INFORMATION TO USERS

This manuscript has b een reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy subm itted. Broken o r indistinct print, colored or poor quality illustrations and photographs, print t>leedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that tMe author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to b o removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, begrinning at the upper left-hand comer and continuing from left to right in equal sanctions with small overlaps.

Photographs included in the original manuscript have tieen reproduced xerographicaily in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Bell & How«ll Information and beaming 300 North Zeeb Road, Ann Art)or, Ml 48106-1346 USA 800-521-0600 UMI’

SEDIMENTOLOGY, MINERALOGY AND GEOCHEMISTRY OF THE SIRIUS GROUP AND OTHER CENOZOIC GLACIGENIC FROM : IMPLICATIONS FOR CLIMATE AND SHEET HISTORY

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of the

By

Sandra Passchier, M.S.

*****

The Ohio State University 2000

Dissertation Committee: Approved by: Professor P.-N. Webb, Advisor Professor L. A. Krissek Professor G. Faure Professor G. D. McKenzie Advisor Department of Geological Sciences UMI Number: 9994919

___ ® UMI

UMI Microform 9994919 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Copyright by Sandra Passchier 2000 ABSTRACT

Sedimentological, mineralogical and geochemical studies are applied to Cenozoic glacial sediments in Antarctica to investigate their provenance and paleoenvironment of deposition. The age and origin of terrestrial glacial deposits of the Sirius Group in the are widely debated. X-ray diffraction studies, and analyses of heavy minerals, chemistry, and grain-size identified two provenance end members: one consisting of deposits on high summit plateaus with a provenance of chemically weathered sedimentary rocks; another consisting of deposits in glacial troughs, the present drainage corridors of the East Antarctic , with a provenance of igneous and metamorphic basement rocks. The different geomorphological settings and compositions of the Sirius Group are caused by emplacement of the Sirius Group at different stages of glacial denudation in the Transantarctic Mountains, which requires that the deposits have different ages.

Climatic records from the basin in the sector of Antarctica show that a decrease in chemical weathering started at the Eocene/Oligocene boundary. Glaciotectonic features in the Cape Roberts drillcores CRP-1 and CRP-2/2A in the Victoria Land basin were studied and suggest that ice first reached the continental shelf in the Ross Sea sector in the mid-

Oligocene, which indicates that the decrease in chemical weathering was associated with climatic cooling and was followed by ice-sheet growth and expansion. Deposition of part of the Sirius

Group may be related to these earlier glacial phases. Because of their widely varying ages, correlation between Sirius Group outcrops to

reconstruct late Neogene paleoclimatic conditions is not recommended. The Oliver Bluffs succession is Pliocene in age and forms an important terrestrial archive of a relatively recent interval of global warming. However, it should not be regarded as representative of the entire

Sirius Group. Although the chemical index of alteration suggests that weathering was limited, the

Pliocene Sirius Group at Oliver Bluffs shows evidence of considerable meltwater reworking, suggesting that surface melting was more important than under the present dry polar climate conditions in continental Antarctica.

I ll To my parents I dedicate this dissertation.

IV ACKNOWLEDGMENTS

I would like to express my sincere thanks to a number of people for continuous support,

guidance, feed back and encouragement during this research. First, I would like to thank my

advisor Peter Webb for guidance through many aspects of the graduate education and for

allowing me to explore the research topics of my choice and in a way I thought was best.

Committee members Peter Webb, Larry Krissek, Gunter Faure, and Garry McKenzie are thanked

for their critical review of the text, and Larry Krissek is also recognized for advice on various

aspects regarding the study of the composition of sediments. Gunter Faure, David Harwood, and

Gary Wilson provided additional samples of the Sirius Group and are thanked for discussion on

the topic in the initial stages of the research. Jason Whitehead is thanked for providing

information about, and samples from the Pagodroma Group of .

I thank Mary Davis for carrying out the Coulter Counter measurements at the Byrd Polar

Research Center, and Tumara Withers for lab assistance with the DVDP-11 pilot study. David

Elliot is recognized for his interest in the Sirius Group, some valuable discussions, and for help

with the identification of minerals and rock fragments. Rosemary Askin shared unpublished data

on pollen and spores in the Sirius Group, which is greatly appreciated. Fellow graduate students here at OSU, including my colleagues Wojciech Majewski and Michael Sperling are thanked for support and reviewing early versions of papers and proposals. Gerhard Schmiedl provided much inspiration and contagious enthusiasm during his year here at OSU.

Peter Barrett gave me the opportunity to participate in the Cape Roberts Drilling Project, and I thank him for his encouragements in the breccia studies. Terry Wilson and Tim Paulsen are also thanked for discussions and collaboration on the study of breccias in the CRP-1 core. Chris

Fielding, Mike Hambrey, Larry Krissek, Ross Powell and Laura De Santis are thanked for

answering questions when observing the Cape Roberts cores in McMurdo Station. Being a part of

the Cape Roberts Science Team has been essential in broadening my horizons and all members

are thanked for sharing an exciting scientific and international experience.

The Ocean Drilling Program is thanked for providing the opportunity to participate in

drilling on the East Antarctic margin. Co-chiefs Alan Cooper and Phil O'Brien, and the Leg 188

shipboard scientific party, in particular Steve Bohaty, Fabio Florindo, Carl Fredrik Forsberg, Pat

Quilty, Michele Rebesco, Kari Strand, Dietz Wamke, and Jason Whitehead are acknowledged for

discussions on various topics of Arctic and Antarctic science.

Jaap van der Meer is thanked for his continuing support and inspiration beyond the

Master's level. I am also indebted to Dick van der Wateren and Anja Verbers who introduced me

to the and provided the opportunity to work on the Sirius Group when I

was still in the Netherlands.

Annemieke and Sandra are thanked for visiting me here and providing moral support as

well as a good opportunity for some leisure time. The social gatherings with my American and

international friends here in Columbus are also much appreciated. My sister Ellen is thanked for

her love and hospitality during my visits to Amsterdam. Lastly, I am greatly indebted to my best

friend and partner Wobbe for his artistic contributions, and for his support for, and understanding

of my dedication to science during the course of the dissertation.

Funding for this research was provided by the National Science Foundation Office of

Polar Programs grants GPP 93-17979, GPP 94-19054 and GPP 94-20475 to Peter-N. Webb

(Principal Investigator), the Lois Jones Fellowship, a Geological Society of America student research grant, and the Friends of Grton Hall.

VI VITA

November 6, 1968 Bom - Leiden, The Netherlands

1994 M.S. Physical Geography University of Amsterdam The Netherlands

1994-1995 Office Coordinator Institute for Geo-ecological Research, University of Amsterdam, The Netherlands

1995 Research Assistant Netherlands Institute For Applied Geoscience TNG Haarlem, The Netherlands

1996 Research Assistant Free University Amsterdam, The Netherlands

Fall 1996 Teaching Assistant The Ohio State University Columbus, Ohio

1997-2000 Research Assistant The Ohio State University Columbus, Ohio

Fall 1997 Technician, Cape Roberts drilling Project, Antarctica

Fall 1998 Technician, Cape Roberts drilling Project, Antarctica

Winter 1999 Lecturer Historical Geology The Ohio State University Columbus, Ohio

Spring 2000 Sedimentologist Ocean Drilling Program Leg 188

VII PUBLICATIONS

Van Tatenhove, F.G.M. and Passchier, S., 1995, Morphological and sedimentological characteristics of past and present ice dammed lakes in west Greenland. In: Abstracts of the International Union for Quaternary Research, XIV International Congress, 3-10 August 1995, Berlin, Germany, p. 284.

Passchier, S., Uscinowicz, S. and Laban, C., 1997, supply and transport directions in the Gulf of Gdansk as observed from SEM analysis of quartz grain surface textures. Prace Panstwowego Instytutu Geologicznego CLVHI, Sopot - Warszawa, Poland, 27p.

Passchier, S., Verbers, A.L.L.M. and Van der Wateren, P.M. 1997, Late-Cenozoic glacial history of the Southern Prince Albert Mountains, northern Victoria Land. In: Antarctica and Global Change: Interactions and Impacts. Symposium Programme & Abstracts, No. 0375P. Antarctic CRC, Hobart, Tasmania.

Passchier, S., 1998, Sedimentology and mineralogy of the Late Cenozoic Sirius Group in the southern Shackleton Region, Transantarctic Mountains. GSA Abstracts with Programs, Vol. 30, No. 2, Abstract No. 1078.

Passchier, S., Verbers, A.L.L.M., Van der Wateren, F.M. and Vermeulen, F.J.M., 1998, Provenance, geochemistry, and grain-sizes of glacigene sediments, including the Sirius Group, and Late Cenozoic glacial history of the southern Prince Albert Mountains, Victoria Land. Annals of , 27, 290-296.

Cape Roberts Science Team (Sandra Passchier), 1998. Initial Report on CRP-l, Cape Roberts Project, Antarctica. Terra Antartica, 5(1), 1-187.

Passchier, S., Wilson, T.J. & T.S. Paulsen, 1998, Origin of breccias in the CRP-1 core. Terra Antartica 5(3), 401-409.

Cape Roberts Science Team (Sandra Passchier), 1999. Studies from the Cape Roberts Project, Ross Sea, Antarctica-Initial Report on CRP-2/2A. Terra Antartica, 6(1/2), 1-173.

Passchier, S., 1999. Sedimentology and petrology of the Sirius Group. 8* International Symposium on Antarctic Earth Sciences, Wellington, New Zealand, Programme & Abstracts p. 236.

Van der Wateren, F.M., Dunai, T. J., Van Balen, R.T., Klas, W., Verbers, A.L.L.M., Passchier, S., Herpers, U., 1999, Contrasting Neogene denudation histories of different structural regions in the Transantarctic Mountains rift flank constrained by cosmogenic isotope measurements. Global and Planetary Change 23(1-4), 145-172.

Leg 188 Shipboard Scientific Party (Sandra Passchier), 2000. Leg 188 Preliminary Report, Prydz Bay-Cooperation Sea, Antarctica: Glacial History and Paleoceanography. Ocean Drilling Program, College Station, TX, U.S.A.

V lll Passchier, S., 2000 (in press). Soft-sediment deformation features in core from CRP-2/2A, Victoria Land Basin, Antarctica. Terra Antartica, 7 (3/4).

Passchier, S., in press. Provenance and depositional environments of Neogene Sirius Group deposits from the Shackleton and areas, central Transantarctic Mountains, Proceedings of the 8* International Symposium on Antarctic Earth Sciences, Wellington, New Zealand.

Passchier, S., in revision. Provenance of the Sirius Group and other Upper Cenozoic deposits from the Transantarctic Mountains: relation to landscape evolution and ice-sheet history. Sedimentary Geology.

FIELDS OF STUDY

Major Field: Geological Sciences

IX TABLE OF CONTENTS

ABSTRACT...... ii

DEDICATION...... iv

ACKNOWLEDGMENTS...... v

VITA...... vü

LIST OF TABLES...... xv

LIST OF FIGURES...... xviii

CHAPTER 1 INTRODUCTION...... 1 LI Introduction...... 1 1.2 Glacigenic deposits as paleoenvironmental indicators ...... 2 1.3 Purpose of research and objectives ...... 6 1.3.1 The Sirius Group...... 6 1.3.2 The “hothouse” to “icehouse” transition in the McMurdo Sound region ...... 10 1.4 Background and previous work...... 11 1.4.1 Cenozoic Antarctic climates...... 11 1.4.2 Antarctic and stratigraphie records ...... 13 1.4.3 “Warm” climates of the Pliocene ...... 18 1.5 Outline of dissertation and summary of results...... 21

CHAPTER 2 LITHOSTRATIGRAPHY AND GEOLOGICAL SETTING OF THE SIRIUS GROUP AND RELATED UPPER CENOZOIC DEPOSITS...... 24 2.1 Introduction...... 24 2.1.1 Structural framework and basement geology of the Transantarctic Mountains...... 24 2.2 Stratigraphy and geological setting of the Sirius Group ...... 27 2.2.1 Mount Sirius type locality ...... 29 2.2.2 Beardmore Glacier...... 31 2.2.3 ...... 35 2.2.4 Reedy Glacier ...... 38 2.2.5 Victoria Land ...... 42 2.2.6 DVDP-11 and Prospect Mesa ...... 45

CHAPTER 3 DEPOSITIONAL ENVIRONMENTS OF THE SIRIUS GROUP...... 48 3.1 Introduction and purpose ...... 48 3.2 Methods ...... 49 3.3 Results of grain-size and bulk density analysis ...... 50 3.3.1 Mount Sirius...... 52 3.3.2 Beardmore Glacier area ...... 55 3.3.3 Shackleton Glacier area...... 56 3.4 Micromorphological observations ...... 58 3.5 Discussion...... 61 3.6 Glacial thermal regime of the Sirius Group ...... 67 3.6.1 Subglacial conditions ...... 67 3.6.2 Proglacial depositional environments...... 71 3.7 Conclusions...... 74

CHAPTER 4 GEOCHEMISTRY AND GRAIN-SIZES OF GLACIGENIC SEDIMENTS FROM THE SOUTHERN PRINCE ALBERT MOUNTAINS, VICTORIA LAND...... 75 4.1 Introduction...... 76 4.2 Method ...... 78 4.3 Results...... 80 4.3.1 Geochemistry ...... 80 4.3.2 Grain-size analysis ...... 81 4.4 Discussion...... 87 4.4.1 Stratigraphy ...... 87 4.4.2 Glacial history ...... 89 4.5 Conclusions...... 92

CHAPTER 5 MINERALOGY OF THE SIRIUS GROUP AND RELATED UPPER CENOZOIC GLACIAL DEPOSITS FROM THE TRANSANTARCTIC MOUNTAINS...... 94 5.1 Introduction...... 94 5.2 Methods ...... 95 5.2.1 X-ray diffraction analysis (XRD) ...... 95 5.2.2 Sand mineralogy ...... 96 5.3 Results...... 97 5.3.1 X-ray bulk mineralogy ...... 97

XI 5.3.2 Heavy mineral content, composition and distribution ...... 103 5.3.3 Thin section description and detrital modes ...... 112 5.4 Source rocks...... 116 5.4.1 Sirius Group of Victoria Land...... 118 5.4.2 DVDP-11 and Prospect Mesa ...... 119 5.4.3 Mount Sirius...... 121 5.4.4 Beardmore and Shackleton Glacier areas ...... 123 5.4.5 Reedy Glacier area ...... 125 5.5 Provenance and paleodrainage of the Sirius Group ...... 127 5.5.1 Central Transantarctic Mountains...... 129 5.5.2 Victoria Land ...... 130 5.5.3 Reedy Glacier area ...... 133 5.6 Weathering and diagenesis ...... 134 5.7 Discussion...... 137 5.8 Summary and concluding remarks ...... 139

CHAPTER 6 CHEMICAL COMPOSITION OF THE SIRIUS GROUP...... 141 6.1 Introduction...... 141 6.2 Methods ...... 143 6.3 Results...... 144 6.4 Provenance...... 150 6.4.1 Grain-size effects...... 150 6.4.2 AlzTi ratios of diam icts ...... 155 6.4.3 Recycling of sedimentary rocks ...... 158 6.5 Weathering ...... 162 6.6 Discussion...... 166 6.7 Summary and concluding remarks ...... 170

CHAPTER 7 CHEMICAL COMPOSITION OF THE UPPER PLIOCENE BARDIN BLUFFS FORMATION OF THE PAGODROMA GROUP, LAMBERT GRABEN, EAST-ANTARCTICA...... 172 7.1 Introduction...... 172 7.2 Methods ...... 178 7.3 Results...... 178 7.4 Discussion...... 184 7.4.1 Provenance...... 184 7.4.2 Weathering ...... 189

CHAPTER 8 MINERALOGY AND CHEMICAL COMPOSITION OF EOCENE AND PLIOCENE ERRATICS, MOUNT DISCOVERY AREA, MCMURDO SOUND...... 195 8.1 Introduction ...... 195

Xll 8.2 Methods ...... 198 8.2.1 X-ray diffraction ...... 198 8.2.2 Chemical composition ...... 198 8.3 Results...... 199 8.4 Discussion...... 201 8.4.1 Volcaniclastic erratics ...... 201 8.4.2 Eocene mudstone erratic ...... 202

CHAPTER 9 SOFT-SEDIMENT DEFORMATION OF CRP-1 AND CRP-2/2A...... 204 9.1 Introduction ...... 204 9.2 Origin of breccias in CRP-1 ...... 206 9.2.1 Introduction ...... 206 9.2.2 Methodology ...... 207 9.2.3 Breccia characteristics ...... 208 9.2.4 Breccia sequence and stratigraphie boundaries ...... 217 9.2.5 Brecciation mechanisms ...... 219 9.2.6 Models for breccia formation ...... 221 9.2.7 Summary and conclusions ...... 227 9.3 Soft-sediment deformation and brecciation in CRP-2/2A ...... 229 9.3.1 Introduction ...... 229 9.3.2 Characterization of the deformation ...... 231 9.3.3 Stratigraphie distribution of sediment deformation features ...... 238 9.3.4 Discussion...... 241 9.3.5 Comparison to other Cenozoic paleoenvironmental records ...... 251 9.3.6 Conclusions...... 253 9.4 Concluding remarks ...... 254

CHAPTER 10 SYNTHESIS AND CONCLUDING REMARKS...... 257 10.1 Introduction...... 257 10.2 Initiation of glaciation in the Ross Sea sector ...... 257 10.3 Relation of the Sirius Group to glacial and tectonic history ...... 260 10.3.1 Integration of Sirius Group compositional data ...... 260 10.3.2 Deposition of the Sirius Group: timing, paleotopography and paleodrainage ...... 263 10.3.3 Landscape evolution of the Transantarctic Mountains ...... 269 10.4 Pliocene paleoclimate deduced from the Sirius Group and the Bardin Bluffs Formation...... 273 10.4.1 The Bardin Bluffs Formation ...... 273 10.4.2 The Sirius Group...... 273 10.5 Neogene climates and ice-volume in the Antarctic region ...... 275 10.6 Summary and concluding remarks ...... 278

X lll APPENDICES...... 281

A. GRAINSIZE ANALYSIS...... 281 Methods and reproducibility ...... 281 Complete results...... 287

B. BULK X-RAY DIFFRACTION...... 300

C. HEAVY MINERAL ANALYSIS...... 320 Reproducibility ...... 320 Complete results...... 322

D- CHEMICAL ANALYSIS...... 327 ICP-OES analysis and complete results ...... 327 Carbonate analysis methods and complete results ...... 332

BIBLIOGRAPHY...... 336

XIV LIST OF TABLES

Table Page

1.1 Genetic classification and descriptive characteristics of ice-proxintial, terrestrial depositional systems (Position, Processes and Facies modified after Dreimanis 1988). Fabric refers to both clast fabric (Hambrey, 1994) and micnrofabric (Van der Meer, 1993). ‘ As defined by Schack-Pederson (1988) ...... 4

1.2 Number of samples analysed with each laboratory method per location ...... 7

3.1 Micromorphological observations on thin sections of selected Sirims Group sediments. Skelsepic plasmic fabric consists of domains of phyllosilicates oriented parallel to grain surfaces (Van der Meer, 1993) ...... 60

3.2 Classification of and their deposits (after: Boulton, 1972; Shaw, 1977; Paterson, 1994; Evans 1989 and Fitzsimons, 1990). Note that the boundaries between types are gradational and that in reality one glacier may mot necessarily have all the characteristics of one type. From Polar to Temperate tfliere is a decreased preservation of the primary structure of the ice-contact sediments due to increasing efficiency of subglacial erosion and deformation in a deformable bed as well as reworking due to the post-depositional -saturated comdition of the sediment. At the same time, the increasing influence of meltwater oresults in greater thicknesses of sorted sediments and more complex stratigraphies...... 68

4.1 Chemical composition of Sirius Group samples from the Kirkpatrick basalt summit plateaus (>2000 m) compared to geochemical data from th=e Kirkpatrick basalt lava flows (main elements in wt % and trace elements in ppim) ...... 79

4.2 Chemical composition of a selection of till samples from terraces CSirius Group), valleys and ice-cored . (Main elements in wt % and trace elements in ppm) ...... 80

5.1 D-spacing (Â) of peaks used to calculate relative abundances of minerals from X-ray diffractograms ...... 96

5.2 Percentages of heavy minerals in fine-sand heavy-mineral fractions ...... 104

5.3 Detrital modes in 22 thin sections of the Sirius Group ...... 113

6.1 Major element data for the Sirius Group as determined by ICP-OES ...... 145

XV 6.2 Trace element data for the Sirius Group, determined by ICP-OES ...... 146

6.3 Chemical index of alteration (CIA), AIzOs/TiOz ratio and calcium carbonate concentration of Sirius Group sediments. For the definition of the CIA see section 6.5...... 149

6.4 Chemical composition of Sirius Group source rocks...... 153

7.1 Stratigraphy of the Pagodroma Group, northern (After: Hambrey & McKelvey, 2000) ...... 174

7.2 Major element data as determined by ICP-OES analysis for the Bardin Bluffs Formation, Prince Charles Mountains...... 179

7.3 Trace elements as determined by ICP-OES in the Bardin Bluffs Formation, Prince Charles Mountains...... 180

7.4 Carbonate concentration, chemical index of alteration (CIA), and AhOs/TiOn ratio for the Bardin Bluffs Formation, Prince Charles Mountains ...... 181

8.1 Chemical composition of middle to late Eocene mudstone erratic D-1 ...... 200

9.1 Characterization of breccias in CRP-1 ...... 216

9.2 List of prominent deformation features and interpretation in CRP-2/2A ...... 233

10.1 Dominant source rock types in the Sirius Group and the relation to géomorphologie setting ...... 261

10.2 Relation between dominant source rock types, presence of indurated sediment clasts, and microfossil data (youngest age) ...... 262

A1 Variability of sand, mud and clay % induced by sample heterogeneities and method. Five bulk splits of sample PNW95-066 were processes separately and sieved over 63 j 0.m, two bulk splits of PNW95-074 and PNW95-002 were processed separately, and three aliquots of the < 63 p.m fraction of sample PNW95-003 were pipetted ...... 282

A2 Sieve results for the sand fraction and sand-silt-clay results. Clay % determined by pipet analysis ...... 287

A3 Sieve and pipet data (%) for Sirius Group samples...... 288

A4 Coulter Counter test data, sample PNW95-066 ...... 289

AS Complete Coulter Counter data for the Sirius Group...... 291

XVI B1 Peaks used to calculate relative abundances of minerals from X-ray diffractograms— 300

B2 Based on three duplicate analyses the peaks used for the XRD mineralogy studies were chosen. The duplicates were aliquots of the same ground sample prepared, glycolated, and scanned at a different times, so the variability between duplicates is a product of sample heterogeneities, variability in glycolation and instrument drift ...... 301

B3 Results of Intensity (I) and Area (A) counts of bulk X-ray mineralogy analysis...... 304

Cl Complete results of heavy mineral counts ...... 322

D1 Results of ICP-OES analysis for the Sirius Group, Transantarctic Mountains, major elements...... 329

D2 Results of ICP-OES analysis for the Sirius Group, Transantarctic Mountains, trace elements...... 330

D3 Results of ICP-OES analysis for the Pagodroma Group, Prince Charles Mountains, major elements...... 331

D4 Results of ICP-OES analysis for the Pagodroma Group, Prince Charles Mountains, trace elements...... 332

D5 Results of ICP-OES analysis for Eocene erratic D-I from the Mount Discovery coastal ...... 333

D6 Results of carbonate analysis of the Sirius Group...... 335

D7 Results carbonate analysis of the Bardin Bluffs Formation, and D-I(Eocene erratic)...... 336

