields cool ice age tropics, Paleoceanography, 14, 350-359,1999a. Mix, A. C , D. C. Lund, N . G.Pisias, P. Boden, L. Bornmalm, M . Lyle, and J. Pike, Rapid climate oscillations in the northeast Pacific during the last deglaciation reflect northern and southern hemisphere sources, in Mechanisms of Global Climate Change at Millenial Time Scales, edited by Clark, P. U., R. S. Webb and L. D. Keigwin, pp. 127-148, American Geophysical Union, 1999b. Nurnberg, D., J. Bijma, and C. Hemleben, Assessing the reliability of magnesium in foraminiferal calcite as a proxy for water mass temperature, Geochimica et Cosmochimica Acta, 60, 803-814,1996. Nurnberg, D., A. Muller, and R. R. Schneider, Paleo-sea surface temperature calculations in the equatorial east Atlantic from M g / C a ratios in planktic foraminifera: A comparison to sea surface temperature estimates from U 3 7 , oxygen isotopes, and foraminiferal trasnfer function, Paleoceanography, 15,124-134, 2000. Pelejero, C , and J. O. Grimalt, The correlation between the U K ' 3 7 index and sea surface temperatures in the warm boundary: the South China Sea, Geochimica et Cosmochimica Acta, 61, 4789-4797,1997. Pelejero, C , J. O. Grimalt, S. Heilig, M . Kienast, and L. Wang, High resolution U K 3 7 -temperature reconstructions in the South China Sea over the past 220 kyr, Paleoceanography, 14, 224-231,1999. Pflaumann, U., and Z. Jian, Modern distribution patterns of planktonic foraminifera in the South China Sea and western Pacific: a new transfer technique to estimate regional sea-surface temperatures, Marine Geology, 156, 41-83,1999. Pierrehumbert, R. T., Huascaran 5 l sO as an indicator of tropical climate during the Last Glacial Maximum, Geophysical Research Letters, 26,1345-1348,1999. Porter, S. C , Snowline depression in the tropics during the Last Glaciation, Quaternary Science Reviews, 20,1067-1091, 2001. 36 Rosenthal, Y., E. A. Boyle, and N . Slowey, Temperature control on the incorporation fo magnesium, strontium, fluorine, and cadmium into benthic foraminiferal shells from Little Bahama Bank: Prospects for thermocline paleoceanography, Geochimica et Cosmochimica Acta, 61, 3633-3643,1997. Rosenthal, Y., G. P. Lohmann, K. C. Lohmann, and R. M . Sherrell, Incorporation and preservation of Mg in Globigerinoides sacculifer: Implications for reconstructing the temperature and 1 8 0 / 1 6 0 of seawater, Paleoceanography, 15,135-145, 2000a. Rosenthal, Y., D. W. Oppo, S. Dannemann, and B. K. Linsley, Millennial-scale variations in western equatorial Pacific sea surface temperature during glacial and Holocene climates, EOS - Transactions of the American Geophysical Union, 81, F657, 2000b. Ruhlemann, C., S. Mulitza, P. J. Miiller, G. Wefer, and R. Zahn, Warming of the tropical Atlantic Ocean and slowdown of thermohaline circulation during the last deglaciation, Nature, 402, 511-514,1999. Sachs, J. P., R. R. Schneider, T. I. Eglinton, K. H . Freeman, G. Ganssen, J. F. McManus, and D. W. Oppo, Alkenones as paleoceanographic proxies, Geochemistry, Geophysics, Geosystems, 1, 2000GC000059, 2000. Sonzogni, C , E. Bard, and F. Rostek, Tropical sea-surface temperatures during the last glacial period: a view based on alkenones in Indian Ocean sediments, Quaternary Science Reviews, 17,1185-1201,1998. Steinke, S., Sedimentological and climatic changes during the last deglaciation recorded in cores from the Sunda Shelf margin and continental slope (Southern South China Sea). Ph.D. Thesis, Univ. Kiel, Germany, 2001. Steinke, S., M . Kienast, U. Pflaumann, M . Weinelt, and K. Stattegger, A high resolution sea surface temperature record from the tropical South China Sea (16,500 - 3000 yr B.P.), Quaternary Research, 55, 352-362, 2001. Stocker, T. F., The seesaw effect, Science, 282, 61-62,1998. Stacker, T. F., Past and future reorganizations in the climate system, Quaternary Science Reviews, 19, 301-319, 2000. Stuiver, M. , and P. M . Grootes, GISP2 oxygen isotope ratios, Quaternary Research, 53, 277-284, 2000. Stuiver, M. , P. M . Grootes, and T. F. Braziunas, The GISP2 6 l sO climate record of the past 16,500 years and the role of the sun, ocean, and volcanoes, Quaternary Research, 44, 341-354,1995. Stuiver, M. , P. J. Reimer, E. Bard, J. W. Beck, G. S. Burr, K. A. Hughen, B. Kromer, G. McCormac, J. van der Pflicht, and M . Spurk, INTCAL98 radiocarbon age calibration, 24,000-0 cal BP, Radiocarbon, 40,1041-1083,1998. Stute, M . , M . Forster, H . Frischkorn, A. Serejo, J. F. Clark, P. Schlosser, W. S. Broecker, and G. Bonani, Cooling of tropical Brazil (5° C) during the last glacial maximum, Science, 269, 379-383,1995. Stute, M. , L. Van Khoi, P. Schlosser, M . Tobin, and S. Higgins, Glacial paleotemperature records for the tropics derived from noble gases dissolved in groundwater, EOS -Transactions of the American Geophysical Union, 78, F44,1997. Thompson, L. G., E. Mosley-Thompson, M . E. Davis, P.-N. Lin, K. A. Henderson, J. Cole-Dai, J. F. Bolzan, and K.-b. Liu, Late glacial stage and Holocene tropical ice core records from Huascaran, Peru, Science, 269, 46-50, 1995. Thompson, P. R., Planktonic foraminifera in the western North Pacific during the past 150 000 years: comparison of modern and fossil assemblages, Palaeogeography, Palaeoclimatology, Palaeoecology, 35, 241-279,1981. Thunell, R. C , and Q. Miao, Sea surface temperature of the western equatorial Pacific Ocean during the Younger Dryas, Quaternary Research, 46, 72-77,1996. Tudhope, A. W., C. P. Chilcott, M . T. McCulloch, E. R. Cook, J. Chappell, R. M . Ellam, D. W. Lea, J. M . Lough, and G. B. Shimmield, Variability in the El Nino-Southern Oscillation through a glacial-interglacial cycle, Science, 291,1511-1517, 2001. 37 von Grafenstein, U., H . Erlenkeuser, A. Brauer, J. Jouzel, and S. J. Johnsen, A mid-European decadal isotope-climate record from 15,500 to 5000 years B.P., Science, 284, 1654-1657,1999. Waelbroeck, C., J.-C. Duplessy, E. Michel, L. Labeyrie, D. Paillard, and J. Duprat, The timing of the last deglaciation in North Atlantic climate records, Nature, All, 724-727, 2001. Waelbroeck, C., L. Labeyrie, J.-C. Duplessy, J. Guiot, M . Labracherie, H . Leclaire, and J. Duprat, Improving past sea surface temperature estimates based on planktonic fossil faunas, Paleoceanography, 13, 272-283,1998. Wang, L., M . Sarnthein, H . Erlenkeuser, J. Grimalt, P. Grootes, S. Heilig, E. Ivanova, M . Kienast, C. Pelejero, and U. Pflaumann, East Asian monsoon climate during the late Quaternary: High-resolution sediment records from the South China Sea, Marine Geology, 156, 245-284,1999. Weyhenmeyer, C , S. J. Burns, H . N . Waber, W. Aeschbach-Hertig, R. Kipfer, H . H . Loosli, and A. Matter, Cool glacial temperatures and changes in moisture source recorded in Oman groundwaters, Science, 287, 842-845, 2000. 38 3. Sedimentary biogeochemistry 3.1 Introductory remarks The carbon and nitrogen isotopic composition of bulk organic matter (8 1 3C o r g , 5 1 5 N ; see Appendix II C, D for analytical details) is widely used in marine biogeochemical studies to assess the present and past cycling and inventories of these elements on local [e.g., Mariotti et al, 1984; Altabet, 1988, 1996; Nakatsuka et al, 1997; Bertrand et al, 2000], regional [Degens et al. 1969; Peters et al. 1978; Calvert et al, 1992; Altabet and Frangois, 1994a; Farrell et al, 1995; Frangois et al. 1997; Struck et al, 2001; S. Kienast et al., 2002], and global [Rau et al. 1991a; Ganeshram et al. 1995, 2000; Altabet et al. 1995, 1999a, 2002; Suthhof et al. 2001] scales. Here I use downcore records of 5 1 5 N and 5 1 3 C o r g from the SCS to hindcast the past isotopic composition of nitrate in the SCS and the adjacent western Pacific, and to assess the credibility of low-latitude pC02 estimates based on 5 1 3C o r g , respectively. As detailed in chapter 1, the SCS experienced dramatic palaeoceanographic changes concordant with glacial-interglacial sea level fluctuations. Thus, during glacial times, vast shelf areas in the northern and especially in the southern SCS emerged, generally leading to an increased terrigenous input into the basin of up to five-fold [Schdnfeld and Kudrass, 1993; Huang and Wang, 1998; Wang, 1999] and the development of additional drainage systems on the emergent shelves [Molengraaff, 1921; Gupta et al, 1987; see also Fig. 1.1]. Glacial primary production, as inferred from organic carbon and opal accumulation rates [Sarnthein et al, 1988; Thunell et al, 1992; Lin et al, 1999] and the species assemblages of planktonic and benthic foraminifera [Huang et al, 1997; Kuhnt et al, 1999; Miao and Thunell, 1996; Jian et al, 2001], as well as nutrient concentrations, as inferred from the carbon isotopic composition [Wang et al, 1999] and Cd /Ca ratios [Lin et al, 1999] of planktonic foraminifera, were considerably higher compared to Holocene levels, supposedly favoring oxygen depletion in intermediate and bottom waters [Wang et al, 1999]. Climatology-driven circulation models predict seasonal upwelling to the southeast of Vietnam [Chao et al, 1996; Wiesner et al, 1996], off the Sunda Shelf [Chao et al, 1996], and to the northwest of the Philippines [Shaw and Chao, 1994] during interglacials which might have changed during glacials in response to fluctuations in the intensity of the summer monsoon [Wang et al, 1999]. Marked changes in the SCS nutrient budget and productivity regime are inferred from these diverse reconstructions. Their effect on the sedimentary records of 6 1 5 N and 5 1 3 C o r g will be evaluated in this chapter. 39 Extrapolating the findings from the SCS, a semi-isolated basin, especially during glacial sea-level lowstand, to the adjacent Pacific and beyond is justified by the continuous exchange of surface and intermediate waters between the SCS and the western Pacific. Thus, the hydrography of the SCS is mainly controlled by inflow of Pacific Intermediate Water and Kuroshio surface water through the Bashi Strait [sill depth -2000 m; Wyrtki, 1961; La Violette and Frontenac, 1967; Shaw, 1989, 1991; Chao et al, 1996]. According to radiocarbon analyses on both sides of the Bashi Strait [Broecker et al, 1986], the renewal of deep water in the SCS is rapid (<100 years). Similar age differences between planktonic and benthic foraminifera during glacials and interglacials [Andree et al, 1986; Broecker et al, 1988, 1990] imply that this rapid exchange of water masses persisted during the last climatic cycle. It is therefore assumed that the biogeochemical cycles in the SCS are closely coupled to those of the global ocean, and that possible glacial-interglacial changes should be sensitively monitored in SCS sediments. 3.2 Nitrogen isotope records 3.2.1 Introduction The availability of nutrients in the surface ocean acts as an important control on primary production, whose variability may have significant causal influence on glacial-interglacial climatic cycles [Broecker, 1982; McElroy, 1983; Altabet et al, 1995, 2002; Ganeshram et al, 1995, 2000, 2002; Broecker and Henderson, 1998]. Analysis of the nitrogen isotopic composition of sedimentary organic matter has been developed as a technique for reconstructing local/regional changes in nitrate (N03~) availability versus nitrate uptake by phytoplankton. A strong linear relationship between surface water N0 3 " depletion and an increase in 5 1 5 N of the underlying surface sediments has been clearly demonstrated from studies in the North Atlantic [Altabet and McCarthy, 1985; Altabet et al, 1991], the Southern Ocean [Altabet and Francois, 1994a,b; Francois et al, 1992, 1997], the central [Altabet and Frangois, 1994a,b] and eastern [Farrell et al, 1995] equatorial Pacific, the Angola Basin [Holmes et al, 1996], and the Benguela Current region [Holmes et al, 1998], with 6 1 5 N values ranging between 1 and 16 %o. This inverse correlation between nitrate utilization and sedimentary 5 1 5 N is dictated by first-order Rayleigh fractionation kinetics caused by the preferential uptake of 1 4 N0 3 " relative to 1 5 N0 3 " by phytoplankton [Wada and Hattori, 1978; Mariotti et al, 1981]. Accordingly, sedimentary 6 1 5 N values can 40 be used to reconstruct past changes in relative surface ocean nitrate utilization, i.e. changes in the ratio of biological uptake versus physical supply of nitrate [Calvert et al, 1992; Frangois et al, 1992,1997; Altabet and Frangois, 1994b; Farrell et al, 1995; Holmes et al, 1997]. In addition to varying degrees of nitrate utilization, sedimentary 6 1 5 N is governed by the initial isotopic signature of the nitrate utilized by phytoplankton. Deep water N0 3", the principal source of new nitrogen to the surface ocean [Eppley and Peterson, 1979], has a global average 6 1 5 N signature in oxygenated deep waters of 5±1 %o [Altabet, 1988; Wada et al, 1975; Liu and Kaplan, 1989; Liu et al, 1996; Sigman, 1997; Sigman et al, 1997; Wu et al, 1997; Brandes et al, 1998]. In suboxic/anoxic water masses, denitrification, the sequential reduction of nitrate to N z O or N 2 gas [Goering, 1968], produces a 5 1 5 N of the residual N0 3 " that is significantly higher [up to 19 %0; Cline and Kaplan, 1975; Brandes et al, 1998; Altabet et al, 1999b; Voss et al, 2001] than the mean deep water value. In contrast, the lighter 6 1 5 N signature of nitrogen fixing cyanobacteria [-1.7 to 0.5 %0; Wada and Hattori, 1976, 1978; Carpenter et al, 1997] is assumed to lower the 5 1 5 N of the surface ocean nitrate pool significantly [Carpenter et al, 1997; Karl et al, 1997; Brandes et al, 1998]. The uptake of recycled bacterially fixed nitrogen by other phytoplankton should therefore lower the 5 1 5N of settling organic particles. Owing to the dominance of local/regional factors (e.g., degree of utilization and/or 5 1 5 N of nitrate used during primary production) in governing the sedimentary nitrogen isotopic composition, the interpretation of sedimentary nitrogen isotope records to date have been limited to the reconstruction of local /regional phenomena. 5 1 5 N records from the South China Sea that potentially overcome these local constraints are presented here. 3.2.2 Materials The six sediment cores selected for nitrogen isotope analyses (Fig. 3.1, see also Appendix I) were chosen from contrasting settings to enable a detailed study of changes in the various parameters that might affect the nitrogen budget of the SCS. Core 17940-2, located 300 km off the Pearl River mouth, is used to monitor changes in riverine input, which can be contrasted to a more oceanic setting (core 17924-3). The influence of postulated temporal changes in the upwelling intensity off Vietnam on the local nutrient budget should be evident in core 17954-2 at the margin of the upwelling field predicted by Wiesner et al. [1996]. Cores 18284-3, 17964-3, and 17961-2 constitute a profile off the Sunda Shelf and off the glacial river system that drained the glacially emergent shelf and 41 105 110 115 120 125 °E Fig. 3.1. Map of the South China Sea (SCS) and 8 1 5N surface sediment distribution. Core locations are marked with arrows; SHI and SH2 refer to sampling stations 5 and 6, respectively, of Saino and Hattori [1987]. Note that the 100 m isobath approximates the coast line during glacial sea level low stands. 42 will record the influence of the emergence and drowning of this shelf region on glacial-interglacial timescales. To fingerprint isotopically the terrigenous input to the SCS, 14 samples were taken from gravity and vibro cores from the glacially emergent Sunda Shelf [Stattegger et al, 1997]. According to microscopic analyses (T. Hanebuth and S. Steinke, pers. com., 1999), these samples cover a wide range of depositional environments, including mangrove swamps, peat deposits, and soils, and should therefore provide a representative terrigenous isotopic end-member in the southern SCS. 3.2.3 Results and discussion The 5 1 5 N surface sediment distribution The 6 1 5 N values in SCS surface sediments range from 3.4 to 6.6 %0 (Fig. 3.1), similar to & 1 5 N values of surface sediments from the East China Sea [Kao et al, 2001; Minagawa et al, 1999] and the Sulu Sea (M. Kienast, unpubl. data, 1999). Recent studies by Chen et al [1998], Saino and Hattori [1987] and Takahashi and Hori [1984] show nitrate concentrations in SCS surface waters (<50 m) of 0.0 to 0.2 \iM (see also Fig. 3.2), suggesting complete nitrate consumption during primary production. In addition, vertical profiles of suspended particulate organic nitrogen (SPON) in the northern and central SCS [Saino and Hattori, 1987; location of profiles given in Fig. 3.1] show 6 1 5 N values approaching 0%o, which is consistent with N recycling in nitrate-limited environments [Checkley and Miller, 1989; Altabet, 1996; Bertrand et al, 2000]. Moreover, there is no consistent 5 1 5 N gradient with expected nitrate gradients, such as off the river mouths and/or within the inferred upwelling areas, and no correlation with annual average primary production as derived from satellite images [T. Piatt, unpublished data taken from Kuhnt et al, 1999], making different utilization factors seem unlikely. Finally, Waser et al [1999] report a nitrogen isotope fractionation factor E of 0 %o during ammonium uptake following N starvation for a natural assemblage of coastal phytoplankton. The 6 1 5 N surface sediment values therefore most likely reflect the complete utilization of SCS subsurface nitrate with an average isotopic composition of 5.6 %o (Fig. 3.2; see further discussion below and Appendix II C for analytical details), which is ultimately derived from the North Pacific Intermediate Water [5 1 5 N n i t r a t e = 5.7 %o; Liu et al, 1996]. The variability of the 5 1 5 N of SCS surface sediments would then reflect the variable importance of nitrogen fixation and minor differences in diagenetic alteration (see discussion below). Nitrogen-fixing cyanobacteria evidently constitute an 43 500 1000 I * 1500 as 3 2000 2500 southern South China Sea 6°39.2N/113°52.1 E • 4 3000 10 20 30 [ iM N03" 40 Fig. 3.2. Water column profiles of nitrate concentration ([NCy], red circles) and its isotopic composition (blue circles) in the southern SCS. 44 important part of the phytoplankton community in the SCS [Carpenter, 1983b], and this could lead to a lighter S 1 5N of dissolved nitrate (see discussion below). Alternatively, the surface sediment 8 1 5N could be interpreted as a reflection of the isotopic composition of Pacific thermocline nitrate [61 5N =~3 %0; Liu et al, 1996] modified by a ~2 %o diagenetic overprint. In the absence of high-resolution 6 1 5 N n j t r a t e profiles or 5 1 5 N measurements on settling particulate samples it is not possible to differentiate between these two possibilities. However, detailed profiles at the sediment-water interface (Fig. 3.4; see discussion below) suggest a diagenetic overprint of <1.2 %o. Down-core records In agreement with the surface sediment values, the 5 1 5 N values of the sediment cores spanning the last 20-200 kyr generally range between 3.7 and 6.2 %>, with no systematic correlation with sediment depth, glacial-interglacial stages, or glacial terminations (Fig. 3.3). Lower values of 2.9-4.2 %o are only observed in the shallowest, southernmost core 18284 (Fig. 3.3; see separate discussion below). Overall, the values are similar to the Pacific Intermediate Water average 5 1 5 N of nitrate of ~5.5 %o. The 1.5 %o amplitude variations in the upper 7 kyr of core 17940 are not correlated with any other palaeoceanographic proxy and are yet to be explained. The comparatively high variability of 8 1 5 N in core 17964-3, especially during and shortly after the deglaciation (i.e., flooding of the shelf), could be due to changes in the ratio of nutrients exported and nutrients recycled at this site close to the continental shelf edge, as hypothesized by Bertrand et al, [2000] for similar records off NW Africa. Alternatively, it could be interpreted as a deglacial perturbation associated with glacial-interglacial changes in the ocean's nitrogen cycle as modeled by Altabet and Curry [1989]. However, the absence of any similar pattern associated with terminations in the other core records questions this interpretation. Nitrate utilization According to the Rayleigh fractionation model [Mariotti et ah, 1981] and assuming an average phytoplankton fractionation factor of 5 %o [Altabet et ai, 1991; Altabet and Frangois, 1994b; Voss et al., 1996; Wu et al., 1997; Waser et al., 1998a,b], the down-core values can be interpreted as reflecting an instantaneous product with a constant nitrate utilization factor of roughly 0.6 due to, for example, micronutrient control on fractional N0 3 " utilization [e.g., Altabet, 2001]. Alternatively, they could be interpreted as the accumulated product of the total nitrate available [Wada and Hattori, 1978; Altabet, 1988], i.e., with a utilization factor of 1. Given the strongly contrasting *•* 3 o IB -V) CO w -00 CM OO T -CO -CM " 00 • Q. v. Q> •O 3 o OO ON o> ^ O Ol O CO 3 ^ > o oi Ml in O to O £ g ^ P <N £ CO to 4-* to T 3 ^£ ° o> is fc- to O J £ £ in H cu O) S £ U OJ 4-1 1 / 5 O c (0 9 tfl 4-< 01 T 3 gj "O a o c £ £ § oi H s .-01 ' coo ; to 60 • ^ « CN £ 3 O -UN (N Ol > o C to : to 1 to « ' w Ol , 01 IH ; a, o ; O " 4-* — ! O i « O 6 c ffl NO ON X t/1 i n l-H oo 03 T—I tO £ & § C ON . 3 ON <T> W ON ON ON C o -a 01 ON T3 01 OH •J3 UN 0) 3 C . o r- 3 « J , > 5 (0 to U) T 3 1-O u 01 >-Ol O u I C o „ D "S.s « v_£ m I-H +3 • >e c 60 ON UI • P4 tr^ to ft H l / l 6 o u c 01 in 01 cJO to +4 O UN s- — CL, to 4-< tO "I 3 § g to ^ 5 ui £ age [ka] 9 46 setting of the cores and the major changes in primary production, terrigenous input, and surface and deep water circulation patterns that occurred in the SCS on a glacial-interglacial timescale [Wang and Wang, 1990; Thunell et al, 1992; Schdnfeld and Kudrass, 1993; Wang et al, 1995, 1999; Miao and Thunell, 1996; Huang et al, 1997], as well as the low nitrate concentrations in SCS surface waters, the assumption of a constant and basin-wide homogenous utilization factor significantly lower than 1 seems unlikely. On the other hand, the interpretation of the data as recording a constant complete utilization of the nitrate available is supported by the 5 1 5N in surface sediments of the present SCS (see discussion above), the low nitrate concentrations in SCS surface waters, and is compatible with temporal changes in primary production as reconstructed from other geochemical proxies. Moreover, nitrate supply to the SCS surface is most likely dominated by monsonally driven vertical mixing [Bauer et al, 1991; Kinkade et al, 1997], therefore requiring the use of the accumulated product equation [Altabet and Frangois, 1994a,b; Sigman et al, 1999]. If this interpretation holds true, it would imply an almost invariant isotopic composition of SCS nitrate, and thus potentially the global oceanic nitrate pool, on a glacial-interglacial timescale. Diagenesis Surface sediment samples (n=83) covering the entire SCS (Fig. 3.1) show no systematic 5 1 5 N variation with water depth (21-4309 m; r=0.37). This argues against diagenetic modification of the 5 1 5 N signal in the water column and contrasts with findings from depth transects (45-4601 m water depth) in the Angola Basin [Holmes et al, 1996], in the Benguela Current region [Holmes et al, 1998] and from the NW African margin [Martinez et al, 1999] that show a 1.6-3.0 %0 and a ~2 %0 1 5N-enrichment, respectively, with depth and distance from the coast. Detailed box core profiles of the sediment-water interface and the uppermost sediment column (Fig. 3.4) show that despite an expected [Miiller, 1977] significant loss (10-36 %) of total nitrogen between the fluff layer and the uppermost sediment sample the 8 1 5 N values show only a minor, albeit consistent, offset (AS1 5N = +0.3 - +1.2 %o). This 5 1 5N increase between the fluff and the surface sediment sample is lower than the 4-5 %0 increase observed by Altabet [1996] in samples from the equatorial Pacific, but similar to the smaller offset (within 1 %o) between sinking particles and surficial sediments from the eastern North Pacific margin [Altabet et al, 1999b; S. Kienast et al, 2002] and at the Mexican margin [Ganeshram, 1996]. The low A6 1 5N in SCS sediments is consistent with the notion of lower degradation rates of labile organic material due to higher sedimentation/ accumulation rates in marginal environments compared to the ON ON bp" oi 1—1 Q) 0) CD b H « X -rt is ^ c g W S±/ — ™ .55 in ca co-48 equatorial Pacific. Moreover, the slightly higher 6 1 5 N values of surface sediments in the central SCS (Fig. 3.1), where sedimentation rates are lowest, could be due to this effect. However, in the absence of 6 1 5 N values of the sinking particulate material, a diagenetic overprint >1.2 %o cannot be excluded. Furthermore, it is noted that in a recent study on equatorial Pacific nitrate utilization, Altabet [2001] in fact argues for a diagenetic decrease in sedimentary 6 1 5 N on a timescale of several 100 kyrs. Obviously, this assertion contrasts markedly with the previously held notion of increasing 6 1 5 N with progressive diagenesis [Altabet, 1996; Holmes et al, 1999; Sigman et al, 1999]. Furthermore, secular decreases in 5 1 5 N on these long time scales, which have been described in several other studies [e.g., Roux et al, 2001], could also be driven by long-term changes in the marine nitrogen cycle, such as, for example, an overall decrease in water column denitrification relative to nitrogen fixation. Clearly, further detailed studies of the diagenetic modification of sedimentary 6 1 5 N records, as well as on the long-term evolution of the marine nitrogen cycle are warranted. Saino and Hattori [1987] report 6 1 5 N values of SPON from two stations in the northern central SCS (see Fig. 3.1 for location) ranging from 0.6 to 8 %o (with one single value at 12.9 %o). The lower values (<1 %o) are observed in the nitrate-depleted upper part of the euphotic zone. Below 200 m, the 6 1 5 N of SPON averages between 6 and 8 %o with no obvious trend with depth. The depth distribution and the depletion in 1 5 N of sinking particles relative to SPON, as inferred from the 5 1 5 N of surface sediment samples, are consistent with observations from the Sargasso Sea [Altabet, 1988], the North Atlantic [Altabet et al, 1991], and the NE Pacific [Wu, 1997]. This 6 1 5 N offset between SPON and surface sediment samples is further evidence against a major diagenetic 1 4 N depletion at the sediment-water interface. Two lines of evidence argue against any major diagenetic modification of the down-core records. First, the 5 1 5 N records show very similar and almost constant absolute values, despite the major differences in local setting (Fig. 3.1), sedimentation rates (Table 3.1), and sedimentological parameters between the different sites and with time at each site [Wang et al, 1999]. Second, the correlation of increasing 5 1 5 N values with decreasing percent TN observed at the sediment-water interface (Fig. 3.4) is absent in the down-core records. For example, in core 17964 (Fig. 3.5a), there is no systematic 6 1 5 N down-core trend despite an apparent (possibly diagenetic) decrease in percent TN. The coupling between T N and 5 1 5 N observed in core 17954-2 (Fig. 3.5b) is likewise inconsistent with a diagenetic trend because both percent T N and 6 1 5 N decrease with depth. The origin of the "step" in 5 1 5N at 125-140 kyr is currently not understood. 49 age [kyrs] 17964 515N (%cvs.air) o-o 3 4 5 6 7 0.12 0.14 % T N >-* 0.16 100 H 150 age [kyrs] -j 17954 615N (%«vs.air) o—o 3 4 5 6 7 0.12 Fig. 3.5. Downcore records of bulk sedimentary 8 1 5 N and percent T N of cores 17954 and 17964. The record of core 17964 is a composite profile of a gravity (17964-3) and a piston (17964-2) core [see Kienast, 2000 for details]. Table 3.1 core # location water depth recovery sed.rates [cm/kyr] * samples ° N • E [m] [m] Interglacial glacial Interval [cm] 17924-3 19:24.7 118:50.8 3438 19.89 - 1 0 * * - 2 0 * * 20 17940-2 20:07.0 117:23.0 1728 13.30 45-85 1 9 10 17954-2 14:45.5 111:31.6 1517 11.52 6-7 5-9 10 17961-2 8:30.4 112:19.9 1795 10.30 3-8 10 10 17964-2/-3 6:09.5 112:12.8 1556 9.12 - 3 0 >55 20 18284-3 5:32.5 110:32.4 226 8.04 10-60*** >100*** 10 (30-40) •from Wang et al. 1999 "M. Sarnthein, unpubl. ""from Steinke et al., 2002 in press 50 Denitrification and upwelling Denitrification in suboxic/anoxic water columns results in elevated sedimentary 6 1 5 N values of 8-12%0 [Altabet et al, 1995,1999a,b; Ganeshram et al, 1995, 2000; Ganeshram, 1996; Pride et al, 1999; Suthhof et al, 2001]. The low uniform sedimentary 5 1 5 N values throughout the SCS suggest that denitrification is unlikely and therefore that inferred suboxia/anoxia of SCS intermediate and/or deep waters during the last climatic cycle [Wang et al, 1999] can be ruled out. Moreover, there are no consistent changes in 5 1 5 N at the sites of model-inferred upwelling (17954 and 17964). This finding corroborates earlier results [Wang et al, 1999] that show no indications of significant upwelling at these sites. Nitrogen fixation Nitrogen-fixing bacteria can play an important role in the nitrogen budget of oligotrophic, i.e., nitrate-limited waters [Carpenter and Romans, 1991; Walsh, 1996; Karl et al, 1997; Capone et al, 1997]. Given an average 6 1 5 N of fluff samples in the SCS of 4.5 %0 (Fig. 3.4) and assuming an initial 5 1 5 N of nitrate of 5.5 %o (=& 1 5N n i t r a t e of SCS subsurface and Pacific Intermediate [Liu et al, 1996] waters), nitrogen fixation could account for as much as 20 % of the total nitrate available for phytoplankton in the SCS. Accordingly, the decrease in S 1 5 N n i t r a t e in the uppermost samples of the SCS water column profile (Fig. 3.2) could be taken as an indication of the addition of isotopically light nitrate from nitrogen fixation at the sea surface, by analogy with the interpretation of water column profiles from the Arabian Sea by Brandes et al. [1998]. Furthermore, the assertion of a significant contribution of bacterially fixed nitrogen to the overall nitrate budget in the SCS is in good agreement with high Trichodesmium abundances observed in the southern and central SCS [Carpenter, 1983a], the East China Sea [Minagawa and Wada, 1986] and the Kuroshio area [Saino and Hattori, 1980], as well as high nitrogen fixation rates in the SCS [Saino and Hattori, 1987], and the Kuroshio waters [Liu et al, 1996; Nakatsuka et al, 1997] as inferred from 5 1 5N analyses of sinking particulate organic matter. Terrigenous input Input of terrigenous organic matter can potentially contribute significantly to the 5 1 5 N signature of bulk sedimentary organic matter [Mariotti et al, 1984; Minoura et al, 1997; Wada et al, 1987]. In order to establish a terrigenous 6 1 5 N end-member for the SCS the nitrogen isotopic composition of samples from various terrestrial depositional environments (see Chapter 3.2.2) in sediment cores from the shallow Sunda Shelf were determined. The 5 1 5 N values range from 0.7 to 5.1 %o, with an average of 3.3 % 0 («=14). 51 This value is only slightly lower than the o 1 5 N t e r r i g e n o u s derived from bulk sediments of core 18284 (see discussion below). A terrigenous end-member of 3.3 %o is too close to a 8 1 5 N value of the marine end-member of 5.5 %o to enable the detection of minor admixtures of terrigenous organic matter in the bulk sediment. Changes in the terrigenous contribution by 10 % only result in a 0.2 %o difference of the bulk 8 1 5 N, which is the precision of the analytical method. Carbon isotope analyses on the same suite of surface sediments and sediment cores (see Chapter 3.3) do not show any significant correlation (r=0.09-0.48) between light 8 1 5 N values and light 8 1 3 C o r g values which would be indicative of the presence of significant amounts of terrigenous organic matter in the sediments [Peters et al, 1978]. Indeed, only the 8 1 3 C o r g data from the southern SCS cores (see discussion below) show any admixture of terrigenous organic matter (see Chapters 3.3 and 4.1), consistent with the close proximity of large river systems coupled with large changes in emergent shelves due to sea level changes, and yet there is no concordant variation in 6 1 5 N in these cores. Further evidence for the lack of a significant terrestrial signal in the SCS comes from core 18284-3 which was raised from the shelf-slope transition at 226 m water depth (Fig. 3.1). This location should sensitively monitor changes associated with the deglacial flooding of the Sunda Shelf and its drainage system [Stattegger et al, 1997; see also Chapter 4.1]. The carbon isotopic composition clearly characterizes the organic material in the lower part of the core (8-17 kyr) as being derived from terrestrial C 3 sources (Fig. 3.6; see also Chapter 4.1.6). The corresponding 8 1 5 N values have a mean value of 3.5 %0, which is within the range of 8 1 5 N values reported for plant remains from the Sunda Shelf discussed previously, further corroborating the similarity between the inferred marine and terrigenous 8 1 5 N signal. A transition to mainly marine organic matter occurs within the interval 8-3.5 kyr (Fig. 3.6). The 5 % 0 rise in 8 1 3 C o r g is paralleled by a very small rise of <1 %o in 8 1 5 N, consistent with the notion of a slightly lighter terrigenous 8 1 5 N signature. The overall lower 8 1 5 N values in this core compared to the other hemipelagic records is interpreted to reflect higher rates of nutrient recycling at the shelf edge, which leads to decreasing sedimentary 8 1 5 N [Checkley and Miller, 1989; Altabet, 1996]. Cancellation of effects It is possible, of course, that some of the influences on the nitrogen isotopic composition of bulk sedimentary organic matter in the SCS cancel each other so that we see little change in 8 1 5 N over glacial-interglacial timescales. Two facts, however, argue against this possibility: (1) Despite the significantly different setting of the cores, none of 52 Fig. 3.6. 6 1 5 N and 5 1 3 C o r g of core 18284-3 versus age. Please note that the age model of core 18284-3 [Steinke et al. 2002] has been revised significantly from the one presented in M. Kienast [2000]. 