XVII LIST OF FIGURES

Figure Page

1.1 Laboratory methods and research strategy of compositional studies on the Sirius Group...... 9

1.2 Proxy records of Antarctic ice-volume and sedimentological records in drillcores. Note that the reduction in smectite content, and the first occurrence of ice-rafted debris roughly coincides with a shift of oxygen-isotope ratios near the Eocene/Oligocene boundary. Data from McKelvey (1981), Hambrey et al. (1989a), Hayes & Frakes et al. (1975), Hambrey et al. (1991), Ehrmann (199i;I997), Wise et al. (1992), Robert & Chamley (1992), Wamke et al. (1992). Time scale of Berggren et al. (1985) ...... 12

1.3 Continental break-up of Gondwanaland and development of the circum- Antarctic current, which led to the progressive thermal isolation of Antarctica and subsequent cooling of the region. (After Kennett, 1977) ...... 14

1.4 Map of Antarctica and the location of drill holes with marine records of Cenozoic glaciation and paleoclimates mentioned in the text. The numbers refer to Deep Sea Drilling Project and Ocean Drilling Program legs. Stars represent the locations of recent drillings (1997-2000) by the Ocean Drilling Program and the Cape Roberts drilling Project (CRP) ...... 16

2.1 Devonian-Permian Beacon Supergroup intruded by Jurassic Ferrar Group Rocks, Upper Taylor , Antarctica ...... 25

2.2 Location of Sirius Group outcrops and related Upper Cenozoic deposits (A), and stratigraphy of the most important outcrops and drillholes (B). Sirius Group deposits studied in detail in this work are indicated by a black dot. Small black squares indicate other important but less well studied Sirius Group localities (after Mayewski, 1975 and Stroeven, 1997) ...... 26

2.3 Location of Sirius Group outcrops in the central Transantarctic Mountains (a). Samples studied here are from outcrops indicated by dots. Squares indicate further outcrops not studied here, (b) Geological profile parallel to the Transantarctic Mountain front...... 28

X V I11 2.4 Lithostratigraphy of the type locality of the Sirius Group at Mount Sirius, Bowden Névé. Based on descriptions of Harwood (1986) and Webb et al. (1986b) ...... - ...... 30

2.5 Lithostratigraphy and facies of three sections along an outcrop of the Sirius Group at Oliver Bluffs, Dominion Range (Webb, pers. comm., 1999) ...... 32

2.6 Lithostratigraphy and facies interpretation of the Sirius Group in the Meyer Desert (Dominion Range) and at Bennett Platfoirm (Shackleton Glacier area). Meyer Desert section 13 after McKelvey et al. (1991) and Harwood, pers. comm. (1998). Bennett Platform after Hambrey

2.7 Sirius Group outcrop at Bennett Platform in the Shackleton Glacier area, central Transantarctic Mountains (a). Note laminated facies interbedded with diamictons (a) and dropstones in laminated facies (b). Photos: David Harwood ...... 36

2.8 Dolerite erosion surface with glacial grooves at Roberts Massif (Shackleton Glacier area). Note fault which runs through grooved rock surface and clearly post-dates the erosional phase ...... 37

2.9 Location of Sirius Group deposits in the Reedy Glacier area. Dots indicate outcrops studied in this work, squares indicate additional Sirius Group outcrops...... 39

2.10 Lithostratigraphy and facies interpretation of the Sirius Group at Tillite Spur, Reedy Glacier area. (After Mercer, 1968, and W ilson et al., 1998a). Units in column (A) correspond to Mercer (1968), units in column (B) correspond to Wilson et al. (1998a) ...... 40

2.11 Lithostratigraphy and facies interpretation of the Quartz Hills section of the Sirius Group. (After Mercer, 1968, and Wilson e t al., 1998a) ...... 41

2.12 Location of the Sirius Group and related Upper Cenozoic deposits in Victoria Land that were investigated in this study. Sample: locations are indicated by black dots...... 43

2.13 Outcrops of the Sirius Group at Table Mountain, area. Dry Valleys. Photo: Steve Bohaty ...... 44

2.14 Lithostratigraphy and facies of Dry Valleys Drilling Project hole 11 (DVDP-11) in Taylor Valley. After McKelvey (1981) and Po'well (1981) ...... 46

3.1 Grain-size data from diamictons and stratified facies (a) and bulk density of a selection of Sirius Group diamictons (b) from the central Transantarctic Mountains...... 51

XIX 3.2 Principal lithofacies in the Sirius Group from the central Transantarctic Mountains 52

3.3 Grain-size distributions of Sirius Group sediments from Mount Sirius...... 53

3.4 Grain-size distributions of the Sirius Group at Oliver Bluffs...... 54

3.5 Grain-size distributions of the Sirius Group at section 13, Meyer Desert ...... 55

3.6 Grain-size distributions of Sirius Group diamictons from Roberts Massif...... 56

3.7 Grain-size distributions of the Sirius Group at Bennett Platform. Last three digits of sample numbers are plotted to the right of the lithological column ...... 57

3.8 Thin section micrographs of sample 13-2 of the Meyer Desert Formation, section 13 on the upper bench of the Dominion Range (a) and from the Mount Fleming Upper Till (b). Note the aligned biréfringent bands of phyllosilicates, dolerite fragments, coal and poly-crystalline quartz (Qp) in (a) and biréfringent bands of phyllosilicates and fractured grains in (b). Crossed polarizers. Scale bar is 100 micrometer ...... 59

3.9 Sand-Silt-Clay percentages of Sirius Group diamictons from the central Transantarctic Mountains compared to other diamictons from Antarctica ...... 62

3.10 Sand-Silt-Clay percentages of Sirius Group diamictons from the central Transantarctic Mountains compared to Pleistocene from North America and Europe ...... 63

3.11 (a) Frontal apron of in the Dry Valleys, Antarctica, which is a polar outlet glacier with minor surface melting (b) and sub-freezing basal conditions. Photo (c) shows the basal debris zone, which consists of attenuated ice and debris, in a tunnel at ca. 500 m upglacier (Principal Investigator: Sean Fitzsimons, Univ. of Otago, New Zealand) ...... 66

3.12 Grain-size distributions of Holocene and Recent sediments from the Greenland ice margin. Unpublished data from Koomen et al. (1993) ...... 70

3.13 Glacial processes and vertical profile models for three different types of glacial thermal regimes. General idea based on Eyles et al., 1983, (a) based on Shaw (1977), Lewis et al. (1999), (b) based on Evans (1989), Fitzsimons (1990), Boulton (1972), (c) based on Shaw (1987), Ehlers & Grube (1983), Fyffe (1990), Brennand & Sharpe (1993), Van der Wateren (1994). Note the transition from diamict-dominated tomeltwater-dominated systems ...... 72

4.1 Map of the southern Prince Albert Mountains. BP, Brimstone Peak; FP, Ford Peak; GN Griffin Nunatak; HB, Hughes Bluff; MB a, Morris Basin; MBi, Mount Billing; ON, Outpost Nunatak; RH, Ricker Hills...... 76

XX 4.2 Wt. % TiOn plotted versus wt. % SiOz of till blankets and ice-cored moraines in the Prince Albert Mountains— ...... 81

4.3 Grain-size data of tills from Griffin Nunatak. Note difference between tills from summit and high terraces (Sirius Group) and ice-cored moraines ...... 83

4.4 Grain-size data of tills from Mount Billing, Mount Howard, and Ford Peak (all Sirius Group)...... 84

4.5 Grain-size data of tills from Ricker Hills/ Morris Basin (all Sirius Group)...... 85

4.6 Grain-size data of one glaciofluvial sample and five glaciolacustrine samples from Valley (Ricker Hills)...... 86

5.1 Examples of processed X-ray diffractograms of powdered bulk samples of Sirius Group sediments. Q=quartz; Fsp=feldspar; Cpx=clinopyroxene; Z=zeolite ...... 98

5.2 Relative abundance of quartz, feldspar and pyroxene in Sirius Group and DVDP-11 sediments as determined by bulk XRD analysis: (a) Dry Valleys, (b) and (c)near the Beardmore Glacier, (d) and (e) near the Shackleton Glacier, (f) near the Reedy Glacier. Quartz=4.25 Â; Pyroxene and Feldspar abundances are based on the averages of the peaks in Table 5.1 ...... 99

5.3 Mineral/quartz ratios from bulk XRD analysis from Sirius Group outcrops in the central Transantarctic Mountains...... 100

5.4 Mineral/quartz ratios from XRD bulk mineralogy in Sirius Group successions at Tillite Spur and Quartz Hills, Reedy Glacier area ...... 101

5.5 Mineral/Quartz ratios from XRD bulk mineralogy for the Neogene part of DVDP-11...... 102

5.6 X-ray diffraction analyses of fine-sand heavy-mineral fractions from four Sirius Group locations in the central Transantarctic Mountains and from four depths in the DVDP-11 core. C= corundum (standard), M=mica, Pyx=pyroxene, Cpx=clinopyroxene, Pig=pigeonite, Ca=calcite, Alm=almandine, Hb/Tr=homblende/tremolite, Ac=actinolite ...... 107

5.7 Garnet grains of different origins: (a) rounded garnet grain probably recycled through sedimentary rocks, (b) etched garnet affected by either burial metamorphism or thermal alteration. Both grains are from the Mount Sirius succession from the size fraction 125-250 micrometer, (a) from sample S-9, and (b) from sample S-6 ...... 108

5.8 Relative abundances of clinopyroxenes, amphiboles, and stable minerals plus garnet for Sirius Group and DVDP-11 sediments ...... 109

XXI 5.9 Stratigraphie distribution of the dominant heavy minerals in Oliver Bluffs and Bennett Platform successions of the Sirius Group, central Transantarctic Mountains...... 110

5.10 Stratigraphie distribution of dominant heavy minerals in DVDP-Il, Dry Valleys. Cpx=clinopyroxene ...... I l l

5.11 Thin section micrographs of sample DR5-10 of Oliver Bluffs section 5. (a) In plain light, (b) with crossed polarizers. Note numerous coal fragments, volcanic rock fragments and dolerite fragments. Scale in lower right comer is 50 micrometers...... 114

5.12 Thinsection micrographs of poorly sorted sandstone from Table Mountain, sample TM-7a (a) and the Upper Fleming Till, sample MtFlem#2 (b). The Mount Fleming Upper Till has a high abundance of phyllosilicates, which show no preferred orientation. Note the difference in grain shape and quartz composition between (a) and fb): (a) contains more rounded quartz of various varieties, whereas mineral grains in (b) are mainly angular monocrystalline quartz and feldspar. Qm=monocrystalline quartz, Qp=polycrystalline quartz, Qs=strained quartz, Fsp=feldpar. Scale bar is 100 micrometers ...... 115

5.13 Relative abundances of quartz, feldspar, and pyroxene in Sirius Group and DVDP-11 sediments as determined by XRD analysis, and provenance interpretation. Quartz=4.25 Â; Pyroxene and Feldspar abundances are based on averages of the peaks in Table 5.1 ...... 117

5.14 Distribution of minerals and average grain-sizes along the Transantarctic Mountain front and relation to géomorphologie setting of stratigraphie sections. Heavy minerals are given as percentages of the heavy mineral fraction. Values of Reedy Glacier Sirius Group are calculated from detrital modes ...... 126

5.15 Simplified geologic map of the central Transantarctic Mountains. After Elliot et al. (1996), and Barrett (1972). Ice-flow indicators from: Mayewski (1975); Mayewski & Goldthwait (1985); McKelvey et al. (1991) ...... 128

5.16 Simplified geological map of south Victoria Land (After: Lopatin, 1972). Paleo-ice flow indicators are from: Denton et al. (1984); Brady & McKelvey (1979) and (1983); Stroeven & Prentice (1997) ...... 131

5.17 Stratigraphie distribution of Calcite, Illite-Montmorillonite and Chabazite from XRD bulk mineralogy, expressed as mineral/quartz ratios ...... 135

6.1 Locations of Sirius Group samples selected for ICP-OES chemical analysis are indicated by a black dot. Small black squares indicate other important Sirius Group localities (after Mayewski, 1975 and Stroeven, 1997) ...... 142

xxii 6 . 2 Plots of concentrations of major element versus S1O2. Regression lines are based on “other Sirius Group” data only. Tillite Spur, Quartz Hills and Table Mountain data are excluded, because their compositions form outliers to the general pattern ...... 147

6.3 Plots of concentrations of trace elements versus SiO?. Linear regression lines are based on “other Sirius Group” data only, because Tillite Spur, Quartz Hills and Table Mountain data are outliers to the general pattern ...... 148

6.4 Plots of selected elements versus wt. % in grain-size fractions. The outlier (Cr and Y) is a sample from a siltstone from Oliver Bluffs (sample 5-16). Grain-size data from Chapter 3 (only central Transantarctic Mountains)...... 151

6.5 Ti-Al ratios of the Sirius Group compared to those of source rocks in the Transantarctic Mountains (data from Faure et al., 1974; Siders & Elliot, 1985; Kyle, 1980; Homer, 1992; Roser & Pyne, 1989). (a) source rocks including McMurdo Volcanics, (b) source rocks excluding McMurdo Volcanics...... 155

6 . 6 Plot of wt. % AI2O3 versus wt. % T102 to compare provenance of the Sirius Group to diamictons from ClROS-1 (McMurdo Sound) and to Canadian Pleistocene tills. Values of CIROS-1 diamictons from Roser & Pyne (1989), values for Canadian tills and Average Upper Crust from Young & Nesbitt (1998) and Nesbitt & Young (1984) ...... 156

6.7 Immobile elements in the Sirius Group. Note that the Upper Fleming Till is more enriched in Ba and K 2 0 ...... 158

6.8 Major and trace element ratios indicating grain-size effects and provenance variability in the Sirius Group. Source rock data from Roser & Pyne (1989) ...... 160

6.9 Chemical Index of Alteration (CIA) of Sirius Group deposits compared to Pleistocene and Proterozoic glacial and interglacial sediments of the Canadian Shield (data from Nesbitt & Young, 1984) ...... 162

6.10 Changes in weathering indices and Ti/Al ratios along stratigraphie sections at Bennett Platform (a) and Oliver Bluffs (b)...... 164

6.11 Ternary plots of Ca0*+Na20 - AI 2O3- K2O (CN-A-K) and Ca0*+Na20 +K2O - AI2O3 - FeO+MgO (CNK-A-FM) systems. In the upper diagrams the Sirius Group is plotted, in the lower diagrams source rocks within the Transantarctic Mountains are plotted with theoretical weathering trends (Nesbitt & Young, 1984; 1989). The linear trends in the upper diagrams are inconsistent with weathering being the sole control of the variability in the chemical composition of the Sirius Group. Recycling of sedimentary rocks with a more evolved chemical composition is a possibility ...... 165

xxiu 6.12 Decreasing CIA with time in sediments from the Victoria Land basin. Data from Roser & Pyne (1989) and Krissek & Kyle (1998). Figure after Krissek & Kyle (1998) ...... 167

6.13 Range of CIAs within the Sirius Group. The spread in CIAs suggests that significant differences in degree of weathering occurred in the parent rocks of the sediments. The presence of reworked sediment clasts and Eocene foraminifera suggest that the highest CIAs are partially caused by reworking of Eocene marine sediments. Slightly lower CIAs occur in sediments of Mount Sirius where Oligocene marine microfossils were found. Only Neogene microfossils occur in the Sirius Group sediments with the lowest CIAs. Microfossil data from Webb et al. (1984) and Harwood (1986) ...... 169

7.1 Location of the Pagodroma Gorge in the Prince Charles Mountains ...... 173

7.2 Map of Pagodroma Gorge with sample localities. Samples in this study were taken from locations 1 through 6. See Figure 7.3 for detailed stratigraphie sections of sample localities. From: Whitehead & McKelvey, in prep ...... 175

7.3 Detailed stratigraphie logs of sample locations within the Bardin Bluffs Formation. From Whitehead & McKelvey, in prep ...... 177

7.4 Plots of major elements versus wt. % SiOi for the Bardin Bluffs Formation. Note grain-size effect on composition ...... 182

7.5 Plots of trace elements versus wt. % SiOz for the Bardin Bluffs Formation. Note grain-size effect on trace element composition ...... 183

7.6 Plot of wt. % CaCOs vs. Sr in ppm. The good correlation suggests that Sr resides in the Ca-carbonate ...... 184

7.7 Plot illustrating the uniform Al/Ti ratios. Minor variability is mainly cause by grain-size effects (cf. Young and Nesbitt, 1998). Solid circles are mudstones, open squares are sandy mudstones and sandstones. A) shows the relationship between Bardin Bluffs Fm Ti-Al contents and Ti-Al contents of basement rocks from the northern Prince Charles Mountains (data from Munksgaard et al, 1992; Sheraton et al., 1995; Kinny et al., 1997; Zhao et al., 1997). B) illustrates the effects of weathering and hydraulic sorting on the Ti-Al contents of the sediments Samples plotting to the right of the dashed line are enriched in A1 with respect to the source rocks due to weathering, and concentration of weathering products in the mudstones by hydraulic sorting ...... 186

7.8 Plots illustrating the relation between immobile elements Ti, Al and Cr. A) Cr-Ti plot of Bardin Bluffs Fm and basement rocks from the northern Prince Charles Mountains. Note that the vertical scale is logarithmic. B) The Bardin Bluffs Fm forms a tight cluster on a 3 variable plot, suggesting that Ti, Al, and Cr

XXIV are derived from the same source in all sediments. Note the similarity in immobile element composition of the Bardin Bluffs Fm and pellitic gneiss. (Basement chemical data from: Munksgaard et al., 1992; Sheraton et al., 1995; Kinny et al., 1997; Zhao et al., 1997) ...... 187

7.9 Plot illustrating concentration of Zr in the sand fraction and concentration of Ti and Al in the finer fractions ...... 188

7.10 Comparison of CIAs of Member 1 (dark gray) and Member 2 (light gray). Most CIAs fall between 71 and 75. There appears to be no significant difference between members, although the number of samples is too low to draw any definite conclusions...... 190

7.11 Plots of CIAs vs. ALOg/TiOz ratios show the variability in CIAs with relatively constant AlzOz/TiOz ratios ...... 190

7.12 Ternary plots of Ca0*4-Na%0 - AI 2O3- K%0 (CN-A-K) and CaO^+NaoO +K2O - AI2O3 - FeO+MgO (CNK-A-FM) systems. In the upper diagrams the Bardin Bluffs Fm is plotted, in the lower diagrams source rocks within the northern Prince Charles Mountains are plotted with estimated weathering trends (Nesbitt & Young, 1984; 1989) ...... 192

8 .1 Location of the Mount Discovery area ...... 196

8.2 Eocene sandstone erratic from the Mount Discovery coastal moraine. Note mollusc fauna...... 197

8.3 Bulk X-ray diffractogram of a volcaniclastic diamict erratic of Pliocene age. Unprocessed diffractogram in upper right comer ...... 199

8.4 Bulk X-ray diffractogram of a fissile mudstone of Eocene age (erratic D-1). Fsp=feldspars ...... 200

9.1 Breccia types in the CRP-1 core, (a) Rubble breccia; characterized by small, subrounded fragments of a variety of grain sizes. Note relatively abrupt change to smaller grain size of rubble breccia across planar boundaries with chaotic breccia above and below; (b) Chaotic breccia ; with large, fractured clasts intermixed with fine-grained brecciated material; (c) Crackle breccia; characterized by network of cracks separating angular fragments that have not been displaced; (d) Matrix-supported breccia, with pebble-sized fragments surrounded by a fine-grained matrix. Core diameter is 61 mm ...... 210

9.2 Boundaries between breccia types in the CRP-1 core, (a) Inclined boundary (white dashed line); (b) Irregular boundary (white dashed line); (c) Gradational and Sharp boundaries. Most boundaries between breccia textures are sharp and planar. Note stacked nature of breccia types in 2c, where crackle breccia (cr) at base of interval passes upward into chaotic (ch) and rubble (r) breccias. Core diameter is 61 mm ...... 2 1 1

XXV 9.3 Detailed 1:20 log of breccia 2, below the Quatemary-Miocene boundary in CRP-1. (Key: cr=crackle, mo=mosaic, ch=chaotic, and r= rubble breccia.) Two types of associations are particularly cononon: 1) crackle breccias adjacent to rubble breccias with a sharp boundary in between the types; or, 2) craclde breccias grading upwards or downwards into chaotic textures with large fractured clasts in a rubbly groundmass, which in turn pass into rubble breccias. No changes in lithology occur at the boundaries between textures. The crackle breccias clearly represent in situ brecciation with minimal movement of clasts. The smaller and more rounded fragments in the rubble breccias point to grain size reduction and abrasion of clasts, suggesting some degree of transport ...... 212

9.4 Detailed 1:20 log of breccia 4, CRP-1. (Key: cr=crackle, mo=mosaic, ch=chaotic, and r= rubble breccia.) Crackle breccias, indicative ofin situ brecciation, are of minor importance. Local crackle breccias pass upwards into rubble and chaotic breccias, which grade into soft-sediment deformation features. Boundaries between breccia types in this interval are mainly gradational ...... 213

9.5 Thin section photomicrograph of sediment-filled crack at 44.49 - 44.64 mbsf. The sediment-filled crack is ~1 mm across. Note wall-parallel stratification and change in grain size of clastic fill along the length of the crack ...... 214

9.6 Soft-sediment deformation and brecciation at -67 mbsf. Contorted lamination may represent soft-sediment folding, which is overprinted by brecciation (crackle breccia). The upper picture is a positive CoreScan image, the lower picture is a negative image of the same interval, where open cracks appear white, highlighting the breccia texture...... 215

9.7 Downcore distribution of breccias and other deformation features in the CRP-1 core. Numbers denote breccia units discussed in the text and described in Table 9.1. Lithologie log and sequence boundaries after Cape Roberts Science Team (1998). Note: single breccia in Quaternary section; association of thickest breccia units (2, 5) with sequence boundaries overlain by diamictites; and the association of brecciated intervals with other deformation features from 55 mbsf to the base of the core ...... 218

9.8 Map of the Ross Sea area, showing the location of the CRP drillsites, the CIROS-1 drillsite and the ice-free Transantarctic Mountains ...... 229

9.9 Breccias from CRP-2/2A: a) crackle breccia at ca. 56 mbsf; b) mosaic breccia at ca. 505 mbsf; and c) deformed breccia at ca. 45 mbsf. Note the thrusted fabric in (c)...... 232

9.10 Clastic dikes from CRP-2/2A: a) wedge-shaped dyke with irregular margins, filled with poorly-sorted sandstone at 363 mbsf; b) pyrite and carbonate cemented sandstone dyke with planar margins at 446 mbsf; c) brecciated medium sandstone dyke at 145 mbsf. The fabric of the brecciated mudstone (b) suggests injection occurred from below ...... 234

xxvi 9.11 Clastic dikes from CRP-1: a) thin section (left) and interpretation (right) of clastic dyke at 133 mbsf; and b) thin section (left) and interpretation (right) of clastic dyke at 133 mbsf. The dikes are filled with concentrically-zoned carbonate microconcretions of the types encountered in fractures in other parts of the core (Baker and Fielding, 1998). Microconcretions in sediments of the upper dyke are smaller than those in the lower dyke. The presence of a diamictite intraclast with small micro-concretions in the upper dyke (a) suggests that the pebble was derived from a lower level in the stratigraphie column. Downlapping laminae in the lower dyke suggest settling of sediment occurred as water pressures diminished. However, the downlapping laminae are truncated, suggesting that the dyke was reactivated after deposition of the laminae. Rapid fluctuations in water pressures in dyke systems are characteristic of subglacial and proglacial environments (Van der Meer et al., 1994, Von Brunn and Talbot, 1986) ...... 235

9.12 Microfaults from CRP-2/2A: a) micro-block faulting in laminated sediments at 319 mbsf; b) reverse microfaulting in subglacially deformed sediments at 363 mbsf; c) normal microfaulting with carbonate cemented fault plane at 543 mbsf. Faults a) and c) are compatible with an extensional stress regime, however do not necessarily share the same origin. Fault b) may have formed in a compressional stress-regime associated with glacial overriding ...... 236

9.13 Shear zones from CRP-2/2A: a) small-scale cataclastic shear zone, with an intraclast of a microlaminated facies and other intraformational pebbles and granules, in situ crackle breccia below the shear zone; b) sheared lower part of diamictite at 306 mbsf with boudinage and intraclast at the top; c) anastomosing dilated laminae at 614 mbsf...... 237

9.14. Contorted bedding from CRP-2/2A at a) 513 mbsf; and b) 612 mbsf...... 238

9.15 Downcore distribution of ductile and brittle sediment deformation features. Clastic dikes and microfaults are recorded as the number of features per 10 m intervals of core. In the interpretation column d, p, and g refer to the proximity of the grounding line with respect to the drillsite: d=distal, p=proximal, g=grounded. Lithological column and chronology from Cape Roberts Science Team (1999). Ar/Ar ages are from McIntosh (in press). Eustatic curve from Haq et al. (1987). Mass-gravity flows in the early Oligocene correspond to sea level highstands of the Haq et al. (1987) eustatic curve. The early Oligocene lowstand corresponds to the initiation of abundant diamictite deposition culminating into grounding of ice on the drillsite. The upper Oligocene section does not show any evidence of grounded ice near the drillsite, but a sea level lowstand does occur. A highstand characterizes the early Miocene, but grounded ice occasionally reached the drillsite ...... 240

9.16 Hypotheses for the formation of meltwater-related glaciotectonic structures described from the Cape Roberts drillcores (partly after Boulton and Caban,

XXVII 1995, Rijsdijk et al., 1999). (a) Free flow of meltwater, no deformation, (b) Hydrofracturing of subglacial and proglacial sand and silt. Subglacial sediment mobilisation may also occur into existing tensile fractures, formed due to ice loading. In the proglacial area low permeability of proglacial sediments or the presence of permafrost may impede free drainage when water depths are shallow. (c) Subglacial deformation of fractured sediments and cataclastic shearing. (d) Proglacial hydrofracturing and redeposition of mudstone breccia. (e) Subglacial and proglacial hydrofracturing in relation to diamictites. The high tensile strength of diamicites apparently leads to the formation of thicker clastic dikes rather than small hairline fractures as observed in sand and mud (b and d) ...... 245

9.17 Simplified stratigraphie distribution of breccias in the Oligocene- Miocene sections of McMurdo Sound cores. Note that near in situ breccias predominate in the late early Miocene, whereas mudstone breccias are more abundant in the earliest Miocene to Oligocene. CRP-1 data are from Passchier et al. (1998), CIROS-1 data are from Hambrey et al. (1989a) and Fielding et al. (1997) ...... 246

9.18 Cartoon explaining the development of unconformities by subglacial erosion and deformation. Left: complete sequence with diamicts at the bottom, which are either directly deposited by ice or by , fining upward into sandstones and mudstones deposited in progressively more distal depositional environments. Right: characteristics of a sequence truncated by glacial erosion. Note that the diamict overlying the unconformity may have been deposited upon glacial retreat, so that fabric studies on diamicts fail to identify ice-grounding episodes ...... 255

10.1 The Eocene/Oligocene transition in (A) the clay mineralogy and (B) the bulk chemistry record of the ClROS-1 drillcore, Victoria Land basin, McMurdo Sound. (A) from Ehrmann, 1997; (B) data from Krissek & Kyle, 1998 ...... 258

10.2 Subglacial topography of Antarctica. Simplified after BEDMAP (British Antarctic Survey, 2000) ...... 264

10.3 (A) First phase of ice-sheet expansion and Sirius Group deposition incorporating open marine microfaunas and floras and chemically weathered sedimentary rocks. (B) represents a later phase of Sirius Group deposition in the Neogene with a higher , lower contemporary chemical weathering rates, and deposition in glacial troughs. TAM= Transantarctic Mountains...... 268

10.4 Model of landscape evolution in the Transantarctic Mountains. A=?Mid- Oligocene/early Miocene overriding of the Transantarctic Mountains. B=Development of overdeepened glacial troughs due to uplift and structural segmentation of the Transantarctic Mountains, deposition in terrestrial glacial environments as well as in Qords. C=Topographic distribution of the Sirius

XXVIU Group in relation to the present morphology of the Transantarctic Mountains ...... 271

10.5 Peat marsh vegetation and dwarf willow (Betiila nana) covering higher ground near the Greenland ice margin, ca. 67oN (a). Dwarf willow grows on the lee side of topographic features near the ice margin because of harsh cold and dry conditions caused by the katabatic wind (b). Nothofagus from the Sirius Group may have grown in a similar enviroment (cf. Francis &Hill, 1996) ...... 274

10.5 Pliocene deposits of Antarctica (data from Webb & Harwood, 1991; Pickard et al., 1988; Hambrey & McKelvey, 2000; Fleming & Barron, 1996; Ishman & Rieck, 1992; Prentice et al., 1993; 1999) compared to the oxygen isotope stratigraphy of planktonic foraminifera at Site 704 in the (Hodell & Venz, 1992) and a eustatic sea level model from the Atlantic coastal plain (Krantz, 1991). The bars representing Pliocene Antarctic sediments delimit the possible time range of each deposit and do not indicate continuous deposition. Note that evidence for significant wanning and deglaciation and marine incursion is concentrated in the interval between 4.5 and 4.0 Ma..Other evidence of warming and high eustatic sea level from proxy records occurs between 3.9 and 3.2 M a ...... 276

Al Logarithmic plot of particle diameter (ixm) versus number of particles in a test run of four subsamples of the same sample...... 284

A2 Three splits of the same subsample were run three times to evaluate the precision of the Coulter Counter measurements...... 285

A3 Grain-size histogram after running three subsamples of sample PNW95-066 by Coulter Counter. The < 4 phi size fractions were determined by sieve analysis, the >9 phi fraction by pipet analysis (only one subsample). Note poor reproducibility in the 4-5 phi grain-size class by Coulter Counter in the test...... 285

Cl From a test count on sample PNW95-057 it was concluded that 300-350 counts would be sufficient to obtain reproducible data...... 321

D1 Calibration curve for the reaction time of CaCO] in contact with 10% HCl ...... 333

xxix CHAPTER 1

INTRODUCTION

1.1 Introduction

In recent years much interest has developed into the effects of global wanning on the stability of polar ice-sheets. Ice-core records show that the polar regions are susceptible to abrupt short-term changes in oceanic and atmospheric circulation patterns (e.g. Johnson et al., 1992).

The history of high-latitude climates and glaciation is essential in understanding the nature and causes of Cenozoic global climate change due to the effects of Antarctic glaciation on deep-water formation and latitudinal thermal gradients.

During the Cenozoic the global climate system experienced several profound changes.

The early Eocene is characterized as a “hothouse world”, an interval of global warmth with high

CO2 levels, elevated deep-sea temperatures, and low latitudinal thermal gradients. In contrast, the

Late Cenozoic “icehouse world” experienced continental glaciation in the polar regions, cool deep-sea temperatures, and high latitudinal thermal gradients. The role of the development of the

East Antarctic Ice Sheet in global cooling is a topic of present interest. In addition, much controversy exists about the stability of high-latitude climates and ice-sheets in the Neogene icehouse world (e.g. Kennett & Hodell, 1993). Although evidence of global warmth in the Pliocene is accumulating (Dowsett et aL, 1996), the effects of global warmth on Antarctic

climates and ice-sheet stability are debated (Webb & Harwood, 1991; Denton et al., 1993).

Here the lithostratigraphy, sedimentology, and composition of glacial sediments from the

Antarctic continental margin are examined in order to reconstruct high-latitude paleoclimatic

conditions in critical intervals of the Cenozoic. The main focus of this study is on the

reconstruction of Antarctic climates and ice extent during the hothouse to icehouse transition

(Eocene-Oligocene transition) and the Pliocene period of global warmth.

1.2. Glacigenic deposits as paleoenvironmental indicators

Glacial sediments have been the subject of geological studies since the 18th century, because large areas of mid-latitude land surfaces of Europe and North America are covered by till-sheets. Initially, ore prospecting sparked an interest in the lithological and raineralogical composition of till, and the physical properties of tills were investigated to evaluate their suitability as foundations for buildings and roads. Due to the growing interest in climate change large-scale mapping of the ice-marginal extent of the Pleistocene ice-sheets and the classification of genetic varieties of tills became the primary research objectives of the scientific community

(Goldthwait, 1971). Compositional studies were undertaken and the data were used to correlate stratigraphie successions across wide geographic areas (e.g. Christiansen, 1971; Johnson et al.,

1971; Kempton et al., 1971). Since subglacial deposition is not directly observable, geologists relied on the properties of ancient subglacial deposits to extract information about the depositional processes underneath glaciers. It is obvious that this approach might lead to unlimited speculations (Carol, 1943). Therefore, studies of modem glacial environments attempt to constrain the depositional processes and to define the physical parameters of glacial flow (e.g. Boulton, 1972; Paterson, 1994). Excavations of tunnels and cavities into glaciers and drilling through glaciers and ice-sheets has provided evidence for the presence of water-saturated deformable sediment beds beneath wet-based glaciers (Alley, 1986; Boulton and Hindmarsh,

1987; Tulaczyket al., 1998) and for the occurrence of basal freezing and limited brittle bed deformation beneath dry-based glaciers (Echelmeyer & Wang, 1987; Fitzsimons et al., 1999).

At the same time, a comprehensive genetic terminology and classification system was developed for ice-contact deposits (Dreimanis, 1988). Although a genetic classification may be useful for mapping glacial deposits in terrestrial lowland areas where till sheets occur in geographic relation to other glacigenic deposits, it is of limited use in stratigraphie interpretations of complex subglacial and proglacial facies associations with limited geographic extent, or in the study of drillcores.

The greatest challenge in the classification of ice-proximal sediments is to distinguish primary tills from diamicts with a different origin, such as non-glacial debris flows and redeposited (secondary) tills. Terminology and criteria used in this dissertation to distinguish varieties of glacigenic sediments are listed in Table 1.1. The term diamicton will be used as a descriptive, non-genetic term for unlithified poorly sorted, matrix-supported, gravel-rich facies.

The term diamictite designates the lithified variety, and the term diamict represents both lithified and unlithified sediments. Eyles et al. (1983) used vertical facies associations for the interpretation of thick stratigraphie sections of glacigenic sediments. In their classification of glacigenic sediments the genetic term till is replaced by the descriptive term diamict. Since some of the deposits of the Sirius Group represent thick successions with limited geographic coverage, a basin analysis approach, such as that of Eyles et al. (1983), is appropriate in these cases. Lithological Position Processes Facies characteristics Fabric

ICE-MARGINAL melting melt-out till massive diamict random sublimation sublimation till stratified diamict ? -frontal debris flow flow till stratified/massive weak to strong -lateral diamict orientation meltwater fluvlo-glacial stratified sands reworking sediment and gravels

melting melt-out till stratified diamict random SUPRAGLACIAL sublimation sublimation till stratified diamict ? debris flow flow till stratlfled/masslve weak to strong diamict orientation rock fall breccia random

erosion SUBGLACIAL lodgement lodgement till massive diamict strong orientation deformation glacltectoniteV brecciated sediment’ deformation till / massive diamict strong orientation melting melt-out till massive diamict random

Table 1.1. Genetic classification and descriptive eharaeteristies of ice-proximal, terrestrial, depositional systems (Position, Processes and Facies modified after Dreimanis 1988). Fabric refers to both clast fabric (Hambrey, 1994) and microfabrie (Van der Meer, 1993). ' As defined by Scliaek-Pedersen (1988). In this study, ice-marginal and subglacial deposits provide the bases for paleoenvironmental reconstructions. Since the facies architecture and sediment properties of ice- proximal and ice-contact deposits are strongly controlled by subglacial thermal regime and proglacial climate, glacigenic deposits provide good paleoclimate indicators. Reference sections of vertical facies profiles and structural criteria are available for a number of terrestrial ice- marginal paleoclimatic conditions (e.g. Shaw, 1977; Eyles et al., 1983; Fitzsimons, 1990) and an extensive body of knowledge has been built on the chemical and physical properties of different types of glacigenic sediments (e.g., Boulton, 1978; Nesbitt & Young, 1996; Young and Nesbitt,

1998).

Grain-size data are useful for characterizing depositional environments of proglacial

(stratified) deposits (e.g., Ehlers & Grube, 1983), whereas the grain-size of a till matrix provides evidence of the depositional setting (Boulton, 1978) and recycling of sedimentary rocks

(Haidorsen, 1981; 1989). In determining the paleo-weathering regime in pre-glacial and glacial environments, weathering indices such as the Chemical Index of Alteration (CIA), calculated from chemical analyses of whole-rock samples, are effective (Nesbitt & Young, 1982; Krissek &

Kyle, 1998).

The composition of sand-sized heavy mineral assemblages has proven to be especially useful for the determining the provenance of glacigenic deposits (Gravenor, 1979; Gwyn &

Dreimanis, 1979; Riezebos, 1983; Hofer & Szabo, 1993; Ehrmann & Polozek, 1998). In combination with detrital modes from point-counted thin-sections, the source rock-type of the sand-sized fraction can be readily identified. Bulk X-ray mineralogy and whole-rock chemistry are not standard techniques used in the investigation of terrestrial glacial deposits, but these methods have been applied successfully to marine sediments from high-latitude regions that are subject to high terrigenous input (Krissek, 1989; Roser & Pyne, 1989; Bohrmann & Ehrmann, 1991, Saito, 1998; Forsberg et al., 1999). The chemical index of alteration (CIA) calculated from

the concentrations of major elements is especially useful in determining weathering intensities

prior to glacial erosion and after deposition of the sediments (e.g. Nesbitt & Young, 1982).

1.3 Purpose of research and objectives

1.3.1 The Sirius Group

The current East-Antarctic ice cover is a stable, dominantly dry-based terrestrial ice

sheet. Climatic conditions at sea level are arid-polar, resulting in limited surface melting at the

termini of glaciers (Shaw, 1977; Lewis et al., 1999). Tills, and glaciofluvial and glaciolacustrine

deposits of the Sirius Group are indicative of less extreme polar climate conditions, and are found

at high elevations in the Transantarctic Mountains (Mercer, 1968, 1972; Mayewski & Goldthwait,

1985; Webb et al., 1984, McKelvey et al., 1991). Although the age of these deposits is controversial (Stroeven et al., 1998; Harwood & Webb, 1998), the sediments provide a window into a period in Antarctica’s Neogene glacial history when climates were warmer than present.

The Sirius Group crops out at high elevation in the Transantarctic Mountains and may have experienced considerable tectonic uplift after deposition (e.g. Webb et al., 1996a).

Compositional studies of the sediments can be used to evaluate the relation between the tectonic and the glacial histories of the Transantarctic Mountains. Recent studies show that tectonic processes have a considerably effect on the drainage patterns of glaciers in Victoria Land (Salvini

& Storti, 1999; Wilson, 1999), and modeling experiments suggest that the development of glacial troughs prohibited Antarctic ice from overflowing the mountain range (Kerr & Huybrechts,

1999). Most of the Sirius Group outcrops have been regarded as the product of one phase of glaciation (Webb et al., 1984; Webb & Harwood, 1991; Denton et al., 1984, 1993). However, a SIRIUS GROUP localities Dry Valieys PAMs*PCMs TOTAL

DVDP Bardin (this IS QH RM BP MD OB MS MFI TM MFe Corel 1 PM Bluffs Fm study)

Bulk XRD 20 16 20 21 9 34 12 2 4 1 11 1 151 Chemical analysis 3 2 2 8 3 6 4 1 2 1 35 20 87 CaCOa 3 2 2 8 3 6 4 1 2 1 20 52 Graln-slze x’ x’ 8 18 4 10 12 x' x' X^ X® X® 35 87 Bulk density 5 7 1 2 15 Heavy mineral analysis 6 9 3 9 4 1 1 8 1 X^ 42 Thin section analysis 2 2 7 4 2 1 2 1 1 X® 22

X= available from other studies

’ Wilson et al., 1998 ^ Stroeven & Prentice, 1996 ® Barrett & Powell, 1982 Barrett, 1998 ® Powell, 1981 ® Prentice et al., 1993 ^Bardin, 1982 ® Dickinson, 1998; Van der Meer et al., 1998

* 116 samples were analyzed for this study but only Sirius Group deposits and other glacigenic deposits relevant to this study are discussed

Table 1.2. Number of samples analyzed with each laboratory method per location. TS=Tillite Spur; QH=Quai1z Hills; RM=Robeils Massif; BP=Bennett Platform; MD= Meyer Desert, section 13; 0B=01iver Bluffs; MS=Mount Sirius; MFl=Mount Fleming; TM=Table Mountain; MFe=Mount Feather; DVDP=Dry Valley Drilling Project; PM=Prospect Mesa; PAMs=Prince Albert Mountains in Victoria Land. PCMs=Prince Charles Mountains in the Prydz Bay/Lambeit Graben area of East Antarctica. systematic study to determine the origin and depositional environments of the Sirius Group has

not yet been conducted. One of the possibilities investigated in this dissertation is whether the

Sirius Group may represent deposits of widely varying ages, origins and depositional settings.

As part of this dissertation a systematic laboratory study of the grain-size, mineralogical

and chemical composition of the Sirius Group and related deposits is undertaken (Table 1.2). The

chemical composition of the Pliocene Bardin Bluffs Formation, the East-Antarctic equivalent of

the Sirius Group from the Lambert Graben, was determined to provide Pliocene paleo­

environmental data from another part of the Antarctic continental margin for comparison with the

Sirius Group.

The data are integrated to address the following issues (Figure 1.1):

1) Depositional enviromnents and glacial thermal regime recorded by the Sirius Group;

2) Paleoenviromnental conditions and paleoclimate before and during deposition of the

Sirius Group;

3) Provenance of the Sirius Group and related Upper Cenozoic deposits and their relation to

ice-sheet drainage patterns and the tectonic history of the Transantarctic Mountains;

4) Cratonic sediment sources and recycling of sediments from East Antarctica;

5) Pliocene Antarctic paleoclimates;

6) Comparison of the Sirius Group to other deposits resulting from continental glaciations,

such as Northern Hemisphere Pleistocene glaciations;

Attempts to establish a chronostratigraphic framework for the Sirius Group have been solely based on paleontological data, yet microfossils are rare in the sediments (Harwood, 1986;

Webb and Harwood, 1991). The main source of data on the paleoenvironment and origin of the

Sirius Group is the Oliver Bluffs succession in the central Transantarctic Mountains (Harwood & bulk ball X-ray sam ple mill m ineralogy

sand bulk sieve thin prov enance provenance 4 phi 1 section

SOURCE ROCK mud sand 1 phi rotap I m ineral ICP chemical AND SEDIMENT fraction fraction sieving C a C O separation analysis TRANSPORT index V d analysis X

silt couiter fraction counter

PALEOENVIRONMENT OF DEPOSITION

Figure 1.1. Laboratory methods and research strategy of compositional studies on the Sirius Group. Webb, 1998). If the diatoms are recycled from marine basins behind the Transantarctic

Mountains the Oliver Bluffs depositional environment represents ice-marginal conditions some

time since 3.8 Ma, whereas deglaciation occurred before 3.8 Ma. The idea of a Sirius Group

deposited during a single glacial phase stems from Mayewski and Goldthwait (1985) who

commented on tlie uniform nature of the deposits and attributed the Sirius Group (then Sirius

Formation) to the Queen Maud glaciation (> 4.2 Ma). Few studies regarding lithostratigraphy and

provenance of the Sirius Group (Faure et al., 1981, 1983, 1995; Wilson et al., 1998a; Hambrey et

al., in prep) have been conducted to test whether the Sirius Group is indeed the result of one

phase of continental glaciation.

1.3.2 The hothouse to icehouse transition in the McMurdo Sound region

Records of the Eocene-Oligocene hothouse to icehouse transition are sparse on the

Antarctic margin and it has been recovered in only one drillhole (CIROS-1) in McMurdo Sound

(Barrett et al., 1989; Wilson et al., 1998b). Therefore, Eocene erratics from McMurdo Sound

coastal moraines provide an important additional source of information about Antarctic

continental paleoclimates before the establishment of the Oligocene to Recent icehouse Earth

(Harwood & Levy, 2000). In this study the chemical composition of a mid-late Eocene mudstone erratic is determined and used to estimate continental weathering intensities prior to the development of large ice-sheets during the Eocene to Oligocene transition.

Recent drilling sited on fast-ice at Cape Roberts in McMurdo Sound also aimed at recovering strata spanning the hothouse to icehouse transition. The Cape Roberts Drilling Project had the following specific objectives: 1) To date the initiation of East Antarctic glaciation and to define the link between glacial fluctuations and sea level in the Ross Sea sector, and 2) To date

1 0 the initiation of rifting and the development of the Victoria Land basin (Barrett, 1997). An

integrated sequence stratigraphy was developed for Holes CRP-1 (1998) and CRP-2/2A (1999)

based on the lithostratigraphy, biostratigraphy, magnetostratigraphy, and wire-line logs.

As part of this dissertation, the soft-sediment deformation features in Oligocene through

Miocene strata of the CRP-1 and CRP-2/2A drillholes were described and interpreted in terms of

depositional environments (Passchier et al., 1998; Passchier, in press). One of the main issues in

the interpretation of glacial sequences is the identification of episodes of grounded ice, and in

particular when grounded ice first reached the Ross Sea continental shelf. A combination of

vertical facies associations and diagnostic deformation structures was used to identify evidence of grounded ice in the CRP-1 and CRP-2/2A drillholes.

1.4 Background and previous work

1.4.1 Cenozoic Antarctic climates

Until the mid-1980s the history of East Antarctic glaciation was mainly described by the use of climate proxies derived from low-latitude marine records (Matthews, 1984). Benthic foraminifera precipitate carbonate skeletons, and the changing isotopic composition of seawater is recorded in the stable isotope composition of calcareous tests. The oceans are enriched in the heavy isotope of oxygen, when much of the light isotope of oxygen, is stored in ice- sheets. The stable isotope ratio Ô ‘*0 is higher during glacial times, but decreases when much is released to the oceans by the melting of glaciers and ice-sheets. Therefore, variations of 5 ‘*0 through time can be interpreted in terms of changing ice-volume. However, the use of this

II Ross Sea South Indian South Ocean Atlantic Eustatic change Oxygen isotope DVDP CIROS Leg Leg Leg Leg Leg Leg from coastal oniap ratios 11 1 28 119 119 120 113 114

200 100 m' 0.0 ppm TiëiE (OC,

Ma

-CZ)

high ? low CZ) %

> (Atlantic Benthic Foraminifera) I

no record (Ico-freo tomperaturo scalo “C) ice-proximal sediments

SEA LEVEL •'f l smectite dominant ciay mineral 1,0 ppm ft"0 2,0 first occurrence of IRD

Figure 1,2. Proxy records of Antarctic ice-volume and sedimentological records in drillcores. Note that the reduction in smectite content, and the first occurrence of ice-rafted debris roughly coincides with a shift of oxygen-isotope ratios near the Eocene- Oligocene boundary. Data from McKelvey (1981), Hambrey et al. (1989), Hayes and Frakes et al. (1975), Hambrey et al. (1991), Ehmiann (1991; 1997), Wise et al. (1992), Robert and Chamley ( 1992),Warnkc ct al.(1992). Time scale of Berggren et al. (1985). parameter has limitations in that it is also sensitive to changes in water temperature and salinity of

the oceans. The other proxy of Cenozoic ice-volume change that has been widely used is the

eustatic curve based on the relative coastal onlap at passive continental margins of Europe and the

United States. For the new generation of eustatic curves, the magnitude of global coastal onlap

has been reconstructed from the sequence stratigraphy of both outcrop and subsurface data (Haq

et al., 1987). However, the chronology of eustatic events reconstructed from isotope records and

sea level curves is inconsistent, and the magnitude of glacio-eustatic change estimated from them

varies significantly (Figure 1.2). Therefore, it is important to compare these proxies with the more

direct geological record of Antarctic glaciation.