53 the 5 1 5 N records shows a glacial-interglacial pattern that accords with other palaeoceanographic proxies [Wang et al, 1999], and the absolute 5 1 5 N values of the records are very similar. (2) The time resolution of the sampling from core 17940-2 (Fig. 3.3) is significantly less than the ~3000 years residence time of N in the ocean. It is most unlikely that the processes discussed previously act on exactly the same timescales, with precisely the amplitude needed to cancel each other. Global implications The sedimentary 8 1 5 N records from the SCS suggest a constant nitrogen isotopic signature of nitrate in the SCS and, by inference, in the Pacific Intermediate (and/or thermocline) Water because of the continuous deep water connection between the SCS and the Pacific during the last climatic cycle (see section 3.1). A constant 8 1 5 N signature of the western Pacific Intermediate Water nitrate reservoir on glacial-interglacial time-scales, in turn, sets important limitations on the global impact of changes in the rate of denitrification in the eastern tropical Pacific and the Arabian Sea as postulated by Altabet et al. [1995; 1999a; 2002], and Ganeshram et al. [1995; 2000]. Four scenarios are offered to illustrate possible implications that the maintenance of a constant isotopic composition of oceanic nitrate suggested by the SCS data could have, given both sources' (riverine and atmospheric input, and N fixation) and sinks (sedimentary and water column denitrification and sediment burial) of oceanic nitrate. First, according to a modeling study by Brandes [1996], if water column denitrification in the Arabian Sea and the eastern tropical Pacific were to influence the oceanic N inventory globally, a decrease in water column denitrification by 50 % (to 40 Tg N/yr ) during glacials would need to be accompanied by a decline in sedimentary denitrification of 75 %. Such changes in sedimentary denitrification are well within the range of estimates by Christensen et al. [1987]. However, the timing of changes in water column and sedimentary denitrification is significantly different. Broecker and Henderson [1998] and Petit et al. [1999] constrain the rise in atmospheric C 0 2 as preceding the change in sea level. Accordingly, if changes in water column denitrification were to govern oceanic productivity and thus atmospheric C 0 2 , they would lead changes in sea level and thus sedimentary denitrification. Sedimentary 5 1 5 N records from the Arabian Sea in fact show that 5 1 5 N leads global sea level by up to 6 kyr [Altabet et al, 1999a], which is significantly longer than the mean residence time of oceanic nitrate of ~3kyr [Gruber and Sarmiento, 1997]. This difference in phasing should thus lead to perturbations in the nitrogen isotopic composition of oceanic nitrate during transitional periods [Altabet and Curry, 1989] that are not observed in the SCS records. Moreover, high-54 resolution studies show short-term changes in the 6 1 5 N record [Pride et al, 1999; Ganeshram et ah, 2000; Suthhofet al, 2001; Altabet et al, 2002] that are not accompanied by changes in sea level. The second possible implication is a glacial decrease in S 1 5 N n i t r a t e due to less water column denitrification that is isotopically compensated by a synchronous decrease in N fixation, a coupling hypothesized to occur locally in the Cariaco Basin [Haug et al, 1998], and predicted for the global ocean based on a simple model [Tyrrell, 1999]. A glacial decrease in global oceanic N fixation, however, is the exact opposite of the scenario hypothesized by Falkowski [1997] and Broecker and Henderson [1998] of an increased fixation rate due to dust (i.e., iron) fertilization. The latter scenarios have been questioned by Ganeshram et al. [2002], however, on the basis of a likely glacial increase of the oceanic N / P ratio, which would subjugate nitrogen fixation due to P limitation. Thirdly, the increased oceanic productivity that is hypothesized to result from budgetary changes due to decreased denitrification in low oxygen oceanic environments could lead to a compensating increased denitrification in the oxygenated open ocean. This process was first postulated by Tsunogai [1971], but has yet to be proven. Finally, as proposed by Suthhof et al. [2001], the occurrence of high-frequency intervals of denitrification during interstadials in the Arabian Sea (and elsewhere) could influence the sources and sinks of nitrate in such a way as to help maintain a rather constant 6 1 5 N n i t r a t e . In contrast to this elegant solution to the conflicting evidence, however, Altabet et al. [2002] argue that rectifying the record of millennial changes in water column denitrification in the Arabian Sea with a 3-kyr (i.e., the residence time of marine nitrogen) moving average would produce a time series of changes in the marine nitrogen inventory with a series of maxima and minima "which are strikingly similar to those found in Antarctic ice-core temperature and C 0 2 records" [Altabet et al, 2002, p. 162]. Given the paucity of information on the distribution of 8 1 5 N n i t r a t e in the global ocean [Sigman, 1997] and the SCS, it is not yet possible (and beyond the scope of this thesis) to quantify a budgetary constraint imposed by an unchanged 5 1 5 N n i t r a t e in the SCS and the western Pacific on the global impact of changes in denitrification elsewhere. However, the data presented here suggest that the 5 1 5 N records from key areas of denitrification might reflect local phenomena to a greater extent than previously assumed and caution their extrapolation to the global oceanic N inventory. 55 3.2.4 Summary In summary, the 6 1 5 N values of six sediment cores from throughout the SCS spanning the last climatic cycle range only from ~3 to ~6 %o, with no systematic correlation to glacial-interglacial stages. The variability within every one of the cores is <2 %o and in one core (17961-2) is only 1 %o. The records are not significantly affected by terrigenous input and/or diagenesis and are interpreted to reflect the complete consumption of nitrate during primary production. This interpretation implies an unchanged isotopic composition of nitrate in the SCS and by inference in the western Pacific subsurface waters during the last 200 kyr. This, in turn, places constraints on models that call upon major changes in the oceanic nitrate budget to explain the glacial-interglacial atmospheric C 0 2 changes observed in ice cores. Clearly, more high resolution sedimentary records from oligotrophic regions confirming the absence of (major) changes in global oceanic 8 1 5 N n i , r a t e on glacial-interglacial timescales as well as studies of the phasing of changes in the marine N cycle are needed. 56 3.3 Carbon isotope record's 3.3.1 Introduction Simple mass balance considerations of the global carbon cycle point to the ocean as major player in glacial-interglacial changes in atmospheric C 0 2 concentrations as recorded in ice cores. In principle, two competing schools of thought that try to characterize and quantify the ocean's impact on atmospheric C 0 2 have evolved during the last two decades. One invokes changes in marine production to account for the observed variability [Broecker, 1982; McElroy, 1983; Sarnthein et al, 1988; Berger et al, 1989; Martin, 1990; Broecker and Henderson, 1998], while the other attributes the ocean's role mainly to physical processes, such as changes in stratification and/ or redistribution of C 0 2 within the different oceanic reservoirs [Boyle, 1988; Keir, 1993; Frangois et al, 1997; Toggweiler, 1999; Stephens and Keeling, 2000]. Changes in upwelling intensity with time have been proposed as a mediatory mechanism [Pedersen et al, 1991; Jasper and Hayes, 1994]. Here, increases in marine production are directly coupled to physical processes that impact on the exchange of C 0 2 between the atmospheric and the oceanic reservoirs, the net effect, however, being oceanic degassing of C 0 2 to the atmosphere. In order to differentiate between these different scenarios and to gain a better understanding of the ocean's contribution to changes in atmospheric C 0 2 , a clearer delimitation and quantification of oceanic palaeosources and sinks of C 0 2 is desirable. The carbon isotopic composition of marine organic matter (S 1 3C o r g) is one method that has shown promise in monitoring the ocean's contribution to changes in atmospheric C 0 2 . Although the exact biological processes determining o 1 3 C o r g are still subject to considerable debate, a variety of laboratory and field studies have shown that the 6 1 3C of the bulk organic fraction of marine plankton generally decreases as the dissolved molecular C 0 2 concentration [C0 2 ( a q ) ] increases [Degens et al, 1968; Degens, 1969; Rau et al, 1989, 1991b, 1992, 1997; Popp et al, 1989; Freeman and Hayes, 1992; Frangois et al, 1993a; Goericke and Fry, 1994; Rau, 1994; Fischer et al, 1998]. In the inferred absence of a significant contribution of terrigenous organic matter to the sediment, which would obscure the bulk sedimentary isotopic signature, this correlation has been widely used to convert bulk sedimentary o 1 3 C o r g records to palaeo-pC02 estimates [Rau et al, 1991a; Pedersen et al, 1991; Fontugne and Calvert, 1992; Thunell et al, 1992; Midler et al, 1994; Westerhausen et al, 1994; Bentaleb et al, 1996;]. Some of these studies have specifically tried to unravel the antagonistic influences of changes in upwelling intensity and in marine production on the direction of C 0 2 exchangebetween the ocean and the 57 atmosphere, a key parameter in understanding the ocean's contribution to changes in atmospheric C 0 2 concentrations. In the first part of this chapter, new records of the S 1 3 C o r g in a suite of sediment cores from the open oceanic SCS spanning the last climatic cycle are presented. These S 1 3 C o r g records are converted to local pC0 2 estimates, following established procedures, and after considering potential biases by terrigenous input, temperature and other effects on the C isotope fractionation as well as possible diagenetic modification of the signal. In the second part, the strengths and limitations of low-latitude marine sedimentary pC0 2 estimates based on 6 1 3 C o r g are discussed. The discussion is focused on three key criteria for evaluating 6 1 3 C o r g -pC0 2 estimates, (1) absolute values of the estimates, (2) correlation between pC0 2 estimates and indicators of upwelling strength, and (3) a comparison of the temporal evolution of pC0 2 estimates with the ice core (Vostok) record of atmospheric C 0 2 . 3.3.2 Results and interpretation The 6 1 3 C o r g in the four sediment cores located in the southern (core 17961-2), western (core 17954-2) and northern (cores 17924-3 and 17940-2) reaches of the SCS (Fig. 3.7) varies between -22.5 and -19.5 %o (Figs. 3.8-3.11). Higher values mark glacial stages, while lower values are characteristic of interglacials. In the following paragraphs, evidence is adduced that this variability is not due to variable terrigenous input to the sediments, direct effects of temperature on planktonic isotope fractionation, changes in marine production, and other biological factors, nor to diagenetic modification of the signal. Marine versus terrigenous organic matter As expected from the palaeogeographic setting of the SCS (see Chapter 1), concentrations of the terrigenous marker n-nonacosane [see M . Kienast et al. 2001 for analytical details] are significantly higher during glacial stages (Figs. 3.8-3.10) when vast shelf areas in the northern and especially in the southern SCS were emergent (see Fig. 1.1). The bulk sedimentary 6 1 3 C o r g signal of cores 17940-2, 17954-2, and 17961-2 is inversely correlated with this molecular indicator of terrigenous input, ruling out an overwhelming contribution of terrigenous C 3 organic matter to the bulk 5 1 3 C o r g signal at these three sites because C 3 plant organic matter is characterized by light 5 1 3 C o r g values [-25 to -27%0; Deines, 1980]. Furthermore, since the absolute values of 6 1 3 C o r g in all four sediment cores (Figs. 3.8-3.11) are quite similar, irrespective of their proximity to 58 ra u (/) co c !E O o CO CO SB 'S XI ~ d  u 2 ON — CO O ON ca 00 l-l 13 . cu -»-» C cu CO CU ' £ s >< CO . 5 CO o c -a CO V 3 co "3 cu _s; 5 £ o _<o •22 „ <N _ £ o o m ON CO bo « CO as as —^  H <N v a \ , , . s cu a, -a c CS £ PQ .2 o CS U T 3 S <3 <u C O - M CO , 3 CO I > \ co £ o •a be cr"-* 5 o x 4-» CO c x CO v CO >^  co O 1 J 9^3 CO s x i ^ £ ~ , , o as ~C as </> CU Xf O 0) cj 0> CU 01 co u CD TfH c CO CU so c O CO u - S O £ H H ^ g as -« • co as w ^ C N >< axj CO C u CJ co e n 2 c as xis O U a. 'co CO co 'co O bO co O £ CL, g cfl 5 CS CO CO p a CO £ CO S cu CO — a fee - • co -a r-1 . . . to P H N as 4-1 | H OS s • • O i-< -s o SH CL, CL, CO CL, CO 6 tn U CD CU 5 2 £ ^ * cu CS ^ 60 2 ° c co .a £ c o 5 •rj > co "to £ O .S u o> 59 core 17940-2 H500 h300 =? 3 O 3 0> O O (0 0) 3 (D S" 100 £ 28 - C - 26 u 0) - 0) H - 24 o o 22 Or. _i UI Z ? < (0 -i—i—i—r 200 400 600 800 1000 1200 cm core depth Fig. 3.8. Bulk sedimentary 8 1 3 C o r g / concentration of n-nonacosane in ng/g [from Kienast et al. 2001], percent clay of the siliciclastic fraction [from Wang et al, 1999], U K ' 3 7 SST estimates [from Pelejero et al, 1999], concentration of C37-alkenones in ng/g [from Kienast et al. 2001], bulk sedimentary 5 1 5 N, and bwOG.ntber white [from Wang et al, 1999] of core 17940-2 in the northern SCS versus core depth. 60 core 17954-2 Fig. 3.9. Bulk sedimentary 8 1 3 C o r g / concentration of n-nonacosane in ng/g'[from Kienast et al. 2001], U K < 3 7 SST estimates [from Pelejero et al., 1999], concentration of C^-alkenones in ng/g [from Kienast et al. 2001], bulk sedimentary 6 1 5 N, and 5 1 8 O G m b e r w h i t e [from Wang et al, 1999] of core 17954-2 in the western central SCS versus core depth. Vertical stippled lines denote the onset of terminations I and II, respectively, as defined by the 5 1 8 O G r u b e r w h i t e and SST records. 61 core 17961-2 — i — ' — ' — ' — i — i — i — 600 800 cm core depth Fig. 3.10. Bulk sedimentary 5 1 3 C o r g / concentration of n-nonacosane in ng/g [from Kienast et al. 2001], percent clay of the siliciclastic fraction [from Wang et al, 1999], U K ' 3 7 SST estimates [from Pelejero et al, 1999], concentration of Cgy-alkenones in ng/g [from Kienast et al. 2001], bulk sedimentary 6 1 5 N, and 5 1 8 O G m b e r w h i t e [from Wang et al, 1999] of core 17961-2 in the southern SCS versus core depth. Vertical stippled lines denote the onset of terminations I and II, respectively, as defined by the SST and benthic 5 L S O [not shown, from Wang et al, 1999] records. Fig. 3.11. Bulk sedimentary S 1 3 C o r g and bulk sedimentary 6 1 5 N of core 17924-3 in the northeastern SCS versus core depth. Vertical stippled lines denote the onset of terminations I and II, respectively, as defined by the 8 1 8 O G m b e r record ( M . Sarnthein, unpublished data, 1999). 63 continental sources and since there is no north-south gradient that could be attributed to changing vegetation patterns, C 4 plant organic matter (5 1 3C o r g: -12 to -15 %0; ibid.) most likely constitutes a relatively minor fraction of the organic matter in the cores. According to Hinrichs et al. [1999], up to 30 % of the bulk sedimentary C o r g would need to be exclusively of C 4 plant origin in order to explain the observed 5 1 3 C o r g variability. This amount is much larger than any possible C 4 contribution as indicated by pollen records from SCS sediments [Sun and Li, 1999; Sun et al, 1999]. Moreover, the interglacial concentrations of n-alkanes in the SCS cores 17961-2 and 17954-2 (-125 and 175 ng/g) are comparable to Holocene values from the central N Atlantic [30-150 ng/g; Villanueva et al, 1997] and the central low-latitude Pacific [30-300 ng/g; Ohkouchi et al, 1997], suggesting comparatively low terrigenous input at least during interglacial periods. In addition, the higher concentrations of n-alkanes/nonacosanes in core 17940-2 (~700 ng/g) as compared with cores 17954-2 and 17961-2 are not associated, as would be expected, with lower 6 1 3 C o r g values in this core. Finally, both the surface sediment distribution of 6 1 3 C o r g itself (Fig. 3.12), and the absence of any correlation of the 6 1 3 C o r g surface sediment values in the SCS with water depth (Fig. 3.13; taking water depth as a first order approximation of distance from the coast) suggest that there is no overwhelming importance of terrigenous organic matter on the distribution of modern bulk sedimentary 8 1 3 C o r g . In the absence of a reliable quantitative proxy for terrestrial input to marine sediments, (for example, an n-alkane/C o r g ratio for the terrigenous end-member [Prahl et al, 1994]), however, it is not possible to conclusively quantify the subdued, time-varying influence of terrestrial organic matter on the 5 1 3 C o r g signal, most importantly the increased input of terrigenous organic matter during glacials (see implications and further discussion below). Keil et al. [1994] and Goni et al. [1998] reported a grain size-dependence of the carbon isotopic composition of surficial sediment particles on the Washington Margin and in the Gulf of Mexico. Both studies showed heavier 5 1 3 C o r g values associated with finer-grained particles. In the two SCS core records where grain size data are available, however, lower 6 1 3 C o r g values are associated with higher % clay (core 17940-2; Fig. 3.8), or there is no correlation between 6 1 3 C o r g and % clay (core 17961-2; Fig. 3.10), excluding the possibility that the carbon isotopic signal in the SCS is governed by effects associated with the hydrodynamic sorting of particles with a different, grain size dependent, carbon isotopic composition. The distance from continental sources and the similarity in the 5 1 3 C o r g record of core 17924-3 (Fig. 3.11) suggest that the interpretation of a subordinate influence of 64 105 110 115 120 125°E Fig. 3.12. Map of the S 1 3 C o r g surface sediment distribution in the SCS. Filled circles: this study; filled squares: 6 1 3 C o r g data from Calvert et al. [1993]. Note that the 100 m isobath approximates the coast line during glacial sea level low stands. 65 terrigenous organic matter on the carbon isotopic composition holds true also for this site, where no biomarker and grain size data are available. Temperature effect Fontugne and Duplessy [1981] observed a positive correlation between increasing sea surface temperature (SST) and increasing 5 1 3 C o r g values in marine plankton from the Atlantic and Indian Oceans. This trend is opposite to that observed for the SCS sediment records (Figs. 3.8-3.10). Accordingly, 6 1 3 C o r g in the SCS is not affected by a thermal effect on C isotope fractionation as postulated by Fontugne and Duplessy [1981]. Moreover, Fontugne and Duplessy [1981] state that in fact a correlation (between SST and 5 1 3 C o r g ) does not exist for the temperature range 15-31 °C, the range of SST observed in the SCS. In light of recent understanding [Rau et al, 1989, 1996, 1997; Freeman and Hayes, 1992; Goericke and Fry, 1994; Raven et al, 1993] the SST dependency of 8 1 3 C o r g claimed by these authors is more likely to be indirect, such as the SST-dependent solubility of C 0 2 in sea water, the temperature dependence of 51 3C of C0 2( a q ) , changes in the diffusivity of C 0 2 ( a q ) with SST, variable average growth rates as a function of SST, the induction of a carbon concentrating mechanism, or any combination of these factors. Marine production, upwelling and biological factors In order to evaluate any possible influence of changes in marine production on 5 1 3C o r g , we use the sedimentary concentration of C37-alkenones as production tracer (Figs. 3.8-3.10; see M . Kienast et al, 2001 for analytical details). This selection is justified by the fact that bulk C o r g in the SCS is affected by time-varying changes in the admixture of terrigenous organic matter (see discussion above). Given the overriding control of linear sedimentation rates on accumulation rates of sedimentary components [Middelburg et al, 1997], we chose not to calculate alkenone accumulation rates but rather incorporate changes in sedimentation rate in a more qualitative way into the discussion. Thus, the higher glacial concentrations of alkenones in core 17940-2 (Fig. 3.8) are offset by significantly lower sedimentation rates during this period (see Table 3.1), suggesting unchanged levels of haptophyte and, by inference, overall marine production on a glacial-interglacial timescale at this site, in agreement with unchanged nutrient levels/utilization (see discussion below in this section). In contrast, the alkenone record of core 17954-2 (Fig. 3.9) shows highest alkenone concentrations paralleled by higher sedimentation rates during the later part of marine isotope stage (MIS) 5, and during MIS 6, suggesting elevated marine production during these periods, decoupled from glacial or interglacial boundary conditions. This reconstruction of temporal changes in 66 marine production at site 17954 is in agreement with benthic foraminiferal evidence from the same core [Jian et al, 2001]. In core 17961-2 (Fig. 3.10), lower alkenone concentrations occur during both glacial terminations, paralleled by decreased sedimentation rates [Wang et al, 1999; see Table 3.1]. Throughout the rest of the core record, alkenone-derived marine production is more or less uniform. Given the similarity of the 5 1 3 C o r g records throughout the SCS and the absence of an obvious correlation between S 1 3 C o r g and variable marine production at any site, 8 1 3 C o r g in the SCS seems not to be significantly affected by changes in marine production, as inferred from the concentration of alkenones. Recent surface seawater C 0 2 measurements throughout the SCS [Chen et al, 1998; Rehder and Suess, 2001] and maps of global air-sea pC0 2 differences [Takahashi et al, 1997, 2002] show that the SCS surface waters are in equilibrium with atmospheric C 0 2 . Similarly, palaeoceanographic studies to date do not show any indication of significant upwelling at any site in the SCS during the last 220 kyr, including the locations sampled in this study [see Wang et al, 1999, and references therein]. A n exception to this general observation might be an area off Vietnam, marginally including site 17954 in the present study. Here, upwelling is predicted based on a climatology-driven circulation model in response to summer monsoonal winds [Wiesner et al, 1996], and planktonic foraminiferal 6 1 3C data have been interpreted to reflect changes in upwelling intensity with time at site 17954 [Wang et al, 1999]. However, the postulated upwelling at site 17954 is neither reflected in the nitrogen isotopic record (Fig. 3.3 and 3.9) nor is the 5 1 3 C o r g record of core 17954-2 significantly different from the other records presented here. Moreover, if site 17954 was influenced by changes in upwelling induced by variable summer monsoonal winds, one would expect a consistent glacial-interglacial contrast in marine production in accord with glacial-interglacial changes in the intensity of the summer monsoons as reconstructed from various sedimentary proxies [Wang et al, 1999]. The alkenone-based reconstruction of marine production at this site (see above in this section) does not conclusively support this scenario. Accordingly, current understanding of SCS palaeoceanography suggests that the SCS surface ocean pC0 2 at the core sites is more or less in thermodynamic equilibrium with the atmosphere both today and during the last 220 kyr. A constant nitrogen isotopic composition of bulk sedimentary organic matter (4-6 %0) at all sites and throughout the records (Figs. 3.8-3.11) has been interpreted to reflect complete consumption of nitrate during primary production (see Chapter 3.2), suggesting nitrate-limited production in the SCS throughout the last 220 kyr. Similarly, foraminiferal Cd /Ca ratios of cores 17940-2 and 17950-2 [Lin et al, 1999; see Fig. 3.7 for 67 location of core 17950-2] indicate no systematic glacial-interglacial variability, which is interpreted to reflect relatively unchanged HP0 4 2 " concentrations in SCS surface waters on a glacial-interglacial timescale. Although this evidence is not direct, such records of nutrient availability/utilization strongly suggest that 8 1 3 C o r g in the SCS has not been significantly affected by changes in growth rate, cell volume/geometry, and/or nutritional status as observed by Laws et al. [1995], Bidigare et al. [1997], Pancost et al. [1997], Popp et al. [1998], Kukert and Riebesell [1998], and Burkhardt et al. [1999]. In addition, invariably low opal concentrations (<5 %, i.e., close to the precision of the method [Mortlock and Froelich, 1989]) throughout the SCS core records (not shown) could be interpreted to indicate a relatively unchanged phytoplankton community structure, suggesting a negligible effect of changes in species composition on the 8 1 3 C o r g signal [Pancost et al, 1999; Popp et al, 1999] in the SCS. It is noted, however, that the low opal concentrations could also be due to opal solution in the sediments. Diagenesis The preferential degradation of 1 2 C relative to 1 3 C in the water column inferred from the 6 1 3 C o r g distribution in plankton samples, sediment trap material, and surface sediments in the equatorial Atlantic [Westerhausen et al, 1993] and the Angola Basin [Milller et al, 1994] is not evident from the 8 1 3 C o r g surface sediment distribution with respect to water depth in the SCS (Fig. 3.13). This implies that there is no discernible differential preburial diagenesis of the C isotopes in the SCS water column. A comparison of the 8 1 3 C o r g signatures of fluff versus surface sediment samples shows a constant, albeit significant, 1 2 C depletion of around 1 %o at the water-sediment interface (Fig. 3.14). This could be interpreted as a reflection of early diagenetic modification of the 8 1 3 C o r g signal. However, it is more likely that the offset is due to the anthropogenic C 0 2 invasion into surface waters. A 8 1 3C o r g decrease of several %o owing to anthropogenic C 0 2 emission has been postulated by Bentaleb and Fontugne [1996] and is substantiated by a comparison of sinking organic matter and surface sediments by Fischer et al. [1997]. This interpretation is corroborated by a similar [Quay et al, 1992; Beveridge and Shackleton, 1994] decrease of 8 1 3 C G r u b e r values in the uppermost samples (n=5; 0-5 cm) of box core 17940-1 compared to the Holocene average [see Wang et al, 1999 for 8 1 3 C G r u b e r record]. In light of these new findings, previous interpretations of pre-burial diagenetic modification of the 8 1 3 C o r g signal (see discussion above in this section) probably need to be reevaluated. The 8 1 3 C o r g down-core records do not display any consistent secular trend with depth which would be indicative of diagenesis [cf. Fontugne and Calvert, 1992]. 68 1000 ~ 2000 V "2 3000 CU (B 5 4000 5000 . it i . -23 -22 -21 -20 -19 -18 ° 1 3 C o r g surface sediment ( * ° V 8 - P D B ) Fig. 3.13. Bulk sedimentary 5 1 3 C o r g of surface sediments versus water depth (n=83, filled circles: this study, filled squares: from [Calvert et al, 1993]) in the SCS. Note the absence of any consistent trend of 5 C o r g with increasing water depth. Fig. 3.14. Bulk sedimentary S 1 3 C o r g of fluff samples versus surface sediments. Note that the fluff samples are consistently lighter by 1%> on average compared to the surface sediment. Thick black line shows 1:1 correlation; the thin line is the actual correlation of the data (n=31), with gradient = 0.85 and r=0.68. 69 Moreover, given the different linear sedimentation rates (Table 3.1), and therefore different C o r g burial rates at the four coring sites, and the different bulk sediment compositions, the similarity of the 8 1 3 C o r g records at all sites suggests that the carbon isotopic composition has not been significantly modified by diagenesis. Thunell et al. [1992] reported 6 1 3 C o r g records from two core sites in the eastern SCS off Palawan (GGC 4 and GGC 11; see Fig. 3.7 for location) showing glacial values that are consistently heavier by up to 2 %o than in the four records presented in the present study. Replication of the analyses on the 1992 study material in our laboratory indicate that this deviation is not due to any analytical offset or differences in sample treatment. These records could indicate decreased terrigenous input to the SCS off Palawan during sea level low stand compared to the records presented here. 3.3.3 Estimating pC02 from sedimentary 8 1 3C o r g We have applied the approach of Rau et al. [1991a] to convert bulk sedimentary 6 1 3 C o r g values to pC02 estimates (Fig. 3.15). This is justified by the narrow range of C0 2( a q) values in SCS surface waters [Chen et al, 1998] that obviates the possibility of using a site-specific & 1 3 C o r g - C 0 2 ( a q ) relationship and, even though there is no a priori reason to assume a linear relationship [Frangois et al, 1993b], allows inter-comparison with other low-latitude 6 1 3 C o r g -pC0 2 estimates (see discussion below). Using the approach of Popp et al. [1989], moreover, yields almost identical estimates, as has been noted earlier by Mtiller et al [1994]. The 5 1 3 C o r g records have been corrected for changes in 6 1 3C of dissolved inorganic carbon (DIC) in seawater by subtracting the difference between the 5 1 3 C G . m b e r downcore values and the average latest Holocene 6 1 3 C G r u b e r value of 1.5%o [613CG.ruber data from Wang et al, 1999]. The corrected 6 1 3 C o r g records were converted to C 0 2 ( a q ) values according to the linear relationship between 8 1 3 C o r g and C 0 2 ( a q ) derived by Rau et al. [1989]: [C0 2 ( a q )] = (5 1 3C o r g +12.6)/-0.8, (1) and the C 0 2 ( a q ) values were converted to partial pressure of C 0 2 (pC02) using Henry's Law: pC0 2 = C 0 2 ( a q ) / a , (2) where a is a temperature dependent solubility constant. This constant was calculated according to Weiss [1974] using the U K 3 7 SST estimates [Figs. 3.8-3.10; Pelejero et al, 1999]. In cases where o 1 3 C G r u b e r and U K ' 3 7 SST estimates are not available at the same depths as the 5 1 3 C o r g data, the former are linearly interpolated (never exceeding distances of 10 cm) to the depth of the 8 1 3C o r g values. 70 Fig. 3.15. The pC02 estimates of cores 17940-2, 17954-2, and 17961-2 from the northern, western central, and southern SCS versus age (see Fig. caption 3.3 for details on the age models). The pC02 of core 17940-2, that has the best age control of the three records \Wang et al, 1999], is enlarged to highlight the timing of changes during the last deglaciation. 71 The selection of 6 1 3 C G r u b e r and U K 3 7 SST estimates to correct for temporal changes in 8 1 3 C D I C , and to calculate a, respectively, makes the SCS records most closely comparable to a similar record from the Angola current region reported by Midler et al. [1994; see Fig. 3.16] who used an identical approach. It is noted, however, that G. ruber, a shallow-dwelling foraminifera, might not optimally monitor temporal changes in 8 1 3 C D I C at the depth where the majority of the marine organic matter is produced, i.e., at the chlorophyll maximum. 3.3.4 Discussion If 6 1 3 C o r g were a reliable proxy for local pCO^ i.e., a measurable descriptor for the desired variable [Wefer et al, 1999], it should meet three key criteria: (1) 5 1 3 C o r g -pC0 2 estimates should yield reliable absolute values; (2) the temporal development of these estimates should correlate with reconstructions of processes that have been shown to influence local pC0 2 in the present ocean, i.e., upwelling intensity; and finally, (3) the temporal evolution of local estimates should be consistent with atmospheric C 0 2 concentrations as recorded in ice cores, specifically in regions of the ocean where equilibrium between atmospheric and oceanic C 0 2 is expected. In the following sections these three premises will be tested. Absolute 6 1 3 C o r g -pC0 2 values Taken at face value, i.e., based on the assumption of 0 % terrigenous organic matter contribution to the bulk C o r g throughout the records, the SCS pCO z estimates presented in Fig. 3.15 (see also Fig. 3.19) suggest that sea surface pC0 2 values were persistently higher than the ice core C 0 2 record by ~100 ppm, and, accordingly, that the SCS has been a constant, significant source of C 0 2 to the atmosphere for the last 220 kyr, irrespective of glacial or interglacial boundary conditions. Given the time-varying presence of terrigenous organic matter in SCS sediments throughout the records, these SCS p C 0 2 estimates are maximum values. In addition, given the evidence from n-nonacosane abundances of significantly increased terrigenous organic matter input during glacials as compared to interglacials (Figs. 3.8-3.10), the reconstructed glacial-interglacial difference at each core site is a minimum estimate. An admixture of only 10-15 % terrigenous C o r g (8 1 3C o r g ca. -26.5 %>) to the bulk C o r g would be enough to decrease the concentration estimates from supersaturated to equilibrium values. The increased glacial contribution of terrigenous organic matter, however, implies even lower glacial pC0 2 estimates, i.e., an even larger glacial-interglacial contrast in excess of the ~80 ppm 72 observed in ice cores. In turn, this implies that the SCS at the core sites was either a considerable C 0 2 sink during glacial times or, alternatively, that it was a strong source during interglacials. Neither of these scenarios, however, is compatible with the current understanding of SCS palaeoceanography. A l l 5 1 3 C o r g -pC0 2 estimates published to date are based on the (tacit) assumption of 0 % terrigenous C o r g contribution throughout the records, seriously qualifying the absolute values of the pC0 2 estimates. Thus, the inability to quantify the terrigenous fraction of the bulk C o r g precludes a quantitative C 0 2 estimate based on & 1 3C of bulk organic matter. This pitfall can be overcome, however, by using the 51 3C composition of a single, uniquely marine biomarker, e.g., alkenones [Andersen et al, 1999; Jasper and Hayes, 1990,1994; see further discussion below]. The pCO z estimates and upwelling intensity In today's ocean, low latitude upwelling regions are the major sources of C 0 2 from the ocean to the atmosphere, with supersaturation ranging up to 140 ^atm [Tans et al, 1990; Takahashi et al, 1997, 2002; see also Fig. 3.7]. Accordingly, 8 1 3C 0 r g-pCO 2 records from upwelling regions, which all show Holocene pC0 2 values 40-100 ppm higher than preindustrial atmospheric C 0 2 values, have been attributed to this supersaturation [Pedersen et al, 1991; Midler et al, 1994; Westerhausen et al, 1994; Raymo and Horowitz, 1996]. As discussed in the previous section, the reconstructed supersaturation could also be due to the admixture of terrigenous C o r g . However, the pC0 2 estimates from bulk sedimentary 6 1 3 C o r g records from major upwelling sites do not appear to reflect changes in upwelling intensity with time. For example, off Angola, times of maximal differences between the local p C 0 2 and the Vostok ice core record, which imply periods of maximum upwelling [Mtiller et al, 1994], do not coincide with minima in sedimentary 5 1 5 N [Holmes et al, 1997; Fig. 3.16]. This relationship should be expected if the interpretation of the 6 1 5 N variations from the same core is correct. The interpretation of the 8 1 5 N records as reflecting upwelling strength is corroborated by synchronous changes in percent C o r g and SST [Holmes et al, 1997], biogenic Ba [Schneider et al, 1997], and organic biomarker compound abundances [Hinrichs et al, 1999] from the same core, all pointing to lower 5 1 5 N during periods of high production, i.e., intense upwelling; A similar independence of o 1 3 C o r g variations and changes in upwelling intensity can be observed in records from the NW Mexican Margin [core NH22P, Ganeshram et al, 1999], the Panama Basin [core P7GC, Pedersen et al, 1991] and the western Indian Margin [core MD76-131, Ganeshram et al, 2000]. In all these cases, sedimentary nitrogen isotope data [Ganeshram et al. 1999, Farrell et al. 1995, and Ganeshram et al. 2000, respectively], 73 maximum maximum upwelling CO source GeoB 1016-3 age ka Fig. 3.16. Bulk sedimentary 6 1 5 N data from Holmes et al. [1997] and pC0 2 reconstruction based on bulk 6 1 3 C o r g [Midler et ai, 1994] and on 6 1 3 C a l k e n o n e [Andersen et al, 1999] of core GeoB 1016-3 from the Angola Basin (8 in Fig. 3.7). Red shaded vertical bars: 6 1 5 N minima are upwelling maxima labeled according to Holmes et al. [1997]; black bars on upper x-axis denote periods of maximum air-sea C 0 2 difference (ApC02) based on S 1 3 C o r g - pC0 2 estimates, according to Mtiller et al. [1994]. Note the lack of an expected consistent correlation between upwelling maxima and ApC0 2 maxima. 