The oxygen isotope records of benthic foraminifera are characterized by three stepwise

shifts in ô (Figure 1.2) since the early Eocene indicating three episodes of rapid cooling

superimposed on a longterm cooling trend (Flower, 1999). The three cooling steps take place at

the Eocene/Oligocene boundary (ca. 33.6 Ma), in the middle Miocene (ca. 13.8 Ma) and in the

late Pliocene (ca. 2.7 Ma). The initial growth of polar ice-sheets in Antarctica was probably

related to the effects of plate tectonics on ocean circulation. The separation of Antarctica from

Australia and South America opened a pathway for a circum-polar current, which may have led to

the progressive thermal isolation of Antarctica (Kennett, 1977; Figure 1.3).

1.4.2 Antarctic and subantarctic stratigraphie records

Direct evidence of Antarctic glaciation can be found in a number of on-shore and off­

shore records in the subantarctic and Antarctic region (Figure 1.2). Both proximal and distal xecords of Antarctic glaciation exist, each of which has its specific advantages and limitations.

13 Earliest Miocene 21 million years ago

Late Eocene 36 million years ago

Cretaceous-Tertiary boundary 65 million years ago

Figure 1.3. Continental break-up of Gondawanaland and development of the circum-Antarctic current, which led to the progressive thermal isolation of Antarctica and subsequent cooling o f the region. (After Kennett, 1977).

14 Sedimentological evidence of glaciation on the Antarctic continent in distal records consists of detrital clays supplied from poorly weathered continental areas and ice rafted debris (IRD). Distal records are influenced by changing ocean circulation, however, so that the source of the material may be difficult to establish. Assessing the link between the concentration of IRD in sediments and ice volume is also complicated because IRD merely indicates that glaciers reached sea level somewhere on the continent. For example, increased abundances of IRD in Neogene sediments of the Southern Indian Ocean have been interpreted as both expansion of the East Antarctic Ice

Sheet (Wamke et al., 1992) and as major episodes of deglaciation (Breza, 1992).

Terrestrial sediments in land-based outcrops and sediments recovered from drillholes on the continental margin provide in situ evidence of glaciation and paleoenvironmental conditions on the continent. However hiatuses occur due to glacial erosion, and sequences may be difficult to date using biostratigraphic markers because of dilution of biogenic material by high terrigenous sedimentation rates. Despite these limitations, the combination of proximal and distal records has produced better constraints on the timing of the initiation of East-Antarctic climatic cooling and glaciation (Figure 1.2), as well as on the dynamics of the East-Antarctic ice-sheet throughout the Cenozoic Era.

The occurrence of ice-rafted quartz sand and the low abundance of calcareous foraminifera in subantarctic Pacific deep-sea cores suggested to Margolis & Kennett (1970) that

Antarctic continental glaciation occurred as far back as the early Eocene (ca. 50 Ma). The first

Antarctic drilling by the Deep Sea Drilling Project (Leg 28, Site 270) confirmed the presence of glaciers along the Ross Sea margin of Antarctica in the mid-late Oligocene (Hayes & Frakes et al., 1975). In the mid-1980s. Ocean Drilling Program (ODP) Legs 113 and 114 drilled a number of holes in the South Atlantic Ocean (Figure 1.4). The late Eocene to early Oligocene records of

15 Atlantic Ocean

120 LPaci India CIROS-1 DVDP-11

180 150°

Figure 1.4. Map of Antarctica and the location of drill holes with marine records of Cenozoic glaciation and paleocUmates mentioned in the text. The numbers refer to Deep Sea Drilling Project and Ocean Drilling Program legs. Stars represent the locations of recent drilling (1997- 2000) by the Ocean Drilling Program and the Cape Roberts drilling Project (CRP).

16 the South Atlantic Ocean show a stepwise shift from more smectite-dominated clay mineral

assemblages to more iliite-dominated assemblages (Robert & Chamley, 1992), roughly

coinciding with the first occurrence of IRD in deep-marine environments (Figure 1.2). The

continental margin of Antarctica was sampled during drilling projects in the late 1980s in the

Ross Sea Basin (MSSTS-1 and CIROS-1), and in the South Indian Ocean near Prydz Bay

(ODPLeg 119). These drilling campaigns provided additional evidence for late Eocene to early

Oligocene East-Antarctic glaciation in the form of diamicts deposited beneath, or near the margin

of, ice grounded on the continental shelves (Figure 1.2). Ice-proximal sediments on the shelf in

Prydz Bay and CIROS-1 occur as early as the late Eocene (Hambrey & Barrett, 1989; Hambrey et

al., 1991). Paleo-environmental studies of recently drilled cores in upper Eocene (?) through

Quaternary strata of the Ross Sea (Cape Roberts Science Team, 1998; 1999; 2000) and Prydz Bay

continental shelves (Cooper & O'Brien et al., 2000), will contribute more data in the near future.

There are indications that the late Oligocene and early Miocene were characterized by

somewhat warmer climates in East Antarctica than those of the late Eocene and early Oligocene.

Clay mineral assemblages at Site 744 show a slight recovery between 35 and 20 Ma to higher

smectite abundances, but not as high as smectite abundances before the initiation of the cooling

(Ehrmann, 1991). Minor amounts of IRD reached the area just south of the present-day Polar

Front Zone (location of Leg 114) from 23 Ma, but the record of ice-rafting does not become continuous until the late Miocene, at ca. 6 Ma (Wamke et al., 1992).

Parts of the early and middle Miocene stratigraphie records on the Antarctic continental shelves are represented by an hiatus, which is interpreted as a major expansion of the East-

Antarctic ice-sheet to the shelf break (Cooper et al., 1991). Much controversy exists about the stability of the Antarctic ice-sheets since the Miocene (e.g. Webb & Harwood, 1991; Kennett &

Hodell, 1993; Oppenheimer, 1998). Due to erosion during Neogene and Quaternary expansions

17 of the ice-sheets. Neogene strata are under-represented in sediment records from the Antarctic continental shelves. Two recent drilling campaigns by the Ocean Drilling Program will partially fill this gap in Antarctic margin records with more stratigraphically complete drillcores from the continental rises of the (Leg 178) and the Prydz-Bay Cooperation Sea (Leg

188).

Several episodes of modest warming in the early Pliocene, between ca. 4.6 and 3.5 Ma are indicated by biogenic carbonate peak values in sediments from Site 737 in the Southern

Ocean (Burckle et al., 1996) and relatively low oxygen isotope values at Site 704 (Hodell &

Venz, 1992). In the continental record from DVDP-11, extremely low oxygen isotope values in benthic foraminifera suggest increased meltwater release from Qord glaciers between 4.5 and 4.0

Ma (Prentice et al., 1999). In situ marine sediments at Marine Plain in East Antarctica are dated at

4.0-4.5 Ma, and contain a diverse marine flora and fauna including vertebrate fossils (Harwood,

1986, Quilty, 1992). Although there is a general consensus about the existence of early Pliocene wanning in the southern high latitudes, considerable debate exists about the degree of that warming (e.g. Kennett & Hodell, 1993).

In the late Pliocene (ca. 2.6 to 1.6 Ma) a significant increase in the abundance of ERD in the Southern Ocean and a northward migration of the Polar Front Zone, expressed as higher siliceous productivity and dominance of the foraminifer N. pachyderma, mark another climate amelioration (Wamke et al., 1992).

1.4.3 “Warm” climates of the Pliocene

The occurrence of present-day global warming has increased the interest of the paleoclimate research community in studying records of past warm global climates. The effects of

18 future global warming can be simulated using general circulation models (GCMs), which are

based on present-day measurements of the interactions between various components of the

Earth’s climate system (Oglesby, 1999). However, many processes of climate change occur on

time scales that exceed the human life span and the historical record of measurements. In GCMs

assumptions are made about the boundary conditions of the components of the climate system,

and geological records should be used to test the validity of these assumptions.

The middle Pliocene was the last interval in Earth history when global climates were

significantly warmer than present under similar plate tectonic configurations, and with a similar

composition and structure of ecosystems as today. In the early 1990s the United States Geological

Survey started a project aimed at producing a quantitative reconstruction of middle Pliocene

conditions, which could then be used to calibrate the boundary conditions for GCM experiments.

The PRISM (Pliocene Research, Interpretation, and Synoptic Mapping) project has since documented orographic changes, and changes in ice extent and in vegetation, as well as produced

a global reconstruction of middle Pliocene sea surface temperatures (Dowsett, et al., 1992; 1996).

According to that reconstruction, sea surface temperatures in the tropics were not much different firom the present values, whereas high latitudes had higher sea surface temperatures during the mid-Pliocene (Dowsett et al., 1996). High-latitude arctic sea surface temperatures in the mid-

Pliocene were reconstructed at 3-5°C warmer than at present (Cronin & Dowsett, 1995), but data are not sufficient to make quantitative statements about the Southern Ocean. The objective of the

PRISM program is to predict the effects of global warming on the geographic distribution of paleotemperatures and humidity through modeling experiments (Sloan et al., 1996; Haywood et al., 2000), and to speculate on the causes of Pliocene warmth (Crowley, 1996).

The results of modeling studies using the PRISM database demonstrate the importance of good records for both low and high-latitude sea-surface temperatures. Future modeling

19 experiments must include ice-sheet sensitivity studies and better constraints on high-latitude ice-

sheet configurations. Proxy data suggest that a large Antarctic ice sheet grew in the mid-Miocene, but also show Pliocene sea levels that are higher than those of today (Haq et al., 1987). The stable isotope stratigraphy of ODP Site 846 in the Pacific Ocean (Shackleton et al., 1995) indicates decreasing global ice volumes between 4.5 and 3 Ma if the assumption of constant cool deep- ocean temperatures is valid. Oxygen-isotope data from ODP Site 607 in the Atlantic Ocean

(Raymo, 1992) show that climate oscillations became progressively colder from 3.1 to 2.6 Ma, but the cold extremes between 3.1 and 2.95 Ma produced more negative 6 ’^O values than modem values. The latter means that either bottom were 3.5°C warmer than today, and/or that Antarctic ice-volumes were significantly smaller. Ice-sheet models predict the response of the Antarctic ice-sheet to increasing mean annual temperature and show that a warming of 9°C is sufficient to melt the , but that a warming of 15°C is needed to have a significant effect on the extent of the (Huybrechts, 1993). However, such a high degree of warming is not supported by the limited paleontological and geochemical evidence available for the mid-Pliocene of the Southern Ocean (Kennett & Hodell, 1993; Dowsett et al., 1996) and the Antarctic continent (Marchant et al. 1993a+b).

Webb et al. (1984) discovered recycled mid Pliocene marine diatoms in some Sirius

Group deposits in the Transantarctic Mountains and dated them to < 3.0 Ma. The diatom-based ages were later confirmed and calibrated by a date on a volcanic ash deposit (2.77 Ma) occurring in the same Pliocene diatom zone in the CIROS-2 drill hole in Ferrar (Barrett et al., 1992).

Webb et al. (1984) suggested that the diatom-bearing sediments were deposited by a temperate ice sheet that expanded towards the Transantarctic Mountains from the continental interior of

Antarctica, and that the marine diatoms were derived from the subglacial Wilkes and Pensacola

2 0 basins after significant deglaciation in the mid Pliocene, when a marine transgression into the

Antarctic interior basins had occurred.

However, the age of the Sirius Group is under considerable debate, because of the

apparent antiquity of landscapes (mid-Miocene) in the Dry Valleys block of the Transantarctic

Mountains (Sugden et al., 1993). Analyses of unconsolidated sediments and in situ tephras from

the Asgard Range (Marchant et al., 1993b) and the Quartermain Mountains (Marchant et al.,

1993a) favor a mid-Miocene overriding episode by the East Antarctic Ice Sheet and persistent

cold and dry conditions since that time. The Miocene tephras from high elevations in the Dry

Valleys are unaltered by chemical weathering, whereas low elevation records do show evidence

of modest early-mid Pliocene warming (Prentice et al., 1993; Prentice et al., 1999). Isotopic ages

of >14 Ma for in situ, unweathered, ash deposits from the higher elevated regions of the Dry

Valleys are in apparent disagreement with major deglaciation in the mid-Pliocene as suggested

on the basis of marine diatoms of this age in the Sirius Group (Webb & Harwood, 1991). The age

debate at present focuses on how marine diatoms were deposited within the Sirius Group

(Burckle and Potter, 1996; Kellogg and Kellogg, 1996; Stroeven and Prentice, 1997; Barrett et

al., 1997; Gersonde et al., 1997; Stroeven et al., 1998).

1.5 Outline of dissertation and summary of results

This dissertation discusses the results of four different projects, which employ

sedimentological properties of glacigenic sediments to propose paleoenvironmental reconstructions of the Antarctic region. Most of the dissertation (Chapters 1-6) concerns compositional studies of Upper Cenozoic glacigenic sediments from the Transantarctic

Mountains in the Ross Sea area. Sediments of the Sirius Group (Mercer, 1972; McKelvey et al..

2 1 1991) show evidence of considerable meltwater reworking, indicating less extreme polar climate

conditions than today (Chapter 2 and 3). Initial results of studies of Sirius Group diamicts from

the Prince Albert Mountains show that compositional studies can be used to define the relation of

the Sirius Group to East-Antarctic glacial history and the tectonic evolution of the Transantarctic

Mountains (Chapter 4).

Mineralogical studies on a number of Sirius Group deposits from well-separated areas in

the Transantarctic Mountains show that the Sirius Group may be subdivided into two or more

petrofacies, based on differences in heavy mineral composition and bulk X-ray mineralogy

(Chapter 5).

The chemical composition of the Sirius Group is highly variable. The CIAs of the

sediments are affected by recycling of weathered Proterozoic, Paleozoic and Lower Cenozoic sedimentary rocks. However, recycling of weathered material is limited in some deposits, such as those at Oliver Bluffs. The CIAs of proglacial facies also indicate very limited post-depositional alteration, suggesting climates that were only slightly warmer and humid than the present-day

(Chapter 6). Geochemical evidence from the Pliocene Bardin Bluffs Formation also supports the persistence of glacial conditions in Antarctica during the mid-Pliocene, with a slightly more temperate climate than today’s during a brief time interval in the middle Pliocene (Chapter 7).

Chapter 8 discusses the mineralogical and chemical composition of Eocene and Pliocene erratics collected from the Mount Discovery coastal moraine. The chemical index of alteration

(CIA) of an Eocene mudstone sample is in agreement with a cool-temperate climate, as determined by other paleoenvironmental indicators (Harwood & Levy, 2000). Chapter 9 discusses soft-sediment deformation studies on two Cape Roberts drillholes from McMurdo

Sound in the Ross Sea area. The first account of grounded ice on the continental shelf is provided by glacio-tectonic features in the mid-Oligocene succession of the CRP-2/2A core.

2 2 Chapter 10 mainly focuses on the integration of compositional data from the Sirius

Group, and the importance of all data collected in this research with regard to the two main research topics: initiation of glaciation in the Ross Sea sector and climate stability in the Neogene of continental Antarctica. It becomes clear that the stratigraphie fiamework of the Sirius Group should be revised, because different outcrops may represent different glacial phases, which may have occurred up to tens of millions of years apart. A new hypothesis is proposed, which is significantly different from previous interpretations of Sirius Group deposition during one relatively short-lived glacial phase.

23 CHAPTER!

LITHOSTRATIGRAPHY AND GEOLOGICAL SETTING OF THE SIRIUS GROUP AND

RELATED UPPER CENOZOIC DEPOSITS

2.1 Introduction

This chapter summarizes the geological setting and outcrop-scale sedimentological observations made by numerous authors, in order to characterize the stratigraphy and facies of the

Sirius Group at different localities. The successions are subdivided into lithostratigraphic units, where available, as proposed by field parties. Only outcrops and formations from which samples were derived for this study are discussed. For a more comprehensive list of known Sirius Group outcrops and their characteristics, one is referred to Stroeven (1997).

2.1.1 Structural framework and basement geology of the Transantarctic Mountains

At present, the East Antarctic Ice Sheet drains through several large outlet glaciers flowing perpendicular to the Transantarctic Mountains front. Between 74° and 87° South, the present drainage system is characterized by 32 overdeepened glacial troughs, which dissect the

Transantarctic Mountains between the East Antarctic polar plateau and the Ross Sea basin

(Webb, 1994). The Transantarctic Mountains form the uplifted margin of the Ross-Sea rift system and separate the East Antarctic craton from the younger West Antarctic crustal blocks. Rapid

24 denudation, as constrained by fission track dates (Fitzgerald, 1992; 1994), began around 50-55

Ma. Uplift rates differ along the Transantarctic Mountains suggesting that the eastern Ross Sea

margin consists of several crustal blocks separated by cross faults now occupied by major outlet

glaciers (Fitzgerald, 1994; Van der Wateren and Verbers, 1994; Tessensohn, 1994).

Figure 2.1 Devonian-Permian Beacon Supergroup intruded by Jurassic Ferrar Group rocks. Upper Taylor Valley, Antarctica.

The basement rocks of the Transantarctic Mountains are composed of metasediments of

Precambrian and Cambrian age, which were intruded by Cambro-Ordovician granites of the

Granite Harbour Intrusives and by the Devonian Admiralty Intrusives (Stump, 1995). The Kukri

Peneplain is an Ordovician-Silurian erosion surface, which is overlain by the Devonian-Triassic

Beacon Supergroup, consisting of glacial, fluvial, and shallow marine sediments (Barrett, 1991).

The Beacon Supergroup was intruded and capped by tholeiitic basalts of the Ferrar Supergroup during Jurassic time, forming dolerite sills (Figure 2.1) and flood basalts (Tingey, 1991). Sirius

25 165°E RH = Ricker Hills GN = Griffin Nunatak North PM = Prospect Mesa Victoria MPI = Mount Fleming Land MFe= Mount Feather TM = Table Mountain MR = Miller Range Ross MS = Mount Sirius RH. Sea CL = Cloudmaker 75° S OB = Oliver Bluffs DVDP-11 Dry Valleys MD = Meyer Desert MFe, BP = Bennett Platform TM RM = Roberts Massif QH = Quartz Hills TS = Tillite Spur Ross

MR- 80°S JW

Beardmore OB MD RM / Glacier jQH 85° S R eedy Glacier- TS

SOUTH NORTH „ j Beard- Reedy Shackleton more G lader Glacier Glaciei Dry Valleys o a v y I Glacier V i i Jr 3r t TS QH RM - --- — G MR §3 C c C TM RH PM OB Scale: c 400 km CL B I 40 m DVDP-11

Figure 2.2. Location of Sirius Group outcrops and related Upper Cenozoic deposits (A), and stratigraphy of the most important outcrops and drillholes (B). Sirius Group deposits studied in detail in this work are indicated by a black dot. Small black squares indicate other important but less well studied Sirius Group localities (after Mayewski, 1975 and Stroeven, 1997).

26 Group deposits unconformably overlie metamorphic basement, deformed Beacon Supergroup,

and striated or grooved Ferrar Dolerite surfaces (Mabin, 1986).

2.2 Stratigraphy and geological setting of the Sirius Group

Outcrops of the Sirius Group are found between the Reedy Glacier in the southern

Transantarctic Mountains and David Glacier in central Victoria Land (Figure 2.2). The thickness

of outcrops varies along the mountain range, with the thickest successions in the central

Transantarctic Mountains along the Beardmore Glacier. The successions include massive

diamictites with striated clasts interbedded with stratified units.

The Sirius Formation was originally described by Mercer (1972) as "compact glacial drift

imconformably overlying pre-Tertiary rocks" at Mount Sirius and the northern end of the

Dominion Range along the Beardmore Glacier. Mayewski (1975) and Mayewski and Goldthwait

(1985) later investigated 24 outcrops of glacigenic sediments and assigned them all to the Sirius

Formation based on the uniform appearance of the deposits. They concluded that the Sirius

Formation was deposited by a wet-based East Antarctic Ice Sheet on a pre-existing glacially

sculptured topography. Detailed studies by McKelvey et al. (1991) and Webb et al. (1996a) of a

succession of diamictons and stratified sediments at Dominion Range and Cloudmaker within the

Beardmore Valley showed that the Sirius Formation records both marine and terrestrial depositional environments. The of the erosion surfaces in the Dominion Range further suggested that deposition occurred concurrently with stepwise erosional lowering of the

Beardmore Glacier valley floor. McKelvey et al. (1991) argued that the morphological complexity and compound stratigraphy justified raising the Sirius Formation to group status.

2 7 ~ r 175° Ross Ice I Shelf 20 km

D orrirhion

V 170 RM ^175° (a)

altitude N W SE a.s.l. Beardmore Shackleton 3000 m Glacier Glacier MD OB 2000 m

Queen i 1000 m Mount Alexandra Dominion Bennett R o b erts Sirius Range Range Platform M assif

Sirius Group Kirkpatrick Basalt lllllllllll Ferrar Dolerite Beacon Supergroup

(b)

Figure 2.3. Location of Sirius Group outcrops in the central Transantarctic M ountains (a). Samples studied here are from outcrops indicated by dots. Squares indicate further outcrops not studied here, (b) Geological profile parallel to the Transantarctic Mountain front.

28 The Sirius Group is geographically widespread and occurs as scattered outcrops, which

are exposed in two typical settings: as thin erosional remnants on flat mountain summits, and in

successions of > 100-m stratigraphie thickness along the walls of broad trunk valleys draining

the East Antarctic Ice Sheet. In Victoria Land (which includes the Dry Valleys), the Sirius Group

crops out as relatively thin diamictons on mountain summits at high elevation. In the central and

southern Transantarctic Mountains the deposits are thicker and comprise diamictons and stratified

facies. The diamictons contain striated and facetted pebbles indicative of basal transport beneath

glacier ice.

2.2.1 Mount Sirius type locality

In the central Transantarctic Mountains the Sirius Group crops out within a NNW-SSE oriented post-Jurassic syncline, extending from near the to the Shackleton

Glacier area (Barrett, 1972; Webb et al., 1986a). The thickest successions of the Sirius Group are preserved close to the axis of the syncline. The type locality of the Sirius Group is at Mount

Sirius in the Bowden Névé drainage basin, to the north of the Beardmore Glacier (Figure 2.3).

The sediments were first described by Mercer (1972) as comprising about 100 m of compact diamictite with lenses of sorted material containing angular fragments of shale and coal.

McKelvey et al. (1984), Harwood (1986), and Webb et al. (1986b) described the Mount

Sirius locality in more detail. The Sirius Group at Mount Sirius predominantly consists of indistinctly stratified diamictites with boulders up to 4 m in diameter, interbedded with pebble breccia or conglomerate lenses, and a few laminated siltstone beds. The sediments overlie a striated pavement with directional data indicating transport between 022° and 105°, with a predominant direction of 075°. The succession is subdivided into three units. The lowermost imit.

29 Mount Sirius

(m ) m I S I g I Weathered sand • • - *^ • 1 '------' C7 Diamictite Dlamlcton Conglomerate.

10 - Stratified diamictite Stratified CO diamicton *E 20 - - 3 Breccia Boulder diamictite G ravel/ 30 - 0 Conglomerate Sand and 40 — gravel Massive diamictite CsJ containing thin Sand 50 - ■ 'E breccia and 3 conglomerate beds Mudstone

60 - Paleosol

Boulder Gravel lag 70 - diamictite with 'E clasts up to 4 m 3 80 - 1 ? i nil = 1111 = Striated Ferrar 90 - = III! = nil : : nil = Mil — Dolerite

Figure 2.4. Lithostratigraphy of the type locality of the Sirius Group at Mount Sirius, Bowden Neve. Based on descriptions of Harwood (1986) and Webb et al. (1986).

30 Unit 1, is 22 meters thick and consists of texturally variable, locally bedded boulder diamictite

(Figure 2.4). The overlying unit. Unit 2, is approximately 30 meters thick and consists of finer

diamictite containing breccia and conglomerate beds up to 6 m thick (Webb et al., 1986b). The

uppermost unit. Unit 3, consists of coarse diamictite with striated clasts, similar to Unit 1. Pebble

breccia and conglomerate beds, some of which are highly deformed, occur more frequently in the

upper part of the succession (McKelvey et al., 1984; Webb et al., 1986b).

Near the Mount Sirius outcrop, diamictites have also been reported by Prentice et al.

(1986) from high-elevation outcrops (>3500m) at Mount Falla, , and Mount

Mackellar, and from the Markham Plateau. The diamictites are yellow to gray, massive and contains striated gravel. The diamictite on Mount Falla contains clasts of limestone, and bedrock striations at this site suggest that the ice flowed toward the northeast.

2.2.2 Beardmore Glacier

The stratigraphy of the Sirius Group in the Beardmore Glacier area was described by

Mercer (1972), Webb et al. (1986b), McKelvey et al. (1991), and Webb et al. (1996a). Three formations are recognized: the Cloudmaker Formation, the Meyer Desert Formation and the

Mount Mills Formation. The stratigraphically lowest unit is the Cloudmaker Formation, which is divided into six members separated by disconformities. The total stratigraphie thickness of the

Cloudmaker Formation is 141 m and it unconformably overlies the Precambrian Goldie

Formation at ca. 1060 m.a.s.l (Webb et al., 1996a; Figure 2.3). The presence of agglutinated foraminifera suggests a glacio-marine depositional environment for the Cloudmaker Formation.

A disconformity separates the Cloudmaker Formation from the overlying terrestrial Meyer Desert

Formation. The Meyer Desert Formation has been described in detail from the Dominion Range,

31 S ection 5 S ection 7

2 7 5 m Oliver Bluffs

10 m n

Section 8 Faciès Interpretation

0 m 175 m

7 3 Diamicton Interbedded Unit o Stratified O diamictites and ? 4 cp pebbly diamicton Unit 4 sandstones Breccia o„\C Gravel/ l a Conglomerate Sand and W gravel ro Basal breccia/ Wateriain conglomerate sediments and Mudstone overlain by lodgement diamictites tiiilte Paieoso Unit 3b Gravel Silty sandstones coarsening upwards to pebble \VV^Y\A\ Unit 3 a conglomerates Fluvial sediments J J < Unit Unit 2 Massive and Wateriain 3b stratified diamictite, sediments and minor laminated lodgement siltstone and pebble tillite Sa beds

Figure 2.5. Lithostratigraphy and facies of three sections along an outcrop of the Sirius Group at Oliver Bluffs, Dominion Range (Webb, pers. comm, 1999). further landward from the Cloudmaker outcrops. The outcrops on the Cloudmaker and in the

Dominion Range have been correlated based on the similar assemblages of agglutinated

foraminifera in the lowest member of the Meyer Desert Formation and the uppermost member of

the Cloudmaker Formation (Webb et al., 1996a).

In the Dominion Range, the Meyer Desert Formation has a maximum thickness of 185 m

and consists predominantly of diamictons interbedded with fluvial and lacustrine sediments. The

Mount Mills Formation of the Dominion Range is a coarse diamictite, breccia and conglomerate

facies of the Sirius Group. The stratigraphie position of the Mount Mills Formation with respect

to the Meyer Desert Formation is unknown.

The Meyer Desert Formation crops out on a stepped erosion surface and consists of a

lower bench at 1800-1900 m.a.s.l. and an upper bench at 2200-2500 m.a.s.I. (McKelvey et al.,

1991). The Meyer Desert Formation is divided into four members based on detailed analysis of

three sections at Oliver Bluffs (Figure 2.5). The outcrop at Oliver Bluffs has been eroded by the

present Beardmore Glacier, exposes the lower part of the Meyer Desert Formation at 1800-1900

m.a.s.l, and has a maximum thickness of ca.lOO m. Thicker successions occur on the higher

bench at 2200-2500 m.a.s.l. Section 13 of McKelvey et al. (1991) represents the lower 138 m of

the sediments found on the upper bench (Figure 2.6). Paleosols are present, especially in the

upper levels of section 13. The stratigraphie relationship between the sediments exposed on the upper bench (section 13) and those exposed in the Oliver Bluffs (lower bench) is unclear. Based on the geomorphology of the erosion surfaces, McKelvey et al. (1991) suggested that the upper bench succession may be older than the Oliver Bluffs Sirius Group.

33 Meyer Desert - Section 13 Bennett Platform

(m) m I S I g I Facies Interpretation units] m I s I g I Facies Interpretation (m) 0-1 0 DIamlct/aravel Glaclal/glaclo- fluvlal UJllllllL Stratified Lodgement Massive diamlct 10 ■ II iiTTii II diamictites tlllltes and 1 0 - and subaerial Lodgement tills 20 - weathered exposure Weakly stratified with sporadic Intervals with surfaces 2 0 - diamlct and winnowing 30 - conglomerate with gravel gravel of fines by o fluvial action 40 - and lags and 3 0 - CP ^ sandstone eollan deposits O 50 - CP 4 0 - 60 - SI Massive diamlct Lodgement tills 70 - 5 0 - 3 . DIamict/gravel/ Glaciolacustrine Stratified lamlnlte and deltaic 80 boulder Lodgement Gravel/lamlnlte association diamictite tlllltes and 60 90 Interbedded wateriain Massive diamlct Lodgement tills with slltstones, sediments 70 sandstones, 100 C 7 Weakly stratified Lodgemen tills/ and granule diamlct/gravel glaclofluvlal beds 110 • 8 0 - O p \ C CP 120 - Massive diamlct Lodgement tills 9 0 - 130 - o S: Massive diamlct, Resedlmented 140 - 100 - Fremouw Fm o gravel breccia, glaclal/glacio- ' (basement) CP slumping lacustrine 110 - HIgh-rellef, Glacially eroded striated dolerite bedrock surface

Figure 2.6. Lithostatigraphy and facies interpretation of the Sirius Group in the Meyer Desert (Dominion Range) and at Bennett Platform (Shackleton Glacier area). Meyer Desert, section 13 after McKelvey et al. (1991) and Harwood, pcrs.comm. (1998). Bennett Platform after Hambrcy et al. (in prep.). For legend see Figure 2.4. 2.2.3 Shackleton Glacier

The Shackleton Glacier area is situated SW of the Beardmore Glacier area (Figure 2.3).

Here, three glacigenic stratigraphie units are recognized, which will likely be accorded formation

status (Webb et al., 1996d). The oldest unit consists of erratic boulders, up to 4 m across, in

Quaternary moraines, and of clasts in younger Sirius Group deposits, which are composed of moderately lithified diamictite, sandstone and conglomerate. No in situ outcrop of this unit has been identified so far. The two younger units (Formations 1 and 2) occur in cliff exposures at

Bennett Platform and are of considerable thickness (Figure 2.6). Formation 1 (thickness 96-100 m) is subdivided into eight members, with lithofacies ranging from massive diamictites to laminites affected by slumping (Figure 2.7). Formation 2 unconformably overlies Formation 1 and has a thickness of 6-46 m, thickening southward along Bennett Platform. It is subdivided into two members: a lower massive diamictite and an overlying weathered diamictite and conglomerate. It is capped by a lag deposit of dolerite boulders. The diamictites in both

Formation 1 and 2 have a distinct clast orientation fabric and many clasts are striated, suggesting deposition by basal ice (Hambrey et al., in prep.).

Other outcrops of the Sirius Group have been described from Roberts Massif, to the south of the Bennett Platform locality (Webb et al., 1996d). The relief of the Roberts Massif nunatak represents a relict glacial topography. In the northern lowland of the nunatak a 25 km" area of

Ferrar Dolerite pavement with ridges and grooves is preserved (Webb et al., 1996c), indicative of large-scale areal scouring (Figure 2.8). Thin remnants of Sirius Group diamictites are found directly above this pavement. Outcrops of diamictites are also found at the southern side of the nunatak, facing the polar plateau. A section of diamictites and stratified sediments, ca. 50 m thick, is exposed above the Shackleton glacier on the northwest side of Roberts Massif.

35 Figure 2.7. Sirius Group outcrop at Bennett Platform in the Shackleton Glacier area, central Transantarctic Mountains (a). Note laminated facies interbedded with diamictons (a) and dropstones in laminated facies (b). Photos: David Harwood.

36 àerialN- 85“30’S photo

Robert? Massif:

5 km

178"W Polar Plateau 176°W

%

Figure 2.8. Dolerite erosion surface with glacial grooves at Roberts Massif (Shackleton Glacier area). Note fault which runs through grooved rock surface and clearly post-dates the erosional phase.

37 2.2.4 Reedy Glacier

The Reedy Glacier area is the most southern and eastern location of Sirius Group

sediments in the Transantarctic Mountains (Figure 2.1 and 2.9). The stratigraphy of the Sirius

Group in the Reedy Glacier area was described by Mercer (1968) and Wilson et al. (1998a). A

section at Tillite Spur consists of 35 m of glacigene sediments overlying fine-grained sandstones

of the Permian Weaver Formation (Figure 2.10). Mercer (1968) subdivided the succession into 6

lithostratigraphic units. The lithostratigraphic subdivision was later refined to 10 units by Wilson et al. (1998a) (Figure 2.10). The lower members consist of lithologically homogenous sediments overlain by fine-grained stratified deposits, interpreted as overridden endmoraines and glacio­ lacustrine sediments. These members are overlain by diamictons, which are more compact and contain striated pebbles. Silty diamicts are found about 2700 m a.s.l. at Metavolcanic Mountain and Mount LeSchack. Wilson et al. (1998a) proposed the name Tillite Spur Formation for the

Sirius Group sediments that crop out at Tillite Spur, Metavolcanic Mountain and Mount

LeSchack.

At Quartz Hills, a 90 m high sequence of stratified sediments is exposed (Figure 2.11), for which Wilson et al. (1998a) proposed the name Quartz Hills Formation. The succession consists of diamictons and laminated horizontally bedded mudstones, alternating with micaceous sand and containing many pebbles and boulders, embedded as dropstones. Both Mercer (1968) and Wilson et al. (1998a) interpreted most of this section as glacio-lacustrine. Wilson et al.

(1998a) argued that the uppermost members may have a marine origin, based on the presence of decapod burrows and bioturbation.

38 87 100 km

o f 'W ^

N

Metavolcanic % Mountain Tillite Spur Mount \ LeShack

Range A

8 6 ' Æ. _i__ Æ:

Figure 2.9. Location of Sirius Group deposits in the Reedy Glacier area. Dots indicate outcrops studied in this work, squares indicate additional Sirius Group outcrops.

39 Units m I s I g I (A) (B) Facies Interpretation (m) N N Conglomerate till 30 _ 6 10 5 9 o Diamictite Lodgement till 25 _

20 _ Lodgement 8 Diamictite till

15 _ ~i(_ 4 7 Conglomerate 6 Diamictite 10 _ 5 Diamict/mudstone 4 Conglomerate/ o diamictite 5 _ o 3 Diamictite Lodgement till CZ)' 2+3 2 Conglomerate Glaclofluvlal? 0 _ ^Diamictite ^ Redeposited l / Folded sandstone sandstone

Figure 2.10. Lithostratigraphy and facies interpretation of the Sirius Group at Tillite Spur, Reedy Glacier area. (After Mercer, 1968, and Wilson et al., 1998a). Units in column (A) proposed by Mercer (1968), units in column (B) proposed by Wilson et al. (1998a).

40 m I s I g Facies Interpretation 100 (m) Diamictite Lodgement O tills 90 — C i F 3 : 80 — Interbedded Turbidites in o diamictites, proglacial mudstones and lake 70 — conglomerates

60 — o

Diamictite Lodgement 50 - tills o CZ) 40 — Interbedded Turbidites in mudstones and proglacial 30 — sandstones with lake dropstones

20 — Interbedded Lodgement CZ diamictites, — tills and pro­ 10 sandstones, and glacial deposits conglomerates 0 _

Figure 2.11. Lithostratigraphy and facies interpretation of the Quartz HiUs section of the Sirius Group. (After Mercer, 1968, and Wilson et al., 1998a).

41 2 .2 .5 V ic to r ia L a n d

Cenozoic glacial deposits in the Dry Valleys were first assigned to the Sirius Formation

by Mayewski (1975). The most extensively studied Sirius Group deposit in the Dry Valleys is the

Mount Feather Diamicton (Brady & McKelvey, 1979; 1983; Barrett et al., 1997; Wilson &

Barron, 1998), which is ca. 40 m thick and unconformably overlies the Mount Feather

Conglomerate and the Lashly Formation of the Beacon Supergroup at an elevation of ca. 2700

m.a.s.l (see Figure 2.12 for location). The diamictite is a poorly sorted, rather uniform mixture of

boulders (up to 2 m across), gravel, sand and mud, with faint local stratification (Wilson &

Barron, 1998). It contains many striated and facetted clasts, and it exhibits a distinct clast fabric that parallels features on underlying striated pavement (Brady & McKelvey, 1979). The striated surface underlying the diamictite is considered to be a fossil valley floor, which must have been carved before the Ferrar and Taylor Valleys existed.

At Mount Fleming, diamictites crop out at two locations (Stroeven & Prentice, 1996).

The more extensive of the two deposits is the Upper Fleming Till, which was assigned to the

Sirius Formation by Mayewski (1975) and Harwood & Webb (1986). The Upper Fleming Till is exposed at an elevation of ca. 2000 m.a.s.l. on a glacially scoured and molded sandstone surface of the Fleming Formation of the Beacon Supergroup. The Upper Fleming Till is a more than one meter thick, poorly sorted, pebbly muddy medium sand with striated clasts. Stroeven & Prentice

(1997) proposed, based on directional data and clast composition, that the Upper Fleming Till was deposited by an alpine glacier.

42 David Glacier Mount 20 km Prince Billing a a % g ♦ - gT A lb ert g f a GrifRn *N un. a ^ M ountains \ 2 Ross Sea

Allan ê/)

Mackay Gl. a Prospect Mesa Mt-Fleming DVDP 11 Taylor Valleys Island: Glacier McMurdo Sound Mt. Feather Table Mtrf

Figure 2.12. Location of Sirius Group and related Upper Cenozoic deposits in Victoria Land that were investigated in this study. Sample locations are indicated by black dots.

43 Glacial sediments from Table Mountain (Figure 2.13) were first described by Barrett &

Powell (1982). The deposits are exposed at an elevation of ca. 1800 m.a.s.l. in a paleovalley

predating the cutting of the Ferrar Glacier valley (see Figure 2.12 for location). The preservation

of the deposits is poor and it forms a veneer, originally reaching a thickness of ca. 15 m.

Additional information has been obtained from a core drilled into the Table Mountain deposit

(Hicock et al., 1999). Diamictite is only a minor facies at Table Mountain, because the Sirius

Group here is dominated by a sandy fluvial facies. The clast composition of the diamictites

suggests that it is a local deposit, not far-travelled. The diamictites at Table Mountain are very

light grey muddy sandstones with scattered pebbles, cobbles and few boulders. The diamictites

have a low mud content, between 8 and 30 % (Barrett & Powell, 1982). No striated pavements

have been found, but a substantial proportion of the clasts in the diamictites are striated. Clast

fabric studies from the Table Mountain drill core (Hicock et al., 1999) suggest that most of the

diamictites have a subglacial origin.

Figure 2.13. Outcrop of the Sirius Group at Table Mountain, Ferrar Glacier area. Dry Valleys. Photo: Steve Bohaty

4 4 The northernmost outcrops of the Siruis Group occur in the southern Prince Albert

Mountains, between David Glacier and Mawson Glacier (Figure 2.12). Diamict blankets more

than a meter thick are found on summits and on glacial terraces of nunataks in this area (Verbers

& Van der Wateren, 1992). At Griffin Nunatak (2260 m) the diamict matrix is coarse-grained and

reddish in color. At Mount Billing and Ricker Hills the diamict matrix is finer-grained and brown

to gray in color (Verbers & Van der Wateren, 1992). The tills contain subrounded, sometimes

striated Ferrar Group volcanic clasts of up to 40 cm across. Diamicts from glacially eroded

terraces at lower elevation in the Prince Albert Mountains are dominated by clasts from the Ferrar

Group, but also contain granite and Beacon Supergroup clasts.

2.2.6 DVDP-11 and Prospect Mesa

Besides the Sirius Group, other Upper Cenozoic glacial deposits from the Dry Valleys

were also examined to establish their relationships to East Antarctic glaciation. Prospect Mesa is

a flat-topped sediment body, ca. 18 m high, on the floor of Wright Valley (Prentice et al., 1993).

A sample from the lower water-laid diamictiton was examined in this study. The presence of dropstones and stratification suggests that the lower diamicton was deposited from floating ice.

Clast lithologies are granite, dolerite, and various dike rocks. The diamicton is overlain by the

Prospect Mesa gravels, which contain in situ pecten shells and foraminifera, suggesting deposition in a Qordal setting during the late Miocene-early Pliocene.

DVDP-11 was drilled in Taylor Valley during the Dry Valley Drilling Project in 1974

(McKelvey, 1981; 1982). Its sediment record extends to ca. 6.1 Ma (Ishman and Rieck, 1992).

Sedimentation occurred mainly in a glaciomarine Qordal setting (McKelvey, 1982;

45 DVDP-11

,1 m i s Facies Interpretation (m) 0 Massive medium sandstone Terrestrial to 25 - shallow marine 0) Pebbly coarse deltaic c sandstone with 0) intraformational 50 - o mudstone debris % 75 - _0) 0_ Medium sandstone Terrestrial to and conglomerate shallow marine with distinct deltaic

1 0 0 - stratification and intraformational mudstone debris 125 - Diamictite inter­ Waterlaid till and bedded with sandy mass-movement 150 - mudstone and deposits conglomerate

Poorly sorted 1 7 5 - granule-pebble Shallow-marine conglomerate lag deposit

200 - Diamictite inter­ bedded with Waterlaid laminated sandy till 225 - to silty m udstone CZ)

250 -

Diamictite with 275 - CD Waterlaid minor sandy till c z mudstone and conglomerate. 300 - c z 325

Figure 2.14. Lithostratigraphy and faciès of Dry Valley Drilling Project hole 11 (DVDP-11) in Taylor Valley. After McKelvey (1981) and PoweU (1981).

46 Powell, 1981; Webb and Wrenn, 1982), although the basal 22 m of the core is a barren

homogenous diamicton devoid of any sedimentological structures indicating marine reworking

(Figure 2.14). The analysis of clast composition shows that ice approached the drillsite from both

the west and the east during the time represented by the core. An important hiatus was identified

at 240 mbsf where the Miocene-Pliocene boundary has been placed (Ishman and Rieck, 1992).

Another hiatus is present at 202 mbsf, where provenance changes, and the abundance of clasts derived from the McMurdo Volcanics gradually increases upsection.

47 CHAPTERS

DEPOSrnONAL ENVIRONMENTS OF THE SIRIUS GROUP

3.1 Introduction and purpose

Terrestrial glacigenic sediments are the products of complex environmental conditions.

The basal thermal regime and proglacial climate of a terrestrial ice-sheet or glacier are the main controls on subglacial and proglacial depositional processes. These assertions are based both on observations of glaciers in a variety of environmental settings (e.g. Shaw, 1977; Eyles et al.,

1983; Fitzsimons, 1990), and on theoretical considerations (Paterson, 1994). This dependence of depositional environments on the thermal regime makes glacigenic sediments important monitors of past glacial conditions. Criteria have been developed to reconstruct paleoenvironmental conditions using both macroscopic field observations, and microscopic and analytical laboratory techniques. Vertical facies profiles and clast orientation fabrics can be used to characterize and interprété glacigenic lithofacies associations in the field (e.g. Eyles et al., 1983; Ehlers &

Stephan, 1983). Grain-size distributions, clast properties, matrix composition, and micro fabrics have been used succesfully to reconstruct the transport conditions of diamictic ice-contact deposits (e.g. Boulton, 1978; Van der Meer, 1993; Kjær, 1999). In this chapter the grain-size characteristics of diamictons and stratified proglacial deposits will be discussed and the

48 micromorphology and bulk density of some diamictons will be described. These results will be

compared to the lithofacies distributions described by field parties to constrain the environmental

conditions during deposition of the Sirius Group successions.

Grain-size data have been collected previously for Sirius Group deposits at Table

Mountain (Barrett & Powell, 1982), Mount Feather (Barrett, 1998), Mount Fleming (Stroeven &

Prentice, 1996), the Prince Albert Mountains (Chapter 4) and the Reedy Glacier area (Wilson et

al., 1998a). No grain-size data were available for key deposits from the Central Transantarctic

Mountains, including the Oliver Bluffs and Bennett Platform sections. Therefore, 52 samples were selected from eight Sirius Group outcrops near the Shackleton and Beardmore glaciers.

3.2. Methods

For grain-size analysis, 50-100 g of sample were disaggregated in water and sieved over a 2 mm mesh, dried in an oven at 105° C and weighed. Samples represent four principal lithofacies: massive diamicton, stratified diamicton, pebbly sand(stone), and mud(stone). Because of the small sample size, the fraction >2 mm was not analyzed. Samples were pretreated with

H2O2, and Na 2P2Û7 was used as a dispersant. Samples were wet-sieved over a 63 p.m (4 phi) mesh and grain-size distributions with 1 phi intervals were obtained through rotap sieving (sand fraction), pipette analysis (silt and clay fraction) and Coulter Counter analysis (silt fraction).

Traditional phi-class grain-size histograms were constructed for all samples, by combining sieve data. Coulter Counter data, and pipette data for the clay fraction. For further details about the methodology see Appendix A.

Bulk density of a selection of samples was determined by measuring weight and volume of coherent diamict fragments. Fragments were weighed and coated with paraffin. After drying of

49 the paraffin, the fragments were weighed again to determine the weight and volume of the

paraffin (spec. grav. =0.86). The fragments were then submerged in water in a measuring cylinder

and the volume displacement of water (ml) was determined by substracting the initial water level

from the final water level on the wall of the cylinder. The calculated volume of the paraffin

coating was subtracted from the volume of water displaced to determine the sediment volume.

The initial weight of the sediment clumps was divided by the volume of the sediment clumps to

determine the bulk density.

To assess the subglacial conditions during deposition of massive diamicts of the Sirius

Group, many of which had been interpreted as basal tills based on field criteria, 22 samples were

selected for thin section microfabric analysis. Thin section fabrics of diamicts were described

following methods simplified from those developed by Van der Meer (1993). Sedimentological

properties that were recorded were: grain and clay fabric, fractures, veins and voids, clay

illuviation and cementation.

3.3. Results of grain-size and bulk density analysis

Sand (63-2000 pm), silt (2-63 pm), and clay (< 2 pm) percentages of the 52 samples are

represented in Figure 3.1. Grain-size histograms for the principal lithofacies are shown in Figure

3.2. The Sirius Group from the central Transantarctic Mountains consists mainly of silty

diamictons, with the exception of some diamictons on Mount Sirius. Stratified sands and gravels

have a sandy matrix with more than 75 % sand. A structureless mudstone from Oliver Bluffs contains more than 80 % silt (Figure 3.1). The bulk density of Sirius Group diamictons generally fluctuates between 2.2 and 2.5 g/cm^. Two samples from the same outcrop at Roberts Massif form an exception, with considerably lower bulk densities of 1.4 and 1.8 g/cm^.

50 Silt (%) Roberts Massif

Bennett Platform

Meyer Desert

Oliver Bluffs

Mount Sirius

4+

Sand C°/ Clay (°/

g/cm^ Bulk density 3.5

2.5

2 - - 1.5

0.5 CO I 9 OJ CL CL Û. Ü. ma . Q. Q. ë? s s o> o> s I Roberts Bennett Beardmore (b) Massif Platform area

Figure 3.1.Grain-size data from diamictons and stratified facies (a) and bulk density of a selection of Sirius Group diamictons (b) from the central Transantarctic Mountains.

51 %40 % 40 - stratrfied diamicton sand and grave! 20 0 ^ . - 1— ■ r %40 %40 - massive diamicton siltstone 20 20 0 0 - %40 %40 -1 massive diamicton muddy sandstone 20 20

^Ov-cMco^iocor^ooI t • * • I • I t a rj-oActicn^i-uicôiieo''OVCMCO (O (O 03 0303 phi phi

Figure 3.2 Principal lithofacies in the Sirius Group from the central Transantarctic Mountains.