74 in agreement with C o r g concentrations and various other indicators of marine production, indicate significant changes in upwelling intensity during the last glacial-interglacial cycle that are not reflected in the sedimentary 8 1 3 C o r g records and the pCC»2 values derived from them. A similar conclusion can be tentatively drawn from a set of sediment surface 5 1 3 C o r g values from the eastern equatorial Pacific (Fig. 3.17; T.F. Pedersen, unpubl. data, 2000). Here, 5 1 5 N values [Farrell et al, 1995; Fig. 3.17] show a clear response to equatorial upwelling, with lowest values centered on the equator and increasing 5 1 5 N values toward the north, reflecting increasing relative nutrient utilization [Farrell et al, 1995]. Recent studies clearly demonstrate that the equatorial Pacific upwelling is not only associated with increased levels of macronutrients but also with high C 0 2 concentrations in surface waters (up to 500 ^atm), i.e., with fluxes of CO z from the ocean to the atmosphere that rapidly diminish to the north and south of the equator where ocean-atmosphere equilibrium is reached with respect to C 0 2 [Chavez et al, 1999; Feely et al, 1999]. However, the surface sediment 5 1 3 C o r g distribution determined on the same set of samples as 5 1 5 N does not mirror the ~100 ^atm gradient in C0 2 ( a q ) governed by equatorial upwelling (Fig. 3.17). Given the lower SSTs associated with the core of equatorial upwelling, pC0 2 estimates derived from these bulk sedimentary 5 1 3 C o r g values would be, if anything, inversely correlated with pC02. In view of these considerations, we suggest that the postulation of an inverse coupling between pC0 2 and upwelling, i.e., a decreased local pC0 2 due to increased upwelling-driven production in the eastern Atlantic equatorial upwelling belt [Struck et al, 1993; Westerhausen et al, 1994; site 7 in Fig. 3.7] should be reevaluated. The correlation between decreased pC02 estimates with increased productivity of this equatorial Atlantic record in these studies was caused, to a significant extent, by the correction of the 6 1 3 C o r g values for winter SST variations that are, in turn, coupled to upwelling intensity [Westerhausen et al, 1994]. The annual average SSTs [U K ' 3 7 estimates; 23.2-27.5°C; Westerhausen et al, 1994] and even the foraminiferal transfer function winter SSTs in this region are higher than 15°C throughout the record (except for a few single values during stage 8), where, as Fontugne and Duplessy [1981] and Freeman and Hayes [1992] have shown, there is very little dependency of 8 1 3 C o r g on SST. Thus, the SST correction seems unjustified. Without the SST correction, there is no significant correlation (r=0.32) between 5 1 3 C o r g and the concentration of C o r g or derived export productivity (Fig. 3.18), in accord with the records summarized above. In addition, the postulation of an inverse coupling between upwelling intensity and marine productivity is at odds with all observations on modern equatorial upwelling systems that clearly demonstrate that 75 15 0 I I I I I I I I I I 1 1 1 1 1 1 1 1 1 a i i i i i i i 1 1 1 1 b • • • • • • • • • • • • • • • equator • • • • • • • , i . . . » i . . . i . . . i , , , i . i , , , , i • i , , , , 5 7 9 11 13 15 -23 -22 -21 -20 -19 6 1 5 N (%. vs. air) ° 1 3 C o r q (%. vs. PDB) Fig. 3.17. (a) Bulk sedimentary 5 1 5 N [Farrell et al, 1995], and (b) bulk sedimentary 81 3Co r„ surface sediment distribution (S 1 3C o r g data of T.F. Pedersen, unpublished data, 2000) along two transects (~ 110°W, filled circles; and ~ 90-95°W, filled squares) across the eastern equatorial Pacific upwelling zone. Note the clear reflection of increasing relative nitrate utilization, i.e., the decreasing supply of nitrate by upwelling from the equator toward the north and the absence in the 5*3Co r g distribution of any obvious reflection of the equatorial upwelling. core 16772, eastern equat. Atlantic 70 > 60 a 2 50 •> t j 40 3 1 O . 3 0 r o a « 20 10 « • '\.. \ • • • • . • . • * •• • s *• i . . . . i -21 -19 -17 -15 81 3C corrected for 61 3C only Fig. 3.18. Bulk sedimentary 5 1 3 C o r g (corrected for 6 1 3 C D I C using 5 1 3 C G n i t e r ) versus estimated export production (in gCm"2 yr"1) of core 16772 in the eastern equatorial Atlantic (n=153, r=0.32; all data from Westerhausen et al [1994]). 76 changes in the direction and intensity of C 0 2 exchange between ocean and atmosphere on interannual and longer timescales are governed by physical, rather than biological, processes [Feely et al, 1987,1999; Chavez et al, 1999]. It is possible that some of the observed down-core and surface sediment variability of o 1 3 C o r g in upwelling areas is obscured by changes in various biological factors, most importantly phytoplankton growth rate [Laws et al, 1995; Bidigare et al, 1997]. Increased average phytoplankton growth rates are associated with decreased C isotope fractionation [Laws et al, 1995; Bidigare et al, 1997]. Thus, the influence of upwelling-related higher concentrations of C0 2( a q) on the 5 1 3 C o r g of phytoplankton is potentially counteracted by increased growth rate associated with higher supply of nutrients due to the same upwelling. To overcome any potential bias due to this counteracting control on 5 1 3C o r g , future studies of sedimentary & 1 3 C o r g should include some measure of average phytoplankton growth rate [Andersen et al, 1999; Stoll and Schrag, 2000]. The first attempt to correct for the effect of variable growth rates on 8 1 3C o r g-pC02 estimates [Andersen et al, 1999], however, does not lead to an improved correlation of temporal changes in upwelling intensity and pC02 estimates either (Fig. 3.16). Moreover, if recent findings by Riebesell et al. [2000] are corroborated, carbon isotope fractionation may be affected by the growth-limiting resource (e.g., light levels, NCy versus N H 4 + N-source, N versus P or Fe as the growth-limiting nutrient), in addition to growth rate and C 0 2 concentration, making it even more difficult to reliably extract the influence of changes in C 0 2 ( a q ) on the sedimentary 5 1 3C o r g . Marine b13Cotg-pC02 estimates in comparison to the ice core CO z record Ignoring for the moment the uncertainty in absolute values of the low-latitude marine isotopic pC02 estimates and the apparent decoupling of upwelling intensity and pCOy most of the marine 6 1 3 C o r g records that we have discussed here record a ~2 %o glacial-interglacial difference, irrespective of the local physico-environmental setting. This is precisely the amplitude expected from the ~80 ppm change in atmospheric C 0 2 observed in ice cores [Rau, 1994]. Two of these records, however, deviate significantly from this general trend, possibly due to the effect of changes in phytoplankton growth rate discussed above. Pedersen et al. [1991; see also Pedersen and Bertrand, 2000] report a negligible deglacial 5 1 3 C o r g change in the Panama Basin, and the (uncorrected) 5 1 3 C o r g data of Westerhausen et al. [1994] from the equatorial Atlantic seem to indicate, if at all, a reversed trend, i.e., a & 1 3 C o r g increase associated with Termination I. The other low-latitude marine S 1 3 C o r g records (Fig. 3.7) show a very similar change to lighter values after Termination I, suggesting a common driving mechanism. 77 We evaluate the possible role of C02( a q) in determining these changes in 5 1 3 C o r g by examining the phasing of changes in the Vostok C 0 2 record and the low latitude marine 6 1 3 C o r g records/pC0 2 estimates, focusing on the MIS 5/4 transition and the last deglaciation. Thus, the Vostok C 0 2 record shows a more or less steady decrease in atmospheric C 0 2 concentration starting immediately after the penultimate (MIS 5e) C 0 2 maximum at ~130 to ~115 ka (Fig. 3.19). In sharp contrast, low-latitude marine pC02 estimates from a wide range of environmental settings [Fontugne and Calvert, 1992; Milller et al, 1994; Raymo et al., 1996; this study], including the molecular record of Jasper and Hayes [1994], indicate persistently high pC0 2 values from ~130 ka until the MIS 5/4 transition (~70 ka) when there is a sharp pC02 drop (the assertion of synchroneity of these estimates is constrained by the uncertainty in the age models and the low resolution of some of these records). None of the published low-latitude pC02 estimates follows the Vostok C 0 2 record during this time period, including the records from the SCS and the eastern Mediterranean where, based on current understanding, C 0 2 equilibrium between the ocean and the atmosphere is expected. If 5 1 3 C o r g is indeed driven by changes in C 0 2 ( a q ) , then there are two important corollaries: (1) persistently high surface pC0 2 in low-latitude upwelling regions during the later part of MIS 5 would need to be compensated for by even larger sinks elsewhere and (2) there would need to be a mechanism to drive those areas of the ocean that presumably are in equilibrium out of equilibrium synchronously with atmospheric C 0 2 during the later part of MIS 5. In the case of the SCS, the high late MIS 5 pC0 2 estimates could be due to the increased admixture of terrigenous C o r g as indicated by the n-nonacosane records (Figs. 3.9, 3.10). However, the similarity in most of the low-latitude 5 1 3 C o r g records during this time interval and the corresponding conclusion from pC0 2 estimates derived from marine biomarkers [Jasper and Hayes, 1994] seem to argue against this possibility. In contrast to the discrepancy during the MIS 5/4 transition, the timing of changes in 5 1 3 C o r g in the SCS cores during the last (and penultimate) deglaciation (Figs. 3.8-3.11) seems to parallel findings from the Vostok and Byrd C 0 2 records that show increases in atmospheric C 0 2 leading global ice volume [Petit et ah, 1999] and the Boiling/Allerod warming in the Northern Hemisphere [Blunier et al, 1997; Stocker, 2000], respectively. Despite this agreement in timing of the onset of deglacial change, SCS pC02 reaches Holocene levels only at ~5 ka following a continuous rise (Fig. 3.15), i.e., significantly later than the ice core records of atmospheric C 0 2 that indicate Holocene levels were reached at ~10 ka [e.g., Neftel et al, 1988]. Note that these discrepancies in the timing of deglacial events are much larger than any uncertainty in stratigraphic correlation of ice core and SCS sedimentary records, and in the absolute age control of 78 79 the SCS cores, particularly of core 17940-2 [Wang et al, 1999]. Moreover, this offset in timing is most likely not due to variable admixture of terrigenous C o r g at site 17940, given the unchanged concentrations of n-nonacosanes in core 17940-2 during the last ca. 10 ka (Fig. 3.8). In summary, the discrepancy in timing of marine S 1 3 C o r g -pC0 2 estimates and the ice core C 0 2 record suggests that C 0 2 ( a q ) may not be the main/ single driving mechanism of low-latitude 8 1 3 C o r g variability. The remarkably similar low-latitude 6 1 3 C o r g records from a wide range of environmental settings, particularly the similar timing of changes during the MIS 5/4 transition, however, suggests that a common mechanism governs 6 1 3 C o r g . This similarity in both timing and amplitude is even more surprising given the wide range of biological factors that have been shown to influence 6 1 3 C o r g (see discussion above), as well as the temporal and spatial variability in the admixture of terrigenous C In this context published power spectra of the ice core C 0 2 record, on the one hand, and the frequency distribution of variability in low-latitude upwelling intensity/marine production, on the other hand, seem to indicate a negligible role of changes in the low-latitude ocean on atmospheric C 0 2 . Thus, whereas low-latitude upwelling/marine production is significantly controlled by precessional periodicities [Lyle, 1988; Clemens et al, 1991; Beaufort et al, 1997; Holmes et al, 1997; Reichart et al, 1997; Schubert et al, 1998], the Vostok C 0 2 record is dominated by the 100-kyr component, with almost no power at lower periodicities [Petit et al, 1999]. Based on the presence of a minor 30 kyr period both in coccolith productivity records from the equatorial Indo-Pacific and in the Vostok C 0 2 record, Beaufort et al. [2001], however, recently argued for the importance of biological carbon fixation in those regions in influencing atmospheric co2. 3.3.5 Conclusions Sedimentary records from the SCS and other low-latitude locales caution the use of bulk sedimentary 6 1 3 C o r g records as a proxy indicator of local C 0 2 ( a q ) . In addition, the weakness/absence of a correlation between 6 1 3 C o r g /pC0 2 estimates and upwelling indicators suggests that previous evaluations of the importance of changes in upwelling-related oceanic degassing on glacial-interglacial atmospheric C 0 2 levels need to be reassessed, possibly using sedimentary 5 1 5 N as a more reliable proxy of nutrient utilization and/or supply. 80 The similarity in timing and absolute values of marine 6 1 3 C o r g records from very different settings could be taken as evidence for a common driving force for determining 6 l 3 Q r g - Previous studies concluded that all models that try to explain the biological/physiological processes governing the 6 1 3C of phytoplankton must be able to reproduce the latitudinal trend in 6 1 3 C o r g observed in the present-day ocean. We further suggest that all these models should be able to reproduce similar (in timing and amplitude) global-scale 5 1 3 C o r g variations on glacial-interglacial timescales, irrespective of the local setting. This conundrum provides a unique opportunity for combined plankton physiological and palaeoceanographic studies to determine the mechanism(s) governing the isotopic composition of marine organic matter. As with all palaeoceanographic reconstructions, pC02 estimates based on 5 1 3 C o r g and/or 81 3Can<e r i o n e s should be corroborated by independent evidence, i.e., an independent sedimentary proxy. 81 3.4 References Altabet, M . A., Variations in nitrogen isotopic composition between sinking and suspended particles: implications for nitrogen cycling and particle tranformation in the open ocean, Deep-Sea Research, 35, 535-554,1988. Altabet, M . A., Nitrogen and carbon isotopic tracers of the source and transformation of particles in the deep sea. in Particle flux in the ocean, edited by Ittekot, V., P. Schafer, S. Honjo and P. J. Depetris, pp. 155-184, John Wiley & Sons, 1996. Altabet, M . A., Nitrogen isotopic evidence for micronutrient control of fractional N03-utilization in the equatorial Pacific, Limnology and Oceanography, 46, 368-380, 2001. Altabet, M . A., and J. J. McCarthy, Temporal and spatial variations in the natural abundance of 1 5 N in PON from a warm-core ring, Deep-Sea Research, 32, 755-772, 1985. Altabet, M . A., and W. B. Curry, Testing models of past ocean chemistry using foraminiferal N / N , Global Biogeochemical Cycles, 3,107-119,1989. Altabet, M . A., and R. Francois, Sedimentary nitrogen isotopic ratio as a recorder for surface ocean nitrate utilization, Global Biogeochemical Cycles, 8,103-116,1994a. Altabet, M . A., and R. Francois, The use of nitrogen isotopic ratio for reconstruction of past changes in surface ocean nutrient utilization, in Carbon cycling in the glacial ocean: constraints on the ocean's role in global change, edited by Zahn, R., T. F. Pedersen, M . A . Kaminski and L. Labeyrie, pp. 281-306, Springer Verlag, 1994b. Altabet, M . A., M . J. Higginson, and D. W. Murray, The effect of millennial-scale changes in Arabian Sea denitrification on atmospheric C0 2 , Nature, 415,159-162, 2002. Altabet, M . A., W. G. Deuser, S. Honjo, and C. Stienen, Seasonal and depth-related changes in the source of sinking particles in the North Atlantic, Nature, 354, 136-139, 1991. Altabet, M . A., R. Francois, D. W. Murray, and W. L. Prell, Climate-related variations in denitrification in the Arabian Sea from sediment 1 5 N / W N ratios, Nature, 373, 506-509, 1995. Altabet, M . A., D. W. Murray, and W. L. Prell, Climate-linked oscillations in Arabian Sea denitrification over the last 1 Ma: Implications for the marine N cycle, Paleoceanography, 14, 732-743,1999a. Altabet, M . A., C. Pilskaln, R. Thunell, C. Pride, D. Sigman, F. Chavez, and R. Francois, The nitrogen isotope biogeochemistry of sinking particles from the margin of the Eastern North Pacific, Deep Sea Research, 46, 655-679,1999b. Andersen, N . , P. J. Mtiller, G. Kirst, and R. R. Schneider, Alkenone 6 1 3C as a proxy for past P C 0 2 in surface waters: results from the late Quaternary Angola Current, in Use of proxies in paleoceanography: Examples from the South Atlantic, edited by Fischer, G. and G. Wefer, pp. 469-488, Springer-Verlag, 1999. Andree, M . , H . Oeschger, W. Broecker, N . Beavan, M . Klas, A. Mix, G. Bonani, H . J. Hofmann, M . Suter, W. Woelfli, and T.-H. Peng, Limits of the ventilation rate for the deep ocean over the last 12 000 years, Climate Dynamics, 1, 53-62,1986. Bauer, S., G. L. Hitchcock, and D. B. Olson, Influence of monsoonally-forced Ekman dynamics upon surface layer depth and plankton biomass distribution in the Arabian Sea, Deep-Sea Research, 38, 531-553,1991. Beaufort, L., T. de Garidel-Thoron, A. Mix, and N . G. Pisias, ENSO-like forcing on oceanic primary production during the Late Pleistocene, Science, 293, 2440-2444, 2001. Beaufort, L., Y. Lancelot, P. Chamberlin, O. Cayre, E. Vincent, F. Bassinot, and L. Labeyrie, Insolation cycles as a major control of equatorial Indian Ocean primary production, Science, 278,1451-1454,1997. Bentaleb, I., M . Fontugne, C. Descolas-Gros, C. Girardin, A . Mariotti, C. Pierre, C. Brunet, and A. Poisson, Organic carbon isotopic composition of phytoplankton and 82 sea-surface pC0 2 reconstructions in the southern Indian Ocean during the last 50,000 yr, Organic Geochemistry, 24, 399-410,1996. Berger, W. H., V. S. Smetacek, and G. Wefer, Ocean productivity and paleoproductivity -an overview, in Productivity of the ocean - present and past, edited by Berger, W. H., V. S. Smetacek and G. Wefer, pp. 1-34, John Wiley & Sons, 1989. Bertrand, P., T. F. Pedersen, P. Martinez, S. E. Calvert, and G. Shimmield, Sea level impact on nutrient cycling in coastal upwelling areas during deglaciation: Evidence from nitrogen isotopes, Global Biogeochemical Cycles, 14, 341-355, 2000. Beveridge, N . A . S., and N . J. Shackleton, Carbon isotopes in recent planktonic foraminifera: A record of anthropogenic C 0 2 invasion of the surface ocean, Earth and Planetary Science Letters, 126, 259-273,1994. Bidigare, R. R., A. Fluegge, K. H . Freeman, K. L. Hanson, J. M . Hayes, D. Hollander, J. P. Jasper, L. L. King, E. A. Laws, J. Milder, F. J. Millero, R. Pancost, B. N . Popp, P. A . Steinberg, and S. G. Wakeham, Consistent fractionation of 1 3 C in nature and in the laboratory: growth-rate effects in some haptophyte algae, Global Biogeochemical Cycles, 11, 279-292,1997. Blunier, T., J. Schwander, B. Stauffer, T. Stocker, A . Dallenbach, A . Indermuhle, J. Tschumi, J. Chappellaz, D. Raynaud, and J.-M. Barnola, Timing of the Antarctic Cold Reversal and the atmospheric C 0 2 increase with respect to the Younger Dyas event, Geophysical Research Letters, 24, 2683-2686,1997. Boyle, E. A., Vertical ocean nutrient fractionation and glacial / interglacial C 0 2 cycles, Nature, 331, 55-56,1988. Brandes, J. A. G., Isotopic effects of denitrification in the marine environment, Ph.D., University of Washington, 1996. Brandes, J. A., A. H . Devol, T. Yoshinari, D. A. Jayakumar, and S. W. A. Naqvi, Isotopic composition of nitrate in the central Arabian Sea and eastern tropical North Pacific: a tracer for mixing and nitrogen cycles, Limnology and Oceanography, 43, 1680-1689, 1998. Broecker, W. S., Glacial to Interglacial Changes in Ocean Chemistry, Prog. Oceanography, 11,151-197,1982. Broecker, W. S., and G. M . Henderson, The sequence of events surrounding the Termination II and their implications for the cause of glacial-interglacial C 0 2 changes, Paleoceanography, 13, 352-364,1998. Broecker, W. S., and T.-H. Peng, Greenhouse Puzzles, 277 pp., Eldigio Press, 1998. Broecker, W. S., W. C. Patzert, J. R. Toggweiler, and M . Stuiver, Hydrography, chemistry, and radioisotopes in the Southeast Asian basins, Journal of Geophysical Research, 91,14345-14354,1986. Broecker, W. S., T. H . Peng, S. Trumbore, G. Bonani, and W. Wolfli, The distribution of radiocarbon in the glacial ocean, Global Biogeochemical Cycles, 4,103-117,1990. Broecker, W., M . Klas, N . Ragano-Beaven, G. Mathieu, A. Mix, M . Andree, H . Oeschger, W. Wolfli, M . Suter, G. Bonani, H . J. Hofmann, M . Nessi, and E. Morenzoni, Accelerator mass spectrometry radiocarbon measurements on marine carbonate samples from deep sea cores and sediment traps, Radiocarbon, 30, 261-295,1988. Burkhardt, S., U. Riebesell, and I. Zondervan, Effects of growth rate, C02 concentration, and cell size on the stable carbon isotope fractionation in marine phytoplankton, Geochimica et Cosmochimica Acta, 63, 3729-3741,1999. Calvert, S. E., B. Nielsen, and M . R. Fontugne, Evidence from nitrogen isotope ratios for enhanced productivity during formation of eastern Mediterranean sapropels, Nature, 359, 223-225,1992. Calvert, S. E., T. F. Pedersen, and R. C. Thunell, Geochemistry of surface sediments of the Sulu and South China Seas, Marine Geology, 114, 207-231, 1993. Capone, D. G., J. P. Zehr, H . W. Paerl, B. Bergmann, and E. J. Carpenter, Trichodesmium, a globally significant marine cyanobacterium, Science, 276,1221-1229,1997. 83 Carpenter, E. J., Nitrogen fixation by marine Oscillatoria (Trichodesmium) in the world's ocean, in Nitrogen in the marine environment, edited by Carpenter, E. J. and D. G. Capone, pp. 65-103, Academic Press, 1983a. Carpenter, E. J., Physiology and ecology of marine planktonic Oscillatoria (Trichodesmium), Marine Biology Letters, 4, 69-85,1983b. Carpenter, E. J., and K. Romans, Major role of the cyanobacterium Trichodesmium in nutrient cycling in the North Atlantic Ocean, Science, 254,1356-1358,1991. Carpenter, E. J., H . R. Harvey, B. Fry, and D. G. Capone, Biogeochemical tracers of the marine cyanobacterium Trichodesmium, Deep Sea Research, 44, 27-38,1997. Chao, S.-Y., P.-T. Shaw, and S. Y. Wu, Deep water ventilation in the South China Sea, Deep Sea Research, 43, 445-466,1996. Chavez, F. P., P. G. Strutto, G. E. Friederich, R. A. Feely, G. C. Feldman, D. G. foley, and M . J. McPhaden, Biological and chemical response of the equatorial Pacific Ocean to the 1997-98 El Nino, Science, 286, 2126-2131,1999. Checkley, D. M . , and C. A . Miller, Nitrogen isotope fractionation by oceanic zooplankton, Deep Sea Research, 36,1449-1456,1989. Chen, C. T. A., S. L. Wang, B. J. Wang, Y. P. Lin, W. H . Liu, F. C. Li , and L. C. Chen, Preliminary data report of Ocean Researcher I cruises 508 and Ocean Researcher III 403, Rep. 33, Institute of Marine Geology and Chemistry, National Su Yat-sen University, 1998. Christensen, J. P., J. W. Murray, A . H . Devol, and L. A. Codispoti, Denitrification in continental shelf sediments has major impact on the oceanic nitrogen budget, Global Biogeochemical Cycles, 1, 97-116,1987. Clemens, S., W. Prell, D. Murray, G. Shimmield, and G. Weedon, Forcing mechanisms of the Indian Ocean monsoon, Nature, 353, 720-725,1991. Cline, J. D., and I. R. Kaplan, Isotopic fractionation of dissolved nitrate during denitrification in the eastern tropical north Pacific Ocean, Marine Chemistry, 3, 271-299,1975. Degens, E. T., Biogeochemistry of stable carbon isotopes, in Organic Geochemistry, edited by Eglinton, G. and M . T. J. Murphy, pp. 304-329, Elsevier, New York, 1969. Degens, E. T., M . Behrendt, B. Gotthard, and E. Reppmann, Metabolic fractionation of carbon isotopes in marine plankton. II. Data on samples collected off the coast of Peru and Ecuador, Deep-Sea Research, 15,11-20,1968. Deines, P., The isotope composition of reduced organic carbon, in Handbook of environmental isotope geochemistry, edited by Fritz, P. and J. C. Fontes, pp. 329-406, Elsevier, 1980. Eppley, R., and B. J. Peterson, Particulate organic matter flux and planktonic new production in the deep ocean, Nature, 282, 677-680,1979. Falkowski, P. G., Evolution of the nitrogen cycle and its influence on the biological sequestration of C 0 2 in the ocean, Nature, 387, 272-275,1997. Farrell, J. W., T. F. Pedersen, S. E. Calvert, and B. Nielsen, Glacial-interglacial changes in nutrient utilization in the equatorial Pacific Ocean, Nature, 377, 514-517,1995. Feely, R. A., R. H . Gammon, B. A. Taft, P. E. Pullen, L. S. Waterman, T. J. Conway, J. F. Gendron, and D. P. Wisegarver, Distribution of chemical tracers in the eastern equatorial Pacific during and after the 1982-1983 El Nino/Southern Oscillation Event, Journal of Geophysical Research, 92, 6545-6558,1987. Feely, R. A., R. Wanninkhof, T. Takahashi, and P. Tans, Influence of El Nino on the equatorial Pacific contribution to atmospheric C 0 2 accumulation, Nature, 398, 597-601,1999. Fischer, G., P. J. Mtiller, and G. Wefer, Latitudinal 5 1 3 C o r g variations in sinking matter and sediments from the South Atlantic: effects of anthropogenic C 0 2 and implications for paleo-PC02 reconstructions, Journal of Marine Systems, 17, 471-495,1998. 84 Fischer, G., R. Schneider, P. J. Muller, and G. Wefer, Anthropogenic C 0 2 in Southern Ocean surface waters: evidence from stabel organic carbon isotopes, Terra Nova, 9, 153-157,1997. Fontugne, M . R., and S. E. Calvert, Late Pleistocene variability of the carbon isotopic composition of organic matter in the eastern Mediterranean: monitor of changes in carbon sources and atmosperic C 0 2 concentrations, Paleoceanography, 7,1-20,1992. Fontugne, M . R., and J.-C. Duplessy, Organic carbon isotopic fractionation by marine plankton in the temperature range -1° to 31°C, Oceanologica Acta, 4, 85-89,1981. Frangois, R., M . A. Altabet, and L. H . Burckle, Glacial to interglacial changes in surface nitrate utilization in the Indian sector of the Southern Ocean as recorded by sediment 8 1 5N, Paleoceanography, 7, 589-606,1992. Francois, R., M . A. Altabet, R. Goericke, D. C. McCorkle, C. Brunet, and A. Poisson, Changes in the 61 3C of surface water particulate organic matter across the subtropical convergence in the S.W. Indian Ocean, Global Biogeochemical Cycles, 7, 627-644,1993a. Francois, R., M . P. Bacon, M . A. Altabet, and L. D. Labeyrie, Glacial/interglacial changes in sediment rain rate in the SW Indian sector of subantarctic waters as recorded by 2 3 0Th, 2 3 1Pa, U, and 5 1 5N, Paleoceanography, 8, 611-629,1993b. Francois, R., M . A. Altabet, E.-F. Yu, D. M . Sigman, M . P. Bacon, M . Frank, G. Bohrmann, G. Bareille, and L. D. Labeyrie, Contribution of Southern Ocean surface-water stratification to low atmospheric C 0 2 concentrations during the last glacial period, Nature, 389, 929-935,1997. Freeman, K. H., and J. M . Hayes, Fractionation of carbon isotopes by phytoplankton and estimates of ancient C 0 2 levels, Global Biogeochemical Cycles, 6,185-198,1992. Ganeshram, R. S., On the glacial-interglacial variability of upwelling, carbon burial and denitrification on the northwestern Mexican continental margin, Ph.D., University of British Columbia, 1996. Ganeshram, R. S., T. F. Pedersen, S. E. Calvert, and J. W. Murray, Large changes in oceanic nutrient inventories from glacial to interglacial periods, Nature, 376, 755-758, 1995. Ganeshram, R. S., S. E. Calvert, T. F. Pedersen, and G. L. Cowie, Factors controlling the burial of organic carbon in laminated and bioturbated sediments off NW Mexico: Implications for hydrocarbon preservation, Geochimica et Cosmochimica Acta, 63, 1723-1734,1999. Ganeshram, R., T. F. Pedersen, S. E. Calvert, G. W. McNeill, and M . R. Fontugne, Glacial-interglacial variability in denitrification in the World's Oceans: causes and consequnces, Paleoceanography, 15, 361-376, 2000. Ganeshram, R. S., T. F. P. Pedersen, S. E. Calvert, and R. Francois, Reduced nitrogen fixation in the glacial ocean inferred from changes in marine nitrogen and phosphorus inventories, Nature, 415,156-159, 2002. Goericke, R., and B. Fry, Variations of marine plankton o 1 3C with latitude, temperature, and dissolved C 0 2 in the world oceans, Global Biogeochemical Cycles, 8, 85-90,1994. Goering, J. J., Denitrification in the oxygen minimum layer of the eastern tropical Pacific Ocean, Deep Sea Research, 15,157-164,1968. GofVi, M . A., K. C. Ruttenberg, and T. I. Eglinton, A reassessment of the sources and importance of land-derived oragnic matter in surface sediments from the Gulf of Mexico, Geochimica et Cosmochimica Acta, 62, 3055-3075,1998. Gruber, N . , and J. L. Sarmiento, Global patterns of marine nitrogen fixation and denitrification, Global Biogeochemical Cycles, 11, 235-266,1997. Gupta, A., A. Rahman, P. P. Wong, and J. Pitts, The Old Alluvium of Singapore and the extinct drainage system to the South China Sea, Earth Surface Processes and Landforms, 12, 259-275,1987. 85 Haug, G., T. F. Pedersen, D. M . Sigman, S. E. Calvert, B. Nielsen, and L. C. Peterson, Glacial/interglacial variations in production and nitrogen fixation in the Cariaco Basin during the last 580 kyr, Paleoceanography, 13, 427-432,1998. Hinrichs, K.-U., R. R. Schneider, P. J. Mtiller, and J. Rullkotter, A biomarker perspective on paleoproductivity variations in two Late Quaternary sediment sections from the southeast Atlantic Ocean, Organic Geochemistry, 30, 341-366,1999. Holmes, M . E., C. Eichner, U . Struck, and G. Wefer, Reconstruction of surface ocean nitrate utilization using stable nitrogen isotopes in sinking particles and sediments. in Use of proxies in paleoceanography: Examples from the South Atlantic, edited by Fischer, G. and G. Wefer, pp. 447-468, Springer-Verlag, 1999. Holmes, M . E., P. J. Muller, R. R. Schneider, M . Segl, J. Patzold, and G. Wefer, Stable nitrogen isotopes in Angola Basin sediments, Marine Geology, 134,1-12,1996. Holmes, M . E., R. R. Schneider, P. J. Muller, M . Segl, and G. Wefer, Reconstruction of past nutrient utilization in the eastern Angola basin based on sedimentary 1 5 N / W N ratios, Paleoceanography, 12, 604-614,1997. Holmes, M . E., P. J. Muller, R. R. Schneider, M . Segl, and G. Wefer, Spatial variations in euphotic zone nitrate utilization based on & N in surface sediments, Geo-Marine Letters, 18, 58-65,1998. Huang, C.-Y., S.-F. Wu, M . Zhao, M.-T. Chen, C.-H. Wang, X. Tu, and P. B. Yuan, Surface ocean and monsoon climate variability in the South China Sea since the last glaciation, Marine Micropaleontology, 32, 71-94,1997. Huang, W., and P. Wang, A quantitative approach to deep-water sedimentation in the South China Sea: changes since the last glaciation, Science in China (Series D), 41, 195-201,1998. Jasper, J. P., and J. M . Hayes, A carbon isotope record of C 0 2 levels during the late Quaternary, Nature, 347, 462-464,1990. Jasper, J. P., and J. M . Hayes, Reconstruction of paleoceanic pC0 2 levels from carbon isotopic compositions of sedimentary biogenic components, in Carbon cycling in the glacial ocean: constraints on the ocean's role in global change, edited by Zahn, R., T. F. Pedersen, M . A. Kaminski and L. Labeyrie, pp. 323-341, Springer Verlag, 1994. Jian, Z., B. Huang, W. Kuhnt, and H.-L. Lin, Late Quaternary upwelling intensity and East Asian Monsoon forcing in the South China Sea, Quaternary Research, 55, 363-370, 2001. Kao, S. J., S. Lin, C. H . Wang, I. T. Lin, and K. K. Liu, Sedimentary organic carbon and nitrogen contents and their isotopic compositions along two transects on the East China Sea shelf, Geochemical Journal, 35, 237-244, 2001. Karl, D., R. Letelier, L. Tupas, J. Dore, J. Christian, and D. Hebel, The role of nitrogen fixation in biogeochemical cycling in the subtropical North Pacific Ocean, Nature, 388, 533-538,1997. Keil, R. G., E. Tsamakis, C. B. Fuh, J. C. Giddings, and J. I. Hedges, Mineralogical and textural controls on the organic composition of coastal marine sediments: hydrodynamic separation using SPLITT-fractionation, Geochimica et Cosmochimica Acta, 58, 879-893,1994. Keir, R. S., Cold surface ocean ventilation and its effect on atmospheric C 0 2 , Journal of Geophysical Research, 98, 849-856,1993. Kienast, M . , Constant nitrogen isotopic composition of organic matter in the South. China Sea during the last climatic cycle: Global implications, Paleoceanography, 15, 244-253, 2000. Kienast, M . , S. E. Calvert, C. Pelejero, and J. O. Grimalt, A critical review of marine sedimentary 6 1 3C o r g-pC0 2 estimates: New paleo-records from the South China Sea ' and a revisit of other low-latitude 6 1 3C o r g-pC0 2 records, Global Biogeochemical Cycles, 15,113-127, 2001. 86 Kienast, S. S., S. E. Calvert, and T. F. P. Pedersen, Nitrogen Isotope and Productivity Variations along the North East Pacific Margin over the last 120 kyr: Surface and Subsurface Palaeoceanography, Paleoceanography, 2002, in press. Kinkade, C , J. Marra, C. Langdon, C. Knudson, and A. G. Ilahude, Monsoonal differences in phytoplankton biomass and production in the Indonesian seas: tracing vertical mixing using temperature, Deep Sea Research, 44, 581-592,1997. Kuhnt, W., S. Hess, and Z. Jian, Quantitative composition of benthic foraminiferal assemblages as a proxy indicator for organic carbon flux rates in the South China Sea, Marine Geology, 156,123-157,1999. Kukert, H . , and U . Riebesell, Phytoplankton carbon isotope fractionation during a diatom spring bloom in a Norwegian fjord, Marine Ecology Progress Series, 173, 127-137,1998. La Violette, P. E., and T. R. Frontenac, Temperature, salinity, and density of the World's seas: South China Sea and adjacent gulfs, 1967. Laws, E. A., B. N . Popp, R. R. Bidigare, M . C. Kennicutt, and S. A. Macko, Dependence of phytoplankton carbon isotopic composition on growth rate and [C02]aq: Theoretical considerations and experimental results, Geochimica et Cosmochimica Acta, 59, 1131-1138,1995. Lin, H.