3.3.1 Mount Sirius

The matrices of Mount Sirius diamictons have a fine sand mode and a silt mode (Figure

3.3). The diamictons contain relatively high amounts of clay-sized material (13-22%). Diamictons from the lower 22 meters of the Mount Sirius succession (Unit 1) have a sandy matrix with a fine sand mode. The overlying lithostatigraphic unit (Unit 2) contains stratified diamictons with higher clay content (> 20 %) and less sand than the massive diamictons in Unit 1. Gravel and sand beds in the upper part of Unit 2 are moderately sorted with a medium or fine sand mode.

Massive diamictons in Unit 3 have grain-size distributions similar to those of the massive diamictons in Unit 1, but are even more sand-rich. A stratified diamicton ca. 10 m below the top of the section is very sand-rich. There is a general increase of sand upsection in the diamictons. In

Units 1 and 2, diamictons contain < 40 % sand, whereas diamictons of Unit 3, overlying the gravel and sand units, contain > 50 % sand. The uppermost ca. 6 m of the section consist of poorly sorted medium sand and gravel.

52 (m) m I S I g 20 15 T o p ____ «e 10 Top 5 ^ ------C7 0 m 20 phf 15 S-6 se 10 10 - S 20 s-s 0 CO S-5 ? 5 5 phi ' c 0 20 ZD S-4 S-4 1 20 - se 1015 phi 5 S-3 20 ssëSSBB B î5sW k ® s 8 S-3 0 se 15-10 30 - S-2 5 ss^sM phi s-2 S-1 phi 40 ■ 1 k l i l i i • > - • - e 30 40 — S-8 se 20 2 S 2 s 5 S 5 2 S “ 10 phi S-9 0 s-8 CSJ

50 - ' c 1 1 -r U3 a ei <» 3 2 5 2 s m ^ wS (6 pL «6 ^

60 -

70 - - c : ' Z c 3 80 se 10

-■— till . U ll

11111=1111=11 se 10 90 - =1111=1111= =■ fill = Itll = M

Figure 3.3. Grain-size distributions o f Sirius Group sediments from Mount Sirius.

53 450 m

Section 5 20 15 10 5 10 m n

rilTTîHMI 5-18 0 m “ 5-16 15 DRS-1 1 0 • ____ ii?3E3LÜfeiiiJi;Jy Section 8 5-14

DR5-12 C' afî 10

Unit 5-12 OR5-10 4 ;< 10 Unit 4

0R 5-S b o æ 10 5 1 1

o f 10 Unit 3b

;p 10 Unit Unit 3a 3b ; DR5-2 5-4 3a a %«* 5-2 Unit 2

DMH051 Unit 1

Figure 3.4. Grain-size distributions of the Sirius Group at Oliver Blufifs.

54 3.3.2 Beardmore Glacier area

Grain-size distributions of the Oliver Bluffs succession are highly variable. A diamicton from the base of the Oliver Bluffs section 8 (Unit 1) is moderately sorted and has a very fine sand mode (Figure 3.4.). Diamictons in section 5 (Unit 2, 3 and 4) at Oliver Bluffs have variable grain- size distributions, with low clay contents between 3 and 15 %. A sand-mode is characteristic of diamictons in Unit 3b. A sample from a stratified diamicton from the middle of Unit 4 (sample 5-

12) has a coarse sand matrix, and a very low mud content. A structureless gray mudstone from the upper part of the section (sample DR5-I6) contains much medium and fine silt.

(m) 0 111(1111 10 - I II II I I I

20 - II Til I r I

30 - 40 50 -

i t 10 60 - m m m 70 - LLLLUII

80 90 •

100 -

110

120 - 130 - 10 140 -

Figure 3.5. Grain-size distributions of the Sirius Group at section 13, Meyer Desert.

Stratified diamictons from Section 13 (Meyer Desert) have variable grain-size distributions with high clay contents (11-24 %) and a fine sand mode (Figure 3.5), similar to stratified diamictons from Mount Sirius. A sample from a weathered horizon (DR13-6) has a low

55 weight percentage in the coarse silt fraction. This may be in part due to inaccuracies of the

method, but it is possible that it has a natural cause (Appendix A).

3.3.3 Shackleton Glacier area

At Roberts Massif grain-size distributions of diamictons vary between locations (Figure

3.6). Brown, friable diamicts (samples 002 and 006) immediately above the dolerite erosion

surface contain more clay (14-15 %) than a massive, grey diamicton (sample 003). Diamictons

from one outcrop facing the polar plateau (samples 012 and 013) have high mud contents,

whereas diamictons from another outcrop facing the polar plateau have a more distinct fine sand

mode and lower clay contents.

15 f 105 ' 00 : oos< 00 23 15 ROBERTS MASSIF fO^-023 ^ 105

23 ______phf PNW954303 5 km 15 ^ 10 5 sMI 176 w 0

p h i

PNUV95-022 PMW954)12

PNW95-023 PNW95-013

Figure 3.6. Grain-size distributions of Sirius Group diamictons from Roberts Massif.

56 (m ) m I s I g 20 2 0 C\i 15 PNW9M53 15 PNW9W54 0 - aff 10 i a# 10 E 05 5 2 5 2 % % : % S 10 - p h l

30 20 1 PM W 95-aS6 IS PNW95055 20 ae ^ af 10 20 - 5 8 0 % : S S p w 30 - 20 15 PN W 95-058 15 æ 10 at 10 o m 5 40 - c? 054 ? 5 2 S 5 053 \ o 055 20 15 PNW 95-06D P N W 95-059 056 a* 10 at 10 50 - 057 5 0 c 3 058 S 5 2 o r 059 p h i 60 - CP 060 20 15 PN W 9S-062 ■ 15 Ê o 061 ac 10 aP 10 o 1 5 u_ G 70 - 062 s 5 2 s 063 p h l O 064 PN W 9S-064 PNW9M63 80 — C 7 i i a i s S

90 o \ = PNWas^fiE PNW95-065 C b 067 068 100 SSSSSSS^

20 20 15 15 PNW-067 = 1111 = T l i T 110 H J* 10 ae 10 0s 5 gfSê - Mil = nir =

Figure 3.7. Grain-size distributions of the Sirius Group at Bennett Platform. Last three digits of sample numbers are plotted to the right of the lithological column.

57 Diamictons from Formation 1 of the Bennett Platform succession have a fine-grained

matrix with high silt contents (Figure 3.7). Mudstones of the stratified units (members 5 and 6)

have coarse sût modes and a graded sandstone (laminite) has a fine sand mode (PNW95-059). A

sample at the bottom of the overlying massive diamicton (PNW95-056) has a high clay content

and a fine sand mode. The overlying stratified diamicton beds are silty with Low clay contents (<

10 %).

3.4. Micromorphological observations

The microscopic characteristics of glacial diamicts from temperate glacial environments

have been described from many locations (Van der Meer, 1993; Hiemstra & Van der Meer,

1997). Distinct features associated with subglacial deformation under temperate conditions

include rotational clay and grain fabrics, alignment of phyllosilicates, fractured grains, and evidence for the mobility of water and clay minerals in the form of argillaceous cutans. The micromorphology of diamicts from polar terrestrial environments is not as well studied (Van der

Meer et al., 1994). However, regarding the difference in subglacial processes between temperate and (sub)-polar glaciers, differences in the micromorphological characteristics of their respective diamicts are to be expected.

One diamicton from Dominion Range, Section 13, and the Mount Fleming Upper Till show a strong clay fabric with aligned phyllosilicates (Figure 3.8). The effects of grain-fracturing and abrasion are clearly visible in the Mount Fleming Upper Till sample. Diamictons from Mount

Sirius, Roberts Massif, Bennett Platform and Oliver Bluffs lack a distinct fabric, with only vague rotational features observed (Table 3.1). Two Roberts Massif samples from the same locality have high porosity and contain mammilated vugs.

58 FFTP

K/ . m - M

Figure 3.8. Thin section micrographs of sample 13-2 of the Meyer Desert Formation, section 13 on the upper bench of the Dominion Range (a) and from the Mount Fleming Upper Till (b). Note the aligned biréfringent bands of phyllosilicates, dolerite fragments, coal and poly-crystalline quartz (Qp) in (a) and biréfringent bands of phyllosilicates and fractured grains in (b). Crossed polarizers. Scale bar is 100 micrometer.

59 Sam ple Fabric Voids Cement/Clay illuviation MtFlem#2 skelsepic plasmic fabric, fractured grains, aligned phyllosilicates TM-7a dense grain fabric, fractured grains microcracks zeolite filling pores and microcracks 8-3 skelsepic plasmic fabric patchy carbonate cement S-9 skelsepic plasmic fabric, rotational grain fabric 13-2 skelsepic plasmic fabric, aligned phyllosilicates microcracks secondary silicates, argillaceous cutans 13-8 skelsepic plasmic fabric patchy carbonate cement 5-10 dense grain fabric microcracks patchy carbonate cement 95DMH143 strongly developed rotational grain fabric PNW95-055 rotational grain fabric PNW95-061 skelsepic plasmic fabric, vague rotational grain fabric oval-shaped vugs zeolite In vugs and patchy gypsum cement PNW95-067 vague rotational grain fabric few microcracks carbonate cement In mlcrocracks,arglll. cutans PNW95-073 random microcracks carbonate cement In microcracks and around grains PNW95-021 rotational grain fabric microcracks PNW95-003 random microcracks carbonate cement In pores and microcracks PNW95-0Q6 skelsepic plasmic fabric, vague rotational grain fabric microcracks patchy carbonate cement PNW95-G13 skelsepic plasmic fabric, vague rotational grain fabric mammilated vugs zeolite In pores, patchy carbonate cement PNW95-014 skelsepic plasmic fabric, vague rotational grain fabric mammilated vugs zeolite In pores, secondary silicates a\o PNW95-037 random microcracks carbonate cement In microcracks, zeollte-fllled pores 94QH27 dense grain fabric, in situ fractured grains 94QH42 stratification patchy carbonate cement 94TS14 obscured by Iron staining brown-red oxidized matrix 94TS27 skelsepic plasmic fabric

FABRIC DESCRIPTIONS

m . «D C P „ Q 0 / * skelsepic alligned rotational fractured grains plasmic fabric phyllosilicates grain fabric

Table 3.1. Micromorphological observations on thin sections of selected Sirius Group sediments. Skelsepic plasmic fabric consists of domains of phyllosilicates oriented parallel to grain surfaces (Van der Meer, 1993). 3.5. Discussion

Sirius Group diamictons from the central Transantarctic Mountains are consistently more

silt-rich than glacial deposits from various depositional envirorunents in Victoria Land or the

Victoria Land basin (Figure 3.9), which may be caused by a difference in the grain-size of the

source rocks, or differences in depositional environments.

Most of the Sirius Group diamictons in the Transantarctic Mountains have been classified

as lodgement tills, based on the presence of striated pavements, striated and facetted clasts, or the

presence of a distinct clast orientation fabric (McKelvey et al., 1991; Stroeven & Prentice, 1997;

Wilson et al., 1998a; Hambrey et al., in prep.). The grain-size distributions and bulk densities of

most of the massive diamicts from the central Transantarctic Mountains are in agreement with

this interpretation (cf. Boulton, 1978; Haldorsen, 1981; Dreimanis & Vagners, 1971). Although

between locations differences exist in the character of the grain-size distributions, massive

diamictons in the Sirius Group tend to have unimodal silty matrix grain-size distributions.

Stratified diamictons have a sand mode and a silt or clay mode, which means that the

stratification may have resulted from meltwater reworking in a subglacial or proglacial environment.

Figure 3.10 indicates the similarities between the diamictons of the Sirius Group from the central Transantarctic Mountains and other continental ice-sheet deposits. The data are from

Pleistocene tills from the Northern Hemisphere. Source rock, mode of deposition, and pre-or post depositional weathering determine grain-sizes of tills (Dreimanis & Vagners, 1971, 1972;

Boulton, 1978). The grain-size compositions of Pleistocene tills from Europe and North America

(Figure 3.10) illustrate this process. Tills deposited by continental ice-sheets draining mainly

61 Silt (%) 0 100

50

100 0 50 100 Sand (%) Clay (%)

•” Taylor and Mackay glaciers, basal layer (Barrett. 1989) modern CIROS-1, diamicts, upper 366 mbsf (Barrett, 1989) Oligocene - Miocene

CRP-1 diamicts (DeSantis & Barrett, 1998) Lower Miocene and Quaternary

Till beneath ice-stream B (Tulaczyk et al., 1998) modern Sirius Group of the Dry Valleys (Stroeven & Prentice, 1997; Barrett, 1998)

Figure 3.9. Sand-Silt-Clay percentages of Sirius Group diamictons from the central Transantarctic Mountains compared to other diamictons from Antarctica.

62 Silt (%) 0 . 100

50, 50

100 0 50 100 Sand (%) Clay (%)

Pleistocene tills from Saskatchew an (Christiansen. 1971), Wisconsin (Johnson et al., 1971), Illinois (Kempton et al., 1971), and Ohio (Gross and Moran, 1971; Steiger & Holowaychuk, 1971)

Pleistocene tills from New England (Mulholland, 1976)

Pleistocene tills from the Baltic Shield, (Jiargenson, 1977)

Pleistocene tills from Denmark (J0rgenson, 1977)

A B C Pleistocene tills from Ontario (Dreimanis & Vagners, 1972) ’ ' A= tills derived from igneous/metamorphic rocks 8= tills derived from limestone and dolostone C= tills derived from limestone and shale

Figure 3.10. Sand-Silt-Clay percentages of Sirius Group diamictons from the central Transantarctic Mountains compared to Pleistocene tills from North America and Europe.

63 igneous and metamorphic terrains of the Baltic Canadian, and Greenland Shield areas are coarse­

grained consisting mainly of sand and minor silt. In contrast. Pleistocene tills from Denmark and

the Great Lakes region have a high mud content. The latter are derived from fine-grained

sedimentary rocks, mainly limestones, chalk, and mudstones, which may have been added to the

basal debris zone in the later stages of glacial transport, although the mineralogical composition

of some Pleistocene Dutch tills with high clay contents (10-40 %) suggests derivation from

sources ca. 1000 km to the north. The high concentration of clay minerals with a distant source,

however, can be explained by englacial transport which precluded mixing with other materials

along the pathway from the Baltic Shield to the depositional area of The Netherlands (Haldorsen

et al., 1989).

Diamictons from Oliver Bluffs and Bennett Platform have a composition similar to tills

from continental shield areas, suggesting that they were produced by erosion of fresh,

un weathered crystalline rocks. The fine-sand modes in some diamictons from Mount Sirius may

have been caused by grain-size sorting during redeposition and minor meltwater reworking,

which is suggested by the faint stratification. However, a potential source rock for the Sirius

Group diamictons is the Beacon Supergroup and Haldorsen (1981) showed that the grain-size mode of tills derived from clastic sedimentary rocks is similar to the grain-size mode of the source rock. The Triassic sequence of the Beacon Supergroup has a grain-size mode of medium to fine sand (Barrett, 1966; 1969). Therefore, the Triassic Beacon Supergroup may have been a source rock for the Mount Sirius diamictons with distinct fine sand modes (Figure 3.3), which is supported by the presence of recycled Triassic pollen in Mount Sirius deposits (Askin, pers. comm. 1999).

The high clay contents of the successions at Mount Sirius and section 13 in the Meyer

Desert cannot be explained by a source of Triassic sandstones. Haldorsen (1981) showed that tills

64 derived from sandstone bedrock have low clay content (< 4 %), because crushing and abrasion of

sandstone in the subglacial environment does not produce much clay-sized material. Clay-sized

material in glacial sediments can be derived either from erosion and glacial comminution of

phyllites, or from erosion of shales or weathered bedrock with a high amount of clay minerals.

Alternatively clay minerals may have formed in situ due to post-depositional chemical processes.

The Sirius Group at Mount Sirius and in section 13 in the Meyer Desert either had

another fine-grained source, such as Permian or Triassic Beacon Supergroup mudrocks, or that it

is derived from fine-grained marine or lacustrine source rocks now hidden beneath the ice-sheet,

or was derived from fine-grained formations that are no longer preserved. The diamictons on

Mount Sirius, at Meyer Desert Formation Section 13, and the in Shackleton Glacier area also

have a higher mud content than is expected for tills derived directly from igneous and

metamorphic rocks. One of the samples from Meyer Desert, Section 13 is from a paleosol at the

top of the section (sample 13-8, Figure 3.5), and has a higher clay-sized fraction than any of the

other Sirius Group samples (24 %). This higher clay content suggests that post-depositional

weathering may have affected the clay contents, at least in the proglacial facies.

At one locality at Roberts Massif, the bulk density of the diamicts is very low for a basal

till, and a somewhat higher primary porosity is observed in thin section along with rotational

fabrics. The combination of these features suggests that these deposits may have been affected by periglacial processes after deposition, or that they are redeposited diamictons rather than primary basal tills.

65 « I

Figure 3.11. (a) Frontal apron of Taylor Glacier in the Dry Valleys, Antarctica, which is a polar outlet glacier with minor surface melting (b) and sub-freezing basal conditions. Photo (c) shows the basal debris zone, which consists of attenuated ice and debris, in a tunnel at ea. 500 m upglacier (Principal Investigator: Scan Fitzsimons, Univ. of Otago, New Zealand). 3.6 Glacial thermal regime during déposition of the Sirius Group

In recent years a number of studies of modem glacial environments have been conducted

to develop facies models for polar continental, sub-polar and temperate glacial settings (Shaw,

1977; Eyles, 1983; Evans, 1989; Fitzsimons, 1990). The basal thermal regime and

depositional processes of the three environments are summarized in Table 3.2. At present,

Antarctic glaciers are dry-based with limited surface melting (Lewis et al., 1999; Figure 3.11).

Sub-polar glaciers in the high Arctic exhibit surface melting and fluvio-glacial depositional

environments, but these glaciers are entirely or partially frozen to their beds. Ice-marginal

deposits of temperate glaciers and ice-sheets are dominated by stratified facies formed by

abundant meltwater, which is supplied by both surface melting and basal melting (Ehlers &

Grube, 1983; Ruegg, 1983; Fyffe, 1990). To distinguish between sub-polar and temperate glacial

paleoenvironments in ancient deposits, the characteristics of both the subglacially deposited

diamicts and the proglacial facies should be investigated.

3.6.1 Subglacial conditions

The presence of fractured grains in thin sections of diamictons can be regarded as an indicator of subglacial shearing in a deformable bed (Iverson et al., 1996; Hiemstra & Van der

Meer, 1997). Deformable beds are generally regarded as characteristic of temperate glaciers. Care must be taken when interpreting rotational fabrics alone as indicators of subglacial deformation, since these fabrics also form in periglacial environments and may indicate deformation after deposition and exposure to the atmosphere (Van der Meer et al., 1994). The origin of a skelsepic plasmic fabric (phyllosilicates aligned parallel to grain surfaces; Table 3.1) is not known, but is

67 Type Basal conditions Glacial processes Resulting facies

POLAR -frozen to the bed -sublimation massive diamict -movement due to -entrainment and attenuation of ice and debris stratified diamict ice and sediment -minor meitwater reworking of giaciai debris discrete lenses and bands of sorted sediments deformation -supragiacial rock fail into stratified diamict with dropstones

SUBPOLAR -frozen to the bed -surface melting massive diamict -movement due to -supragiacial debris flows massive or stratified diamict ice and sediment -lateral and proglaciai meitwater transport deformation and deposition stratified sands and gravels 0\ -ice-marginal ponding stratified muds 00 -entrainment of proglacial and lateral deposits brecciated/fauited/foided stratified sediments

TEMPERATE -ice-bed interface -subglacial erosion and transport massive diamict or poorly sorted sediment above pressure -supragiacial debris flows massive or stratified diamict melting point -lateral and proglaciai meitwater transport -basai sliding and and deposition stratified sands and gravels deformabie bed -ice-marginal ponding stratified muds

Table 3.2. Classification of glaciers and their deposits (after; Boulton ,1972; Shaw, 1977; Paterson, 1994; Evans 1989 and Fitzsimons et al., 1990). Note that the boundaries between types arc gradational and that in reality one glacier may not necessarily have all the characteristics of one type. From Polar to Temperate there is a decreased preservation of the primary structure of the ice-eontact sediments due to increasing efficiency of subglacial erosion and deformation in a deformable bed as well as reworking due to the post-depositional water-saturated condition of the sediment. At the same time, the increasing influence of meltwatcr results in greater thicknesses of sorted sediments and more complex stratigraphies. thought to also result from rotational movement of grains within a matrix (Van der Meer, 1993).

To interpret the depositional characteristics of diamictons from thin sections, the combination of several types of micromorphological features should be considered.

The presence of clay illuviation in association with rotational fabrics in sample PNW95-

067 (Bennett Platform) is similar to micromorphological observations of recent tills from the

Shackleton Range in the Transantarctic Mountains (Van der Meer et al., 1994). However, these features may have formed in response to similar periglacial processes, rather than as primary properties of the till.

The diamictons from Mount Fleming and Dominion Range Section 13 exhibit both aligned phyllosilicates and rotational features, indicative of transport and. deposition in a deforming subglacial environment. Mount Fleming diamictons also contain crushed and abraded grains. The strong microfabric of both Upper Fleming Till and Section 13 diamictons (Figure

3.7), suggests that most of the stress exerted by the glaciers was taken up by the till under low pore-water pressures (Van der Meer, 1993; Hiemstra and Van der Meer, 1997; Boulton, 1996).

The thin sections from Section 13 and the Upper Fleming Till were made from samples taken close to the base of each outcrop, and both successions overlie Beacon Supergroup sandstones, which may have allowed dewatering of the deforming bed at the base of the glacier (cf. Boulton,

1996).

The absence of clay fabrics in sediments from Table Mountain, Quartz Hills, Oliver

Bluffs, Roberts Massif, and Bennett Platform may be a result of the low clay contents in these deposits, or may be related to the depositional environment (not subglacial) or subglacial porewater conditions. Vague rotational fabrics alone are not reliable indicators of subglacial transport in a deformable bed, but do not argue against such transport either. The absence of crushed grains and well-developed grain and clay fabrics in many Sirius Group deposits may be

69 Silt (%) 0 JO O R ew orked glacigenic sed im en ts Distal glacio- marine and glaciolacustrine D ebris in 50 ice and b asal tills Fluvioglacial and supragiacial deposits 100 0 50 100 S an d (%) C lay (%)

70 % Eolian 30 %

1 mm. wh

50 % 50% Supragiacial Lacustrine

60% 40 % Fluvioglacial Glaciomarine/ lacustrine

.(4 0.1 1.2 20 ].« 4.S 9^ 0.7 70 0.9 >9 UA .1-0 0.t 1-2 20 M 4.5 SO ft-7 70 M >9

Figure 3.12. Grain-size distributions of Holocene and Recent sediments from the Greenland ice margin. Unpublished data from Koomen et al. (1993).

70 attributed to high pore-water pressures, which limited the interaction between particles during

sub-glacial transport (Tulaczyk et al., 1998; Hiemstra & Van der Meer, 1997). Alternatively, the

process of subglacial deformation may have been inhibited by partly frozen conditions at the ice-

bed interface.

3.6.2 Proglaciai depositional environments

Grain-size distributions of principal lithofacies in the Sirius Group from the central

Transantarctic Mountains (Figure 3.2) show similarities to grain-sizes of sediments in the

terrestrial ice-marginal area of west Greenland (Figure 3.12). Figure 3.12 is a plot of recent and

Holocene deposits from the margin of the west Greenland ice-sheet in the area of continuous

permafrost conditions (Koomen et al., 1993, unpubl. data). Sands and gravels from Unit 2 at

Mount Sirius have a low mud content and high degree of sorting, similar to characteristics of

fluvioglacial deposits from Greenland. Similar grain-size distributions also are found for

materials in sandurs from other present-day Arctic ice caps and glaciers (Church & Gilbert,

1975). The gravels and sands at Oliver Bluffs have a moderately sorted matrix with a low mud

content, which indicates selective sorting in a low energy aqueous depositional environment. A

structureless siltstone from Unit 5 in the Oliver Bluffs Section may be an eolian unit formed by

deposition of wind-blown silt in a terrestrial environment, or may represent deposition in a small

aquatic basin or floodplain.