-L., C.-T. Lai, H.-C. Ting, L. Wang, M . Sarnthein, and J.-J. Hung, Late Pleistocene nutrients and sea surface productivity in the South China Sea: a record of teleconnections with Northern hemisphere events, Mar. Geology, 156,197-210,1999. Liu, K.-K., and I. R. Kaplan, The eastern tropical Pacific as a source of 1 5N-enriched nitrate in seawater off southern California, Limnology and Oceanography, 34, 820-830, 1989. Liu, K.-K., M.-J. Su, C.-R. Hsueh, and G.-C. Gong, The nitrogen isotopic composition of nitrate irt the Kuroshio water northeast of Taiwan: evidence for nitrogen fixation as a source of isotopically light nitrate, Marine Chemistry, 54, 273-292, 1996. Lyle, M . , Climatically forced organic carbon burial in equatorial Atlantic and Pacific Oceans, Nature, 335, 529-532,1988. Mariotti, A., J. C. Germon, P. Hubert, P. Kaiser, R. Letolle, A. Tardieux, and P. Tardieux, Experimental determination of nitrogen kinetic isotope fractionation: some principles; illustration for the denitrfication and nitrification processes, Plant and Soil, 62, 413-430,1981. Mariotti, A., C. Lancelot, and G. Billen, Natural isotopic composition of nitrogen as a tracer of origin for suspended organic matter in the Scheldt estuary, Geochim. Cosmochhn. Acta, 48, 549-555,1984. Martin, J. H. , Glacial-interglacial C 0 2 change: The iron hypothesis, Paleoceanography, 5, 1-13,1990. Martinez, J. I., P. De Deckker, and T. T. Barrows, Palaeoceanography of the last glacial maxmum in the eastern Indian Ocean: planktonic foraminiferal evidence, Palaeogeography, Palaeoclimatology, Palaeoecology, 147, 73-99,1999. McElroy, M . B., Marine biological controls on atmospheric C 0 2 and climate, Nature, 302, 328-329,1983. Miao, Q., and R. C. Thunell, Late Pleistocene-Holocene distribution of deep-sea benthic foraminifera in the South China Sea and Sulu Sea: paleoceanographic implications, Journal of Foraminiferal Research, 9-23,1996. Middelburg, J. J., K. Soetaert, and P. M . J. Herman, Empirical relationships for use in global diagenetic models, Deep Sea Research, Part I, 44, 327-344,1997. Minagawa, M . , and E. Wada, Nitrogen isotope ratios of red tide organisms in the East China Sea: a characterization of biological nitrogen fixation, Marine Chemistry, 19, 245-259,1986. 87 Minagawa, M. , Y. Kadota, and T. Kuramoto, Nitrogen and carbon isotopic compositions of settling particles in the East China Sea: seasonality, locality and effect of settling depth, Deep Sea Research, 1999, subm. Minoura, K., K. Hoshino, T. Nakamura, and E. Wada, Late Pleistocene-Holocene paleoproductivity circulation in the Japan Sea: sea-level control on 6 1 3C and 5 1 5 N records of sediment organic material, Palaeogeography, Palaeoclimatology, Palaeoecology, 135, 41-50,1997. Molengraaff, G. A . F., Modern deep-sea research in the East Indian Archipelago, The Geographical Journal, 57, 95-121,1921. Mortlock, R. A., and P. N . Froelich, A simple method for the rapid determination of biogenic opal in pelagic marine sediments, Deep-Sea Research, 36,1415-1426, 1989. Miiller, P. J., C / N ratios in Pacific deep-sea sediments: effect of inorganic and organic nitrogen compounds sorbed by clays, Geochimica et Cosmochimica Acta, 41, 765-776, 1977. Miiller, P. J., R. Schneider, and G. Ruhland, Late Quaternay p C 0 2 variations in the Angola current: evidence form organic carbon 5 1 3C and alkenone temperatures, in Carbon cycling in the glacial ocean: constraints on the ocean's role in global change, edited by Zahn, R., T. F. Pedersen, M . A . Kaminski and L. Labeyrie, pp. 343-367, Springer Verlag, 1994. Nakatsuka, T., N . Handa, N . Harada, T. Sugimoto, and S. Imaizumi, Origin and decomposition of sinking particulate organic matter in the deep water column inferred from the vertical distributions of its 8 1 5N, 6 1 3C and 81 4C, Deep Sea Research I, 44,1957-1979,1997. Neftel, A., H . Oeschger, T. Staffelbach, and B. Stauffer, C 0 2 record in the Byrd ice core 50,000-5,000 years BP, Nature, 331, 609-611,1988. Ohkouchi, N . , K. Kawamura, H . Kawahata, and A. Taira, Latitudinal distribution of terrestrial biomarkers in the sediments from the Central Pacific, Geochimica et Cosmochimica Acta, 61,1911-1918,1997. Pancost, R. D., K. H . Freeman, and S. G. Wakeham, Controls on the carbon-isotope compositions of compounds in Peru surface waters, Organic Geochemistry, 30, 319-340,1999. Pancost, R. D., K. H . Freeman, S. G. Wakeham, and C. Y. Robertson, Controls on carbon isotope fractionation by diatoms in the Peru upwelling region, Geochimica et Cosmochimica Acta, 61, 4983-4991,1997. Pedersen, T. F., and P. Bertrand, Influences of oceanic rheostats and amplifiers on atmospheric C02 content during the late Quaternary, Quaternary Science Reviews, 19, 273-283, 2000. Pedersen, T., B. Nielsen, and M . Pickering, Timing of late Quaternary productivity pulses in the Panama basin and implications for atmospheric CO^ Paleoceanography, 6, 657-677,1991. Pelejero, C , J. O. Grimalt, S. Heilig, M . Kienast, and L. Wang, High resolution U K 3 7 -temperature reconstructions in the South China Sea over the past 220 kyr, Paleoceanography, 14, 224-231,1999. Peters, K. E., R. E. Sweeney, and I. R. Kaplan, Correlation of carbon and nitrogen stable isotope ratios in sedimentary organic matter, Limnology and Oceanography, 23, 598-604,1978. Petit, J. R., J. Jouzel, D. Raynaud, N . I. Barkov, J.-M. Barnola, I. Basile, M . BErnder, J. Chappellaz, M . Davis, G. Delaygue, M . Delmotte, V. M . Kotlyakov, M . Legrand, V. Y. Linpenkov, C. Lorius, L. Pepin, C. Ritz, E. Saltzman, and M . Stievenard, Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica, Nature, 399, 429-436,1999. 88 Popp, B. N . , R. Takigiku, J. M . Hayes, J. W. Louda, and E. W. Baker, The post-Paleozoic chronology and mechanism of 1 3 C depletion in primary marine organic matter, American Journal of Science, 289, 436-454,1989. Popp, B. N . , E. A . Laws, R. R. Bridigare, J. E. Dore, K. L. Hanson, and S. G. Wakeham, Effect of phytoplankton cell geometry on carbon isotope fractionation, Geochimica et Cosmochimica Acta, 62, 69-77,1998. Popp, B. N . , T. Trull, F. Kenig, S. G. Wakeham, T. M . Rust, B. Tilbrook, F. B. Griffiths, S. W. Wright, H . J. Marchant, R. R. Bidigare, and E. A. Laws, Controls on the carbon isotopic composition of Southern Ocean phytoplankton, Global Biogeochemical Cycles, 13, 827-843,1999. Prahl, F. G., J. R. Ertel, M . A. Goni, M . A. Sparrow, and B. Eversmeyer, Terrestrial organic carbon contributions to sediments on the Washington margin, Geochimica et Cosmochimica Acta, 58, 3035-3048,1994. Pride, C , R. Thunell, D. Sigman, L. Keigwin, M . Altabet, and E. Tappa, Nitrogen isotopic variations in the Gulf of California since the last deglaciation: response to global climate changes, Paleoceanography, 14, 397-409,1999. Quay, P. D., B. Tilbrook, and C. S. Wong, Oceanic uptake of fossil fuel C 0 2 : Carbon-13 evidence, Science, 256, 74-79,1992. Rau, G. H., Variations in sedimentary organic 61 3C as a proxy for past changes in ocean and atmospheric C 0 2 concentrations, in Carbon cycling in the glacial ocean: constraints on the ocean's role in global change, edited by Zahn, R., T. F. Pedersen, M . A. Kaminski and L. Labeyrie, pp. 307-321, Springer Verlag, 1994. Rau, G. H. , U . Riebesell, and D. Wolf-Gladrow, A model of photosynthetic 1 3 C fractionation by marine phytoplankton based on diffusive molecular C 0 2 uptake, Marine Ecology Progress Series, 133, 275-285,1996. Rau, G. H. , U . Riebesell, and D. Wolf-Gladrow, C02aq-dependent photosynthetic 1 3 C fractionation in the ocean: a model versus measurements, Global Biogeochemical Cycles, 11, 267-278,1997. Rau, G. H., T. Takahashi, and D. J. Des Marais, Latitudinal variations in plankton 5 1 3C: implications for C 0 2 and productivity in past oceans, Nature, 341, 516-518,1989. Rau, G. H., P. N . Froelich, T. Takahashi, and D. J. Des Marais, Does sedimentary organic 6 1 3C record variations in Quaternary ocean (C02) ( a q ))?, Paleoceanography, 6, 335-347, 1991a. Rau, G. H., T. Takahashi, D. J. Des Marais, and C. W. Sullivan, Particulate organic matter 5 1 3C variations across the Drake Passage, Journal of Geophysical Research, 96, 15131-15135,1991b. Rau, G. H. , T. Takahashi, D. J. Des Marais, D. J. Repeta, and J. H . Martin, The relationship between 5 1 3C of organic matter and [C02(aq)] in ocean surface water: Data from a JGOFS site in the northeast Atlantic Ocean and a model, Geochimica et Cosmochimica Acta, 56,1413-1419,1992. Raven, J. A. , A . M . Johnston, and D. H . Turpin, Influence fo changes in C 0 2 concentration and temperature on marine phytoplankton 1 3 C / 1 2 C ratios: an analysis of possible mechanisms, Global and Planetary Change, 8,1-12,1993. Raymo, M . E., and M . Horowitz, Organic carbon paleo-pC02 and marine-ice core correlations and chronology, Geophysical Research Letters, 23, 367-370,1996. Raymo, M . E., B. Grant, M . Horowitz, and G. H . Rau, Mid-Pliocene warmth: stronger greenhouse and stronger conveyor, Marine Micropaleontology, 27, 313-326,1996. Rehder, G., and E. Suess, Methane and p C 0 2 in the Kuroshio and the South China Sea during maximum summer surface temperatures, Marine Chemistry, 75, 89-108, 2001. Reichart, G. J., M . den Dulk, H . J. Visser, C. H . van der Weijden, and W. J. Zachariasse, A 225 kyr record of dust supply, paleoproductivity and the oxygen minimum zone from Murray Ridge (northern Arabian Sea), Palaeogeography, Palaeoclimatology, Palaeoecology, 134,149-169,1997. 89 Riebesell, U., S. Burkhardt, A. Dauelsberg, and B. Kroon, Carbon isotope fractionation by a marine diatom: dependence on the growth-rate-limiting resource, Marine Ecology Progress Series, 193, 295-303, 2000. Roux, A., P. Martinez, P. Bertrand, T. F. Pedersen, K. Charlier, L. Pichevin, Seaward variations in nutrient utilization and organic sedimentation during the last 200,000 years in the coastal upwelling cell of Liideritz (the Benguela Current), Program&Abstracts 7 t h Internatl. Conf. On Paleoceanography, 176, 2001 Saino, T., and A. Hattori, Nitrogen fixation by Trichodesmium and its significance in nitrogen cycling in the Kuroshio area and adjacent waters, in The Kuroshio IV, edited by Takenouti, A. Y., pp. 697-709,1980. Saino, T., and A. Hattori, Geographical variation of the water column distribution of suspended particulate organic nitrogen and its 1 5 N natural abundance in the Pacific and its marginal seas, Deep-Sea Research, 34, 807-827,1987. Sarnthein, M. , K. Winn, J.-C. Duplessy, and M . R. Fontugne, Global variations of surface ocean productivity in low and mid latitudes: influence on C 0 2 reservoirs of the deep ocean and atmosphere during the last 21,000 years, Paleoceanography, 3, 361-399,1988. Schneider, R. R., B. Price, P. J. Muller, D. Kroon, and I. Alexander, Monsoon related variations in Zaire (Congo) sediment load and influence of fluvial silicate supply on marine productivity in the east equatorial Atlantic during the last 200,000 years, Paleoceanography, 12, 463-481,1997. Schdnfeld, J., and H.-R. Kudrass, Hemipelagic sediment accumulation rates in the South China Sea related to Late Quaternary sea-level changes, Quaternary research, 40, 368-379,1993. Schubert, C. J., J. Villanueva, S. E. Calvert, G. L. Cowie, U . von Rad, U . Berner, and H . Erlenkeuser, Stable phytoplankton community structure in the Arabian Sea over the past 200,000 years, Nature, 394, 563-566,1998. Shaw, P.-T., The intrusion of water masses into the sea southwest of Taiwan, Journal of Geophysical Research, 94,18213-18226,1989. Shaw, P.-T., The seasonal variation of the intrusion of the Philippine sea water into the South China Sea, Journal of Geophysical Research - Part C, 96, 821-827,1991. Shaw, P.-T., and S.-Y. Chao, Surface circulation in the South China Sea, Deep Sea Research, 41,1663-1683,1994. Sigman, D. M . , The role of biological production in Pleistocene atmospheric carbon dioxide variations and the nitrogen isotope dynamics of the Southern Ocean, Ph.D., MIT/WHOI, 97-28:385pp, 1997. Sigman, D. M . , M . A. Altabet, R. Francois, D. C. McCorkle, and J.-F. Gaillard, The isotopic composition of diatom-bound nitrogen in Southern Ocean sediments, Paleoceanography, 14,118-134,1999. Sigman, D. M . , M . A. Altabet, R. Michener, D. C. McCorkle, B. Fry, and R. M . Holmes, Natural abundance-level measurement of the nitrogen isotopic composition of oceanic nitrate: an adaption of the ammonia diffusion method, Marine Chemistry, 57, 227-242,1997. Stattegger, K., W. Kuhnt, H . Wong, et al., Cruise report SONNE 115: Sundaflut - Sequence stratigraphy, late Pleistocene-Holocene sea level fluctuations and high resolution record of the post-Pleistocene transgression on the Sunda Shelf 211 pp., 1997. Steinke, S., M . Kienast, and T. Hanebuth, On the significance of sea-level variations and shelf palaeo-morphology in governing sedimentation in the southern South China Sea during the last deglaciation, Marine Geology, 2002, in press. Stephens, B. B., and R. F. Keeling, The influence of Antarctic sea ice on glacial-interglacial C 0 2 variations, Nature, 404,171-174, 2000. Stocker, T. F., Past and future reorganizations in the climate system, Quaternary Science Reviews, 19, 301-319, 2000. 90 Stoll, H . M . , and D. P. Schrag, Coccolith Sr/Ca as a new indicator of coccolithophorid calcification and growth rate, Geochemistry, Geophysics, Geosystems, 1, 2000. Struck, U. , M . Sarnthein, L. Westerhausen, J. M . Barnola, and D. Raynaud, Ocean-atmosphere carbon exchange: impact of the "biological pump" in the Atlantic equatorial upwelling belt over the last 330,000 years, Palaeogeography, Palaeoclimatology, Palaeoecology, 103, 41-56,1993. Struck, U., K.-C. Emeis, M . Voss, M . D. Krom, and G. H . Rau, Biological productivity during sapropel S5 formation in the eastern Mediterranean Sea: Evidence from stable isotopes of nitrogen and carbon, Geochim. Cosmochim. Acta, 65, 3249-3266, 2001. Sun, X., and X. Li , A pollen record of the last 37 ka in deep sea core 17940 from the northern slope of the South China Sea, Marine Geology, 156, 227-244,1999. Sun, X., X. Li , and H.-J. Beug, Pollen distribution in hemipelagic surface sediments of the South China Sea and relation to modern vegetation distribution, Marine Geology, 156, 211-226,1999. Suthhof, A. , V. Ittekkot, and B. Gaye-Haake, Millennial-scale oscillations of denitrification intensity in the Arabian Sea during the late Quaternary and its potential influence on atmospheric N 2 0 and global climate, Global Biogeochemical Cycles, 15, 637-649, 2001. Takahashi, M . , and T. Hori, Abundance of picophytoplankton in the subsurface chlorophyll maximum layer in subtropical and tropical waters, Marine Biology, 79, 177-186,1984. Takahashi, T., R. A . Feely, R. F. Weiss, R. H . Wanninkhof, D. W. Chipman, S. C. Sutherland, and T. T. Takahashi, Global air-sea lux of C 0 2 : An estimate based on measurements of sea-air pC0 2 differences, Proc. Natl. Acad. SCi. USA, 94, 8292-8299, 1997. Takahashi, T., S. C. Sutherland, C. Sweeney, A. Poisson, N . Metzl, N . Bates, R. H . Wanninkhof, R. A. Feely, C. Sabine, J. Olafsson, and Y. Nojiri, Global sea-air C 0 2 flux based on climatological surface ocean pCO^ and seasonal biological and temperature effects, Deep Sea Research II, 49,1601-1622, 2002. Tans, P. P., I. Y. fung, and T. Takahashi, Observational constraints on the global atmospheric C02 budget, Science, 247,1431-1438,1990. Thunell, R. C , Q. Miao, S. E. Calvert, and T. F. Pedersen, Glacial-Holocene biogenic sedimentation patterns in the South China Sea: productivity variations and surface water pC0 2 , Paleoceanography, 7,143-162,1992. Toggweiler, J. R., Variations of atmospheric C 0 2 by ventilation of the ocean's deepest water, Paleoceanography, 14, 571-588,1999. Tsunogai, S., Ammonia in he oceanic atmosphere and the cycle of itrogen compounds through the atmosphere and the hydrosphere, Geochemical Journal, 5, 57-67,1971. Tyrrell, T., The relative influences o nitrogen and phosphorus on oceanic primary production, Nature, 400, 525-531,1999. Villanueva, J., J. O. Grimalt, E. Cortijo, L. Vidal, and L. Labeyrie, A biomarker approach to the organic matter deposited in the North Atlantic during the last climatic cycle, Geochimica et Cosmochimica Acta, 61, 4633-4646,1997. Voss, M. , M . A . Altabet, and V. Bodungen, 8 1 5 N in sedimenting particles as indicator of euphotic-zone processes, Deep-Sea Research, 43, 33-47,1996. Voss, M . , J. W. Dippner, and J. P. Montoya, Nitrogen isotope patterns in the oxygen-deficient waters of the eatern tropical North Pacific Ocean, Deep Sea Research, Part I, 48,1905-1921, 2001. Wada, E., and A. Hattori, Natural abundance of 1 5 N in particulate organic matter in the North Pacific Ocean, Geochimica et Cosmichimica Acta, 40, 249-251,1976. Wada, E., and A. Hattori, Nitrogen isotope effects in the assimilation of inorganic nitrogenous compounds by marine diatoms, Geomicrobiology Journal, 1, 85-101, 1978. 91 Wada, E., T. Kadonaga, and S. Matsuo, 1 5 N abundance in nitrogen of naturally occurring substances and global assessment of denitrification from isotopic viewpoint, Geochemical Journal, 9,139-148,1975. Wada, E., M . Minagawa, H . Mizutani, T. Tsuji, R. Imaizumi, and K. Karasawa, Biogeochemical studies on the transport of organic matter along the Otsuchi River watershed, Japan, Estuarine Coastal and Shelf Sciences, 25, 321-336,1987. Walsh, J. J., Nitrogen fixation within a tropical upwelling ecosystem: evidence for a Redfield budget of carbon/nitrogen cycling by the total phytoplankton community, Journal of Geophysical Research - Part C, 101, 20607-20616,1996. Wang, L., and P. Wang, Late Quaternary paleoceanography of the South China Sea: glacial-interglacial contrasts in an enclosed basin, Paleoceanography, 5, 77-90,1990. Wang, L., M . Sarnthein, H . Erlenkeuser, J. Grimalt, P. Grootes, S. Heilig, E. Ivanova, M . Kienast, C. Pelejero, and U. Pflaumann, East Asian monsoon climate during the late Quaternary: High-resolution sediment records from the South China Sea, Marine Geology, 156, 245-284,1999. Wang, P., Response of western Pacific marginal seas to glacial cycles: paleoceanographic and sedimentological features, Marine Geology, 156, 5-39,1999. Wang, P., L. Wang, Y. Bian, and Z. Jian, Late Quaternary paleoceanography of the South China Sea: surface circulation and carbonate cycles, Mar. Geol, 127,145-165,1995. Waser, N . A., K. Yin, Z. Yu, K. Tada, P. J. Harrison, D. H . Turpin, and S. E. Calvert, Nitrogen isotope fractionation duing nitrate, ammonium and urea uptake by marine diatoms and coccolithophores under various conitions of N availability, Marine Ecology Progress Series, 169, 29-41,1998a. Waser, N . A. D., P. J. Harrison, B. Nielsen, S. E. Calvert, and D. H . Turpin, Nitrogen isotope fractionation during the uptake and assimilation of nitrate, nitrite, ammonium and urea by a marine diatom, Limnol. Oceanogr., 43, 215-224, 1998b. Waser, N . A., Z. Yu, K. Yin, B. Nielsen, P. J. Harrison, D. H . Turpin, and S. E. Calvert, Nitrogen isotopic fractionation during a simulated diatom spring bloom: importance of N-starvation in controlling fractionation, Marine Ecology Progress Series, 179, 291-296,1999. Wefer, G., W. H . Berger, J. Bijma, and F. Fischer, Clues to ocean history: A brief overview of proxies, in Use of proxies in paleoceanography: Examples from the South Atlantic, edited by Fischer, G. and G. Wefer, pp. 1-68, Springer-Verlag, 1999. Weiss, R. F., Carbon dioxide in seawater: the solubility of a non-ideal gas, Marine Chemistry, 2, 203-215,1974. Westerhausen, L., J. Poynter, G. Eglington, H . Erlenkeuser, and M . Sarnthein, Marine and terrigenous origin of organic matter in modern sediments of the equatorial East Atlantic: the 613C and molecular record, Deep Sea Research, 40,1087-1121,1993. Westerhausen, L., M . Sarnthein, U . Struck, H . Erlenkeuser, and J. Poynter, p C 0 2 variations of equatorial surface water over the last 330,000 years: the 8 C record of organic carbon, in Carbon cycling in the glacial ocean: constraints on the ocean's role in global change, edited by Zahn, R., T. F. Pedersen, M . A. Kaminski and L. Labeyrie, pp. 367-382, Springer Verlag, 1994. Wiesner, M . G., L. Zheng, H . K. Wong, Y. Wang, and W. Chen, Fluxes of particulate matter in the South China Sea. in Particle flux in the ocean, edited by Ittekot, V., P. Schafer, S. Honjo and P. J. Depetris, pp. 293-312, Wiley & Sons, 1996. Wu, J., The carbon and nitrogen isotopic variations of particulate organic matter in the subarctic northeast Pacific, Ph. D., The University of British Columbia, 1997. Wu, J. P., S. E. Calvert, and C. S. Wong, Nitrogen isotope variations in the subarctic northeast Pacific: relationships to nitrate utilization and trophic structure, Deep Sea Research, 44, 287-314,1997. Wyrtki, K., Scientific results of marine investigations of the South China Sea and the Gulf of Thailand, N A G A report, 1961. 92 4. Ocean-continent linkages in the South China Sea 4.1 On the significance of sea-level variations and shelf palaeo-morphology in governing sedimentation in the southern South China Sea during the last deglaciation 4.1.1 Introduction Late Quaternary climate variations combined with eustatic sea-level changes, especially during the last deglaciation, have significantly influenced the rate and composition of terrigenous sediment supply to shelves, continental slopes and deep-sea basins. This is due to increased incision of rivers during lowered sea level, which leads to sediment bypassing of shelves and high rates of sediment supply to continental slopes, modulated by precipitation/runoff changes, and to lowered sedimentation rates on slopes and the aggradation of over-deepened river valleys and the re-establishment of shelf sedimentation during deglacial sea level rise. The influence of glacio-eustatic sea level fluctuations and climate variations (e.g., precipitation) on the terrigenous sediment supply to shelves, continental slopes and deep sea basins has been demonstrated in the classic paper on shelf sediments in the Gulf of Mexico by Curray [1960], as well as by more recent studies of sediment cores from the South China Sea [SCS; Broecker et al, 1988; Schdnfeld and Kudrass, 1993; Pelejero et al, 1999b], the East China Sea [Ujiie et al., 2001], the Japan Sea [Lee et al, 1996], the Arabian Sea [e.g., the Indus shelf and slope, von Rad and Tahir, 1997; Prins et al, 2000a; the Makran continental slope, Prins et al, 2000b; Prins and Postma, 2000], the Bay of Bengal [Ganges-Brahmaputra delta; Goodbred and Kuehl, 1999, 2000a,b], and the northeastern Brazilian continental margin [Arz et al, 1999]. In addition, the shelf width has been shown to play a significant role in the transfer of terrigenous sediments to the continental slope and the deep sea [Prins and Postma, 2000]. Similarly, Martinez et al [2000] and Bertrand et al [2000] have highlighted the effect of eustatic sea-level changes on sedimentary records of primary production and nitrogen isotope dynamics on the NW African margin, demonstrating the dependence of the sedimentary regime at one coring site on its position relative to the locus of upwelling and production, which, in turn, is significantly affected by its location relative to the coast line and the shelf edge during trans- or regression. In the southern SCS, the exposure of the Sunda Shelf at the last glacial maximum (LGM) led to an increased terrigenous input to the basin of up to five-fold when sea-level was ~120 m lower [Schdnfeld and Kudrass, 1993; Pelejero et al, 1999b]. This was caused by erosional processes on the emerged shelf and by the centralised discharge of 93 fluvial sediment via major river systems, e.g. North Sunda River [= 'Molengraaff River; Molengraaff and Weber, 1920; Fig. 4.1], onto the slopes and into the deep sea. The subsequent rise in sea level during the last deglaciation was paralleled by a marked decrease of sediment supply to the continental margin and the deep sea due to a shift of the depositional centres from the shelf break and slope to proximal shelf areas [Schdnfeld and Kudrass, 1993; Wang, 1999]. Isotopic, sedimentological and organic geochemical data from two sediment cores from the southern SCS revealed significant hydrographical and sediment compositional changes in the southern SCS at ca. 15 ka [Pelejero et ah, 1999b] associated with the deglacial flooding of the shelf during Meltwater Pulse (MWP) la. The objective of this chapter is to investigate in detail the temporal and spatial variations of the terrigenous sediment flux along a transect from the shelf margin down to the lower continental slope of the southern SCS in response to the morphological evolution on the Sunda Shelf and the breakdown of the river systems during the post-glacial flooding. For this purpose, we analyzed sedimentary records from late Pleistocene/ Holocene times (the last ca. 20 ka) at four localities on a transect from the outer Sunda Shelf to the lower continental slope (Fig. 4.1). These localities, which represent the most proximal depocenters during shelf exposure, should be the most sensitive recorders of changes in the sedimentation regime in this area, as well as of environmental changes on the shelf during the deglacial transgression. These new records from the shelf margin and continental slope will be compared to one record from the southern SCS published previously by Pelejero et al. [1999b], and, most importantly, to detailed reconstructions of the palaeo-environmental development of the Sunda Shelf based on a large suite of sediment cores recently published by Hanebuth and Stattegger [2001, in press]. 4.1.2 Regional setting of the Sunda Shelf and the continental slope The Sunda Shelf and the land areas of peninsular Malaysia, Sumatra and the north-western part of Kalimantan form the so-called Sundaland [Tjia, 1980], which has probably been a tectonically stable craton since the mid-Tertiary [Stauffer, 1973; Tjia and Liew, 1996]. The Sunda Shelf is the submarine part of the Sundaland, and averages -70 to -80 m in depth. The research area (Fig. 4.1) is characterized by an average low gradient of ca. 0.03° extending from the coast to the shelf break at about 220 m. The gradient reaches 2-3° at the shelf break and decreases again to values around 0.5-0.2° below 500 m water depth [Wong et al, 2001, in press]. Based on the first bathymetric mapping of the Sunda Shelf, Molengraaff and Weber [1920] suggested that the Sunda plain was drained by 94 Fig . 4.1. M a p of the South China Sea and adjacent seas. The position of the palaeo-coastline during the last glacial sea-level lowstand was not far beyond the 100 m isobath (thick line). The inferred courses of the last glacial Sunda plain river system are shown according to Molengraaff and Weber [1920]. The insert map shows the location of the cores (large dots) and coring transect wi th positions of sites on the central Sunda Shelf (gray line; small dots; Stattegger et al, [1997]). The transect follows the palaeo-valley of the North Sunda river. 95 large rivers during the glacial lowstand. The modern topography on the inner and middle shelf is characterized by over-deepened valleys and depressions mainly related to the former river system. The central submarine valley of the North Sunda River is up to 40 m deeper than the bordering submarine plain, and is 10 to 50 km wide [Haile, 1969; Dash et al, 1972; Tjia, 1980; Stattegger et al, 1997]. The depositional system of the modern Sunda Shelf shallower than 130 m water depth reflects several phases of environmental changes during the past 21 ka, which are causatively related to the deglacial sea-level changes. As indicated by the suite of cores from the central Sunda Shelf (SONNE cruise 115, December 1996-January 1997) studied by Hanebuth et al [2000], Hanebuth and Stattegger [2001, in press], and Stattegger et al. [1997] (Fig. 4.1), long-lasting hiatuses alternating with high sedimentation episodes dominated the sedimentary history during the past tens of thousands of years at all locations. The deposits of the Sunda Shelf core transect (Fig. 4.1) show that the environmental changes occurred in three distinct steps. (1) During the last major regression, shallow water sediments prograded basinward in the form of foresets and relocated the coastline into a NW direction. This unit was covered by a widespread soil horizon which developed over the expanding land surface. Erosion and sediment bypassing were dominant processes during this period of emergence. (2) The deglacial sea-level rise, starting at ca. 19 ka (cal. kyr B.P.), led to a gradual inundation of the Sunda plain. The coastline transgressed the area relatively rapidly due to the low shelf gradient, and only thin transgressive deposits of coastal plain and nearshore environments were formed. The outer part of the course of the North Sunda River generally turned to a northern direction after passing Natuna Island [see Fig. 4.1; Wong et al, 2001, in press] and was flooded at around 17-16 ka during the first phase of submergence as fluvial sediment accumulation shifted to a new river mouth southeast of Natuna Island. This process also coincides with a change from deltaic (assumed) to estuarine conditions. The transgressive deposits mainly occur within the palaeo-valley as a consequence of the difference in elevation between the central valley and the surrounding Sunda plain. The mosaic-like pattern of scattered facies records - due to frequent discontinuities and the highly variably sedimentary environments -characterizes the transgressive sedimentary architecture that shifted landwards through time until (3) the modern conditions were established after 13 ka as a consequence of the diluvial flooding of the main Sunda plain (modern depth of about -70 m). This, in turn, caused the irretrievable loss of sediment supply to the shelf and slope by the dissection of the former river system into several coastally restricted lower-order rivers [Hanebuth and Stattegger, 2001, in press]. The ancient valley system of the North Sunda River 96 represents, therefore, the third major factor determining the stratigraphic development of sedimentation on the slopes and in the deep basin of the SCS. 4.1.3 Core stratigraphy The age model of core 18282-2 is based on 5 AMS- 1 4 C ages (Tab. 4.1). According to these data, core 18282-2 contains a Late Pleistocene (~39 ka) to Holocene (~4 ka) succession, which ranges back to marine isotope stage (MIS) 3, interrupted by a hiatus of more than 30 ka (~36 ka to 4 ka). The age models and the stratigraphic correlation of cores 18284-3, 18287-3 [adopted from Steinke et al, 2001], 18294-4 and 17964-3 [adopted from Pelejero et al., 1999b] are based on AMS- 1 4 C dates as well as on planktonic foraminiferal (G. ruber) oxygen isotope stratigraphy (Fig. 4.2, see also Tab. 4.1). An analogue oxygen isotope stratigraphic age of 9.8 ka has been assigned to the end of Termination lb [Winn et al, 1991], 11.6 ka to the end of the Younger Dryas [Alley et al, 1993; Grootes and Stuiver, 1997], and 14.6 ka to the onset of the Boiling-Transition [Meese et al, 1997; Stuiver and Grootes, 2000; see Tab. 4.1]. The assignment of these "North Atlantic standard ages" to the 8 l s O (Figs. 4.2a, b) and/or SST records of the cores presented here is justified by the fact that recently published, radiocarbon dated SST records from the southern SCS [Chapter 2; M . Kienast et al, 2001a; Steinke et al, 2001], including core 18287-3, reveal close synchroneity between deglacial SST variability and Greenland temperatures during the Boiling-Transition and the Younger Dryas. In the early Holocene section of core 18284-3, a 1 4 C age reversal occurs in the depth interval 130-150 cm (Fig. 4.2a; Tab. 4.1). The planktonic ages at 134 cm (6040 reservoir-corrected 1 4 C yr B.P., equiv. 6910 cal. yr B.P.) and 134-137 cm (6900 reservoir-corrected 1 4 C yr B.P., equiv. 7740 cal. yr B.P.) are younger than the age at 120 cm (7780 reservoir-corrected 1 4 C yr B.P., equiv. 8640 cal. yr B.P.). An X-ray photograph of this core section shows many traces of bioturbation, including vertical burrows of up to 20 cm in depth filled with carbonate sand. Accordingly, an admixture of younger sediment is most likely responsible for the age reversal in the core depth between 134-150 cm. Thus, the ages at 134 cm and 134-137 cm were excluded from the final age model (Fig. 4.2, Tab. 4.1). Additionally, an analogue oxygen isotope age of 9.8 ka [the end of Termination lb; Winn et al, 1991] has been tentatively assigned to the oxygen isotope records of G. ruber and G. sacculifer of core 18284-3 at 137 cm core depth. The Younger Dryas climatic rebound is documented in the isotope records by a &lsO-plateau and highly variable values between 240 cm and 280 cm in the planktonic isotope records of G. sacculifer and G. ruber, and is confirmed by a 14C-date of 10,315 1 4 C yr B.P. at 206-210 cm (equiv. 11,490, 11,800, 11,760 97 (b) 0 5 10 15 20 cal. age ka BP Fig. 4.2. (a) Planktonic (G. ruber) foraminiferal 5 1 8 0 depth profiles of cores 18284-3, 18287-3, 18294-4 and 17964-3 [from Wang et al., 1999]. Additionally, the planktonic foraminiferal 5 l s O depth profile of G. sacculifer of core 18284-3 is shown (light gray line). Al l ages are shown as cal. yr B.P. (compare Table 4.1). AMS- 1 4 C ages are corrected by -400 years reservoir age according to Bard [1988] and converted to calendar years according to Stuiver et al. [1998] and Bard et al. [1998]. Ages in core 18284-3, which are not used in the final age model (see also Table 4.1) are given in parentheses (see Chapter 4.1.3 for further details). Analogue oxygen isotope chronostratigraphy ages of 9.8 ka, 11.6 ka and 14.7 ka were assigned to the end of Termination 1 [End Tib; Winn et al, 1991], end of the Younger Dryas [Alley et al, 1993; Grootes and Stuiver, 1997] and to the onset of the Boiling Transition [Stuiver et al, 1995; Stuiver and Grootes, 2000], respectively (stippled lines), (b) Age-depth relation (in calendar years) of the investigated cores. Arrows point to analogue oxygen isotope chronostratigraphy ages (for references see above). 98 Table 4.1. A M S , 4 C ages and cal. ages of cores 18282-2, 18284-1, 18284-3, 18287-3, 18294-4 Depth AMS- ' 4 C age Error l a foraminifera Calibrated Age l a ranges* Sedimentation rate (cm) (yr B.P.)