Sub-polar glacial environments, such as occur at the margin of the Greenland ice-sheet, are diamict-dominated, whereas stratified facies usually dominate temperate ice-marginal depositional systems (Figure 3.13; Ehlers & Grube, 1983; Ruegg, 1983; Fyffe, 1990). In temperate glacial environments large quantities of meitwater enter the proglaciai area through

71 a) arid polar glacial environment

<------debris entrainment and attenuation ■ - X glacial deposition

cliff melting minor frontal meitwater ^ apron reworking —I- • ...... ' c , ^ ^

b) sub-polar glacial environment

glacial erosion and transport • - X glacial deposition

surface melting

FIT debris flows ice-cored moraine end moraine

c) temperate glacial environment

< glacial erosion X ------glacial transport and deposition-

surface melting

debris flows outwash . fan

Figure 3.13. Glacial processes and vertical profile models for three different types of glacial thermal regimes. General idea based on Eyles et al., 1983, (a) based on Shaw (1977), Lewis et al. (1999), (b) based on Evans (1989), Fitzsimons (1990), Boulton (1972), (c) based on Shaw (1987), Ehlers & Grube (1983), Fyffe (1990), Brennand & Sharpe (1993), Van der Wateren (1994). Note the transition from diamict-dominated tomeltwater-dominated systems.

72 subglacial and englacial conduits. The preservation potential of ice-marginal deposits in

temperate glacial environments is lower than in sub-polar environments, because of the efficiency

of subglacial erosion and the erosive power of proglaciai meitwater systems.

Meitwater reworking is recorded by the presence of stratified, moderately sorted facies in

most Sirius Group successions. However, the poorly sorted nature of most sediments and the

limited thickness of stratified facies suggest that meitwater had only a relatively limited role in

the overall depositional system. From the limited information that is available about the facies architecture and the relative proportion of facies preserved in the outcrops, it appears that most

Sirius Group successions are diamict-dominated, which is different than sequences deposited in temperate, meltwater-dominated Pleistocene ice-marginal environments (Figure 3.13; Ehlers &

Grube, 1983; Ruegg, 1983; Fyffe, 1990).

Differences in the abundance of meitwater deposits at different Sirius Group outcrops can be explained by differences in depositional setting (ice-lateral versus ice-frontal), differences in paleoclimate, or differences in preservation of the sedimentary successions. Because some Sirius

Group deposits may have been uplifted after deposition (e.g.Webb et al., 1996a), paleoenvironmental conditions during the time of deposition probably also were controlled by their lower elevation above sea level.

The exact depositional environment (e.g., lateral or frontal ice margin) for each outcrop cannot be established here, since such interpretations require information about the lateral extent and architecture of the facies associations. However, lateral moraines tend to constain more sandy diamictons and breccias due to a higher contribution of supragiacial debris and rock fall (Boulton,

1978). There is no evidence that the Sirius Group consists entirely of lateral moraines (as was proposed by Prentice et al. (1986) for the Oliver Bluffs succession), but it cannot be ruled out that the upper parts of some of the outcrops consist partly of such material.

73 3.7 Conclusions

Samples of the Upper Fleming Till and Meyer Desert Formation (section 13) of the Sirius

Group show microscopic evidence of subglacial deformation under temperate, wet-based

subglacial conditions. It remains difficult to assess the subglacial depositional conditions of the

other Sirius Group diamictons. Although the successions are diamict-dominated, the presence of

stratified and grain-size sorted facies in the Sirius Group of the central Transantarctic Mountains

indicates that significant surface melting was occurring at the time of deposition, more than in the

present dry-polar continental Antarctic climate. However, whether the surface melting was also

associated with significant basal melting with subglacial conduit flow, as is more typical of

temperate glacial environments, remains an open question that needs to be addressed with more

detailed outcrop studies and sampling campaigns.

Depositional environments of the Sirius Group show similarities to depositional

environments observed at the margin of the present-day Greenland ice-sheet, where the ice-sheet

enters the zone of continuous permafrost conditions. However, the facies associations and

proportions of principal lithofacies vary considerably between Sirius Group outcrops. For some

Sirius Group deposits the more maritime polar paleoenvironment of deposition may be explained partially by subsequent uplift of the deposits, from a depositional setting near sea level to their present high elevation.

74 CHAPTER 4

GEOCHEMISTRY AND GRAIN SIZES OF GLACIGENE SEDIMENTS IN THE SOUTHERN

PRINCE ALBERT MOUNTAINS, VICTORIA LAND

This chapter was previously published in Annals of Glaciology Vol. 27 (1998) with Anja

L.L.M. Verbers (Physical Geography Department, University of Utrecht), FrederikM. Van der

Wateren (Department of Earth Sciences, Free University, Amsterdam) and Frans J.M. Vermeulen

(NTTG- TNO, National Geological Survey, Section of Geochemical Mapping, Haarlem, The

Netherlands). The data and interpretations are the result of a pilot study using compositional criteria to establish the relation between glaciation and landscape evolution in the Prince Albert

Mountains. The samples were collected by Verbers. The grain-size analyses were carried out by

Passchier at the Free University, Amsterdam in 1996. The geochemical data were produced by

Vermeulen at the Section of Geochemical Mapping NTTG-TNO. Surface exposure ages were provided by Van der Wateren. Ca. 90% of the paper was written by Passchier, and 10 % by Van der Wateren. Minor modifications have been made to the paper to facilitate incorporation into the dissertation. Grainsize data are presented as histograms rather than cumulative grain-size curves and the "background" section was removed because of duplication of parts of Chapter 1.

75 4.1 Introduction

During GANOVEX expeditions VI and VC in 1991, 1992 and 1993, samples were taken by

Anja Verbers in the southern Prince Albert Mountains for grain size and chemical analyses, to

determine the provenance of glacial sediments in this region. Samples were taken from flat-topped

nunataks rising several 100 m above the present ice surface, up to elevations of 2340 m (Figure

4.1). The basement geology of the region is dominated by Devonian Beacon Supergroup

sandstones and conglomerates intruded by Jurassic Ferrar Group sills, and overlain by the

Kirkpatrick Basalt of the Ferrar Group (Skinner & Ricker, 1968). Stacks of lava flows with

thicknesses varying from tens to hundreds of meters occur in the highest nunataks most distal from

the coast (Womer, 1992). Near the Ross Sea shore, basement is exposed and consists of

Precambrian metamorphic complexes and Ordovician Granite Harbour Intrusives.

MBa MBi Ross Sea

2 5 k m

76°go' s - GN

162°E158 E 160°E162°E158 164°E

Figure 4.1 Map of the southern Prince Albert Mountains. BP, Brimstone Peak; FP, Ford Peak; GN Griffin Nunatak; HB, Hughes Bluff; MBa, Morris Basin; MBi, Mount Billing; ON, Outpost Nunatak; RH, Ricker Hills.

76 The summits of the inland nunataks (> 2,000 m a.s.I.) are flat plateaus consisting of

Kirkpatrick Basalt. The glacially streamlined bedrock (direction SW-NE) is covered with discontinuous till blankets at least 1 m in thickness (Verbers & Van der Wateren, 1992). The coarse ftaction of the tills consists of subrounded, and sometimes striated, basalt clasts up to 40 cm in diameter. The surfaces of the tills are extremely weathered; desert pavement, desert varnish and polygons have developed. The till matrix at Griffin Nunatak is reddish brown.

Glacial terraces and flat mountain summits at lower elevations have thin till layers with a varying intensity of weathering on sometimes striated bedrock. Paleo-ice-flow directions shown by striae from the terraces are SW-NE (Verbers & Van der Wateren, 1992). Till matrix is brown to gray in color and fine-grained. The clast composition of these tills is mainly Ferrar Group dolerite and Beacon Supergroup sandstone, and occasionally granite. In Cirque Valley (Ricker Hills), glaciofluvial and glaciolacustrine deposits are preserved at elevations of -1000 m.

A sample from the summit of Mt. Billing (1600 m) contains marine diatoms (Thalassiosira vulnifica and T. lentiginosd) dating from the mid-Pliocene and early Pliocene to Recent. A coarser till from a high terrace at Mt. Howard (1430 m) contains marine diatoms Actinocyclus actinochilus with a First Appearance Datum (FAD) of 3.1 Ma and A. ingens with a range of mid-Miocene to

620 ka (Harwood, 1992; Van der Wateren et al., 1996). A sample from Hughes Bluff (300 m) contains: Thalassiosira vulnifica Gombos, with a range of 3.1-2.2 Ma. The Hughes Bluff deposit occurs at a much lower elevation than the other diatom-bearing tills, and it is located at the margin of the present David Glacier, which is a large outlet glacier of the East Antarctic Ice Sheet with a valley floor scoured to 1,630 m below sea level (Swithinbank, 1988; Verbers and Damm, 1994).

Ice-cored moraines are developed at the NE side of large nunataks. Clasts in these moraines are mainly derived from the Ferrar Group, but Beacon Supergroup sandstone erratics also occur frequently. A large number of morainic ridges without ice-cores at elevations between the present

77 ice surface and the terraces in the study area were also sampled, but will not be discussed in detail.

The main focus in this paper will be on a broad transect of glacial deposits in the area between

Griffin Nunatak and Hughes Bluff (Figure 4.1).

4.2 Method

A combination of chemical and grain-size analyses was used to determine the provenance of the tills. Geochemical analysis of till matrices (< 2mm) was performed by XRF and ICP-MS

(Vermeulen, 1994). Major elements measured were: SiOz, TiOz, AlzOz, FezOg, MnO, CaO, NazO,

KzO, PzOs; trace elements measured were: As, Ba, Cr, Ga, Nb, Ni, Pb, Rb, Sr, Th, V, Y, Zn, Zr.

The sediments are mainly coarse-grained diamicts. For the determination of characteristic grain sizes the matrices (<2 mm) of 116 samples were analyzed using a Fritsch laser particle sizer

(Analysette 22). Samples were first sieved over 2 mm. Samples were pre-treated by oxidation with

HzOz to remove organic compounds and with HCl to remove inorganic compounds such as CaCOz and salts. To avoid coagulation of particles in suspension by the presence of positive ions an excess 800 mL of distilled water was added. After 24 hours the water with the dissolved salts was removed. Peptisation was supported by boiling the suspensions with Na^PzO?. The Fritsch laser particle sizer determines grain sizes at 0.25 Phi intervals between 0.1 and 1400 p.m by scanning individual grains. Long-time measurements were made (9000 scans per sample) to get the best result for the broad size distributions usually encountered in glacial sediments. The two coarsest sand fractions (up to 2 mm) are extrapolated.

78 Sample BP BP GN GN ON A B

SiOz 56.88 56.58 57.25 56.79 55.74 56.01 56.66 TiOz 2 . 1 2 2.40 1 .8 6 2.13 2.75 2.33 1.93 AlzOz 11.74 11.51 11.78 11.58 10.75 11.56 1 2 . 1 0 FezOz 16.15 16.57 16.24 16.52 18.09 15.31 15.32 MnO 0.18 0.18 0.19 0.18 0.19 0.23 0 . 2 1 MgO 1.25 1.39 1.13 1 .2 0 1.23 2 . 1 2 2.30 CaO 6.76 6.42 6.38 6.45 6.17 7.15 6.92 NazO 2.63 2.61 2.73 2.72 2.64 2.41 2.17 KzO 1.81 1 .8 6 1.93 1.91 1.92 1.77 2.14 P2O5 0.30 0.29 0.33 0.32 0.33 0.26 0.27 Rb 73 72 80 77 77 65 71 Sr 133 127 143 132 118 128 128 V 449 541 394 440 541 391 Cr 18 26 26 25 28 18 Ni 2 0 23 17 18 18 2 2 Ba 552 591 587 597 653 398 Y 53 52 58 55 56 60 Nb 14 16 14 17 18 13 Zr 246 257 263 265 285 252 Th 4 4 8 7 9 1 0

Notes: BP=Brimstone Peak, GN=Griffin Nunatak, ON=Outpost Nunatak

A=fIow 12 (top of Kirkpatrick Lavas) at Storm Peak, Queen Alexandra Range, Faure et al., 1974 B=flow at top Mesa Range lava pile, Siders & Elliot, 1985

Table 4.1. Chemical composition of Sirius Group till samples from the Kirkpatrick Basalt summit plateaus (> 2000 m) compared to geochemical data from the Kirkpatrick Basalt lava flows (main elements in wt % and trace elements in ppm).

79 4.3 Results

4.3.1 Geochemistry

The chemical composition of the tills on the highest summit plateaus (>2,000 m a.s.l.) shows

high values of TiOz and FeiOs, and low values of SiOi (Table 4.1; Figure 4.2). They are

chemically very similar to Kirkpatrick Basalts and have very low regional variability. TiOz and

FezO] concentrations of these tiUs are higher than or comparable to the enriched cap rock covering

the Kirkpatrick Basalt lava flows (Faure et al., 1974; Siders & Elliot, 1985). The chemical

composition of samples from lower summits and terraces (1200-1600 m. a.s.l.) is more variable.

The TiOz and FezOz concentrations are lower than in tills from the highest summit plateaus (Table

4.2; Figure 4.2).

RH RH. RH MBi MH MH HB GN fluvial lacust valley summit summitterrace icm

SiOz 76.18 69.98 74.72 65.78 66.03 72.28 71.53 71.70 TiOz 0.51 0.60 0.51 0.79 0.79 0.72 0.28 0.64 AlzOz 14.41 18.66 15.23 15.16 11.56 11.79 14.70 17.78 FezOz 4.02 4.66 4.13 7.66 9.04 6.42 2.33 3.79 MnO 0.05 0.05 0.05 0 . 1 2 0.14 0.08 0.05 0.05 MgO 0.73 0 . 8 8 0.80 2.08 2.65 1.14 0.40 0.77 CaO 0.69 1.05 0.83 4.36 6.05 3.25 2.46 1 .2 0 NazO 0.80 0.84 1 .0 0 1.43 1.70 1.33 4.04 0.91 KzO 2.41 3.07 2.54 2.38 1.79 2.73 3.93 2.93 P2O5 0.06 0.06 0.06 0 . 1 2 0.14 0 . 1 2 0 . 1 1 0.06 Ba 562 707 601 471 475 516 808 738 Rb 1 1 0 144 118 1 0 0 71 1 1 0 155 131 Sr 105 128 125 117 119 1 0 1 494 159 V 63 79 67 167 2 2 0 148 24 78 Y 27 27 28 29 27 24 17 32

For key o f abbreviations see figure 1

Table 4.2. Chemical composition of a selection of till samples from terraces (Sirius Group), valleys and Pleistocene ice-cored moraines. (Main elements in wt % and trace elements in ppm).

80 Some chemical data show a resemblance to the chemical analyses of Ferrar Group dolerites

(Kyle, 1980). However, the tills from the lower summits and terraces are all S10%-enriched. The

samples from the ice-cored moraines have characteristically high SiO? contents and high levels of

Ba. Hughes Bluff samples have high Na%0, K 2O, Ba and Sr contents (Table 4.2).

3.00

0 □ plateaus > 2000 m 2.50 □ A low e r plateaus □ □ • ice-cored moraines

S 1.50 A

^ A 0.00 . 50.00 60.00 70.00 80.00

S iO ,

Figure 4.2. Wt. % TiOi plotted versus wt. % SiOi of till blankets and ice-cored moraines in the Prince Albert Mountains.

4.3.2 Grain-size analysis

Although chemically very similar, samples from the summits of Brimstone Peak (2340 m).

Outpost Nunatak (2170 m) and Griffin Nunatak (2260 m) have variable grain size distributions.

Sandy tills from Griffin Nunatak (Figure 4.3) and the W-side of Brimstone Peak have high (>25%) silt+clay contents. Samples from Outpost Nunatak and Brimstone Peak are much coarser (< 5% silt+clay). Samples from ice-cored moraines at the NE-side of Griffin Nunatak are fine-grained

81 (>50 % silt+clay) (Figure 4.3), which suggests that the tills from the summit and high terraces of

Griffin Nunatak have quite different provenance compared with the ice-cored moraines.

Bimodal silty sandy tills occur at a high terrace (1370 m) and the summit (1600 m) of Mt.

Billing (Figure 4.4), at Ford Peak (1360 m) and at terraces (1300-1650 m) at Ricker Hills (Figure

4.5). Samples from Mt. Howard (1300-1450 m) are coarser-grained as are some of the Ricker Hills samples.

Grain sizes of presumably glaciolacustrine deposits and a glaciofluvial deposit in valleys at

900-1000 m elevation in Cirque Valley (Ricker Hills) are plotted in Figure 4.6. The glaciolacustrine deposits contain up to 52 % silt and clay, whereas the glaciofluvial deposit contains 80% sand. The age of these glaciolacustrine and glaciofluvial sediments is unknown, but today a small frozen lake covers the valley floor. Lake levels must have been higher, probably during the last glacial maximum, because lacustrine sediments are found up to 25 m above the present lake level.

82 Ice-cored moraines Summit/high terraces

— UTrn UJ rrnTTnTnTrm

10

% 15

10

nTtTt-n I n-rrrrTTn 11111 irrTTTTrrn-re

1 —nrffi Mil Tit TliliTrTiïTrriTrTTTn-rrr>r^ %

- i d M . T? liïmTiïliïTïïTlTnTnTTTTn^ , -1 0 1 2 3 4 5 6 7 8 9 10 11 12 Phi

Figure 4.3. Grain-size data of tills from Griffin Nunatak. Note difference between tills from summit and high terraces (Sirius Group) and ice-cored moraines.

83 25 % - MH 15

f ...Ih-p—I---,---,---r— prrp I r p n p

[qTrTpiniLMii.l|llifmTvn.

lîTTTTTfTTTTl]îlM JhT7TTn7frr7l1l|TU|LLI|iiqxi-n

Figure 4.4. Grain-size data of tills from Mount Billing, Mount Howard, and Ford Peak (all Sirius Group).

84 ITTT#ÎTTTi7n^ E J ü I I I l w

uïTTTTfriT|Tll]]IT]7rrprTT^ lïïTIIIimifTTTi^

m^nprrptiiJiiiiiillfrnTm^

h m rrT TTTnillllllllliïlI.nTrnT^ irnmiii

lll|lll[Ttl|lll|lll|lll|nTrrT>i—,

Figure 4.5. Grain-size data of tills from Ricker Hills/ Mlorris Basin (all Sirius Group).

85 % glaciolacustrine glaciolacustrine

DT

% IS glaciolacustrine glaciolacustrine

10 10 •

5

0

% % 15 15

glaciolacustrine glaciofluvial

1 0 - 10 ■

-1 0 1 2 3 4 5 6 7 8 9 10 11 12 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 Phi Phi

Figure 4.6. Grain-size data of one glaciofluvial sample and five glaciolacustrine samples from Cirque Valley (Ricker Hills).

8 6 4.4 Discussion

4.4.1 Stratigraphy

We distinguish three groups of glacial deposits which we discuss below in order of decreasing

elevation and inferred age. A summary of results and interpretations is given in Table 4.3.

1. The chemical composition of the tills on the high summit plateaus (>2,000 m a.s.l.) is very

similar to the composition of the Kirkpatrick Basalt (Table 4.1). The Kirkpatrick Basalt is

preserved locally in the Transantarctic Mountains and stretches from northern Victoria Land

to the Pensacola Mountains (Kyle, 1980). The presence of glacially sculptured bedrock

overlain by a glacial diamict points to former glaciation of the Kirkpatrick Basalt plateau,

which, since there is no indication of admixture of other lithologies (notably basement),

occurred prior to its dissection by glacial valleys. The highly variable grain sizes may be

associated with variable crystal size of the source rock.

2. The bimodal silty-sandy tills at Mt. Billing (1370 and 1600 m) have a mixed chemical

composition, of Beacon and Ferrar provenance (Table 4.2 and 3). Higher levels of 810% are

caused by the admixture of Beacon Supergroup material. The discovery of diatoms with age

ranges of early and mid-Pliocene to Recent allows correlation with deposits from the Sirius

Group elsewhere in the Transantarctic Mountains, assuming that the diatoms were deposited

by ice. Samples from Mt. Howard are much more coarse-grained, but also contain marine

diatoms and are geochemically similar to the Mt. Billing samples. Glaciolacustrine and

occur at Cirque Valley (Ricker Hills). These probably date from the late

Pleistocene. Meitwater must have been present during part of the summer season to produce

these deposits.

87 The composition of the Hughes Bluff samples is significantly different from that of other

tills within this group. High Na^O, KoO, Ba and Sr contents indicate a granitic basement

provenance. David Glacier eroded a deep trench down to 1,630 m beneath sea level into

basement rocks (Swithinbank, 1988, Verbers and Damm, 1994). Even minor Pleistocene

fluctuations of David Glacier covered Hughes Bluff, which is only 10 m above the present ice

surface. The presence of mid-Pliocene diatoms in the deposit, and the basement provenance,

suggest that during the mid-Pliocene, the David Glacier valley was a fjord from which these

deposits originated. Therefore, the formation of the large outlet glacier valleys, commonly

coinciding with major transverse faults, probably occurred before the Pliocene (Van der

Wateren et al., 1996).

3. Ice-cored moraines in an extensive area on the NE (downwind) side of Griffin

Nunatak are fine-grained (>50 % silt+clay. Figure 4.3). Their chemical composition suggests

a mixed provenance of Beacon Supergroup, Ferrar Group and basement rocks. The

enrichment in SiOi is probably a result of reworking of sediments. The moraines, which occur

up to -10 m above the present ice surface, were deposited by sublimation of the ice. Major

blue ice areas and associated ice-cored moraines were observed NE of Griffin Nunatak,

Ricker Hills and Brimstone Peak. A negative mass balance, resulting from low accumulation

rates and high ablation (sublimation) rates on the leesides of nunataks, is compensated by

upward ice flow (Orheim & Lucchitta, 1990), which produces ice flow opposite to the

regional trend at Griffin Nunatak and Ricker Hills (Morris Basin). This process forces

material from the basal transport zone up into a high level transport zone, and therefore,

material sampled from the ice-cored moraines at Griffin Nunatak (Figure 4.3) represents late

Pleistocene to Recent basal tiU.

8 8 4.4.2 Glacial history

The presence of glacial erosion features amd till blankets on the highest plateaus suggests that

it was covered by temperate ice; i.e. ice sliding over bedrock under conditions of basal melting.

Therefore, either large ice-thicknesses are needled to produce basal ice at the pressure melting

point, or small ice-caps under conditions warm-er than today. The direction of the streamlined

bedrock features (SW-NE), favors continental glaciation, but does not exclude a model with local

glaciers. No marine diatoms were found in the deposits associated with this glacial phase,

suggesting that this glaciation predated the earl]y-mid-PIiocene deglaciation phase inferred from the

diatoms in the Sirius Group sediments.

Burckle et al. (1997) have proposed an alternative mechanism for the emplacement of marine diatoms into glacigene sediments: by eolian tramsport onto the ice-sheet, followed by glacial

transport to the base of the ice-sheet or the ice-sheet margin. According to their model, for diatoms blown onto the ice-sheet 1800 km from the ice-jnargin, the time to reach the bed or be deposited is up to 380,000 yr. This mechanism produces similar age constraints for the Sirius Group as Webb

& Harwood’s (1984) deglaciation hypothesis. Tlherefore, we assume that the late Pliocene age of the Sirius Group is correct and that glaciation o f the highest plateaus in the Prince Albert

Mountains predated the early Pliocene age of th e diatoms in the Sirius Group.