** species (cal. yr B.P.) (cm/kyr) Core 18282-2 30 4330 +40/-40 G. ruber 4430' 4500-4400 -50 32640 +660/-610 A. pulchella 37440 2 38260-36770 -139-141 31400 +560/-520 A. pulchella 36080 2 36690-35500 -260 31680 +580/-540 A. pulchella 36390 2 37020-35790 -377-379 34020 +780/-710 A. pulchella 38950" 39790-38170 -Core 18284-1/3 10 3645 +30/-30 G. sacculifer 3550' 3580-3480 10.4 50 6870 +40/-40 G. sacculifer 7400' 7420-7350 56.5 120 8180 +50/-40 G. sacculifer 8640' 8810-8580 14.6 (134) (6445) (+40/-40) G. sacculifer (69201) (6980-6860) -(134-137) (7300) (+50/-50) G. sacc. + G. rub. (77401) (7790-7680) -137 analogue - - 9800 3 - -(150) (10450) (+60/-60) G. sacculifer (11360') (11900-10880) 57.2 (206-210) (10715) (+55/-55) G. sacc. + G. rub. (11940, 11800, 11760') (12280-11680) -240 analogue - - 116003 - 22.5 310 analogue • - - 147003 - 120.8 (316-319) (12730) (+70/-70) G.sacc. + G. rub. (14150" (15060-14110) 490 13980 +80/-80 Rot alia sp. 16190' 16440-15960 675.0 760 14330 +80/-80 Rotalia sp. 16590' 16850-16350 -Core 18287-3 10 3665 +25/-25 G. ruber 3570' 3605-3540 27.6 140 7805 +40/-40 G. ruber 8280' 8330-8190 32.6 170 8815 +40/-40 G. ruber 9415, 9200, 9129' 9730-9080 50.0 200 analogue - - 9800 3 47.2 285 analogue - - 116003 _ 62.5 290 10600 +50/-50 G.ruber 11680' 12090-11400 49.8 410 12570 +60/-60 G. ruber 14090' 14290-13720 62.8 510 13540 +80/-80 G. sacc. + G. rub. 15680' 15920-15470 Core 18294-4 42.5 analogue - - 9800 3 _ 5.5 52.5 analogue - - 11600 3 _ 12.9 92.5 analogue - - 147003 - 14.2 122,5 14520 +80/-80 G.ruber 16810' 17070-16570 155.1 221 15070 +80/-80 G. ruber 17445' 17720-17190 20.2 351 20690 + 140/-130 G. ruber 23920' 24080-23770 21 .6 520 27490 +310/-300 G. sacc. + G. rub. 31720' 32070-31380 ' AMS- ' 4 C ages have been calibrated using program CALIB 43 (Stuiver et al. 1998); 'converted using the polynomial algorithm of Bard (1998). 3 Analog oxygen isotope stratigraphy ages of 9,8 ka., 11,6 ka and 14,7 ka have been assigned to the end of Termination lb (Winn et al., 1991), end of the Younger Dryas (Alley et al., 1993; Grootes and Stuiver, 1997) and to the onset of the B0lling-transition (Stuiver et al., 1995; Stuiver and Grootes, 2000), respectively. Ages given in brackets were not used in the final age model of core 18284-3 (compare chapter 4.1). * 1 a enclosing 6 8 3 % of probability distribution (Stuiver et al., 1998). * * measured at the Leibniz-Labor, University of Kiel. 99 cal. yr B.P. with an l a error range of 12,280-11,680 cal. yr B.P.). Due to calibration uncertainties caused by a deglacial radiocarbon "plateau" during the latter part of and after the Younger Dryas [Edwards et al, 1993; Wohlfarth, 1996], an 'analogue' age of 11,600 cal. yr B.P. has been assigned to end of the Younger Dryas at 240 cm [Alley et al, 1993; Stuiver et al, 1995]. Due to a "14C-plateau" near the onset of the Boiling period, the conversion of reservoir-corrected radiocarbon ages to calendar ages is problematic during this period [Stuiver et al, 1998], resulting in large l a errors (Tab. 4.1). Therefore, an 'analogue' age of 14,700 cal. yr B.P. [e.g., Stuiver and Grootes, 2000], at 310 cm was included in the final age model (Tab. 4.1). 4.1.4 Sedimentary proxies used in this analysis The reconstruction of sedimentological changes along the transect from the outer Sunda Shelf to the lower continental slope presented here is largely based on two parameters, the relative contribution of terrigenous organic carbon to the bulk sedimentary organic carbon (C o r g), and the average grain size of the sediments. The carbon isotopic composition (6 1 3C o r g) of marine and terrigenous (C3) organic matter are sufficiently different (ca. -20%o versus ca. -27 %o, e.g., Deines [1980]) to allow a semi-quantitative assessment of the relative proportion of both endmembers in sediment samples from near-shore environments. Further offshore, the use of this proxy is complicated by sorting processes, which can lead to a preferential enrichment of fine particles, for example degraded soil C o r g or terrigenous C o r g from C 4 plants (see further discussion in chapter 4.1.5.2). To circumvent this limitation, downcore variations in the concentrations of n-nonacosane, one of the most abundant odd-numbered long chain n-alkanes in SCS sediments, are presented for one site from the lower continental slope [core 17964-3; data from Pelejero et al, 1999b]. Long chain w-alkanes with a predominance of odd-carbon-numbered members are uniquely produced by terrestrial plants [Eglinton and Hamilton, 1967], and are thus widely used in palaeoceanographic studies to monitor terrigenous organic matter input to marine sediments (e.g., Cacho et al, 2000; Zhao et al, 2000; Calvo et al, 2001; see figure caption 4.9 for further details). The reconstruction of temporal variations in the average grain size of the siliciclastic detritus is based on two proxies, S i / A l ratios and percent clay, which are used together and interchangably depending on data availability. Thus, given the low concentrations (<2 %) of biogenic silica throughout the cores, high S i / A l ratios are caused by high quartz contents relative to aluminosilicate phases, i.e, the dominance of sand-sized particles over clay-sized ones. Similarly, the percent clay (defined as the percent siliciclastics <2 (im of the 100 carbonate- and Corg-free sediment <63nm) is a measure of the relative contribution of clay-sized particles versus silt- and sand-sized grains. Furthermore, the sedimentary concentration of CaC0 3 is interpreted to reflect the relative dilution of biogenous carbonate by siliciclastic material, and is thus used as a proxy for the amount of siliciclastic sediment input to a site. This is justified by previous studies of SCS C a C 0 3 records, which clearly demonstrated that carbonate cycles in SCS above the lysocline (ca. 3000 m) are dominated by dilution [see Wang, 1999 for a recent summary]. 4.1.5 Sedimentary records The outer shelf (core 18282-2,151 m water depth) The sedimentary sequence of core 18282-2 consists of 602 cm medium grey clay, which is unconformably overlain by a light olive grey clayey marine carbonate sand. The age model (see chapter 4.1.3 above and Tab. 4.1) indicates that this core contains a Late Pleistocene to Holocene succession, which ranges back to MIS 3, interrupted by a hiatus of more than 30 kyr between late MIS 3 and the mid-Holocene. The lower clay unit contains <5 % bulk carbonate, whereas the Holocene carbonate sand contains up to 30 % CaC0 3 (not shown). Analysis of the >125-1000 nm-size fraction [Schimanski, 1999] of this unit reveals concentrations of planktonic and benthic foraminifera between 20-60 % and a high content of plant fragments (-fibres; 30-80 %), indicating a high terrigenous sediment supply. In contrast, the Holocene clayey carbonate sand is dominated by planktonic and benthic foraminiferal tests and shell fragments, indicating predominantly marine conditions without any substantial input of terrigenous organic material. This carbonate sand appears as a covering of variable thickness in all cores on the modern outer Sunda Shelf, and the lithological succession of the outer shelf core 18282-2 (i.e., a clayey carbonate sand underlain by a grey clay unit) is, thus, characteristic of all outer shelf cores recovered on the shelf transect between 134 and 166 m water depth [Stattegger et al, 1997]. Thus, given the radiocarbon age determinations of core 18282-2 (see chapter 4.1.3 above and Tab. 4.1) and the core lithology, the outer shelf sedimentary record is characterized by a lack of deposition and /or erosion during the glacial sea-level lowstand and the postglacial sea-level rise. This interpretation is corroborated by high resolution seismic (Parasound) profiles that show prominent prograding clinoform-type reflectors, which are gently dipping seaward and locally cropping out [Stattegger et al, 1997; Wong et al, 2001, in press], reflecting erosion of the outer shelf. 101 The shelf margin (core 18284-3,226 m water depth) Core 18284-3 covers a continuous deglacial succession from ~17 to ~3 ka. (Fig. 4.3). A reservoir corrected AMS- 1 4 C age of 3245 yr BP (3550 calendar age B.P.) at 10 cm in box core GBC 18284-1 (see Fig. 4.2a and Tab. 4.1) suggests that erosion prevented sediment accumulation during the last ~3000 years. According to the age model, sedimentation rates were high during the late glacial period (120-675 cm/ka; Fig. 4.2b). After the onset of the Boiling-Transition, sedimentation rates decreased markedly to a Holocene value of ca. 12 cm/kyr. The temporal variability of S 1 3 C o r g in core 18284-3 (Fig. 4.3a; see also Chapter 3.2) shows values of ca. -26.5 %o in the lower part of the core, indicating a predominantly terrestrial origin of the organic matter from 17 to ca. 8.5 ka. A minor increase in 5 1 3 C o r g (to ca. -25.5 %o) occurs at ca. 16.5 cal. kyr B.P., just after a clay minimum and silt/fine sand maximum noted below. Starting at ca. 8.5 ka, & 1 3 C o r g values increase sharply and reach a maximum of -21.5 %o at the core top, indicating a progressively decreasing proportion of terrigenous organic matter in the upper part of the core (Fig. 4.3a). The bulk CaC0 3 content (Fig. 4.3b) largely parallels the 5 1 3 C o r g record. CaC0 3 concentrations are low (<5 wt.%) until ca. 8.7 ka and gradually increase thereafter (leading the change in 8 1 3 C o r g by a few centuries), reaching maximum values of ca. 22 wt. % at the core top (Fig. 4.3b). Core 18284 reveals an upcore decrease of the siliciclastic clay fraction from the late glacial to the Holocene (Fig. 4.3c,d). The clay content shows maximum values of 30-37 % (Si /Al ratios of 5-6) during most of the late glacial period, except for a minimum content of ca. 7 % (Si/Al up to 10) at 16.5 ka (Fig. 4.3c,d), which marks a silt/sand layer. A decrease in the proportion of fine-grained siliciclastic material occurred at ca. 15 ka, ca. 300-400 years (60 cm depth in core) before the sudden decrease in planktonic 5 1 80 in the same core, and the first rapid sea-level rise at 14.6-14.3 ka [Hanebuth et al, 2000; Fig. 4.3c-f]. After this abrupt decrease, both the clay content and the S i / A l ratios show relatively minor fluctuations (Fig. 4.3c,d). The upper continental slope (core 18287-3,598 m water depth) Core 18287-3 covers a continuous deglacial succession from ~16.5 to ~3 ka, and shows highest sedimentation rates of 60 cm/ka during the late glacial, which decrease to ca. 30 cm/ka during the Holocene (Fig. 4.2a, b, see also Tab. 4.1). The 8 1 3 C o r g record (Fig. 4.4a) shows an abrupt increase from values between -24 %o and -22 %o in the lower part of the core to values of -21 %o to -20.5 %o at around 16 ka, ca. 1500 yr before the sudden decrease in planktonic 5 1 80 and the SST rise, indicating a marked decrease in the supply of terrigenous organic carbon well before significant sea level rise. The bulk CaC0 3 102 G C 18284-3 - shelf margin (water depth 226 m) -21 —| ' ' — ' ' ' ' ' • ' ' ' • ta. Cal. Age (ka) Fig. 4.3. Downcore variations in (a) %> 6 1 3 C o r g , (b) CaC0 3 (wt.%), (c) S i / A l , (d) clay content (% <2 um), and (e) oxygen isotope data (8 l sO) of G. ruber (filled circles) and G. sacculifer (open circles) of core 18284-3. (f) Estimated sea-level curve for the Sunda Shelf region compiled from Geyh et al. [1979], Hesp et al. [1998] and Hanebuth et al. [2000]. Shaded bars indicate the Younger Dryas (YD) and Boiling-Altered (B/A) intervals (ages adopted from Stuiver et al, 1995). 103 G C 18287-3 - continental slope (water depth 598 m) -20 H ' ' ' ' 1 ' — ' 1 ' 1—• ' || i r-L Cal. Age (ka) Fig. 4.4. Downcore variations in (a) % 6 1 3 C o r g / (b) CaC0 3 (wt.%), (c) S i / A l , (d) clay content (% <2 um), (e) U K 3 7 SST (0-30 m water depth), and (f) oxygen isotope data (5 l sO) of G. ruber of core 18287-3. For explanation of sea-level curve and shaded bars see Fie 4.3. 5 " 104 content steadily increases from values of ca. 1-2 wt.% during the late glacial to late Holocene values of around 20 wt.% (Fig. 4.4b). The S i / A l ratios and the clay fraction show relatively stable values of ca. 3.1 and 50 %, respectively, throughout most of the record, with periods of slightly finer-grained sediments (lower S i / A l , higher percent clay) between 16.5 to 14 ka and after 5 ka, respectively (Fig. 4.4c,d). The slightly higher clay concentrations during the interval 10.5-8.5 ka are not paralleled by lower S i / A l ratios, possibly due to the fact that S i / A l ratios are determined on the bulk sediment samples, whereas the percent clay is determined on the sediment sample <63 urn only. Thus, any detrital grains >63 |xm would affect the S i / A l ratio but not the percent clay value. The lower continental slope (core 17964-3,1556 m water depth) In core 17964-3, the sedimentation rate decreases from ca. 56 cm/ka during the L G M to ca. 25 cm/ka during the Holocene [Fig. 4.2; Wang et ai, 1999a]. As indicated by the S 1 3 C o r g , H-nonacosane [Pelejero et al, 1999b], and clay records [Pelejero et al, 1999b], the first major shift in the depositional environment occurred at about ~16 ka, ca. 1,5 kyr prior to the sudden decrease in 5 1 8 0 and the SST rise (Fig. 4.5). Prior to this marked change in the sedimentation regime, the S 1 3 C o r g values, carbonate and clay contents, are fairly constant, ranging between -21.8 and -22.5 %o, 2-4 wt.%, and 74-77 %, respectively. This initial phase of changing sedimentation is followed by an accelerated phase, which starts shortly after the sharp decrease in planktonic S 1 8 0 and the SST rise [Fig. 4.5; Pelejero et al, 1999b]. At this time, a sudden decrease in n-nonacosane content and an increase in S 1 3 C o r g values, as well as decreasing clay and an increase in percent CaC0 3 , indicate a significant decrease in the supply of terrigenous organic and siliciclastic material to the lower slope. Fairly uniform Holocene carbonate and clay values of 15-19 wt.% and 67-70 %, respectively, are established around 9-8 ka, whereas the stable organic carbon isotope ratios and rc-nonacosane abundances reach almost modern values around ca. 13-14 ka (Fig. 4.5). A topographic high on the lower slope (core 18294-4,849 m water depth) In contrast to the cores from the shelf margin and the upper continental slope, the sedimentary sequence of core 18294-4 extends back to mid-MIS 3 (for consistency with the other records only the upper 20 ka are shown and discussed here). The sedimentation rate decreases from about 20 cm/ka during glacial MIS stages 3 and 2 to an average Holocene value of 15 cm/ka with a distinct maximum of ca. 155 cm/ka during the late glacial period (ca. 17.5-16.5 ka), which most likely is an artefact of the age 105 G C 17964-3 - continental slope (water depth 1556 m) -20 Cal. Age (ka) Fig. 4.5. Downcore variations in (a) %0 & C o r g , (b) n-nonacosane concentration [ng/g; from Pelejero et al, 1999b], (c) CaC0 3 (wt.%), (d) clay content [% <2 urn; from Pelejero et al, 1999b], (e) U K ' 3 7 SST [0-30 m water depth; taken from Pelejero et al, 1999a], and (f) oxygen isotope data (5 l sO) of G. ruber [Wang et al, 1999a] of core 17964-3. For explanation of sea-level curve and shaded bars see Fig. 4.3. 106 model (Fig. 4.2; see also Tab. 4.1). Relatively constant 5 1 3C o r g values of -21 to -19.5 %> are recorded throughout the record, with a period of slightly lighter values between 16-18 ka and the heaviest values in the uppermost part (Fig. 4.6a). Given an average 8 1 3 C o r g signal throughout the open SCS of -20 to -19.5 %o during MIS 3 and 2 [M. Kienast et al. 2001b, see also Chapter 3.3], this record is interpreted to reflect a minor contribution (ca. 15 %) of terrigenous organic matter to the bulk sedimentary organic carbon during glacial sea level lowstand, and a decrease in this proportion during and after the deglacial flooding of the shelf. The carbonate content shows relatively constant values of ca. 5 wt.% until ca. 15 ka and increases suddenly thereafter, reaching values between 20-26 wt.% during the Holocene (Fig. 4.6b). The sudden increase in CaC0 3 content is paralleled by an increase in the terrigenous sand content (Si/Al ratio; Fig. 4.6). 4.1.6 Interpretation and discussion 4.1.6.1 The sequence of sedimentation on the shelf margin and continental slope during the last deglaciation According to the sedimentary records described above, the shelf margin and the continental slope of the southern SCS are characterized by several distinct temporal as well as spatial patterns of sedimentation (Tab. 4.2, Figs. 4.3-4.7). In order to evaluate their causes, the observed sedimentological changes beyond the shelf margin will be examined in relation to reconstructions of the environmental development of the inner to central Sunda Shelf, which is directly linked to the deglacial sea-level history [Hanebuth and Stattegger, 2001, in press]. The timing and magnitude of sea-level fluctuations in the following discussion are based on the deglacial sea-level curve, derived from the siliciclastic system on the Sunda Shelf [Hanebuth et al, 2000], and on sea-level reconstructions of Geyh et al [1979] and Hesp et al. [1998] for the Holocene. Given the overall normal age succession in the sediment cores from the shelf margin and the continental slope, it is unlikely that the records are significantly affected by small- and large-scale down-slope mass movement. The absence of any distinctly graded, sandy and/ or finely laminated silt or mud layers further underlines the lack of any significant down-slope mass movement. Thus, the sedimentary records should reliably reflect the chronology of changes in terrigenous sediment supply during the last deglaciation. Stage 1 (ca. 20-16,5 ka): During the last glacial sea-level lowstand and the moderate sea-level rise thereafter, the palaeo-river mouth of the North Sunda River was closest to the shelf edge and the core locations shown in figures 4.1 and 4.7. Accordingly, 107 GC 18294-4 - continental slope (water depth 849 m) -19.5H 1 1 1 1 1 1 1 1 — 1 1 —' 1 1 1 Cal. Age (ka) Fig. 4.6. Downcore variations in (a) % 6 1 3C o r g , (b) CaC0 3 (wt.%), (c) S i / A l , and (d) oxygen isotope data (6180) of G. ruber (M. Weinelt, unpubl. data, 1998) of core 18294-4. For explanation of sea-level curve and shaded bars see Fig. 4.3. 108 Stage 1: 20 to 16.5 ka S u n d a S h e l f S o u t h C h i n a S e a I [mbss] Length of section: 700 km Stage 3 : 14.6 to 14 ka S u n d a S h a l f I S o u t h C h l n a S e [mbss] Length of section: 700 km Fig. 4.7. Intervals of depositional changes on the outer Sunda Shelf, the shelf margin and continental slope in response to the flooding of the shelf, and the interrelation between shelf palaeo-physiography and changes in sea-level (compare Table 4.2, Figs. 4.3-4.6). Sketches illustrating the environmental history on the central Sunda Shelf are taken from Hanebuth et al. (in prep.). 109 Stage 5: 8.5 tO 0 ka S u n d a S h e l f S o u t h C h i n a S e a [mbss] Length of section: 700 km Area of deposition J Area of decreasing sediment supply Area of reworking processes — — Area of sediment bypass R Position o* pelaeo-river mouth | Position of palaeo-coastline 53 Sedimentation rate [cnVkal (+) locally restncted episodically high sedimentation (#) marine sedimentation and reworking Ratio of terrigenous to marine HlillUUIlD-Fig. 4.7 contd. 110 c o Q. u a. o . S P c ^ o • O 00 5 u •a ca 5 °-ca E c O '5b3 i _ 2 c E o 5 c 3 i- -a O t/3 ^ 5 S c S 2 c o 2 =» c u •« JO O , CB (i) - G •a « « E -5 » .s& C/l o § ^ • t o .22 -° '5 H > O 3 Ed § Is 5 E c K 5 o-<= "53 E 6 "1 E cc — ool * T) t 1 I (A -ty-00 ~ •* oo - a rj i / , T s>3s •"8 i 2 * E -= o 2 c 09 o u SI o i -a •* •a t O U « V 11" .5 « a - f H < • 8 § .11 W W5 O C D . O " g " .2 S M -5 3 ° c C w ' 3 u— • — — CB ° 00 g "5. D . .i. J3 .5 u ~ u "2 — t <4- C 2 o o = E 5? 3 O a> c «^ n s E U « c c/3 i E <n c 6 -° .5 :lay : of Si ase in c increase ka) al and :cre: ne, i 4.5 1 1 P ;a: First i-nonacc :iclastic , low 8' 2 = f t 1 0 u rained silic ay content T i l < ve c « 2 * ^ u -t u c t iS S P 0 2 S ^ ~ 1 —> o CM 2 *- E E o •5. w c*j| .2? t r C E o O -a c o u u u — j= c« IS I ' E - i 3 C c I* .2 ION 5 o a. CD •a ca * E i-a o IS 1 O; o I - • tl l « l 0) I Q . a JS CJ C o -o o u cs •£ 'C • U CO 4> JS ca = s ca 2 X -g - "ca i l 0 .t; o 1 r- a. 1 T3 u o 53 E S*8 P cfl CO h 3 -S 1 U O. Ib .2 IS I l l the sediment records are characterized by high accumulation rates of fine-grained siliciclastic material and terrigenous organic matter (Figs. 4.3-4.7). Due to hydrodynamic sorting, clay accumulated in lower proportions on the shelf margin (ca. 35 %), whereas the upper and lower continental slope were characterized by clay contents of around 55 % and 75 %, respectively [Pelejero et al, 1999b; Figs. 4.4, 4.5]. Carbonate concentrations are generally low due to dilution by a high riverine sediment discharge, which is indicated by high linear sedimentation rates of cores 18284-3 and 18287-3 (Fig. 4.2b). During the sea-level lowstand, the North Sunda River discharged its sediment load across the emerged shelf [sediment bypassing; Hanebuth and Stattegger, 2001, in press] and most of the fluvially derived terrigenous material was probably trapped near the coastline, which had moved to the present outer shelf at about -120 m. (see Fig. 4.1). During this time, the area in front of the North Sunda River must have formed a low gradient shallow water area with palaeo-depths between -10 m and -40 m, which was subjected to intensive sediment reworking in inner shelf and surf zone environments [Wright, 1995]. The shallow water deposits as well as the formerly deposited delta foresets (Fig. 4.7) were probably remobilized by interactions of shallow water waves and currents [e.g. longshore currents; Einsele, 1996] as a result of the lowered wave base. Additionally, intensified winter monsoon winds during the last glaciation [e.g., Porter and An, 1995; Xiao et al, 1995; Wang et al, 1999a; see also Chapter 4.2] could have increased the wave heights in the coastal zone, further amplifying sediment reworking processes. Resuspension of sediments in this turbulent environment led to long distance transportation [Wright, 1995]. Additionally, a higher proportion of fine-grained fluvial material may have directly bypassed the outer shelf by transport in suspension, e.g. in buoyant sediment-laden river plumes [Wright, 1995], and settled on the continental slope [Schdnfeld and Kudrass, 1993]. These processes probably delivered fine-grained fluvial sediments together with reworked terrigenous material to depositional centers on the continental slope and margin, and most likely prevented the development of clear delta clinoforms on the ancient inner shelf, as would be expected following the concept of Einsele [1996]. This sediment-bypassing is most likely the cause of the absence of deposits on the modern outer shelf during the L G M . As a consequence of these various modes of sediment transport, most of the river-supplied material was deposited on the shelf margin and slope. Stage 2 (16.5-ca. 14.5 ka): The initial change in the depositional environment of the shelf margin and the upper continental slope occurred at ca. 16.5 to 16 ka, as reflected by an increase in 5 1 3 C o r g and a decrease in clay content in cores from these locations (Tab. 4.2; 112 Figs. 4.3, 4.4, 4.7). This could be associated with the flooding of the lowermost part of the North Sunda River between 19 to 15.5 ka [Hanebuth and Stattegger, 2001, in press]. Due to the initial sea-level rise and the first retreat of the river mouth, most of the sediment was probably trapped in front of the river mouth, as indicated by thick delta-front sediments on the shelf around -110 m water depth [Hanebuth and Stattegger, 2001, in press]. The newly created accommodation space on the shelf probably caused a reduction in the sediment supply to the upper continental slope (Fig. 4.4), whereas the shelf margin is characterized by the deposition of coarser material (lower percent clay, increased S i / A l , Fig. 4.3) around 16.5-16 ka, probably due to sediment reworking along the margin during the initial rise of sea level. This initial change in the depositional environment on the shelf margin and upper slope is in phase with the first increases in carbonate content and 5 1 3 C o r g , and decreases in clay content and n-nonacosane concentrations observed at the lower continental slope site [Pelejero et al., 1999b; Fig. 4.5]. In contrast to the slope sites, the next shift in the depositional environment, as inferred from the decrease in clay content, occurred on the shelf margin around 15 ka, ca. 300-400 years before the sudden decrease in planktonic 8 l s O and the rapid rise in sea-level (Fig. 4.3). This sudden decrease in the clay content probably indicates a change from a more fine-grained riverine sediment supply to an increasing input of reworked, coarser material from the former coastal area to the shelf margin around 15 ka (Figs. 4.3, 4.7) caused by the progressive landward displacement of the coastline as indicated by the retreat of the N Sunda River depocenter into the middle river course between 15.5-14.6 ka [Hanebuth and Stattegger, 2001, in press]. However, it seems that the retrogression of the river mouth is not sufficient to significantly reduce and/or completely cut off the upper and lower continental slope sites from any riverine supplied fine-grained material as indicated by the unchanged depositional environments at these sites. Alternatively, the unchanged sedimentology at the continental slope sites could be attributed to a higher contribution of reworked and winnowed fine-grained material, replacing the fine-grained riverine sediment supply. Stage 3 (14.5-14 ka): During the rapid sea-level rise (-96 m to -80 m) between 14.6 and 14.3 ka (=meltwater pulse (MWP) la), the continental slope received a more rapidly decreasing supply of fine-grained siliciclastic material (lower clay content, higher S i / A l ratios), a further decrease in the proportion of terrigenous organic matter (higher 5 1 3C o r g , lower n-alkanes) and an increase in carbonate content (Figs. 4.4, 4.5). During this short period, the middle part of the Sunda plain river system was rapidly flooded, and short-lived episodes of channel incision and subsequent filling on the coastal plain occurred 113 [Hanebuth and Stattegger, 2001, in press]. The onset of this most prominent change in the depositional environment is in phase with the end of the sudden decrease in 5 1 8 O G M B E R and the SST rise at the begining of the Boiling/Allerod period. Two non-exclusive hypotheses are offered to explain this centennial-scale lag of sedimentary parameters (% CaC0 3 , 5 1 3 C o r g and S i / A l ratio in core 18287-3, and n-nonacosane in core 17964-3; Figs. 4.4 and 4.5) versus planktonic 8 1 8O G . m b e r and UK 3 7-SSTs observed in both records from the continental slope in the southern SCS. The lag could be due to the delayed/slower melting of the Northern Hemisphere ice sheets and the associated global sea level rise compared to the extremly rapid (<40 years; Alley et al., 1993) warming during the Boiling Transition. If this suggestions is valid, the records from the southern SCS provide an independent means to tie the timing of MWP l a to the warming during the Boiling-Transition, corroborating the absolute age of MWP la established on the Sunda Shelf by Hanebuth et al. [2000; see also Chapter 4.3]. Alternatively, this subtle lag could be due to a threshold in the Sunda Shelf palaeo-morphology/sea level rise, which leads to sedimentological changes only after an initial increase in sea level. As a consequence of the landward displacement of the river mouth and the early dissection of the riverine system by flooding of the middle course of the river, most of the terrigenous sediment was trapped on the central shelf. Only a minor proportion of fluvially-transported terrigenous material reached the upper continental slope after this phase. Additionally, the extensive mangrove swamps established during this phase acted as an efficient trap for terrigenous sediments [Hanebuth et al, 2000], further amplifying the decrease of terrigenous sediment supply onto the continental slope. Whereas the continental slope is characterized by a decrease in terrigenous sediment supply, the sedimentary record from the shelf margin does not display any significant changes associated with this phase of accelerated sea-level rise, as indicated by a constant supply of terrigenous material, e.g. unchanged low 5 1 3 C o r g values, S i / A l ratios and percent clay (Figs. 4.3, 4.7). This is interpreted to reflect a persistent higher supply of remobilized coarser nearshore deposits from the palaeo-coastline, occurring simultaneously with channel incision. Starting at 14.5 ka, the depositional environment of the shelf margin and upper continental slope is characterized by a change from a more riverine-dominated fine-grained sediment supply to an increasing input of reworked material from the former coastal area (Fig. 4.4, 4.5). Stage 4 (14-8.5 ka): During the following phase of decelerated sea-level rise (14.3-12.5 ka), the plain around the North Sunda River valley was flooded, which caused the disappearance of the riverine system on the central Sunda Shelf after ca. 13.5 ka 114 [Hanebuth and Stattegger, 2001, in press]. This resulted in a gradual decrease of terrigenous material supply onto the slope (Figs. 4.4-4.7), whereas the depositional environment remained unchanged on the shelf margin. The first shift in the sedimentary records of core 18294-4, as indicated by a marked increase in carbonate content and S i / A l ratios at ca. 13.5-13 ka (Fig. 4.6), coincides with the final breakdown of the riverine system on the modern central Sunda Shelf at ca. 13.5 ka [Hanebuth and Stattegger, 2001, in press]. During this period, when sea level reached -60 to -70 m, the central part of the Sunda Shelf was opened to processes of intensive reworking during the transgression of the coastline landwards and the subsequent onset of 'inner shelf conditions. This is reflected by a continuous supply of terrigenous organic matter (5 1 3C o r g) and siliciclastic material to the shelf margin (Figs. 4.3, 4.7). It is also evident from the sediment cores of the central shelf, which are characterized by starvation due to reworking since ca. 13.5 ka [Hanebuth and Stattegger, 2001, in press]. Whereas the cores from the upper and lower continental slope display an almost uniform depositional history, the sedimentary records from the topographic high (site 18294) exhibit a different stratigraphic development. Due to its elevation above the surrounding sea floor, this site was probably unaffected by any near-bottom transport on the slope. Thus, this site most likely received primarily river plume-derived suspended fine-grained material (Figs. 4.6, 4.7), as reflected in the relatively low sedimentation rates (Tab 4.1, Figs. 4.2b, 4.7). In contrast, the other slope sites were probably subjected to near-bottom sediment transport. As shown by Plug and Palanques [1998], fluvial supply of suspended sediment or resuspended material is transferred from the shelf to the slope, and from upper to lower parts of the slope, mainly through a bottom nepheloid layer. This could partly explain why the sites of the upper and lower continental slope show higher sedimentation rates and higher proportions of river-supplied terrigenous organic matter as compared to the site on the topographic high (Fig. 4.7). Stage 5 (8.5-?3 ka): Predominantly marine conditions similar to the present-day conditions were established around ca. 9 ka on the central parts of the shelf [Hanebuth and Stattegger, 2001, in press]. The final change in the depositional environment of the shelf margin and upper continental slope, however, occurred later, at ca. 8.5 and 6 ka, as reflected in decreasing sedimentation rates, a large increase in o 1 3 C o r g and carbonate content on the shelf margin (Figs. 4.2a, 4.3, 4.7) and the establishment of almost modern sedimentary conditions on the slope (Tab. 4.2; Fig. 4.7). However, this apparent delay could also be a reflection of sediment reworking and redeposition. This final change in 115 the depositional environment of the shelf margin and the continental slope probably marks the end of any significant supply of fluvial and /or reworked material to the coring sites during the final transgression of the shelf, when the coastline reached its modern position around ca. 8-7 ka [late Holocene sea-level highstand; Geyh et al, 1979; Hesp et al, 1998]. Thereafter, foraminiferal mud accumulated on the shelf, which unconformably covers pre-Holocene deposits in varying thicknesses on the modern shelf (see core 18282-2). This indicates that the supply of fluvial sediment to the middle and outer shelf was reduced to a minimum and/or completely cut off. Presently, significant fluvial-terrestrial supply onto the middle and the outer shelf does not occur due to the width of the Sunda Shelf. Most of the river-delivered sediments are trapped in the coastal and nearshore-zone and/ or are deposited on the innermost part of the broadened shelf, preventing substantial terrigenous sediment transport to the distal parts of the shelf. 4.1.6.2 Climatic controls on the depositional environment? As the previous chapter has shown, the depositional environments of the outer shelf, the shelf margin, and the continental slope appear to be governed by the interaction between shelf palaeo-physiography and variations in sea level. Thus, major depositional changes coincide with environmental trends in the coastal zone, e.g. phases of retreat and disappearance of the North Sunda River. However, sediment supply not only depends on the morphology of the river basin and sea level, but also on the regional climate (e.g. precipitation), which together affect the hydrology of rivers, as shown, for example, by Mulder and Syvitski [1996]. Prins and Postma [2000] also demonstrated that the depositional environment on the continental slope of the Arabian Sea appears to have been controlled by sea-level fluctuations as well as by climatic variations. In order to unravel the effects of sea level and shelf palaeo-morphology from any potential climatic controls, depositional changes of the shelf margin and continental slope summarized above will be discussed and contrasted with other deglacial marine and terrestrial records from SE Asia. Considerable debate exists on whether the SE Asian region was significantly drier or whether humid conditions prevailed during the last glaciation. General circulation models predict generally drier conditions in southern Asia during the L G M [Prell and Kutzbach, 1987], corroborated by palaeo-ecological studies of Van Campo et al [1993], Flenley [1998], and, more recently, by van der Kaars et al. [2000] from the Banda Sea. According to van der Kaars et al [2000], the altered land-sea configuration had a significant influence on precipitation in the region during the last glaciation. When the 116 Sunda Shelf was fully uncovered, moisture availability was strongly reduced, "regardless of whether monsoonal circulation or intensity changed" [van der Kaars et al, 2000]. In contrast, Crowley [1995] and Petit-Maire [1999] postulate the existence of tropical rainforest for the 'Sundaland' region, a conclusion supported by atmospheric global circulation model (AGCM) simulations [Prentice and Fung, 1990]. Recent pollen data from core 17964-3 corroborate these findings [Li and Sun, 1999; Sun et al, 2000]. According to these studies, the climate was periodically cooler than today, but no evidence of dry conditions could be found. According to Sun et al. [2000], tropical lowland rainforests occupied the glacial Sundaland, which indicates that conditions were similar to those of the modern humid climate. In accordance with this interpretation, model simulations by Bush and Fairbanks (2001) indicate a concentration of atmospheric convergence over the exposed Sunda Shelf during the last glacial sea level lowstand, most likely supporting an increase in precipitation. In contrast to these earlier studies, however, Penny (2001) infers cooler and probably drier glacial conditions in NW Thailand based on the palynological study of a downcore record from a small peat-swamp. Bulk o 1 3 C o r g values in the sedimentary records from the margin and shelf are not sufficiently distinctive to allow identification of the presence of C 3 versus C 4 assimilating plants, which would point of the importance of rainforest versus grasslands, respectively, in the North Sunda River catchment area. Thus, average bulk 6 1 3 C o r g values of -26.5 %o (core 18284-3, Fig. 4.3) could indicate a predominantly C 3 source of organic matter [Tyson, 1995]. However, the estimation of the fraction of C 4 organic matter from bulk 6 1 3 C o r g values are often too low due to hydrodynamic sorting of the different source materials during transportation, as Goni et al. [1997] have shown for a sediment transect in the Gulf of Mexico. Indeed, o 1 3 C o r g analyses of several grain-size fractions of core 18284-3 reveal that the finest fraction (<1 jum; not shown) is significantly isotopically heavier than the coarser fractions, potentially indicating a signal of mixed C 3 versus C 4 organic matter. Alternatively, the heavy isotope values of the finest fraction may indicate the presence of marine organic matter or heavily degraded soil organic matter (Bird and Pousai, 1997). However, bulk 6 1 3 C o r g values of -15.7 %o to -25.7 %> were found in marshy soils on the shelf (mostly dated at ca. 20 1 4 C kyrs, Hanebuth 2000), probably reflecting a mixture of C 3 and C 4 plants without any substantial marine influence, indicating a grass and sedge covered coastal wetland [Hanebuth and Stattegger, 2001, in press]. On the basis of this conflicting evidence, the vegetational history as well as the climatic conditions (dry vs. humid) of glacial SE Asia remains controversial. As inferred 117 from a site in the northern SCS [site 17940; see Fig. 4.1 for core location; Wang et al, 1999a; see also Chapter 4.2], the monsoonal climate was characterized by a series of abrupt changes in summer and/or winter monsoon intensities during the last deglaciation, involving marked changes in monsoon precipitation. Additionally, palaeo-ecological studies of Sun et al. [2000] reveal that the deglacial climate in the southern SCS was comparable to that of the L G M , although the pollen assemblages show rapid fluctuations, probably indicating unstable climatic conditions. Alternatively, these frequent alternations could have been caused by hydrodynamic sorting effects of pollen, masking potential climate variations. Zhou et al. [1999] report that marked climatic variations occurred in eastern Asia during the Younger Dryas (YD) interval. Further indications of unstable climate conditions during the YD, with rapid fluctuations between dry and wet periods, are inferred from the Chinese Midivan swamp-peat deposits [Zhou et al, 1999]. UK' 3 7-SST records reveal that the YD in the tropical southern SCS is associated with a minor, albeit clearly detectable, cooling of ca. 0.5 °C with respect to the Boiling/Allerod period [Pelejero et al, 1999a; Steinke et al. 2001; see also Chapter 2.4]. However, a temperature decrease of ca. 0.5 °C cannot fully explain the increase of planktonic foraminiferal 5 1 8 0 during this interval. Tentative estimations of A5 l s O residuals and local sea surface salinities [Steinke, 2001] suggest that the YD is also associated with a change in surface ocean 6 l sO towards heavier values and, by inference, with higher salinities, possibly due to more arid conditions. As described above, the first change in the depositional environment on the shelf margin and continental slope occurred between ca. 16.5 to 16 ka (Figs. 4.3, 4.4, 4.6), preceding the rapid and abrupt decrease in planktonic foraminiferal 5 1 80, as well as the sharp rise of sea-level at 14.6-14.3 ka by ca. ~1500-2000 years. This initial change in the depositional environment may have been associated with the loss of the lowermost river course of the North Sunda River during the moderate phase of sea-level rise between 19-15.5 ka [Hanebuth and Stattegger, 2001, in press]. On the other hand, this initial change appears to be more or less in phase with the sharp increase in 6 l s O of the stalagmite record from Hulu Cave, which presumably reflects an abrupt decrease in summer monsoonal intensity [Wang et al, 2001]. Thus, the decreased terrigenous sediment supply to the continental slope sites (Figs. 4.4, 4.5, 4.7) at ca. 16.5-16 ka could also have been associated with a reduction in continental river run-off due to weakened summer monsoonal precipitation during this time. According to Wang et al. [2001], however, summer monsoonal intensity increased gradually immediately following this maximum low as well as more rapidly at ca. 14.6 ka and ca. 11.6 ka. Neither of the sedimentary records presented above, however, shows any significant increase in riverine supply 118 (increasing percent clay, decreasing S i / A l ; Figs. 4.4, 4.5) during these times of intensifying summer monsoonal precipitation, clearly suggesting that shelf morphology and sea-level rather than changes in monsoonal climate are the dominant controls on sedimentation off the Sunda Shelf. The final change in the depositional environment of the shelf margin and the upper continental slope occurred between ca. 8.5 and 6 ka, as indicated, for example, by decreasing sedimentation rates, which may have been associated with the final flooding of the shelf. Alternatively, the sedimentological changes could potentially be linked to a decreased riverine discharge due to drier conditions. However, many studies of the SE Asian palaeo-monsoon [e.g., Wang et al, 1999; see also Chapter 4.2] indicate an early Holocene summer monsoon optimum, which is associated with increased precipitation and river runoff. Consequently, decreased river runoff is not likely to be the cause of the sedimentological changes observed on the Sunda Shelf margin and the upper continental slope at ca. 8.5 to 6 ka. 4.1.7 Summary The deglacial sedimentary history of the outer Sunda Shelf (ca. 150 m water depth), the shelf margin (ca. 220 m water depth) and the continental slope (600-1550 m water depth) is closely related to distinct thresholds in the flooding of the emerged shelf and interrelated effects of shelf palaeo-morphology and sea level. Based on five down-core records of sedimentological and geochemical variability along a transect across the Sunda Shelf covering the last 20,000 years, five distinct phases of sediment depositional changes can be reconstructed. 1) (ca. 20-16,5 ka) During sea-level lowstand and the moderate sea-level rise thereafter, the sedimentary records on the shelf margin and slope are characterized by high accumulation rates of fine-grained siliciclastic material and terrigenous organic matter, due to shelf erosion and fluvial runoff, whereas the modern outer shelf was completely reworked due to wave and current actions. 2) (16.5-ca. 14.5) The first change in the depositional environment is indicated by an initial decrease in clay content and terrigenous organic matter concentrations as well as an increase in carbonate content at 16.5 to 16 ka, when the lower river course of the North Sunda Paver was flooded. Due to the initial retreat of the river mouth, most of the riverine sediment was probably trapped in front of the river mouth, as indicated by thick delta-front sediments on the central shelf. The opening accommodation space on 119 the shelf led to a diminuition of the sediment supply to the shelf margin and the continental slope. 3) (14.5-14 ka) The pattern of sedimentation on the continental slope is characterized by an accelerated decrease of terrigenous sediment supply associated with the flooding of the middle part of the palaeo-river valley during the rapid rise of sea-level. In contrast, the sedimentary records from the shelf margin do not display any significant changes associated with this phase of accelerated sea-level rise, as indicated by an unchanged supply of terrigenous siliciclastic and organic material. 4) (14-8.5 ka) A gradual decrease of terrigenous supply to the slope occurred during the following decelerated sea-level rise, when the surrounding plains of the North Sunda River valley were flooded. The shelf margin is still characterized' by a continuous terrigenous supply during this period. 5) (8.5-?3 ka) The final change in the depositional environment occurred between ca. 8.5 and 6 ka, when the coastline reached its modern position and modern sedimentary conditions were established. This final change in the depositional environment probably marks the end of reworking during the final phase of transgression of the Sunda Shelf. Given the concordant temporal development of depositional changes on the southern SCS continental margin and slope and the environmental transformations on the central Sunda Shelf, sedimentation on the shelf margin and continental slope appears to have been mainly controlled by the interrelationship between sea-level and shelf palaeo-morphology. As shown above, the sedimentary records display distinct temporal and spatial sediment compositional changes that are directly associated with different stages of the postglacial transgression of the Sunda Shelf, potentially masking any variability in terrigenous sediment supply due to variable precipitation, vegetation and/or atmospheric circulation. This study cautions against the use of near-shelf sedimentary records as unbiased recorders of local climate change, and highlights the need to include the assessment of shelf morphology/sea-level variations in future studies. Any future unravelling of SE Asian monsoon variability from SCS sediments wil l critically depend on a clear distinction between sea level and climatic influences on changes in sediment composition. 120 4.2 Reconstructing palaeo-monsoon variability: Downcore variations in the major element composition at site 17940 in the northern SCS 4.2.1 Introduction Based on high-resolution records of grain size variability at site 17940-2 in the northern SCS (Fig. 4.8), Wang and Sarnthein [1999] and Wang et al. [1999a,b] argue for a stepwise increase of summer monsoonal intensity accompanied by a decrease in winter monsoonal winds from the last glacial period to the Holocene. Thus, the clay content, defined as the percentage of the acid insoluble grain-size fraction smaller than 6.3 \im (note that this definition of clay also includes the fine silt fraction 2-6.3 \im, and is different from the definition in Chapter 4.1; for consistency with the cited publications, however, it is adopted here), was interpreted as a reflection of runoff from the Pearl River, and was used as a proxy of summer monsoonal precipitation. The silt median grain size (Qm) on the other hand, defined as the grain size at the 50 th percentile of the mass accumulation of the silt fraction (6.3-63 nm), was used as a proxy for winter monsoonal winds, which deliver loess from interior China to this site. The interpretation of the silt grain size in terms of wind intensity was substantiated by a recent proxy evaluation and calibration study from the Arabian Sea [Clemens, 1998]. Similarly, the silt median grain size has been used in previous studies of changes in winter monsoonal wind intensity as recorded in loess-palaeosol sequences from the Chinese loess plateau [e.g., Porter and An, 1995; Xiao et al, 1995]. Superimposed on the general deglacial change from a regional dominance of the winter monsoon regime during the L G M to a summer monsoon dominated Holocene, Wang et al. [1999b] found several high frequency variations, which they interpreted as a reflection of millennial-scale flickerings in monsoon climate linked to the climate of the North Atlantic realm. The major element composition of marine sediments has been shown to be a useful tracer of variations in sediment provenance and transport mechanism(s) due to climate change. For example, Reichart et al. [1997] and Sirocko et al. [2000] used downcore records of T i / A l and M g / A l in the Arabian Sea to infer both changes in the intensity and in the prevalence of winter and/or summer monsoonal winds. Similarly, Calvert and Fontugne [2001] identified key differences in the inorganic chemical composition of the sediments deposited during times of sapropel formation compared to the intercalated marls in the Mediterranean. Accordingly, they argue for fundamentally different climatic conditions during times of deposition of these two contrasting sedimentary fades. 121 Fig. 4.8. Down core records of silt median (black line) and total mode (blue line) diameters (both in urn), percent clay (red line; all grain size data from Wang et al, [1999a,b]), and alkenone sea surface temperature (SST) estimates (blue line; from Pelejero et al, [1999a]) of core 17940-2 versus depth. The thin horizontal line (at 6.3 urn) delineates clay (see Chapter 4.2.1 for definition) from silt-sized total mode diameters. Thin vertical lines mark the onset of the Boiling warming, the end of the Boiling/ Allerod warm period and the end of the Younger Dryas, respectively, as defined by the SST estimates. 122 The goal of this chapter is to assess changes in the major element composition at site 17940 as indices of deglacial variations of monsoonal climate. These new records will be compared with the reconstructions based on grain size parameters from the same site summarized above. To account for variable dilution of the major element concentrations by carbonate, all element concentrations are normalized to A l . This approach of using A l as a conservative tracer is widely used in inorganic geochemical studies, and is justified by the similar concentration of A l in a wide variety of source rocks, the absence of any redox chemistry, as well as its low mobility during weathering. 4.2.2 Results and interpretation The downcore distribution of S i / A l (see Appendix II E for analytical details) shows distinctly different values between the glacial section and the Holocene, with a clear two-step change during the deglaciation, interrupted by a brief plateau (Fig. 4.9). This deglacial pattern is markedly different from the distribution of other major elements discussed below. Biogenic opal concentrations [Lin et al, 1999] are low (ca. 2 to <4 %) thoughout the core, without any clear glacial-interglacial contrast, and are thus not affecting this variability. High sedimentary S i / A l is generally associated with high quartz contents relative to aluminosilicate phases. Moreover, loess typically has a high silica content but is depleted in potassium [Taylor et al, 1983]. Based on the lack of coherence between the large-scale variability of S i / A l and K / A l , which would be expected if the relative proportion of K-feldspar overwhelmingly governed the S i / A l variations, S i / A l is interpreted to reflect primarily the relative contribution of quartz to the sediment. Thus, higher S i / A l during the glacial is probably indicative of higher loess input at site 17940, and, by inference, of intensified winter monsoonal winds. By analogy, the Holocene section is characterized by a weaker winter monsoon. Overall, the Si / A l record and its interpretation is consistent with, and independently corroborates, earlier reconstructions of the deglacial evolution of winter monsoonal winds in the northern SCS based on the downcore variations of the median grain size of silt by Wang and Sarnthein [1999] and Wang et al. [1999a] (see section 4.2.1 as well as further discussion in 4.2.3). Superimposed on this general deglacial trend in S i / A l , however, there are several millennial-scale events, which are also evident in the downcore records of K / A l , T i / A l , M g / A l and Fe /Al (Fig. 4.9). The coherent variations in S i / A l , K / A l and T i / A l at this scale are most likely caused by changes in sediment texture, whereby higher element/Al ratios are indicative of coarser grained sediments. Thus, higher S i / A l is 123 0 200 400 600 800 1000 cm core depth Fig. 4.9. Down core distributions of S i / A l , T i / A l , K / A l , Fe /Al and M g / A l of core 17940-2 versus depth. Due to the high-frequency variability of T i / A l , K / A l , Fe /Al and M g / A l , the 5-point running average is included to better visualize the longterm trends. The vertical lines are adopted from Fig. 4.8 and mark the onset of the Boiling warming, the end of the Boiling/ Allerod warm period and the end of the Younger Dryas, respectively. Selected AMS 1 4 C dates on the x-axis (uncorrected, uncalibrated) are taken from Wang et al. [1999a] (see text for further discussion). 124 associated with an increased proportion of quartz versus clay minerals, whereas higher K / A l suggests more K-feldspar compared to non K-bearing clay minerals. Similarly, Ti resides in the coarser-grained sediment fraction, reflecting the presence of silt-size heavy minerals, such as rutile and ilmenite. Both K / A l and T i / A l show higher and more variable values in the lower part of the record and lower, less variable values in the uppermost part. The long-term pattern of S i / A l discussed above could thus be interpreted as a superposition of loess input on a generally fining upward sequence, which is interrupted by several short-term events of coarser-grained sediment deposition, for example between 450 cm and 620 cm. The largely coherent downcore distributions of Fe/ A l and M g / A l ratios, at least above ~820 cm, differ from the other records of major element variability in that they do not show any significant glacial-interglacial contrast (Fig. 4.9). Nevertheless, the brief maxima in element/Al ratios, indicative of coarser grained sediment deposition (see discussion above) are also evident in the Fe /Al and M g / A l records. F e / A l and M g / A l in SCS sediments are probably a reflection of the distribution of ferromagnesian minerals [Calvert et al., 1993]. Thus, the coarser-grained layers are also enriched in ferromagnesian minerals, whereas the comparatively lower F e / A l and M g / A l ratios in the lower part of the core are caused by the marked depletion of Fe and particularly Mg in loess [Taylor et al, 1983], again evidencing the superposition of a deglacial decrease in loess input on a decrease in average grain size. This interpretation of Fe/ A l and M g / A l as a reflection of ferromagnesian mineral distribution is further corroborated by the lack of correlation between the downcore variations in total sulphur (not shown) and Fe /Al , which would indicate an association of Fe/ A l with pyrite. The interpretation of the downcore variations in major element composition in terms of sediment textural changes is corroborated by a detailed study of SCS surface sediments [Calvert et al, 1993], and is generally consistent with the downcore distribution of siliciclastic grain sizes. Thus, the deglacial increase in the bulk clay content [Fig. 4.8; Wang and Sarnthein, 1999; Wang et al, 1999a,b] is mirrored by an overall decrease in S i / A l , T i / A l and K / A l . However, the marked maxima of element/Al values in the middle section of the core, which are interpreted as an indication of coarser-grained sediments, are not reflected in the percent clay record; the coarser-grained intervals actually coincide with the percent clay maximum between 450 and 620 cm. This discrepancy is probably due to the fact that the percentage of clay reported by Wang and Sarnthein [1999] and Wang et al. [1999a,b] is determined on the fraction smaller than 63 jim of the bulk sediment only. Thus, any detrital material coarser than 63 jam would not affect their percent clay values. These coarser grained layers are tentatively 125 interpreted as turbidite deposits, although they are not reflected in the magnetic susceptibility record of the core, and were not noted in the detailed core description [Sarnthein et al., 1994]. Alternatively, they could indicate winnowing events. This interpretation, however, is unlikely due to the short-lived nature of these deposits. Furthermore, these distinctly coarser grained layers are most likely not volcanic ash layers, given the coherent increases in S i / A l , T i / A l and M g / A l , which contrast with the distinct depletion of Mg and Ti in known ash layers from the SCS [Buhring et al, 2000; Song et al, 2000; Wiesner et al, 1995]. 4.2.3 Discussion Overall, the downcore records of element/Al ratios suggest a distinctly different inorganic geochemical composition of sediments at site 17940 during glacial and interglacial times. Furthermore, they display a marked deglacial decrease in the amplitude of sample-to-sample variability (Fig. 4.9). This high frequency variability is interpreted primarily as a reflection of 'random' sediment inhomogeneity, rather than as a reflection of short-term climate change. The more pronounced glacial sediment inhomogeneity could be caused by a variety of processes, some of which, for example downslope gravity mass movements and /or sediment reworking on the exposed shelf, are probably less important during the Holocene sea level highstand. This assertion is further corroborated by the fact that the deposition of the distinctly coarser grained layers (see 4.2.2) is similarly limited to the pre-Holocene part of the record. The interpretation of the glacial-interglacial contrast in sediment composition, particularly the fining-upward trend, in terms of deglacial palaeoclimatic/-oceanographic changes is complicated by the potentially complementary and/or antagonistic effects of decreasing loess input, increasing river mouth distance due to the deglacial rise in sea level, variations in runoff due to'summer monsoonally-induced precipitation, and the closed sum problem. Furthermore, the assignment of absolute ages and the detailed comparison to other records of palaeo-monsoon variations [e.g., Overpeck et al, 1996; Sirocko et al, 1996] is limited by unknown, and likely variable, reservoir ages at site 17940 (see Fig. 2.7, as well as further discussion in Chapter 4.3, particularly figure caption 4.10). The following discussion of the temporal evolution of inorganic sediment geochemistry and inferred monsoonal climate variability is thus limited to establishing the phasing with respect to major climate events, which are readily identifiable in the SST record from the same core. Furthermore, records are 126 displayed versus depth only (the AMS 1 4 C dates shown on the lower j-axis of Fig. 4.9 are intended to provide a crude stratigraphic orientation only). Accepting the unequivocal decrease in silt median diameters [Fig. 4.8, Wang and Sarnthein, 1999; Wang et al, 1999a], as well as the decrease in S i / A l as an indication of decreasing transporting wind velocity, both parameters suggest a two-step deglacial reduction in winter monsoon intensity. Thus, the first decrease leads the Boiling warming by several kyrs, whereas the second step occurs during the Younger Dryas (YD); the Boiling/Allerod (B/A) warm period, in turn, is accompanied by a plateau in S i / A l values (Fig. 4.9). This pattern and timing of the glacial-interglacial transition of S i / A l and the silt median diameter is reminiscent of the temporal changes in deglacial SST in the low-latitude Pacific [Lea et al, 2000; Koutavas et al, 2002], possibly indicating a mechanistic linkage of longterm changes in the intensity of winter monsoonal winds and low-latitude Pacific SSTs. Even though this mechanism is not yet understood, it seems plausible given the significant influence of Pacific Ocean SSTs on interannual variations of monsoonal climate observed today [Webster et al, 1998]. Assuming that intensified winter monsoonal winds are not only associated with the transport of coarser material but also with higher amounts of transported material, some of the deglacial decrease in average grain size derived from the major element geochemistry (see Chapter 4.2.2), as well as from the increase in percent clay [Fig. 4.8, Wang and Sarnthein, 1999; Wang et al, 1999a] can be interpreted as a result of the deglacial decrease in winter monsoonal winds discussed above. This assertion is corroborated by the downcore variations of the total modal particle diameter. Thus, sediments in the glacial section of the core are characterized by high abundances of particles having diameters between 10-30 \im (Fig. 4.8), a grain size typical of loess deposits in the Chinese Loess Plateau [e.g., Xiao et al, 1995] whereas the Holocene samples are dominated by clay-sized particles. The transition between high glacial and low post-glacial total mode diameters largely parallels the first decrease in S i / A l and silt median diameters (Fig. 4.9). Accordingly, any increase in percent clay accompanied by a decrease in grain size is interpreted here as primarily reflecting decreased loess input. In turn, this implies that the early deglaciation (prior to the end of the YD) is dominated by changes in winter monsoonal climate only, and that there are no obvious sedimentological indications of changes in summer monsoon climate. This interpretation contrasts with the reconstruction of summer monsoonal intensity by Wang and Sarnthein [1999] and Wang et al. [1999a,b], who interpreted the percent clay record solely in terms of riverine 127 material, implying a first increase in summer monsoonal precipitation significantly prior to the Boiling warming (Fig. 4.8). In contrast to the early deglacial increase in percent clay, the marked clay maximum immediately following the YD is not paralleled by any significant decrease in grain size, thus suggesting a true runoff maximum, indicative of intensified summer monsoonal precipitation. This assertion is in line with the interpretation of the percent clay record, as well as the coherent foraminiferal 6 l sO minimum (not shown) by Wang and Sarnthein [1999] and Wang et al. [1999a], and is further corroborated by the major element geochemistry. Thus, the onset of the clay maximum is mirrored by a decrease in T i / A l and K / A l ratios, pointing to a pronounced decrease in average grain size. Furthermore, this clay maximum is paralleled by a marked minimum in percent carbonate (ca. 10 wt.% compared to a rather constant glacial-interglacial average of 15%; not shown), most likely indicative of dilution by peak sediment discharge from the Pearl River. The post-YD increase in summer monsoonal intensity is consistent with previous reconstructions based on marine sedimentary records from the Arabian Sea [e.g., Overpeck et al., 1996; Sirocko et al, 1996], as well as the interpretation of the stalagmite 6 l s O record from Hulu Cave published recently [Wang et al., 2001]. These records demonstrate a close coupling of increases in summer monsoonal intensity with the Boiling and the post-YD warmings in the N-Atlantic realm as recorded in Greenland ice cores. However, neither the major element records nor the siliciclastic grain size indices [Wang and Sarnthein, 1999; Wang et al, 1999a,b] show any pronounced change associated with the Boiling warming. This lack of any significant response in the sediment texture to an inferred first deglacial increase in summer monsoonal intensity is most likely due to the opposing effects of a decrease in clay delivery to site 17940 in response to the retrogression of the Pearl river mouth during MWP la, which occurred synchronously with the Boiling warming (see chapter 4.3), and an increased runoff caused by invigorated summer monsoonal precipitation. 4.2.4 Conclusions The amplitudes of downcore variations in element/Al ratios (except S i /Al ) in core 17940-2 are small compared to the range of values observed in SCS surface sediments [Calvert et al, 1993; M . Kienast, unpubl.] and in downcore records from other basins influenced by monsoonal climate, such as for example the Mediterranean [Calvert and Fontugne, 2001] and the Arabian Sea [Reichart et al, 1997; Sirocko et al, 2000]. This 128 suggests that despite the distinct environmental changes at site 17940 caused by the deglacial rise in sea level and the marked contrast in glacial and interglacial monsoon climate, the sedimentology remained fairly constant. Nevertheless, there are distinct sediment compositional differences, which can be interpreted in terms of variable palaeo-climate. Thus, S i / A l reveals a two-step deglacial weakening of winter monsoonal winds. A first decrease in loess input to site 17940 occurs significantly prior to the Boiling warming, which, in turn, is paralleled by a plateau in S i / A l values; the second decrease coincides with the YD cold period. The pattern and timing of the deglacial decrease of winter monsoonal intensity suggests a mechanistic link to equatorial Pacific SSTs, which needs to be investigated further. In contrast to this deglacial record of weakening winter monsoonal winds, the reconstruction of changes in summer monsoonal precipitation, and thus river runoff, based on major element/Al ratios (and grain size indices) is compromised by antagonistic effects. For example a decrease in sediment discharge due to the deglacial retrogression of the river mouths in response to rapidly rising sea level during MWP la is postulated to have offset any increase in fine-grained fluvial sediments delivery to site 17940 caused by intensified summer monsoonal precipitation during the onset of the Boiling warm period. Nevertheless, an early Holocene peak in summer monsoon intensity can be inferred from a marked clay maximum immediately following the YD. Overall, the interpretation of the major element geochemistry presented here is hampered by the lack of mineralogical and minor/ trace element data, which could prove useful in clearly identifying fluvial and loess endmembers. 129 4.3 Constraints on the timing of MWP la from the SCS 4.3.1 Introduction Following the last glaciation, sea level rose extremely rapidly during meltwater pulse (MWP) l a [Bard et al, 1990; Fairbanks, 1989]. Together with the Boiling warming, MWP la arguably represents the most dramatic event in Earth's climate history during the last 20 kyrs, and determining its precise age is of utmost importance for a mechanistic understanding of the oceanographic, glaciological and climatic changes that occurred during the last deglaciation [Clark et al, 1999; 2001]. Nevertheless, the absolute timing of MWP la and its phase relation with the Boiling warming are still a matter of considerable debate. According to the widely accepted chronology and correlation provided by Bard et al. [1996], MWP la at -14,000 cal yr BP corresponds to the Older Dryas (OD), that is to the first major cooling event following the Boiling warming. This would imply a significant weakening of the thermohaline circulation and its associated heat transport to the North Atlantic region due to the fresh-water input. Various modeling studies [Stocker et al, 1992; Manabe and Stouffer, 1997;] appear to corroborate a causative coupling of MWP la and the Older Dryas, indicating a cessation of deep water formation in the North Atlantic in response to a massive meltwater input. However, Lohmann and Schulz [2000] suggest that previous models frequently underestimate overflow over the Greenland-Scotland Ridge, and show that meltwater discharge and continued deep-water formation in the North Atlantic can be reconciled using a new model, which allows for significant deep-water formation in the Greenland-Iceland-Norwegian Sea. Furthermore, a recent modeling study by Clark et al [2002] suggests that most of the meltwater during MWP la indeed originated from Antarctica, and not, as previously thought, from the Laurentide ice sheet. On the other hand, the dating of MWP l a provided by Hanebuth et al. [2000] shows that the major rise of sea level occurred between 14.6 and 14.3 ka, that is synchronously with the Belling warming in Greenland at 14.6 ± 0.3 ka [Stuiver and Grootes, 2000]. This, in turn, would suggest a continuation of heat transport to the North Atlantic region by the thermohaline conveyor during times of massive meltwater discharge. Here, an independent means of establishing the phase relation between MWP la and the Belling warming is presented, circumventing the inherent uncertainties of quasi-absolute chronologies, and of comparing independently dated records, e.g., ice core with coral reef [Bard et al, 1996] or siliciclastic shelf [Hanebuth et al, 2000] records. 130 4.3.2 Results and discussion Terrestrial organic matter supply to the northern SCS (site 17940, see Fig. 4.1 for core location) decreased rapidly at the end of the last glaciation (Fig. 4.10), a change that is paralleled by a similar abrupt increase in sea-surface temperature (SST) of ca. 1°C. A n identical synchroneity between the drop in terrigenous sediment supply to the open SCS and a significant warming at the end of the last glaciation is also observed at a site from the southern SCS in front of the Sunda Shelf [site 17964, see Fig. 4.1 for core location; Pelejero et al, 1999b; see also Fig. 4.5 and Chapter 4.1], as well as in two lower resolution cores from the open SCS (cores 17954-2 and 17961-2; see Figs. 3.9 and 3.10). This drop in terrigenous supply to the SCS is not the result of decreased summer monsoonal precipitation in the region, which could equally lead to reduced fluvial sediment discharge. On the contrary, a number of studies show that summer monsoonal precipitation increased during the deglaciation [Wang et al, 1999a; Wang et al. 2001; see also Chapter 4.2], which would have led to an increase in terrigenous sediment supply to the SCS. Furthermore, this decrease in terrigenous organic matter supply to the SCS could not have been caused by a decrease in aeolian transport because the decrease in winter monsoonal strength during the last deglaciation has been shown to occur significantly prior to the abrupt deglacial warming [Wang et al, 1999a,b; see also Chapter 4.2]. Accordingly, the abrupt decrease in terrigenous organic matter supply to site 17940 is interpreted to be a result of MWP la. The rapid rise in sea level of 13.5 to 24 m in <290 to 500 years [Fairbanks, 1989; Bard et al, 1990; Blanchon and Shaw, 1995; Hanebuth et al, 2000] during MWP la led to a rapid retreat of shorelines around the world (including river mouths), and a sudden flooding of large parts of exposed shelf areas, and, therefore, the inundation of the lower reaches of rivers close to or at the shelf margin. This rapid transgression of large, shallow areas, most notably in western Pacific marginal seas, instantaneously provided a rapidly landward-extending and deepening accumulation space for terrigenous sediment. These conditions preferentially favoured deposition on the shelves and, in turn, led to an abrupt decrease in the supply of sediment to the outer shelves, the continental slopes and the deep sea. The rapid increase in SST at two sites in the southern SCS (cores 18252-and 18287-3, see Fig. 4.1 for core location) has been constrained by AMS radiocarbon dating to start at 14.6±0.3 ka (see chapter 2.3), synchronously with the Boiling warming in the Greenland GISP2 ice core record. Similarly, we argue that the rapid warming at the end of the last glacial observed at site 17940 in the northern SCS corresponds to the Boiling warming. AMS radiocarbon dates [Fig. 4.10; see also Fig. 2.7; Wang et al, 1999a] yield an 131 500 r i i i I i i i l i i — i — I — i — i — i — i — i — i — i — i — i — T T T T core depth [cm] Fig. 4.10: Alkenone (U K 3 7 ) sea surface temperature estimates (open circles; from Pelejero et al, [1999a]) and M-nonacosane (C 2 9 n-alkane) concentrations (filled circles; ng/g; see Villanueva et al. [1997] for analytical method) of core 17940-2 versus depth. Long chain n-alkanes with a predominance of odd-carbon-numbered members are uniquely produced by terrestrial plants [Eglinton and Hamilton, 1967], and are thus widely used in palaeoceanographic studies to monitor terrigenous organic matter input to marine sediments. For simplicity, only n-nonacosane data are reported here. This approach is justified by the uniform distribution of homologues ranging between C 2 3 and C 3 3 in SCS sediments, and a close linear correlation between n-nonacosane abundances and total odd carbon numbered long chain n-alkanes in selected samples analyzed for the complete set of long chain n-alkanes. AMS radiocarbon dates bracketing the Belling warming/MWP la (at 792.5 and 842.5 cm) are taken from Wang et al. [1999a] (see text for more details). Note the sample-to-sample synchroneity of the sharp SST increase and n-nonacosane decrease at the onset of the Boiling warming. 132 interpolated age of the midpoint of this warming step in core 17940-2 of 15,970 +285/-260 years. Given the unlikelihood of such a significant lead of the Boiling warming in the northern SCS compared to two SST records from the southern SCS (Chapter 2.3), this apparent offset in absolute ages is attributed to locally increased reservoir ages of up to 1,400 +610/-470 years at site 17940, possibly caused by advection of Pacific Inter-mediate waters [Wang et al, 1999a]. Three lines of evidence are adduced to corroborate this assertion. First, the increases in local reservoir age at site 17940-2 are within the range of similar variations recently reported from the North Atlantic [Waelbroeck et al, 2001], which suggests that variations of local reservoir ages in the order of 1,400 years are not necessarily restricted to strong upwelling centers. Second, the sample-to-sample correspondence between the SST increase during the Boiling warming and a planktonic foraminiferal 5 1 80 decrease observed in the southern SCS (sites 18252-3 and 18287-3, see Fig. 2.6) is also evident at site 17940 [Pelejero et al, 1999a; see Fig. 3.8]. Third, the detailed phasing of SST and benthic 5 1 80 from within the same core is identical in cores 17940-2 and 17964-3 from the northern and southern SCS, respectively (see Fig. 4.11). Taken together, this evidence argues against a diachrony of the rapid deglacial warming step within the SCS. The first direct dating of MWP la from the SCS [Hanebuth et al, 2000] shows that it began between 15.0 and 14.4 ka and terminated no later than 14.2 ka (Fig. 4.12). The termination of MWP la at 14.2 ka is constrained by the average of the youngest peak in the 2a probability of the uncalibrated radiocarbon ages [Hanebuth, 2000]. Based on an inferred coupling of the onset of MWP la and the Boiling warming in Greenland, Hanebuth et al. [2000] argued for an onset of MWP la at 14.6 ka (Fig. 4.12). The evidence for the close correspondence between the steep SST increase and the rapid drop in terrigenous sediment supply to the open SCS in sediment cores from the northern and southern SCS strongly implies that the Boiling warming and the onset of MWP la are, indeed, synchronous, therefore constraining the timing of MWP la to the period of 14.6 to 14.3 ka. There are several factors that need to be considered in attempting to reconcile the timing of MWP la based on coral records [Fairbanks, 1989; Bard et al, 1990; Bard et al, 1996] and the older age of MWP la presented here and by Hanebuth et al. [2000]. First, corals (from monospecific reefs indicating a certain position with respect to the sea surface) can only yield maximum and minimum ages of rapid sea level changes because they are drowned during sea-level rises faster than about 1.2 m/100 yr [Buddemeier and Smith, 1988; Montaggioni et al, 1997]. Thus, the bounds of the meltwater pulse are ultimately determined by the proximity of the coral samples to the central interval of 133 CO CO I-28 27 26 25 1 1 1 1 I I I . [" core 17964-3 i i i i I I I -o 2.5 3 3.5 4 4.5 2 2.5 3 3.5 _ I I I I I . 1 , 1 1 1 — i — i — i — i 1 ' 1 1 i 1 1 1 1 j core 17964-3 j *• TO CM -L Is core 17940-2 NA 1 Oi. : maximum aiactal & W Q _ • V \ -~ . . . . i . . . . i . . . . I I -O* P n 2 2.5 3 3.5 4 10 15 age ka 20 25 Fig. 4.11. Alkenone sea surface temperature (SST; red dots, from Pelejero et al. [1999a]) and benthic (C. wuellerstorfi) oxygen isotope (5 1 80 C l„.; blue dots, from Wang et al. [1999a]) records of cores 17964-3 (a) and 17940-2 (b) from the southern (06°09'N, 112°13'E, 1556 m water depth) and northern (20°07'N, 117°23'E, 1727 m water depth) SCS, respectively, versus depth. Stapled vertical lines mark the onset of the abrupt deglacial warming step. Note that in both cores the 6 l sO maximum is paralleled by the glacial SST minimum. The maximum 6 l sO value in core 17964-3 is also the maximum glacial 5 l sO value recorded in the spliced record of both the piston core (17964-2) and the gravity core 17964-3 from this site (see Fig. 3c in Wang et al. [1999a]). (c) bwOCw, records of cores 17964-3 (blue dots) and 17940-2 (open blue circles) versus age. Note that the scale of core 17940-2 on the right y-axis is shifted by 0.15%o to align the glacial 5 1 8 0 maxima of both cores. The vertical stapled lines are adopted from a) and b). The age model of core 17964-3 (adopted from Pelejero et al. [1999b]) gives an age of the onset of the abrupt warming step of 14.6 ka, which is synchronous with the two high-resolution records from neighbouring cores 18252-3 and 18287-3 [M. Kienast et al, 2001], whereas the age model of core 17940-2 is based on the radiocarbon dates published in Wang et al. [1999a]. Note that the benthic 5 1 8 0 record of core 17940-2 leads that of core 17964-3 at the time of the onset of the Boiling warming by ca. 1,400 years. The only plausible way to bring the benthic foraminiferal isotope records of both cores (recovered from comparable water depths) in phase during the deglaciation is adopting an increased reservoir age at the northern site (17940-2) of ca. 1,400 years. This approach, as well as the magnitude of the local increase in reservoir age at site 17940-2 is similar to the record from the North Atlantic published by [Waelbroeck et al, 2001]. It further corroborates our assumption of a synchronous deglacial warming throughout the SCS (see text.) 134 13 14 15 16 " 4 3 1GISP2 0 0 8 0 1 1 ' 1 ' ! "i 'r 13 14 15 16 age (ka) Fig. 4.12. Estimated timing of MWP la based on the Sunda Shelf tidal-organic and Barbados coral-reef records as the only data sets bounding the pulse. (A) Sunda Shelf sea-level reconstruction. Open circles represent age intercepts; note that the numerous data points at 98-96 m are multiples produced by calibrating the 1 4 C ages. Dark grey field: l a probability of calibrated ages; in parts strongly expanded due to a prominent " U C plateau". Vertical lines: 1) 15.0 ka: earliest possible start of MWP la given by the 1 a probability of two dates [Hanebuth et al, 2000]; 2) 14.67 ka: onset of MWP la if causatively linked [Hanebuth et al, 2000] to the Boiling warming in Greenland (GISP2), and as evidenced by the synchroneity of the deglacial (Boiling) warming step and the abrupt decrease in terrigenous organic matter supply to the SCS (this study, see text); 3) 14.2 ka: end of MWP l a based on the last peak of the uncalibrated 1 4 C age 2a probabilities [Hanebuth et al, 2000; Hanebuth, 2000]. Hatched field: l a sea-level probability and an assumed maximum palaeo-tidal range (= vertical uncertainty) of ± 3 m. Thick line: most probable sea-level curve after n-nonacosane concentrations from core 17940-2 (Fig. 4.10) and Hanebuth et al. [2000]. (B) Comparison of the Sunda Shelf sea-level curve from A and the Barbados data set (black triangles and circles) as summarized by Bard et al. [1998]. Error bars indicate 2a probabilities of single (filled triangle) and replicate (filled circles) analyses. Note that the later start of MWP la [Fairbanks, 1989; Bard et al, 1990; Bard et al, 1996], which would coincide with the OD, is based on only one single U / T h date; omitting this sample, the coral record would frame the re-defined MWP la interval (see text), mbss = meters below modern sea surface; Bo = Boiling, OD = Older Dryas. 135 most rapid rise in sea-level (Fig. 4.12). Second, there is a significant difference of 200 to 400 years between the radiocarbon dates corrected for variable reservoir ages and changes in atmospheric 1 4 C (as far as known) and the U / T h ages of the corals used to constrain the timing of MWP la [Stuiver et al, 1998; Hanebuth et al, 2000]. This offset is most likely due to uncertainties in the reconstruction of local reservoir ages, and, thus, usually only the U / T h dates are used for sea-level reconstructions. Consequently, the widely-cited beginning of MWP la at 14.2 ka is constrained by only a single U / T h date (from Barbados) without replicates [Fig. 4.12; Fairbanks, 1989; Bard et al, 1990]. Third, the onset of coral reef growth following land inundation is probably delayed by 200 to 300 years due to the time needed for settlement and consolidation [Montaggioni and et al, 1997]. This delay would automatically reduce any closing age of the pulse determined from coral samples. There are several implications of the phase relation between the Belling warming and MWP la proposed here. First, the synchroneity of both events argues against any weakening of the thermohaline circulation in response to MWP la, an interpretation that is corroborated by recent Pa/Th records from the N Atlantic [McManus et al, 2001], and is expected if one accepts an Antarctic source of the meltwater during MWP la [Clark et al, 2002]. Second, methane concentrations in ice cores [Severinghaus and Brook, 1999; Brook et al, 2000] show that its atmospheric content increased rapidly during the Belling warming. Given the apparent synchroneity of MWP la and the Belling warming, we speculate that the methane increase could be partly due to an abrupt rise in water tables caused by heightened sea level. Third, given the immense impact of MWP la on coastal sedimentation, the proposed synchroneity cautions against the interpretation of changes in sedimentological parameters (e.g., grain size distribution, organic carbon concentration) solely in terms of climatic/biological shifts that are presumed to have occurred synchronously with the Belling warming. Finally, the synchroneity of MWP la and the Belling warming in Greenland presented here is consistent with the general notion of rapid melting of ice sheets in response to dramatic warming events. According to a model study by Marshall and Clarke [1999], for example, the large southward-extending vulnerable Laurentide ice sheet responded within a few years (i.e., synchronously within the resolution of marine sedimentary and coral records) to the air temperature increase in excess of 5 °C at the onset of the Belling period with a profusive melt-water pulse. According to Clark et al. [1996], there is, however, little geological evidence for such a rapid melting of the Laurentide ice sheet. Furthermore, based on a recent modeling experiment, Clark et al. [2002] argue that most of the meltwater during MWP la must have originated from Antarctica. This could explain why there is no slow-136 down of the thermohaline circulation in response to the massive freshwater input. The Antarctic origin of MWP la proposed by Clark et al. [2002], however, poses the even greater conundrum of why there is no response (rapid ice sheet melting) to a large forcing (Boiling warming) in the northern hemisphere, whereas there is an abrupt meltwater discharge from Antarctica, quite possibly without a similarly abrupt warming [Steig et al, 1998; Mulvaney and etal, 2000; Grootes et al, 2001]. In addition, glaciological studies so far indicate only a minor contribution of the Antarctic ice sheet to glacial sea level lowering [Ritz et al, 2001; Denton and Hughes, 2002; Huybrechts, 2002]. This conundrum will have to be resolved in future studies of the Antarctic ice sheet itself, as well as of sedimentary records of ice berg and meltwater discharge from Antactica. 4.3.3 Conclusions A rapid drop in the supply of terrigenous organic matter to the open SCS also corresponds with a rapid increase in sea-surface temperature during the last deglaciation, corresponding with the Boiling warming at 14.7 ka. This is interpreted to reflect a rapid retrogression of local rivers due to rapid sea-level rise, strongly implying that the Boiling warming and the onset of MWP la are synchronous. This phase relation contrasts with the widely cited onset of this MWP la at ca. 14 ka [Bard et al, 1996], and implies that MWP la did not cause a reduction of deepwater formation in the North Atlantic. To the contrary, following modeling studies by Mikolajewicz [1998] and Seidov et al. [2001], Clark et al. [2002, in press] proposed that an Antarctic source of MWP la [Clark et al, 2002] could have actually caused the intensifying deep water formation in the North Atlantic during this time, and the consequent Belling warming. Our results demonstrating synchroneity of MWP 1A with the onset of the Belling warming provide critical support for this hypothesis. 137 4.4 References Alley, R. B., D. A. Meese, C. A. Shuman, A. J. Gow, K. C. Taylor, P. M . Grootes, J. W. C. White, M . Ram, E. D. Waddington, P. A . Mayewski, and G. A. Zielinski, Abrupt increase in Greenland snow accumulation at the end of the Younger Dryas event, Nature, 362, 527-529,1993. Arz, H . W., J. Patzold, and G. Wefer, Climatic changes during the last deglaciation recorded in sediment cores from the northeastern Brazilian continental margin, Geo-Marine Letters, 19, 209-218,1999. Bard, E., Correction of accelerator mass spectrometry 1 4 C ages measured in planktonic foraminifera: paleoceanographic implications, Paleoceanography, 3, 635-645,1988. Bard, E., B. Hamelin, R. G. Fairbanks, and A. Zindler, Calibration of the 1 4 C timescale over the past 30,000 years using mass spectrometric U-Th ages from Barbados corals, Nature, 345, 405-409,1990. Bard, E., B. Hamelin, M . Arnold, L. Montaggioni, G. Cabioch, G. Faure, and F. Rougerie, Deglacial sea-level record from Tahiti corals and the timing of global meltwater discharge, Nature, 382, 241-244,1996. Bard, E., et al., Radiocarbon calibration by means of mass spectrometric 230Th/234U and 14C ages of corals: A n updated database including samples from Barbados, Mururoa and Tahiti, Radiocarbon, 40,1085-1092,1998. Bertrand, P., T. F. Pedersen, P. Martinez, S. E. Calvert, and G. Shimmield, Sea level impact on nutrient cycling in coastal upwelling areas during deglaciation: Evidence from nitrogen isotopes, Global Biogeochemical Cycles, 14, 341-355, 2000. Bird, M . I., Pousai P., Variations of 8 1 3C in the surface soil organic carbon pool. Global Biogeochem. Cycles 11, 313-322,1997. Blanchon, P., and J. Shaw, Reef drowning during the last deglaciation: evidence for catastrophic sea-level rise and ice-sheet collapse, Geology, 23, 4-8,1995. Broecker, W. S., M . Andree, M . Klas, G. Bonani, W. Wolfli, and H . Oeschger, New evidence from the South China Sea for an abrupt termination of the last glacial period, Nature, 333,156-158,1988. Brook, E. J., S. Harder, J. Severinghaus, E. J. Steig, and C. M . Sucher, On the origin and timing of rapid changes in atmospheric methane during the last glacial period, Global Biogeochemical Cycles, 14, 559-572, 2000. Buddemeier, R. W., and S. V. Smith, Coral reef growth in an era of rapidly rising sea level: Predictions and suggestions for long term research, Coral Reefs, 7, 51-56,1988. Buhring, C , M . Sarnthein, and Leg 184 Shipboard Scientific Party, Toba ash layers in the South China Sea: Evidence of contrasting wind directions during eruption ca. 74 ka, Geology, 28, 275-278, 2000. Bush, A . G., R. G. Fairbanks, Exposing the Sunda Shelf: Climatic consequences of eustatic sea-level change, EOS - Transactions of the American Geophysical Union, 82, F794, 2001 Cacho, I., J. O. Grimalt, F. J. Sierro, N . Shackleton, M . Canals, Evidence for enhanced Mediterranean themohaline circulation during rapid climatic coolings, Earth Planet. Sci. Lett. 183, 417-429, 2000. Calvert, S. E., and M . R. Fontugne, On the late Pleistocene-Holocene sapropel record of climatic and oceanographic variability in the eastern Mediterranean, Paleoceanography, 16, 78-94, 2001. Calvert, S. E., T. F. Pedersen, and R. C. Thunell, Geochemistry of surface sediments of the Sulu and South China Seas, Marine Geology, 114, 207-231,1993. Calvo, E., J. Villanueva, J. O. Grimalt, A. Boelaert, L. Labeyrie, New insights into the glacial latitudinal temperature gradients in the North Altalntic. Results from U K 3 7 sea surface temperatures and terrigenous inputs. Earth Planet. Sci. Lett. 188, 509-519, 2001. 138 Clark, P. U., R. B. Alley, and D. Pollard, Northern hemisphere ice-sheet influence on global climate change, Science, 286,1104-1111,1999. Clark, P. U., A . Mix, and E. Bard, Ice sheets and sea level of the last glacial maximum, Eos, Transactions American Geophysical Union, 82, 241-247, 2001. Clark, P. U., J. X. Mitrovica, G. A. Milne, and M . E. Tamisiea, Sea-level fingerprinting as a direct test for the source of global Meltwater Pulse IA, Science, 295, 2438-2441, 2002. Clark, P.U., McCabe, A . M . , Mitrovica, J.X., and Weaver, A., The 19-ka meltwater pulse. Eos, Transactions American Geophysical Union, 2002, in press. Clark, P. U., et al., Origin of the first global meltwater pulse following the last glacial maximum, Paleoceanography, 11, 563-578,1996. Clemens, S. C , Dust response to seasonal atmospheric forcing: proxy evaluation and calibration, Paleoceanography, 13, 471-490,1998. Crowley, T. J., Ice age terrestrial carbon changes revisited, Global Biogeochemical Cycle, 9, 377-389,1995. Curray, J. R., Sediments and history of Holocene transgression, continental shelf, northwest Gulf of Mexico, in Recent sediments, northwest Gulf of Mexico, edited by Shepard, F. P., F. B. Phleger and T. H . van Andel, pp. 221-266,1960. Dash, B. P., C. M . Shepstone, S. Dayal, S. Guru, B. L. A . Hains, G. A. King, and G. A. Ricketts, Seismic investigation on the northern part of the Sunda Shelf south and east of Great Natuna Island, United Natios ECAFE, CCOP Technical Bulletin, 6, 179-196, 1972. Deines, P., The isotope composition of reduced organic carbon, in Handbook of environmental isotope geochemistry, edited by Fritz, P. and J. C. Fontes, pp. 329-406, Elsevier, 1980. Denton, G. H. , and T. J. Hughes, Reconstructing the Antarctic ice sheet at the Last Glacial Maximum, Quaternary Science Reviews, 21,193-202, 2002. Edwards, R. L., J. W. Beck, G. S. Burr, D. J. Donahue, J. M . A. Chappell, A. L. Bloom, E. R. M . Druffel, and F. W. Taylor, A large drop in atmospheric 14C/12C and reduced melting in the Younger Dryas, documented with 230Th ages of corals, Science, 260, 962-968,1993. Eglinton, G., and R. J. Hamilton, Leaf epicuticular waxes, Science, 156,1322-1335,1967. Einsele, G., Event deposits: the role of sediment supply and realtive sea-level changes -Overview, Sedimentary Geology, 104,11-37,1996. Fairbanks, R. G., A 17,000-year glacio-eustatic sea level record: influence of glacial melting rates on the Younger Dryas event and deep-ocean circulation, Nature, 342, 637-642,1989. Flenley, J., Tropical forests under the climates of the last 30,000 years, Climate Change, 39, 177-197,1998. Geyh, M . A., H.-R. Kudrass, and H . Streif, Sea-level changes during the late Pleistocene and Holocene in the Strait of Malacca, Nature, 278, 441-443,1979. Goni, M . A., K. C. Ruttenberg, and T. I. Eglinton, Sources and contribution of terrigenous organic carbon to surface sediments in the Gulf of Mexico, Nature, 389, 275-278,1997. Goodbred Jr., S.L., Kuehl, S.A., Holocene and modern sediment budgets for the Ganges-Brahmaputra river: system: Evidence for highstand dispersal to flood-plain, shelf, and deep-sea depocenters. Geology 27, 559-561,1999. Goodbred Jr., S.L., Kuehl, S.A., The significance of large sediment supply, active tectonism, and eustasy on margin sequence development: Late Quaternary stratigraphy and evolution of the Ganges-Brahmaputra delta. Sed. Geol. 133, 227-248, 2000a. Goodbred Jr., S.L., Kuehl, S.A., Enormous Ganges-Brahmaputra sediment discharge during strengthened early Holocene monsoon. Geology 28,1083-1086, 2000b. 139 Grootes, P. M. , and M . Stuiver, Oxygen 18/16 variability in Greenland snow and ice with 103- to 105-year time resolution, Journal of Geophysical Research, 102, 26,455-426,470,1997. Grootes, P. M . , E. J. Steig, M . Stuiver, E. D. Waddington, and D. L. Morse, The Taylor Dome Antarctic l s O record and globally synchronous changes in climate, Quaternary Research, 56, 289-289, 2001. Haile, N . S., Quaternary deposits and geomorphology of the Sunda Shelf off Malaysian shores, Etudes sur le Quaternaire dans le Monde, VIII, 162-164,1969. Hanebuth, T. J. J., Sea-level changes on the Sunda Shelf during the last 50,000 years, Ph.D., Kiel University, 2000. Hanebuth, T. J. J., and K. Stattegger, The stratigraphic evolution of the Sunda Shelf during the past fifty thousand years, in Deltas of South-east Asia and vicinity -sedimentology, stratigraphy, and petroleum geology, edited by Sidi, F. H., D. Nummedal, H . W. Posamentier, H . Darman and P. Imbert, 2001, in press. Hanebuth, T., K. Stattegger, and P. M . Grootes, Rapid flooding of the Sunda Shelf: A late-glacial sea-level record, Science, 288,1033-1035, 2000. Hesp, P. A., C. C. Hung, M . Hilton, C. L. Ming, and I. M . Turner, A first tentative Holocene sea-level curve for Singapore, Journal of Coastal Research, 14, 308-314,1998. Huybrechts, P., Sea-level changes at the L G M from ice-dynamic reconstructions of the Greenland and Antarctic ice sheets during the glacial cycle, Quaternary Science Reviews, 21, 203-231, 2002. Kienast, M . , Constant nitrogen isotopic composition of organic matter in the South China Sea during the last climatic cycle: Global implications, Paleoceanography, 15, 244-253, 2000. Kienast, M . , S. Steinke, K. Stattegger, and S. E. Calvert, Synchronous tropical South China Sea SST change and Greenland warming during deglaciation., Science, 291, 2132-2134, 2001a. Kienast, M . , S. E. Calvert, C. Pelejero, and J. O. Grimalt, A critical review of marine sedimentary S 1 3C o r g-pC0 2 estimates: New paleo-records from the South China Sea and a revisit of other low-latitude 6 1 3C 0 r g-pCO 2 records, Global Biogeochemical Cycles, 15,113-127, 2001b. Koutavas, A., J. Lynch-Stieglitz, T. M . Marchitto, and J. P. Sachs, El Nino-like pattern in ice age tropical Pacific sea surface temperature, Science, 297, 226-230, 2002. Kudrass, H . R., X. L. Jin, H . Beiersdorf, and P. Cepek, Erosion and sedimentation in the Xisha Trough at the continental margin of southern China, in Marine geology and geophysics of the South China Sea, edited by Jin, X., H . R. Kudrass and G. Pautot, pp. 137-147, China Ocean Press, 1992. Lea, D. W., D. K. Pak, and H . J. Spero, Climate impact of late Quaternary equatorial Pacific sea surface temperature variations, Science, 289,1719-1724, 2000. Lee, H . J., S. K. Chough, and S. H . Yoon, Slope-stability change from late Pleistocene to Holocene in the Ulleung Basin, East Sea (Japan Sea), Sedimentary Geology, 104, 39-51, 1996. Li , X., and X. Sun, Palynological records since the Last Glacial Maximum from a deep sea core in the southern South China Sea, Quaternary Science, 6, 526-535,1999. Lin, H.-L., C.-T. Lai, H.-C. Ting, L. Wang, M . Sarnthein, and J.-J. Hung, Late Pleistocene nutrients and sea surface productivity in the South China Sea: a record of teleconnections with Northern hemisphere events, Marine Geology, 156, 197-210, 1999. Lohmann, G. P., and M . Schulz, Reconciling Boiling warmth with peak deglacial meltwater discharge, Paleoceanography, 15, 537-540, 2000. Maloney, B. K., Evidence for the Younger Dryas climatic event in Southeast Asia, Quaternary Science Reviews, 14, 949-958,1996. 140 Manabe, S., and R. J. Stouffer, Coupled ocean-atmosphere model response to freshwater input: Comparison to Younger Dryas event, Paleoceanography, 12, 321-336,1997. Marshall, S., and G. Clarke, Modeling North American freshwater runoff through the last glacial cycle, Quaternary Research, 52, 300-315,1999. Martinez, P., P. Bertrand, S. E. Calvert, T. F. Pedersen, G. B. Shimmield, E. Lallier-Verges, and M . R. Fontugne, Spatial variations in nutrient utilization, production and diagenesis in sediments of a coastal upwelling regime (NW Africa): Implications for the palaeoceanographic record, Journal of Marine Research, 58, 809-835, 2000. Maxwell, A.L. , Holocene Monsoon changes inferred from lake sediment pollen and carbonate records, northeastern Cambodia. Quat. Res. 56, 390-400, 2001. McManus, J. F., R. Francois, J. Gherardi, J. M . Nuwer, and S. Cosier, Glacial and deglacial rates of thermohlaine circulation, Eos, Trans. AGU, Fall Meet. Sappl., 82, abstract OS31E-11, 2001. Meese, D. A., A. J. Gow, R. B. Alley, G. A. Zielinski, P. M . Grootes, M . Ram, K. C Taylor, P. A . Mayewski, and J. F. Bolzan, The Greenland Ice Sheet Project 2 depth-age scale: Methods and results, Journal of Geophysical Research, 102, 26,411-426,423,1997. Mikolajewicz, U. , Effect of meltwater input from the Antarctic ice sheet on the thermohaline circulation. Annals of Glaciology, 27, 311-315,1998. Molengraaff, G. A . F., and M . Weber, On the relation between the Pleistocene glacial period and the origin of the Sunda Sea (Java- and South China- Sea), and its influence on the distribution of coral reefs and on the land- and freshwater fauna, Proceedings Royal Acad. Amsterdam, 23, 394-428,1920. Montaggioni, L. F., and et al., Continuous record of reef growth over the past 14 k.y. on the Mid-Pacific Island of Tahiti, Geology, 25, 555-558,1997. Mulder, T., and J. P. M . Syvitski, Climatic and morphologic relationships of rivers: implications of sea-level fluctuations on river loads, Journal of Geology, 104, 509-523, 1996. Mulvaney, et al., The transition from the last glacial period inland and near-coastal Antarctica, Geophysical Research Letters, 27, 2673-2676, 2000. Overpeck, J., D. Anderson, S. Trumbore, and W. Prell, The southwest Indian monsoon over the last 18 000 years, Climate Dynamics, 12, 213-225,1996. Pelejero, C , J. O. Grimalt, S. Heilig, M . Kienast, and L. Wang, High resolution U K 3 7 -temperature reconstructions in the South China Sea over the past 220 kyr, Paleoceanography, 14, 224-231,1999a. Pelejero, C , M . Kienast, L. Wang, and J. O. Grimalt, The flooding of Sundaland during the last deglaciation: imprints in hemipelagic sediments from the southern South China Sea, Earth and Planetary Science Letters, 171, 661-671,1999b. Penny, D. A 40,000 year palynological record from north-east Thailand; implications for biogeography and palaeo-environmental reconstructions. Palaeogeogr., Palaeoclimatol., Palaeoecol. 171, 97-128, 2001. Petit-Maire, N . , Natural variability of the Earth's environments: The last two climatic extremes (18000±2000 and 8000±1000 yrs BP), C.R. Acad. Sc. Paris - Earth and Planetary Sciences, 328, 273-279,1999. Porter, S. C , and Z. An, Correlation between climate events in the North Atlantic and China during the last glaciation, Nature, 375, 305-308,1995. Prell, W. L., and J. E. Kutzbach, Monsoon variability over the past 150,000 years, Journal of Geophysical Research - Part D, 92, 8411-8425,1987. Prentice, K. C , and I. Y. Fung, The sensitivity of terrestrial carbon storage to climate change, Nature, 346, 48-51,1990. Prins, M . A., and G. Postma, Effects of climate, sea level, and tectonics unraveled for last deglaciation turbidite records of the Arabian Sea, Geology, 28, 375-378, 2000. 141 Prins, M . A., G. Postma, J. Cleveringa, A . Cramp, and N . H . Kenyon, Controls on terrigenous sediment supply to the Arabian Sea during the late Quaternary: the Indus Fan, Marine Geology, 169, 327-349, 2000a. Prins, M . A., G. Postma, and G. J. Weltje, Controls on terrigenous sediment supply to the Arabian Sea during the late Quaternary: the Makran continental slope, Marine Geology, 169, 351-371, 2000b. Puig, P., and A. Palanques, Nepheloid structure and hydrographic control on the Barcelona continental margin, northwestern Mediterranean, Marine Geology, 149, 39-54,1998. Reichart, G. J., M . den Dulk, H . J. Visser, C. H . van der Weijden, and W. J. Zachariasse, A 225 kyr record of dust supply, paleoproductivity and the oxygen minimum zone from Murray Ridge (northern Arabian Sea), Palaeogeography, Palaeoclimatology, Palaeoecology, 134,149-169,1997. Ritz, C , V . Rommelaere, and C. Dumas, Modeling the evolution of Antarctic ice sheet over the last 420,000 years: Implications for altitude changes in the Vostok region, Journal of Geophysical Research, 106, 31,943-931,964, 2001. Rottman, M . L., Dissolution of planktonic foraminifera and pteropods in South China Sea sediments, Journal of Foraminiferal Research, 9, 41-49,1979. Sarnthein, M . , U . Pflaumann, P. X. Wang, and H . K. Wong, Preliminary report of the Sonne-95 cruise "Monitor Monsoon" to the South China Sea, Rep. 68, Berichte -Reports, Geol.-Palaont. Inst. Univ. Kiel, 1994. Schimanski, A., Diversitat und Entwicklung von Faziesraumen auf dem Sunda Schelf im Verlauf der postpleistozanen Transgression, Diplom, Kiel University, 1999. Schdnfeld, J., and H.-R. Kudrass, Hemipelagic sediment accumulation rates in the South China Sea related to Late Quaternary sea-level changes, Quaternary Research, 40, 368-379,1993. Seidov, D., Barron, E. and Haupt, B.J., Meltwater and the global ocean conveyor: northern versus southern connections. Global and Planet. Change, 30, 257-270, 2001. Severinghaus, J. P., and E. J. Brook, Abrupt climate change at the end of the last glacial period inferred from trapped air in polar ice, Science, 286, 930-934,1999. Sirocko, F., D. Garbe-Schonberg, A. Mclntyre, and B. Molfino, Teleconnections between the subtropical monsoons and high-latitude climates during the last deglaciation, Science, 272, 526-529,1996. Sirocko, F., D. Garbe-Schonberg, and C. Devey, Processes controlling trace element geochemistry of Arabian Sea sediments during the last 25,000 years, Global and Planetary Change, 26, 217-303, 2000. Song, S.-R., C.-H. Chen, M.-Y. Lee, T. F. Yang, Y. Iizuka, and K.-Y. Wei, Newly discovered eastern dispersal of the youngest Toba Tuff, Marine Geology, 167, 303-312, 2000. Stattegger, K., W. Kuhnt, H . Wong, et al., Cndse report SONNE 115: Sundaflut - Sequence stratigraphy, late Pleistocene-Holocene sea level fluctuations and high resolution record of the post-Pleistocene transgression on the Sunda Shelf 211 pp., 1997. Stauffer, P. H., Cenozoic. in Geology of the Malay Peninsular (West Malaysia and Singapore), edited by Gobett, D. J. and C. S. Hutchinson, pp. 143-176, Wiley, 1973. Steig, E. J., E. J. Brook, J. W. C. White, C. M . Sucher, M . L. Bender, S. J. Lehman, D. L. Morse, E. D. Waddington, and G. D. Clow, Synchronous climate changes in Antarctica and the North Atlantic, Science, 282, 92-95,1998. Steinke, S., Sedimentological and climatic changes during the last deglaciation recorded in cores from the Sunda Shelf margin and continental slope (Southern South China Sea). Ph.D. Thesis, Univ. Kiel, Germany, 2001. Steinke, S., M . Kienast, U. Pflaumann, M . Weinelt, and K. Stattegger, A high resolution sea surface temperature record from the tropical South China Sea (16,500 - 3000 yr B.P.), Quaternary Research, 55, 352-362, 2001. 142 Stacker, T. F., D. G. Wrigth, and M . L. A., A zonally averaged, coupled ocean-atmosphere model for paleoclimate studies, Journal Clint., 5, 773-797,1992. Stuiver, M. , and P. M . Grootes, GISP2 oxygen isotope ratios, Quaternary Research, 53, 277-284, 2000. Stuiver, M. , P. M . Grootes, and T. F. Braziunas, The GISP2 6 l sO climate record of the past 16,500 years and the role of the sun, ocean, and volcanoes, Quaternary research, 44, 341-354,1995. Stuiver, M . , P. J. Reimer, E. Bard, J. W. Beck, G. S. Burr, K. A. Hughen, B. Kromer, G. McCormac, J. van der Pflicht, and M . Spurk, INTCAL98 radiocarbon age calibration, 24,000-0 cal BP, Radiocarbon, 40,1041-1083,1998. Sun, X., X. L i , Y. Luo, and X. Chen, The vegetation and climate at the last glaciation on the ermerged continental shelf of the South China Sea, Palaeogeography, Palaeoclimatology, Palaeoecology, 160, 301-316, 2000. Taylor, S. R., S. M . McLennan, and M . T. McCulloch, Geochemistry of loess, continental crustal composition and crustal model ages, Geochimica et Cosmochimica Acta, 47, 1897-1905,1983. Tjia, H . D., The Sunda Shelf, Southeast Asia, Z. Geomorph. N.F., 24, 405-427,1980. Tjia, H . D., and K. K. Liew, Changes in tectonic stress field in northern Sunda Shelf Basin, in Tectonic evolution of Southeast Asia, edited by Hall, R. and D. Blundell, pp. 291-306, Geological Society Special Publication, 1996. Tyson, R. V., Sedimentary organic matter. Organic fades and palynofacies, 615 pp., Chapman & Hall, 1995. Ujiie, H. , Y. Hatakeyama, X. X. Gu, S. Yamamoto, R. Ishiwatari, and L. Maeda, Upward decrease of organic C / N ratios in the Okinawa Trough cores: Proxy for tracing the post-glacial retreat of the continental shore line, Palaeogeography, Palaeoclimatology, Palaeoecology, 165,129-140, 2001. Van Campo, E., J. Guiot, and C. Peng, A data-based re-appraisal of the terrestrial carbon budget at the last glacial maximum, Global and Planetary Change, 8,189-201,1993. van der Kaars, S., X. Wang, P. Kershaw, F. Guichard, and D. A. Setiabudi, A late Quaternary palaeoecological records from the Banda Sea, Indonesia: patterns of vegetation, climate and biomass burning in Indonesia and Australia, Palaeogeography, Palaeoclimatology, Palaeoecology, 155,135-153, 2000. Villanueva, J., C. Pelejero, and J. O. Grimalt, Clean-up procedures for the unbiased estimation of C37 alkenone sea surface temperatures and terrigenous n-alkane inputs in paleoceanography, Journal of Chromatography, 757,145-151,1997. von Rad, U., and M . Tahir, Late Quaternary sedimentation on the outer Indus shelf and slope (Pakistan): evidence from high-resolution seismic data and coring, Marine Geology, 138,193-236,1997. Waelbroeck, C , J.-C. Duplessy, E. Michel, L. Labeyrie, D. Paillard, and J. Duprat, The timing of the last deglaciation in North Atlantic climate records, Nature, Ail, 724-727, 2001. Wang, L., and M . Sarnthein, Long-/short-term variations of monsoon climate and its teleconnection to global change: High resolution grain size records of siliciclastic sediments from the South China Sea. in Reconstructing ocean history: A window into the future, edited by Abrantes, F. and A. Mix, pp. 57-71, Kluwer Academic/Plenum Publisher, 1999. Wang, L., M . Sarnthein, H . Erlenkeuser, J. Grimalt, P. Grootes, S. Heilig, E. Ivanova, M . Kienast, C. Pelejero, and U. Pflaumann, East Asian monsoon climate during the late Quaternary: High-resolution sediment records from the South China Sea, Marine Geology, 156, 245-284,1999a. Wang, L., M . Sarnthein, P. M . Grootes, and H . Erlenkeuser, Millenial reoccurrence of century-scale abrupt events of East Asian monsoon: A possible heat conveyor for the global deglaciation, Paleoceanography, 14, 725-731,1999b. 143 Wang, P., Response of western Pacific marginal seas to glacial cycles: paleoceanographic and sedimentological features, Marine Geology, 156, 5-39,1999. Wang, Y. J., H . Cheng, R. L. Edwards, Z. An, J. Y. Wu, S. C.-C, and J. A. Dorale, A high-resolution absolute-dated late Pleistocene monsoon record from Hulu Cave, China, Science, 294, 2345-2348, 2001. Webster, P. J., V. O. Magana, T. N . Palmer, J. Shukla, R. A. Tomas, M . Yanai, and Y. Yasunari, Monsoons: processes, predictability, and the prospects for prediction, Journal of Geophysical Research, 103,14451-14510,1998. Wiesner, M . G., Y. Wang, and L. Zheng, Fallout of volcanic ash to the deep South China Sea induced by the 1991 eruption of Mount Pinatubo (Philippines), Geology, 23, 885-889,1995. Winn, K., M . Sarnthein, and H . Erlenkeuser, 8lsO stratigraphy and chronology of Kiel sediment cores from the East Atlantic, 99 pp., Berichte - Reports, Geol.-Palaont. Inst. Univ. Kiel, 1991. Wohlfarth, B., The chronology of the last termination: A review of radiocarbon-dated, high-resolution terrestrial stratigraphies, Quaternary Science Reviews, 15, 267-284, 1996. Wong, H . K., C. Haft, A . - M . Paulsen, T. Ludmann, C. Hubscher, and J. Geng, Late Quaternary sedimentation and sea level fluctuations on the northern Sunda Shelf, southern South China Sea. in Deltas of South-east Asia and vicinity - sedimentology, stratigraphy, and petroleum geology, edited by Sidi, F. H. , D. Nummedal, H . W. Posamentier, H . Darman and P. Imbert, 2001, in press. Wright, L. D., Morphodynamics of inner continental shelves, CRC Marine Science Series, CRC Press Inc., 1995. Xiao, J., S. C. Porter, Z. An, H . Kumai, and S. Yoshikawa, Grain size of quartz as an indicator of winter monsoon strength on the Loess Plateau of Central China during the last 130,000 yr, Quaternary research, 43, 22-29,1995. Zhao, M. , G. Eglinton, S. K. Haslett, R. W. Jordan, M . Sarnthein, Z. Zhang, Marine and terrestrial biomarker records for the last 35,000 years at ODP site 658C off NW Africa. Org. Geochem. 31, 919-930, 2000. Zhou, W., M . J. Head, X. Lu, Z. An, A. J. T. Jull, and D. J. Donahue, Teleconnections of climatic events between East Asia and polar, high latitude areas during last deglaciation, Palaeogeography, Palaeoclimatology, Palaeoecology, 152,163-172,1999. 144 5. Closing remarks The sedimentary records examined in this thesis demonstrate a complex oceanographic response of the SCS during the last deglaciation. Thus, the sea surface temperature (SST) records presented in Chapter 2 display a close link to the climate of the circum-North-Atlantic realm, evidenced particularly by a synchronous rapid warming step during the Boiling transition both in the southern SCS and the GISP2 ice core record, as well as by a Younger Dryas (YD)-type cooling. Similarly, the summer monsoon intensified in phase with the post-YD warming of the Northern Hemisphere as shown by the major element composition of a downcore record from the northern SCS (site 17940; Chapter 4.2). In contrast, the record of the deglacial decrease in winter monsoon intensity (Chapter 4.2) deviates from the archetypal pattern and timing of Northern Hemisphere deglaciation, and suggests a mechanistic link to the equatorial Pacific and/or the Southern Hemisphere. This dual-natured behaviour of the SCS is analoguos to records of SST and sea surface salinity (SSS) variability from the western equatorial Pacific [Stott et al. 2002] and the Sulu Sea [Dannemann, 2000; Rosenthal et al, 2001; Oppo et al, 2002], which on the one hand suggest a close link to D / O cycles evident in many Northern Hemisphere climate records, but, on the other hand show a distinctly different deglacial pattern, reminiscent in timing and pattern of the Southern Hemisphere climate change as recorded in (most) Antarctic ice cores. Further high resolution multi-proxy records are needed to investigate this intricate interplay in detail, which may eventually shed light on the temporal evolution of the different forcing mechanisms in this region, such as the SE Asian monsoon, ENSO dynamics, and/or the position of the ITCZ, and their connection to Northern and Southern Hemisphere influences. In addition to these strictly climatological forcings, western Pacific marginal basins, particularly the SCS, are significantly affected by sea-level-related changes in basin configuration. This is not only manifested in variable fluxes of terrigenous material to the SCS basin, which possibly mask or amplify climate-related variations in terrigenous sediment fluxes (Chapters 4.1, 4.2.) and in changes in SST related to the opening/closure of important gateways [Wang et al, 1995; Pelejero et al, 1999a,b], particularly in the southern SCS. A recent modeling study by Bush and Fairbanks [2001] even demonstrated that the exposure/submergence of the Sunda Shelf in response to eustatic changes in sea-level can directly affect regional climate itself. The potentially antagonistic and/or amplifying effects of sea-level related changes in basin 145 configuration on sedimentation and climate in the SCS is exacerbated by the synchroneity (Chapter 4.3) of the most pronounced step-like change in Northern Hemisphere climate during the last deglaciation (the Boiling transition) and melt-water pulse (MWP) la , the largest and most abrupt rise in sea-level during the last deglaciation. The biogeochemistry of the SCS as reflected in the nitrogen and carbon isotopic composition of organic matter (6 1 5N, S 1 3C o r g , respectively; Chapter 3) is less affected by monsoonal climate than similar records from the Arabian Sea. Thus, down-core records of sedimentary 5 1 5 N from throughout the SCS show values similar to the 6 1 5 N of subsurface nitrate in the SCS and the North Pacific Intermediate Water (6 1 5N = ca. 5.5 %o), and do not display any significant changes during the last glacial-interglacial cycle. This probably reflects unchanged complete nitrate utilization during marine primary production, as well as the absence of denitrification in the water column. Accordingly, the SCS is the only location known to date that appears to continuously monitor temporal changes in the 8 1 5 N of nitrate on a regional rather than a local scale. The unchanged sedimentary 6 1 5 N on glacial-interglacial (and shorter) timescales, in turn, implies that the 5 1 5 N of the nitrate used during primary production remained approximately constant during the last climatic cycle, thus constraining proposed changes in the global oceanic N inventory. The records of S 1 3 C o r g variability are similar in amplitude and timing to records from other low-latitude settings, including the oligotrophic Mediterranean and the upwelling system in the Angola Basin. This similarity in spite of the very different physico-environmental settings cautions the use of 8 1 3 C o r g as an unambiguous tracer of dissolved molecular CO z in the surface ocean and calls for a re-evaluation of the role of the low-latitude ocean in temporal changes in atmospheric C 0 2 . It further calls for combined palaeoceanographic and plankton physiological studies to unravel the mechanism(s) governing the carbon isotopic composition of marine organic matter. References Bush, A. G., R. G. Fairbanks, Exposing the Sunda Shelf: Climatic consequences of eustatic sea-level change, EOS - Transactions of the American Geophysical Union, 82, F794, 2001 Dannemann, S., A multi-proxy study of foraminifera to identify past millennial-scale climate variability in the East Asian Monsoon and the Western Pacific Warm Pool, Ph.D. Thesis, Univ. at Albany, State Univ. of New York, 2001. Oppo, D.W., B. K. Linsley, Y. Rosenthal, S. Dannemann, L. Beaufourt, Ice volume and the tropical hydrological cycle, Geochemistry, Geophysics, Geosystems, 2002, in press. 146 Pelejero, C , J. O. Grimalt, S. Heilig, M . Kienast, and L. Wang, High resolution U K 3 7 -temperature reconstructions in the South China Sea over the past 220 kyr, Paleoceanography, 14, 224-231,1999a. Pelejero, C , M . Kienast, L. Wang, and J. O. Grimalt, The flooding of Sundaland during the last deglaciation: imprints in hemipelagic sediments from the southern South China Sea, Earth and Planetary Science Letters, 171, 661-671,1999b. Rosenthal, Y., D. W. Oppo, S. Dannemann, and B. K. Linsley, Millennial-scale variations in western equatorial Pacific sea surface temperature during glacial and Holocene climates, EOS - Transactions of the American Geophysical Union, 81, F657, 2000. Stott, L., C. Poulsen, S. Lund, R. Thunell, Super ENSO and global climate oscillations at millennial time scales, Science, 297, 222-226, 2002. Wang, P., L. Wang, Y. Bian, and Z. Jian, Late Quaternary paleoceanography of the South China Sea: surface circulation and carbonate cycles, Marine Geology, 127, 145-165, 1995. 147 Appendix This Appendix contains a brief description of the sample material used in this thesis (part I) and of the analyses I carried out at UBC (part II, A-E). For analytical methods and further details of data that have not been generated at UBC, for example foraminiferal oxygen isotope ratios, percent clay, and n-nonacosane concentrations, the reader is referred to the original publications referenced in the text, the figure captions, and the preface of this thesis. I. Sample material The surface sediment samples (Table A l ) used in this study were taken from box cores recovered during R / V SONNE cruises 95 (April-June 1994), 115 (December 1996-January 1997) and 140 (April-May 1999) and represent the undisturbed uppermost centimeter of the sediment [Sarnthein et al, 1994; Stattegger et al, 1997; Wiesner et al. 1999]; the overlying fluff layer was carefully sampled using large-volume syringes. In addition, the surface 2 cm of larger-diameter gravity cores along two transects off Palawan [Calvert et al. 1993] were included in the set of surface sediment samples. Down-core subsamples of the box and gravity cores (Table A l ) were taken using 5 or 10 cm 3 syringes. II. Methods A) Alkenone determinations were carried out on freeze-dried, manually ground samples (7 to 11 g) following procedures described in Villanueva et al. [1997], whereby hydrolyzed (with 6% K O H in methanol) dichloromethane sediment extracts (ultrasonicated) are dissolved in toluene (20 u\ final volume) and analyzed by gas chromatography (HP 5880A). Prior to dissolution in toluene, the alkenones (and other neutral compounds) were separated from highly polar compounds using silica gel columns (20x0.5 cm), packed with 2 g of silica gel in a mixture of dichloromethane and n-hexane (8:2), eluting with 10 ml of the same solvent mixture. The C37-alkenones were identified by their gas chromatographic retention times by analogy with an extract of a pure Emiliania huxleyi culture (kindly provided by Joe Needoba, UBC). Selected samples were analyzed by GC-MS for confirmation of compound identification and evaluation of possible coelutions. UK ' 3 7-SST estimates were calculated using the equation developed by 148 Pelejero and Grimalt [1997] specifically for the SCS, which corresponds to annually averaged temperatures of the top 30 m of the water column: SST = (([C37:2] /[C37:2 + C37:3])-0.092)/0.031 (1) where [C37:2] and [C37:3] are the di- and tri-unsaturated alkenones, respectively. Based on multiple analyses of the same sediment extract as well as on repeat extractions of selected sediment samples, the standard deviation of the UK ' 3 7-SST estimates is ± 0.1°C. The U K ' 3 7 -SST SST estimates of cores 18252-3 and 18287-3 are available at ftp://ftp.ngdc.noaa.gov/paleo/paleocean/by_contributor/kienast2001c. B) Total nitrogen (percent TN) was measured on freeze-dried, ground bulk samples using a Carlo Erba NA1500 elemental analyzer with an analytical precision of ±3 % (la relative standard deviation). C) Nitrogen isotope ratios were determined on freeze-dried, ground bulk samples using a Fisons NA1500 element analyzer coupled to a V G prism mass spectrometer in a continuous flow of helium. The 6 1 5 N ( 8 1 5 N [%0] = [ ( 1 5 N / u N S a m P i e -1 5 N/ 1 4 N a t m o s p h . N 2 ) / 1 5 N/ u N a t m o s p h . N 2 ]*10 - 3 ) values are reported relative to air N 2 with an analytical precision of ±0.2 %o. A l l sedimentary 6 1 5 N data reported in this thesis are available at ftp://ftp.ngdc.noaa.gov/paleo/paleocean/by contributor/kienast2000. The nitrogen isotopic composition of nitrate (5 1 5 N N 0 3 . , samples kindly provided by Dr. E.-F. Yu, Taiwan) was determined following a technique developed by Sigman et al. [1997]. Briefly, after the addition of 0.5 g of MgO, 200 ml of sea water (in 250 ml Wheaton bottles with Teflon-lined caps) were incubated at 65 °C for five days to decompose any labile dissolved organic nitrogen to ammonia. After this initial incubation, the sample volume was reduced to 30-40 ml by evaporation at 95 °C (ca. 36 hrs). Following the addition of a 'diffusion package' (a 1-cm G F / D filter acidified with 25 [il 4N sulphuric acid sandwiched between two 2.5-cm diameter, 10-^m pore-size Teflon membranes) and 0.15 g of Devarda's Alloy (an alloy of Cu, A l , Zn that reduces nitrate to ammonia), samples were incubated for 12-14 days in an oven at 65 °C, and for another 14 days on a reciprocating shaker table at 1 rps. After that, the G F / D filters were recovered and analyzed for their nitrogen isotopic composition. Due to significant isotope fractionation during nitrate reduction and the diffusion of ammonia onto the acidified filters, it is imperative to recover all nitrate (i.e., ammonia) quantitatively. This 149 is attested by the highly significant linear correlation between the initial nitrate concentration of the sample and the amount of nitrogen recovered on the filter (Fig. A l ) . Fig. A l . Relationship between nitrate concentrations in SCS sea water samples (^M NCy; kindly analyzed by Joe Needoba, UBC) and u.g N recovered on the filter after ammonia diffusion (note that only 200 ml of each sample were used to determine 6 1 5 N N 0 3 -). The slope of the expected regression (black line; based on the molar weight of nitrogen of 14.00674 g) agrees very well with the observed (red line). The minor x-intercept of the observed regression is most likely associated with the analytical blank of the ammonia diffusion method. D) The carbon isotopic composition of organic matter (6 1 3C o r g) was measured on decalcified (10 % HC1 followed by oven drying without washing) samples using the instruments described above. The 6 1 3 C o r g data are reported relative to Peedee belemnite (PDB), with a precision of ±0.1 %o. A l l 6 1 3 C o r g data reported in this thesis are available at ftp://ftp.ngdc.noaa.gov/paleo/paleocean/by contributor/kienast2001b E) Major element determinations were carried out following methods detailed in Calvert et al. [1985] and Calvert [1990]. Briefly, 400 mg of sediment mixed with lithium tetraborate were fused at 1100 °C and cast into glass beads, which were analyzed on a Philips 2400 wavelength-dispersive sequential automatic X-ray fluorescence spectrometer. Precisions of the analyses were ±3 %. Elemental concentrations of Mg and K have been corrected for dilution by sea salt in the dried samples, using the weight difference of a wet versus dry sediment sample of known volume (data kindly provided by Luejiang Wang), and assuming a salinity of 33 and constant ratios between total salt content and K and Mg. Al l major element data presented here are given in table A2. 150 References Calvert, S. E., Geochemistry and origin of the Holocene sapropel in the Black Sea. in Facets of Modern Biogeochemistry, edited by Ittekkot, V., S. Kempe, W. Michaelis and A. Spitzy, pp. 326-352, Springer-Verlag, Berlin, 1990. Calvert, S. E., B. L. Cousens, and M . Y. S. Soon, An X-ray fluorescence spectrometric method for the determination of major and minor elements in ferromanganese nodules, Chemical Geology, 51, 9-18,1985. Calvert, S. E., T. F. Pedersen, and R. C. Thunell, Geochemistry of surface sediments of the Sulu and South China Seas, Marine Geology, 114, 207-231,1993. Pelejero, C , and J. O. Grimalt, The correlation between.the Uk37 index and sea surface temperatures in the warm boundary: the South China Sea, Geochimica et Cosmochimica Acta, 61, 4789-4797,1997. Sarnthein, M . , U . Pflaumann, P. X. Wang, and H . K. Wong, Preliminary report of the Sonne-95 cruise "Monitor Monsoon" to the South China Sea, Rep. 68, Berichte -Reports, Geol.-Palaont. Inst. Univ. Kiel, 1994. Sigman, D. M . , M . A . Altabet, R. Michener, D. C. McCorkle, B. Fry, and R. M . Holmes, Natural abundance-level measurement of the nitrogen isotopic composition of oceanic nitrate: an adaption of the ammonia diffusion method, Marine Chemistry, 57, 227-242,1997. Stattegger, K., W. Kuhnt, H . Wong, C. Buhring, C. Haft, T. Hanebuth, H . Kawamura, M . Kienast, S. Lorenc, B. Lotz, T. Ludmann, M . Lurati, N . Miihlhan, J. Paulsen, J. Pracht, A. Putar-Roberts, Q. Hung, A. Richter, B. Salomon, A . Schimanski, S. Steinke, R. Szarek, N . Van Nhan, M . Weinelt, and C. Winguth, Cruise report SONNE 115: Sundaflut - Sequence stratigraphy, late Pleistocene-Holocene sea level fluctuations and high resolution record of the post-Pleistocene transgression on the Sunda Shelf, 211 pp., 1997. Villanueva, J., C. Pelejero, and J. O. Grimalt, Clean-up procedures for the unbiased estimation of C37 alkenone sea surface temperatures and terrigenous n-alkane inputs in paleoceanography, Journal of Chromatography, 757,145-151, 1997. Wiesner, M . G., K. Stattegger, W. Kuhnt, and et al., Cruise Report SONNE 140 - Siidmeer III, Rep. 7, Berichte - Reports, Inst, fur Geowiss., Univ. Kiel, 1999. 151 Table A1 STATION latitude °N longitude °E water depth (m) STATION latitude °N longitude °E water depth (m) SONNE 95 SONNE 115 17920 14.35 ' 119.45 2507 18249 9.24 108.55 1 33 17921 14.54 119.32 2507 18250 9.24 108.58 148 17922 15.25 117.27 4221 18252 9.15 109.23 1277 17923 15.08 117.25 1839 18253 9.24 109.3 1479 17924 19.24 118.5 3438 18260 9.24 108.2 74 17925 19.51 119.05 2980 18265 9.23 107.45 47 17926 19 118.44 3761 18266 9.23 107.44 47 17927 17.15 119.27 2800 18277 4.56 109.56 134 17928 18.27 119.75 2486 18284 5.33 110.32 226 17929 20.68 115.7 371 18287 5.39 110.39 598 17930 20.33 115.78 629 18294 6.08 111.18 849 17931 20.1 115.96 1005 18300 4.22 108.39 91 17932 19.95 116.04 1365 18314 3.59 108.59 100 17933 19.53 116.23 1972 18321 2.18 107.25 109 17934 19.03 116.46 2665 18323 2.47 107.53 92 17935 18.88 116.53 3143 17936 18.77 117.12 3809 SONNE 140 17937 19.5 117.67 3428 18376 7.05 108.06 90 17938 19.79 117.54 2835 18391 9.34 108.5 1 22 17939 19.97 117.46 2473 18401 12.13 109.32 134 17940 20.12 117.38 1728 18404 13.3 109.34 169 17941 21.52 118.48 2201 18409 15.41 108.41 40 17942 19.33 1,13.2 329 18414 15.06 108.58 21 17943 18.95 117.55 917 18416 15.02 109.09 63 17944 18.66 113.64 1219 18425 16.35 108.28 94 17945 18.13 113.78 2404 17946 18.13 114.25 3465 17947 18.47 116.03 3765 17948 16.71 114.9 2841 17949 17.35 115.17 2195 17950 16.09 112.9 1868 17951 16.29 113.41 2340 17952 16.67 114.47 2882 17953 14.33 115.08 4309 17954 14.8 111.53 1517 17955 14.12 112.18 2404 17956 13.85 112.59 3387 17957 10.9 115.3 2197 17958 11.62 115.08 2581 17959 11.14 115.29 1957 17960 10.12 115.56 1707 17961 8.51 112.33 1795 17962 7.18 112.08 1970 17963 6.17 112.67 1233 17964 6.16 112.21 1556 17965 6.16 112.55 889 bold numbers: core sites; plain numbers: surface sediment sample only 152 Table A2 17940-2 [cm] Si/Al 0 3.11 6 3.04 8 3.03 10 3.2 12 3.08 14 3.07 16 3.07 18 3.06 20 3.28 22 3.14 24 3.19 26 3.17 28 3.16 30 3.24 32 3.08 34 3.13 36 3.18 38 3.18 40 3.17 42 3.14 44 3.16 46 3.07 48 3.15 50 3.19 52 3.15 54 3.11 56 3.22 58 3.23 60 3.32 62 3.23 64 3.19 66 3.1 68 3.14 70 3.27 72 3.15 74 3.16 76 3.13 78 3.14 80 3.19 82 3.13 84 3.12 86 3.17 88 3.13 90 2.95 92 3.14 94 3.12 96 3.14 99 3.13 100 3.16 102 3.07 104 3.11 TI/AI K/Al 0.0464 0.171 0.0443 0.178 0.0442 0.17 0.0499 0.192 0.0447 0.181 0.0442 0.17 0.044 0.182 0.0441 0.18 0.0503 0.195 0.0434 0.174 0.0441 0.177 0.0442 0.176 0.0448 0.17 0.0512 0.192 0.0439 0.17 0.045 0.178 0.0454 0.176 0.0445 0.17 0.0453 0.165 0.044 0.176 0.0445 0.181 0.0434 0.174 0.0452 0.183 0.0463 0.17 0.0452 0.177 0.0443 0.176 0.0448 0.183 0.0441 0.183 0.0535 0.195 0.0447 0.177 0.0452 0.186 0.0439 0.174 0.045 0.18 0.0508 0.187 0.0449 0.176 0.0446 0.179 0.0441 0.18 0.0451 0.18 0.0468 0.177 0.0459 0.181 0.0447 0.184 0.0449 0.184 0.0447 0.178 0.0433 0.173 0.0443 0.181 0.0442 0.183 0.0449 0.18 0.0448 0.18 0.0474 0.184 0.0442 0.175 0.0445 0.179 Fe/Al Mg/Al 0.403 0.17 0.381 0.196 0.381 0.198 0.437 0.212 0.387 0.197 0.376 0.197 0.373 0.201 0.37 0.203 0.446 0.209 0.389 0.204 0.395 0.208 0.383 0.206 0.38 0.21 0.427 0.216 0.366 0.201 0.387 0.204 0.378 0.203 0.376 0.209 0.393 0.206 0.367 0.206 0.377 0.21 0.369 0.202 0.366 0.205 0.388 0.213 0.374 0.205 0.368 0.205 0.398 0.206 0.453 0.216 0.47 0.225 0.422 0.209 0.415 0.218 0.374 0.202 0.381 0.204 0.422 0.214 0.387 0.205 0.424 0.216 0.386 0.205 0.379 0.203 0.387 0.206 0.381 0.202 0.368 0.203 0.385 0.207 0.371 0.201 0.372 0.195 0.389 0.212 0.385 0.205 0.36 0.204 0.377 0.201 0.4 0.208 0.364 0.199 0.386 0.205 17940-2 [cm] SI/AI 444 3.05 446 2.96 448 2.99 450 3.04 452 3.02 454 3.05 456 3.03 458 3.07 460 3.14 462 3.06 464 3.07 466 3 468 3.05 470 3.04 472 3.01 474 2.96 476 2.99 478 2.99 480 3.11 482 3.04 484 3 486 3.03 488 3.02 490 3.06 492 2.98 494 3 496 2.97 498 3.02 500 3.03 502 3.04 504 3 504 3.11 506 3.19 508 3.17 510 3.08 512 3.39 514 3.33 520 3.09 524 3.02 525 3 526 3.1 528 3.04 532 2.99 534 2.96 535 2.99 542 3.07 544 3.08 545 3 546 3.05 548 3.11 555 3.04 Ti/AI K/Al 0.0446 0.174 0.0442 0.167 0.0442 0.1-75 0.0451 0.182 0.0441 0.177 0.0448 0.179 0.0444 0.182 0.0442 0.181 0.046 0.199 0.0448 0.18 0.0442 0.182 0.0443 0.171 0.0451 0.182 0.0451 0.178 0.0451 0.176 0.0443 0.177 0.0439 0.179 0.0435 0.179 0.046 0.196 0.0442 0.179 0.0438 0.173 0.0443 0.184 0.0442 0.19 0.0457 0.186 0.0429 0.175 0.0435 0.173 0.0426 0.175 0.0438 0.183 0.0452 0.181 0.0445 0.186 0.043 0.178 0.0481 0.191 0.0511 0.238 0.0446 0.193 0.0448 0.178 0.0581 0.271 0.0551 0.245 0.0455 0.183 0.044 0.193 0.0445 0.175 0.0452 0.215 0.0471 0.19 0.0455 0.179 0.0428 0.17 0.042 0.169 0.0475 0.201 0.045 0.192 0.0446 0.182 0.0449 0.186 0.0447 0.217 0.0444 0.186 Fe/Al Mg/Al 0.382 0.197 0.361 0.2 0.366 0.195 0.37 0.199 0.365 0.196 0.387 0.203 0.379 0.201 0.393 0.203 0.38 0.203 0.386 0.203 0.401 0.204 0.37 0.199 0.369 0.198 0.374 0.204 0.375 0.2 0.38 0.202 0.38 0.198 0.387 0.199 0.381 0.208 0.391 0.202 0.376 0.201 0.383 0.204 0.375 0.199 0.364 0.201 0.377 0.2 0.354 0.195 0.397 0.209 0.378 0.2 0.359 0.202 0.374 0.202 0.371 0.2 0.404 0.198 0.428 0.207 0.388 0.203 0.389 0.204 0.491 0.219 0.506 0.223 0.383 0.202 0.356 0.2 0.383 0.198 0.401 0.201 0.42 0.196 0.409 0.196 0.377 0.195 0.349 0.197 0.394 0.2 0.379 0.2 0.376 0.199 0.377 0.201 0.369 0.198 0.416 0.193 153 17940-2 [cm] S i / A l 106 3.1 108 3.14 110 3.17 112 3.14 114 3.16 116 3.12 118 3.11 120 3.16 122 3.03 124 3.1 126 3.1 128 3.12 130 3.22 132 3.09 134 3.15 136 3.15 138 3.09 142 3.12 144 3.11 146 3.11 148 3.08 150 3.17 152 3.11 154 3.12 156 3.1 158 3.11 160 3.25 162 3.09 164 3.16 166 3.13 168 3.12 170 3.17 172 3.13 174 3.14 176 3.14 178 3.1 180 3.3 182 3.16 184 3.14 186 3.18 188 3.15 192 3.21 194 3.12 196 3.12 199 3.16 200 3.14 202 3.27 204 3.16 206 3.12 208 3.09 210 3.17 212 3.13 214 3.1 216 3.15 T i / A l K / A l 0.0443 0.18 0.0448 0.177 0.0475 0.184 0.044 0.184 0.0451 0.187 0.0449 0.18 0.044 0.18 0.0472 0.179 0.0435 0.175 0.045 0.176 0.045 0.18 0.0441 0.179 0.0505 0.2 0.0445 0.18 0.0482 0.194 0.0458 0.185 0.0443 0.18 0.045 0.181 0.0442 0.171 0.0445 0.178 0.0446 0.176 0.0464 0.196 0.0443 0.172 0.0444 0.18 0.0447 0.171 0.0452 0.18 0.0492 0.196 0.0453 0.177 0.0446 0.176 0.0446 0.179 0.0448 0.177 0.0459 0.177 0.0446 0.176 0.0449 0.179 0.0448 0.181 0.0447 0.184 0.0515 0.21 0.045 0.178 0.0448 0.178 0.0441 0.181 0.0457 0.186 0.0447 0.187 0.0454 0.179 0.0448 0.178 0.0446 0.172 0.0457 0.169 0.0502 0.199 0.0447 0.179 0.0439 0.183 0.0446 0.173 0.047 0.174 0.0449 0.182 0.045 0.176 0.0438 0.176 Fe/AI Mg/AI 0.381 0.206 0.351 0.202 0.402 0.207 0.386 0.209 0.377 0.201 0.381 0.205 0.383 0.202 0.445 0.191 0.361 0.198 0.386 0.203 0.38 0.209 0.376 0.205 0.483 0.186 0.373 0.205 0.4 0.203 0.39 0.203 0.364 0.198 0.388 0.204 0.38 0.204 0.374 0.2 0.372 0.202 0.383 0.204 0.387 0.209 0.384 0.207 0.363 0.2 0.376 0.207 0.411 0.209 0.376 0.206 0.39 0.209 0.373 0.207 0.368 0.203 0.391 0.198 0.368 0.203 0.386 0.202 0.389 0.209 0.385 0.207 0.446 0.209 0.383 0.203 0.371 0.2 0.392 0.208 0.361 0.202 0.384 0.205 0.376 0.202 0.385 0.211 0.377 0.208 0.39 0.196 0.444 0.214 0.38 0.21 0.369 0.205 0.366 0.201 0.381 0.205 0.375 0.199 0.362 0.198 0.363 0.197 17940-2 [cm] S i / A l 556 2.99 558 3.28 562 2.99 565 3.13 566 3.03 568 3.01 574 3.02 575 2.99 582 3.01 585 2.97 586 3.02 592 3.02 594 3.04 595 3.04 596 3.01 598 3.07 604 3.17 605 3.12 606 3.19 608 2.98 608 2.99 614 3.13 615 2.99 615 2.98 625 3.08 635 3.02 645 3.08 648 3.33 652 3.27 655 3.27 656 3.29 662 3.24 665 3.29 666 3.37 675 3.33 676 3.32 678 3.31 682 3.36 684 3.28 685 3.28 686 3.26 688 3.52 692 3.31 695 3.48 698 3.38 702 3.51 705 3.41 708 3.45 712 3.46 714 3.54 715 3.42 716 3.26 718 3.7 722 3.64 T i / A l K / A l 0.0441 0.173 0.0556 0.254 0.0437 0.176 0.0518 0.215 0.0449 0.198 0.044 0.196 0.0453 0.211 0.0451 0.196 0.0438 0.195 0.0438 0.184 0.0443 0.201 0.0458 0.188 0.0457 0.2 0.0451 0.206 0.0441 0.195 0.0451 0.217 0.0518 0.24 0.0446 0.171 0.0504 0.215 0.0431 0.175 0.0439 0.175 0.0455 0.22 0.0448 0.178 0.0448 0.178 0.0484 0.198 0.0448 0.164 0.0448 0.172 0.0519 0.232 0.0475 0.205 0.0478 0.214 0.0498 0.214 0.0482 0.207 0.0506 0.216 0.0502 0.232 0.0477 0.222 0.0482 0.221 0.0474 0.226 0.0537 0.222 0.0506 0.195 0.0475 0.207 0.048 0.208 0.0653 0.243 0.0475 0.177 0.0517 0.225 0.0524 0.194 0.0514 0.22 0.0481 0.199 0.0487 0.198 0.0486 0.202 0.0497 0.227 0.0488 0.193 0.047 0.173 0.0667 0.239 0.0584 0.262 Fe/AI Mg/AI 0.379 0.189 0.496 0.219 0.396 0.198 0.455 0.214 0.383 0.199 0.371 0.198 0.374 0.199 0.392 0.199 0.372 0.195 0.381 0.197 0.368 0.195 0.413 0.198 0.406 0.197 0.391 0.2 0.372 0.196 0.388 0.197 0.466 0.207 0.371 0.209 0.449 0.208 0.379 0.193 0.38 0.192 0.388 0.195 0.378 0.189 0.377 0.188 0.408 0.196 0.377 0.192 0.367 0.201 0.425 0.209 0.399 0.2 0.397 0.199 0.408 0.204 0.388 0.204 0.417 0.202 0.418 0.216 0.38 0.2 0.395 0.202 0.38 0.2 0.447 0.203 0.413 0.199 0.387 0.199 0.385 0.196 0.564 0.231 0.377 0.2 0.416 0.206 0.426 0.209 0.424 0.215 0.386 0.204 0.387 0.205 0.387 0.208 0.404 0.207 0.38 0.207 0.377 0.2 0.606 0.237 0.49 0.227 154 17940-2 [cm] Si/Al 218 3.2 222 3.18 224 3.22 226 3.2 228 3.19 230 3.18 232 3.14 234 3.26 236 3.16 238 3.14 240 3.18 242 3.09 244 3.15 246 3.22 248 3.22 250 3.23 252 3.14 254 3.15 256 3.21 258 3.15 260 3.21 262 3.2 264 3.18 266 3.17 268 3.13 270 3.17 272 3.14 274 3.17 276 3.11 278 3.19 280 3.15 282 3.18 284 3.19 286 3.2 288 3.19 290 3.29 292 3.19 294 3.17 296 3.2 299 3.16 302 3.18 304 3.22 306 3.13 308 3.13 310 3.21 312 3.18 314 3.21 316 3.06 318 2.97 320 3.2 322 3.14 324 3.12 326 3.08 328 3.14 Ti/AI K/Al 0.0449 0.177 0.0453 0.182 0.0449 0.18 0.0451 0.176 0.046 0.181 0.0456 0.171 0.0446 0.175 0.0453 0.188 0.0451 0.176 0.0447 0.177 0.0461 0.17 0.0445 0.174 0.045 0.174 0.0452 0.183 0.0461 0.186 0.0472 0.183 0.0446 0.174 0.0457 0.176 0.0469 0.173 0.0453 0.177 0.0473 0.179 0.0475 0.182 0.0478 0.19 0.0453 0.179 0.0456 0.175 0.046 0.171 0.046 0.181 0.0458 0.175 0.0456 0.166 0.0464 0.187 0.0482 0.177 0.0464 0.19 0.046 0.174 0.045 0.185 0.0448 0.172 0.0469 0.2 0.0464 0.178 0.0454 0.173 0.0465 0.191 0.0453 0.168 0.0473 0.175 0.0468 0.184 0.0448 0.167 0.0447 0.171 0.0459 0.186 0.046 0.188 0.0463 0.18 0.0447 0.158 0.0434 0.165 0.046 0.187 0.0459 0.177 0.0466 0.18 0.0454 0.181 0.0466 0.176 Fe/Al Mg/Al 0.377 0.203 0.372 0.202 0.391 0.208 0.375 0.203 0.37 0.202 0.363 0.198 0.376 0.203 0.359 0.205 0.368 0.204 0.367 0.201 0.387 0.191 0.358 0.198 0.375 0.218 0.401 0.208 0.386 0.211 0.391 0.201 0.381 0.202 0.368 0.202 0.379 0.208 0.373 0.213 0.393 0.2 0.373 0.212 0.386 0.212 0.387 0.207 0.369 0.216 0.397 0.192 0.364 0.211 0.366 0.205 0.359 0.204 0.377 0.212 0.403 0.203 0.384 0.21 0.363 0.21 0.354 0.209 0.351 0.209 0.394 0.2 0.359 0.201 0.361 0.202 0.378 0.206 0.362 0.209 0.38 0.212 0.384 0.207 0.366 0.213 0.367 0.209 0.381 0.201 0.376 0.203 0.381 0.206 0.35 0.209 0.364 0.21 0.372 0.201 0.371 0.208 0.382 0.213 0.372 0.215 0.359 0.202 17940-2 [cm] Si/Al 725 3.26 726 3.29 728 3.48 734 3.62 735 3.39 736 3.59 742 3.75 745 3.41 746 3.57 752 3.47 754 3.71 755 3.65 765 3.5 766 3.61 768 3.55 772 3.53 775 3.53 778 3.52 782 3.47 795 3.67 802 3.56 805 3.43 806 3.43 808 3.62 814 3.38 815 3.77 816 3.63 818 3.6 822 3.52 824 3.61 825 3.46 826 3.61 828 3.64 832 3.65 834 3.68 835 3.64 836 3.62 838 3.78 842 3.62 844 3.68 845 3.69 846 3.61 848 3.6 852 3.54 854 3.75 855 3.79 856 3.88 858 3.56 862 3.87 864 3.61 865 3.81 866 3.67 868 3.65 872 3.78 Ti/AI K/Al 0.0461 0.172 0.0463 0.171 0.049 0.221 0.0561 0.249 0.0473 0.184 0.0551 0.214 0.0612 0.246 0.0463 0.168 0.054 0.2 0.0469 0.171 0.0547 0.215 0.0504 0.221 0.0485 0.201 0.0515 0.217 0.0485 0.206 0.0481 0.193 0.0474 0.205 0.0493 0.198 0.0486 0.189 0.0536 0.233 0.0486 0.199 0.0464 0.17 0.0467 0.168 0.0507 0.201 0.0448 0.167 0.0569 0.241 0.0519 0.199 0.0549 0.202 0.0482 0.177 0.0595 0.214 0.0501 0.195 0.057 0.228 0.0559 0.23 0.0526 0.205 0.0506 0.2 0.0489 0.204 0.0482 0.188 0.0545 0.23 0.0486 0.182 0.0511 0.186 0.0498 0.214 0.0477 0.175 0.0485 0.178 0.0477 0.171 0.0552 0.227 0.0539 0.22 0.0637 0.22 0.0448 0.165 0.0563 0.215 0.0483 0.188 0.0492 0.199 0.0461 0.167 0.0477 0.174 0.0486 0.193 Fe/Al Mg/Al 0.371 0.202 0.366 0 198 0.402 0 203 0.44 0 216 0 393 0 193 0 449 0 217 0 486 0 234 0 366 0 199 0 425 0 212 0 369 0 191 0 415 0 214 0 399 0 204 0 374 0 198 0 419 0 205 0 377 0 202 0 378 0 198 0 365 0.2 0 391 0.203 0.41 0.202 0 417 0.2 0 396 0 194 0 374 0 188 0 356 0 187 0 383 0 192 0 344 0 183 0 444 0 208 0 414 0 195 0 447 0 193 0 375 0 183 0 485 0 209 0 375 0 189 0 432 0 209 0 476 0 219 0 409 0 197 0 376 0 195 0 361 0 192 0 361 0 187 0 437 0 207 0 381 0 181 0 381 0 188 0 356 0 188 0 354 0 185 0 366 0 182 0 357 0 184 0 418 0 202 0 401 0 196 0 521 0 192 0 327 0 .18 0 405 0.204 0 343 0 .19 0 364 0 189 0 417 0 194 0.35 0 181 0.354 0.195 155 17940-2 [cm] Si/Al 330 3.26 332 3.08 334 3.19 336 3.14 338 3.19 340 3.13 342 3.13 344 3.15 346 3.11 350 3.17 352 3.14 354 3.15 356 3.2 358 3.2 360 3.27 362 3.12 364 3.11 366 3.14 368 3.08 370 3.15 372 3.13 374 3.12 376 3.15 378 3.13 382 3.15 384 3.11 386 3.14 388 3.13 390 3.18 392 3.01 394 3.08 396 3.05 398 3.08 400 3.19 402 3.1 404 3.12 406 3.07 408 3.07 410 3.21 412 3.05 414 3.1 416 3.03 418 3.01 420 3.13 422 3.02 424 3.01 426 3.1 428 3.05 430 3.12 432 3.08 434 3.08 436 3.51 438 3.06 440 3.04 442 3.08 Ti/Al K/Al 0.0472 0.203 0.045 0.167 0.0462' 0.177 0.0457 0.177 0.0483 0.186 0.0466 0.18 0.0457 0.167 0.0457 0.181 0.0453 0.174 0.0469 0.192 0.0454 0.177 0.0459 0.172 0.0453 0.187 0.0453 0.177 0.0497 0.205 0.0461 0.174 0.0463 0.172 0.0455 0.178 0.0455 0.174 0.0465 0.175 0.0457 0.177 0.0459 0.185 0.0467 0.179 0.0457 0.175 0.0464 0.185 0.0464 0.18 0.0462 0.178 0.0465 0.178 0.0465 0.177 0.0441 0.171 0.0469 0.179 0.0454 0.177 0.0456 0.183 0.0476 0.193 0.0456 0.18 0.0454 0.175 0.0458 0.175 0.0452 0.174 0.0463 0.179 0.0462 0.175 0.045 0.176 0.0453 0.172 0.0451 0.17 0.0463 0.196 0.0447 0.175 0.0452 0.181 0.0464 0.183 0.0466 0.178 0.047 0.19 0.0473 0.182 0.0446 0.176 0.0475 0.18 0.0457 0.175 0.0433 0.169 0.0465 0.187 Fe/AI Mg/AI 0.395 0.204 0.341 0.205 0.37 0.208 0.371 0.203 0.383 0.207 0.382 0.202 0.368 0.201 0.36 0.207 0.377 0.212 0.378 0.2 0.393 0.197 0.372 0.2 0.418 0.214 0.417 0.18 0.412 0.204 0.369 0.201 0.356 0.198 0.397 0.204 0.356 0.197 0.365 0.2 0.358 0.206 0.364 0.198 0.398 0.212 0.363 0.207 0.37 0.201 0.369 0.199 0.368 0.205 0.372 0.203 0.369 0.198 0.35 0.199 0.366 0.204 0.359 0.195 0.365 0.201 0.387 0.203 0.371 0.206 0.396 0.209 0.366 0.198 0.361 0.209 0.389 0.203 0.373 0.215 0.406 0.216 0.371 0.202 0.365 0.198 0.369 0.198 0.367 0.197 0.363 0.199 0.382 0.201 0.37 0.206 0.386 0.205 0.384 0.205 0.434 0.209 0.368 0.192 0.379 0.201 0.35 0.197 0.371 0.204 17940-2 [cm] Si/Al 875 4.08 876 3.72 878 3.74 882 4.02 884 3.85 885 3.95 886 3.93 888 4.26 892 4.03 894 4.07 895 4.17 896 3.99 898 4.09 905 4.17 915 4.26 925 4.14 935 4.18 945 4.15 955 4.08 965 4.08 975 4.09 985 4.11 995 4.18 1005 3.99 1015 3.94 1025 4.14 1035 3.73 1035 3.71 1045 3.8 1055 3.94 1065 4.05 1075 3.99 1085 4.2 1095 4.07 1105 4.01 1115 4.05 1125 3.88 1135 3.76 1145 3.9 1 155 3.92 1165 3.89 1175 3.64 1185 3.9 1195 3.96 1205 3.94 1215 4.01 1225 3.8 1235' 3.79 1235 3.99 1245 3.87 1255 3.97 1275 4.07 1285 4.08 1295 4.03 Ti/Al K/Al 0.0604 0.239 0.049 0.175 0.0485 0.172 0.052 0.203 0.0486 0.184 0.0513 0.189 0.0485 0.184 0.0521 0.211 0.0478 0.175 0.0505 0.184 0.0508 0.206 0.05 0.179 0.0611 0.219 0.0511 0.21 0.0601 0.232 0.0512 0.179 0.0528 0.18 0.0549 0.225 0.0586 0.202 0.0524 0.213 0.0517 0.194 0.0604 0.206 0.0576 0.24 0.051 0.182 0.0521 0.222 0.0605 0.239 0.0487 0.176 0.0488 0.197 0.0508 0.178 0.0512 0.174 0.0532 0.223 0.0517 0.188 0.0591 0.249 0.0452 0.208 0.0561 0.199 0.0629 0.219 0.0505 0.192 0.0494 0.177 0.0547 0.198 0.0517 0.176 0.0513 0.18 0.0488 0.176 0.0519 0.199 0.0515 0.208 0.0525 0.19 0.0517 0.22 0.0504 0.198 0.0525 0.187 0.0586 0.237 0.0517 0.213 0.0513 0.203 0.0465 0.188 0.0624 0.255 0.0595 0.231 Fe/AI Mg/AI 0.451 0.216 0.363 0.182 0.352 0.17 0.371 0.204 0.344 0.192 0.369 0.19 0.345 0.188 0.365 0.194 0.33 0.19 0.342 0.192 0.355 0.196 0.349 0.189 0.426 0.202 0.359 0.195 0.424 0.21 0.346 0.19 0.401 0.187 0.366 0.197 0.445 0.193 0.375 0.197 0.364 0.189 0.432 0.189 0.404 0.205 0.36 0.185 0.376 0.2 0.454 0.208 0.349 0.186 0.346 0.184 0.368 0.177 0.38 0.194 0.361 0.199 0.366 0.197 0.439 0.205 0.311 0.163 0.41 0.184 0.543 0.204 0.359 0.19 0.375 0.191 0.406 0.177 0.364 0.191 0.375 0.194 0.361 0.188 0.37 0.191 0.364 0.193 0.423 0.194 0.357 0.186 0.346 0.18 0.404 0.189 0.445 0.206 0.407 0.192 0.393 0.187 0.328 0.169 0.432 0.21 0.491 0.21 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0052772/manifest

Comment

Related Items