The regional SW-NE trending striation patitems on the lower plateaus and high terraces point to East Antarctic Ice Sheet advance across an area more than ten times wider than the present

David Glacier. Their present high elevation is probably due to a combination of tectonic uplift and isostatic response to valley downcutting (Van d e r Wateren et al., 1996).

Our model of the landscape evolution of th«e Prince Albert block is supported by analyses of in situ cosmogenic nuclides *°Be, "‘’Al and "*Ne (Van der Wateren et al., 1996, 1997). The second

89 and third authors collected quartz samples for ‘°Be and ‘®AI surface exposure dating from summit

plateaus, erosion terraces and lateral moraines. As elsewhere in central and northern Victoria

Land, exposure ages generally increase with elevation (Van der Wateren et al., 1996), suggesting

that lowering of the ice surface gradually exposed the topography. Yet, since in the Prince Albert

block it could be demonstrated that valley erosion postdates deglaciation of the summit plateaus,

we believe that thinning of the East Antarctic Ice Sheet since the Miocene, as suggested by Denton

et al. (1991), was not the main cause for the observed exposure age pattern in this region.

Measurements of cosmogenic ‘°Be, “ A1 and "‘Ne are in agreement with a late Pliocene

depositional age of Sirius Group deposits on the lower summits and high terraces (Van der

Wateren et al., 1996, 1997). The summit of Ricker Hills (1650 m) produced ‘°Be minimum

exposure ages of 1.26 Ma and 0.89 Ma. We modeled ‘°Be concentrations for a number of samples

from summit plateaus and erosion terraces undergoing different uplift rates. Surface uplift rates

between 0.5 and 1 km/My produced exposure ages which are in good agreement with a late

Pliocene age of the supposedly Sirius Group tills. Including ‘®A1 in our model calculations

produces 0.66 km/My since 2.5 Ma as the most likely uplift rate for the measured ^°Be and "®Al

concentrations, including -300 ky of burial of the surface by ice during this period.

Plio/Pleistocene surface uplift becomes even more likely considering that downcutting of more

than 90 per cent of the original summit erosion surface to an average -1000 m lower level would

produce 500 to 600 m of erosional rebound.

Ice-cored moraines on the lee-sides of large nunataks indicate cold climate conditions in the

area since the formation of the moraines, and the intact sediment cover of sublimation till points to

the absence of meitwater during this time. The moraines are probably late Pleistocene in age, which is confirmed by a '°Be exposure age of 29.4 ka from an ice-cored moraine at Morris Basin.

9 0 Van der Wateren & Verbers (1994) proposed that the present landscape of the area is the

result of at least three distinct glacial episodes. The summit plateaus up to 2340 m represent the

original roughly planar surface, which was later dissected by downcutting of valley glaciers along

fault lines in the late Pliocene, to form the highest level of terraces, where the diatom-bearing

Sirius Group tills were deposited. Grain size and chemical data of the sediments, biostratigraphy

and surface exposure ages confirm this model. The terraces are interpreted as the floors of former

valleys. A similar association of high glaciated platforms and diatom-bearing Sirius Group

sediments in high elevated valleys is found in the Shackleton, Reedy and Beardmore Glacier areas

(Mercer, 1968, 1972; Mayewski & Goldthwait, 1985; McKelvey et al., 1991).

Mercer (1972) originally defined tlie Sirius Formation as compact drift that unconformably

overlies pre-Tertiary rocks on Mt. Sirius in the Beardmore Glacier area. According to this

definition both pre-Pliocene plateau tills and Pliocene tills on lower summits and terraces belong

to what was later called the Sirius Group (McKelvey et al., 1991). It is possible that major trunk

valleys already existed when the highest plateaus in the Prince Albert Mountains were glaciated,

perhaps by small ice caps covering the plateaus. The plateaus were later dissected into smaller

nunataks, and the diatom-bearing Sirius tills were deposited on terraces cut into these plateaus.

The current debate considering the age of the Sirius Group developed as a result of interpretations of both these pre-Pliocene tills and the diatom-bearing strata. The result is that, considering glacial deposits at high mountain summits, some interpret the Sirius Group as a thin sequence of pre-

Pliocene glacial deposits, whereas others, considering diatom-bearing deposits in valleys and on terraces, date the Sirius Group as mainly Pliocene. However, McKelvey et al. (1991) already suggested that the Sirius Group may include deposits of significantly different ages.

91 4.5 Conclusions

The results of our till provenance study are in agreement with at least three major glacial

episodes in the southern Prince Albert Mountains. Glacial erosion and deposition of a till blanket

suggest temperate glacial conditions for a glacial phase preceding dissection of the landscape.

Fjord conditions in David Glacier Valley during the early-mid-PIiocene are evidence of a major

deglaciation phase during that period. Granitic basement tills, morainic ridges and ice-cored

moraines are evidence of a higher Pleistocene ice surface. We propose the following model for the

palaeoclimatic history of the southern Prince Albert Mountains:

1. Pre-Pliocene glaciation of the Kirkpatrick Basalt plateau, which implies glaciation of the

Kirkpatrick Basalt by temperate ice without erosion of pre-Jurassic rocks. This episode represents the first phase of Sirius Group deposition.

2. Deglaciation and marine transgression in the David Glacier trench.

3. East Antarctic Ice Sheet readvance and dissection of the landscape followed by deposition of diatom-bearing Sirius Group tills later than mid-Pliocene.

4. Partial deglaciation and cirque glaciation of the nunataks (Verbers & Damm, 1994).

Glaciolacustrine and glaciofluvial deposits are evidence of the presence of meitwater at the

end of this glacial phase.

5. Pleistocene glaciations. Morainic ridges near the present ice surface, ice-cored moraines and

granitic tills are remnants of this phase. The composition of the tills indicates further

dissection of the landscape down into basement.

The stratigraphy and geomorphology of the southern Prince Albert Mountains are very similar to areas south of the Dry Valleys. We believe that the presence of clasts with Pliocene marine

9 2 diatoms in the Sirius Group requires emplacement by a late Pliocene temperate continental ice

sheet, however, some of the diatoms in surface units of thin till blankets may be transported in a

different way. In the southern Prince Albert Mountains the diatom ages are confirmed by

minimum surface exposure ages. A problem is that thick sequences of diatom-bearing Sirius

Group tills are absent in the Dry Valleys. This paper does not address this problem, but it contributes to our understanding of the terrestrial Cenozoic glacial record, showing that the Dry

Valleys form a break in an otherwise continuous sequence of temperate Neogene glacial deposits.

Correlation of the terrestrial glacial geologic record and the record from the Ross Sea Basin may be useful in reconstructing glacial dynamics and tectonic processes. In addition, the sedimentology of the Sirius Group has never been described in detail and correlations have been based on diatoms alone. The Sirius Group comprises a variety of deposits: lodgement tills, glaciolacustrine, glaciofluvial and marine deposits. A detailed sedimentological, mineralogical and geochemical analysis of the Sirius Group from several areas in the Transantarctic Mountains will provide a better basis for correlations and, if the biostratigraphy is correct, may provide information about the nature of Neogene climate in Antarctica.

93 CHAPTERS

MINERALOGY OF THE SIRIUS GROUP AND RELATED UPPER CENOZOIC GLACIAL

DEPOSITS IN THE TRANSANT ARCTIC MOUNTAINS

5.1 Introduction

In this chapter. X-ray diffraction (XRD), analysis of detrital modes, and heavy mineral

analysis, are applied to sediments of the Sirius Group to determine their mineralogical

composition, in order to identify their source. Three techniques were used, because ice-proximal

sediments are generally poorly sorted and have a high matrix content. The sand fraction was

studied by optical microscopic analysis of thin sections in order to quantify the contribution of different types of sedimentary and crystalline source rocks (Pettijohn et al., 1987). Heavy liquid separations were used to isolate diagnostic accessory minerals, which are a powerful tool for identifying specific source rocks of glacial sediments (Gwyn and Dreimanis, 1979; Gravenor,

1979, Polozek & Ehrmann, 1998). In addition, analysis by XRD provides semi-quantitative mineral compositions, especially of the fine fraction of the sediments, which cannot be examined optically. Bulk X-ray mineralogy has provided good results in other provenance studies of high latitude marine successions, such as Krissek (1989) and Bohrmann & Ehrmann (1991).

94 5.2 Methods

5.2.1 X-ray diffraction analysis (XRD)

One hundred and forty-five samples from 11 measured sections and surficial outcrops of the Sirius Group were selected for bulk XRD analysis. In addition, eleven samples from the

DVDP-11 core and one sample from the lower waterlaid diamicton at Propect Mesa were analyzed. The samples were dried at 60°C and powdered in a ball mill for 5 minutes. XRD procedures generally followed those outlined by Krissek (1989), although no internal standard was added, because the data were only used for a qualitative assessment of the dominant minerals, and to evaluate variability in mineral abundances (e.g. Forsberg et al., 1999). The powders were pressed into aluminum sample holders at random orientation and glycolated

(ethylene glycol) for 12 hours prior to the analysis. Samples were analyzed on a Rigaku Miniflex

XRD system with an automated sample changer at the Crary Laboratory in McMurdo Station,

Antarctica. The samples were X-rayed with CuKa radiation (30kV; 15mA) with a scan range from 3 to 65° 20, a step size of 0.02°, and a dwell time of 2 seconds per step. The digital data were interpreted using Jade 3.0 software, which includes a search-match routine based on digital

Powder Diffraction Files (International Center for Diffraction Data). During processing with

Jade 3.0, the X-ray background was removed and the diffractograms were calibrated using known quartz peak positions. Reproducible results (see Appendix B) were obtained for the mineral/quartz ratios listed in Table 5.1. Eight heavy mineral fractions were also X-rayed. The heavy minerals were crushed with a mortar and pestle, and the powder mounts were spiked with alpha-alumina (corundum) for peak calibration purposes. Further procedures were as described for bulk X-ray analysis.

95 Mineral Peaks used Quartz peak Intensity or Area (A) (A) ratio

Feldspar 4.04, 3.19, 3.24, 2.77 3.34 Intensity ratio Pyroxene 2.99, 2.95, 2.89 4.25 Intensity ratio Amphibole 8.40-8.50 4.25 Intensity ratio Mica/niite 10, 4.99 4.25 Intensity ratio Total clays 4.47 4.25 Area ratio Chlorite 7.05-7.20 4.25 Area ratio Calcite 3.03 4.25 Area ratio Chabazite 9.30-9.35 4.25 Area ratio

Table 5.1. D-spacing (Â) of peaks used to calculate relative abundances of minerals from X-ray diffractograms.

5.2.2 Sand mineralogy

Heavy minerals in 42 samples were separated from light minerals by gravity settling in a heavy liquid (Sodium Polytungstate, density 2.89). The samples were selected to maximize geographical and stratigraphie coverage. The Table Mountain Sirius Group was not analyzed, because of very low abundances of heavy minerals in the sediments available. The weight of the heavy mineral fraction was determined, and the magnetic minerals were separated from the heavy mineral fraction and also weighed. Non-magnetic heavy minerals were mounted on glass slides for optical microscope identification using Norland optical adhesive 61 with a refractive index of

1.6. The fine sand fraction (125—250 pm) was used in this study, because identification of minerals using optical properties is very difficult for grains > 250 pm in diameter, and the 63-125 pm fraction of some tills from the Transantarctic Mountains is virtually free of heavy minerals

(Faure et al., 1995). More than 300 grains were counted for each sample.

96 Twenty-two thin sections were examined to characterize the clast petrology and detrital

modes of Sirius Group sediments. For modal analysis at least 300 grains > 60 pm diameter were

counted in each thin section and the percent matrix (< 60 pm) was estimated using visual

comparison charts from Flugel (1978). Secondary minerals and post-depositional sediment

remobilization were also noted.

5.3 Results

5.3.1 X-ray bulk mineralogy

The mineralogy of sediments of the Sirius Group and related deposits is dominated by

quartz, feldspar and pyroxene (Figure 5.1). Pyroxenes found in the Sirius Group are mainly Al-

augite, titanium-rich clinopyroxenes, pigeonite, and hypersthene. Some sediments contain minor

amounts of amphibole, mica, clay minerals and calcite. Chabazite, a Ca-zeolite, was also detected

locally in poorly sorted sandstone at Table Mountain and in stratified facies at Oliver Bluffs and

Bennett Platform.

Figure 5.2 shows the relative abundances of quartz, feldspar and clinopyroxene in bulk samples of the Sirius Group and related deposits based on actual X-ray counts. Sediments from

Bennett Platform, the Dominion Range and the DVDP-11 core contain significant amounts of pyroxenes and feldspar relative to quartz. Sediments from the upper part of the Mount Sirius succession, Roberts Massif, and the Sirius Group from the Dry Valleys are enriched in quartz.

Table Mountain forms an outlier with relatively high quartz contents.

The Mount Sirius and Tillite Spur sections display a stratigraphie trend in composition along the section, whereas the other successions show cyclic changes related to lithological boundaries, but no distinct trends (Figures 5.3 and 5.4).

97 Mount Sirius Mount Fleming 5000 5000 Q S-9 Q MtFlem#2 4000

Fsp

Q 1000 Q F sp 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 .JL L j ^ I i. i . 3 13 23 33 43 53 63 3 13 23 33 43 53 63

Oliver Bluffs Q uartz Hills 2500 3000 ( ^ DR5-10 Q 9 4 Q H 1 3 2000

1500 F sp VO Fsp 00 1000 . Mica 500

---- jiAl k. 1 . !.. IW1.X . j A L .L i . 0 3 13 23 33 43 53 63 3 13 23 33 43 53 63

Bennett Platform Tillite Spur 3000 7000 ( PNW95-065 Q 9 4 T S 1 6

Fsp 1000 Fsp M c a 500 Q 0 -"1 i i ‘i 1 l^r i4iff iVr 3 13 23 33 43 53 63 3 13 23 33 43 53 63

Figure 5.1. Examples of processed X-ray diffractograms of powdered bulk samples of Sirius Group sediments. Q=quartz; Psp=feldspar; Cpx=c]inopyrDxene; Z=zeolite. Feldspar Feldspar

. DVDP-11

Sirius Group

* — Table Mountain Mount Sirius

Quartz Pyroxene Quartz Pyroxene

Feldspar Feldspar

o o

Section 13 Oliver Bluffs Bennett Platform

Quartz Pyroxene Q u a rtz Pyroxene

Feldspar Feldspar

Quartz Hills

• Tillite Spur

Roberts Massif

Quartz Pyroxene Quartz Pyroxene

Figure 5.2. Relative abundance of quartz, feldspar and pyroxene in Sirius Group and DVDP-11 sediments as determined by bulk XRD analysis: (a) Dry Valleys, (b) and (c)near the Beardmore Glacier, (d) and (e) near the Shackleton Glacier, (f) near the Reedy Glacier. Quartz=4.25> A; Pyroxene and Feldspar abundances are based on the averages of the peaks in Table 5.1.

99 Caldte/ Total clays/ Chlorite/ Feldspar/ Pyroxene/ Mica/ % s a n d Mount Quartz Quartz Quartz Quartz Quartz Quartz Sirius 0.5 0.5 0 0.5 0.5 50 100

(m) o

30

50 O 60 -

70

SO

Feldspar/ Pyroxene/ Amphibole/ Calcite/ Total clays/ Chlorite/ Oliver Quartz Quartz Quartz Quartz Quartz Quartz I sand Bluffs 1 0 0.5 1 0 0.5 0 1 2 0 0.5 1 0 0.5 10 SO 100

(ml

O 20

3 0 -

40 ~

Feldspar/ Pyroxene/ Amphibole/ Calcite/ Total clays/ Chlorite/ % sand Bennett Quartz Quartz Quartz Quartz - Quartz_ . Quartz - . Platform 0.5 0.5 0 0.5 0.5 50 100 (m). c

55-

65- O

85-

Figure 5.3. Mineral/quartz ratios firom bulk XRD analysis from Sirius Group outcrops in the central Transantarctic Mountains.

1 0 0 Tillite Fsp/Q Pyx/Q Mica/Q Calc/Q Clays/Q Chl/Q

Spur 0 .1 0 ^ 0.3 0 0.2

> ^ V G

2 0 -

(m )

Fsp/Q Pyx/Q Amph/Q Mica/Q Calc/Q Illite/Q GhI/Q Quartz Hills 0.0 1.5 0.0 0.3 0.0 1.0 0.0 20 0.0 1.0 0.0 0.2 0.0 1.0 CP' c 10 -

^ 30 -

50 -

3 70 (m) o

Figure 5.4. Mineral/quartz ratios from XRD bulk mineralogy in Sirius Group successions at Tillite Spur and Quartz Hills, Reedy Glacier area.

1 0 1 DVDP-11 Fsp/Q Pyx/Q Amph/Q Calcite/Q Clays/Q Chlorite/Q 2.0 0.00.0 1.5 0.0 2.0 0.0 2.0 0.00 0.50 0.0 1.5

(m)

a.

100 -

a. 200 -

300 - '

Figure 5.5. Mineral/Quartz ratios from XRD bulk mineralogy for the Neogene part of DVDP-11.

102 The Neogene record of the DVDP-11 core shows relatively constant feldspar contents

and an upward increase in the relative abundance of pyroxenes (Figure 5.5). At Mount Sirius

pyroxene contents are low throughout the succession, whereas feldspar and total clay contents

display a stratigraphie pattern (Figure 5.3). At Bennett Platform, pyroxene and amphibole contents, which are the most abundant heavy minerals (see below), vary inversely.

5.3.2 Heavy mineral content, composition and distribution

Clinopyroxenes are the dominant heavy minerals in most localities of the Sirius Group and related deposits (Table 5.2). Several different varieties occur. Colorless clinopyroxenes with cleavage are most abundant. Colorless, light-green and brown clinopyroxenes without cleavage also occur. Pigeonite was identified by XRD and is a major constituent in most samples from the central Transantarctic Mountains and in the lowermost part of DVDP-11 (Figures 5.6).

Titanaugite and aegerine were identified optically in DVDP-11. Orthopyroxene with straight extinction is a major constituent of some heavy mineral fractions. Most abundant are large grains of pleochroic hypersthene.

Garnet is the dominant heavy mineral in samples from Mount Sirius, Mount Fleming and

Mount Feather. Both colorless and pink garnets are present, but colorless garnets are more abundant. Some garnet grains are well rounded, indicating abrasion in an aquatic environment, whereas others have characteristic etch features (Figure 5.7). X-ray diffractograms of the heavy mineral fraction of a sample from Mount Sirius confirmed the presence of almandine garnet

(Table 5.2; Figure 5.6). A third colorless variety with numerous inclusions occurs in samples with low overall garnet content.

103 Location/sample Cpxl Cpx2 Cpx3 Opx Ami Am2 Am3 Gt Stab Ep Bio 01 Ox Op Lith other wfi Roberts Massif

PNW95-002 1 11 20 2 19 <1 4 0 1 20 0 0 10 1 6 5 14 PNW95-012 11 5 33 8 3 0 4 <1 1 8 0 0 9 1 15 3 17 PNW95-013 5 3 48 6 0 <1 1 <1 1 6 0 0 15 2 8 5 13 PNW95-022 3 3 45 2 3 0 1 1 1 17 0 0 14 4 5 2 5 PNW95-023 3 3 48 6 2 0 1 1 0 7 0 0 16 2 9 2 7 PNW95-032 6 8 37 4 2 0 0 0 <1 18 0 0 12 2 7 3 7

Bennett Platform

PNW95-053 6 1 38 10 5 0 2 <1 0 8 0 0 11 1 15 3 22 PNW95-055 13 2 32 5 3 0 1 0 0 8 0 0 15 <1 20 3 25 PNW95-057 10 0 28 5 12 0 6 <1 1 7 0 0 15 1 11 5 14 PNW95-059 15 2 31 5 2 0 2 0 0 6 0 0 19 <1 15 3 13 PNW95-061 7 1 42 7 2 0 3 0 <1 7 0 0 8 2 18 4 21 PNW95-063 15 1 36 3 3 0 1 <1 0 3 0 0 20 1 14 3 21 PNW95-065 5 2 47 5 2 0 4^o 2 <1 <1 2 0 0 19 2 10 5 19 PNW95-066 6 0 33 4 12 0 5 1 0 8 0 0 15 2 8 5 12 PNW95-068 23 3 18 10 4 <1 3 0 <1 15 0 3 16 <1 4 3 18

Mount Sirius

S-6 7 0 18 9 9 0 3 26 3 1 0 0 11 4 5 4 2 8-4 9 1 22 8 <1 0 6 21 2 3 1 1 9 2 10 7 1 S-9 6 <1 13 7 7 0 6 23 2 5 1 0 8 2 10 10 3 S-12 14 0 7 21 3 <1 0 15 4 4 0 0 2 8 18 3 2

Meyer Desert Section 13

13-8 3 <1 34 5 5 0 6 2 <1 4 0 0 15 1 22 3 4 13-4 25 0 24 8 0 0 0 1 2 9 0 2 18 1 7 4 10 13-2 10 0 29 10 1 0 2 1 <1 12 0 0 14 2 9 9 8

Table 5.2. Percentages of minerals in fine sand heavy-minerai fractions. (continued) Table 5.2. (continued)

Location/sample Cpxl Cpx2 Cpx3 Opx Ami A m2 Am3 Gt Stab Ep Bio 01 Ox Op Lith other wt% Oliver Bluffs

5-18 9 0 37 7 7 0 <1 2 2 7 0 1 10 3 10 6 13 5-14 8 1 28 7 6 0 2 1 <1 10 0 0 12 1 21 3 27 5-12 4 0 32 2 5 <1 2 1 <1 10 0 <1 19 1 22 3 18 5-10 22 0 11 9 0 0 5 0 <1 7 0 0 13 4 23 6 9 5-8b 5 1 34 2 5 0 1 0 <1 11 0 0 17 1 19 3 7 5-6 7 <1 33 3 6 0 3 2 1 4 0 0 20 2 14 6 7 5-4 27 0 16 8 0 0 2 <1 3 4 0 0 10 0 25 5 26 5-2 6 1 21 2 2 0 <1 0 0 16 0 0 22 1 25 3 17 95DMH051 15 0 24 7 4 0 2 1 1 8 0 1 14 4 17 3 11

Diy Valley Drilling Project Core 11 (DVDP-11)

21.3-23.8 mbsf 9 1 16 7 <1 1 1 0 5 1 0 18 16 9 13 3 19 120.06-120.16 7 1 22 13 1 3 13 <1 8 4 2 0 16 3 4 4 8 156.52-156.62 10 0 8 8 0 0 26 0 4 <1 1 0 18 5 8 13 14 170.35-170.45 5 0 OLA 16 10 4 5 19 1 6 <1 1 0 18 3 7 6 14 221.94-222.04 7 3 12 8 4 5 19 0 7 2 4 0 8 1 16 3 14 279.90-280.00 2 3 24 6 4 4 10 0 4 2 5 0 19 3 13 2 10 311.62-311.72 3 5 25 25 4 5 12 0 5 2 2 1 3 1 2 5 12 324.95-325.05 1 3 29 24 7 12 5 <1 4 2 1 0 4 <1 6 2 10

Diy Valleys Sirius Group

Mt. Feather 0 0 <1 1 0 0 0 54 4 5 1 0 27 2 6 1 <1 Mt. Flem. #2 0 <1 1 <1 0 0 0 20 16 1 0 0 53 <1 7 3 <1

Prospect Mesa 2 1 44 16 15 1 2 0 0 3 0 0 8 1 1 4 23

Cpxl=brown Aml=colorless Gt=garnet Bio=biotite Op=opaque Cpx2=green Am2=green Stab=stablc minerais Oi=o!ivine Lith=rock fragment Cpx3=coIorless Am3=brown Ep=epidotc Ox=oxidized wt%=wt % heavy minerals Amphiboles in samples from the Central Transantarctic Mountains are mainly colorless

and appear to be of tremolite/actinolite and anthophyllite composition (Figure 5.6). In DVDP-11,

brown-green or blue-green pleochroic amphiboles are more abundant. Epidote consists of light

green to light yellow pleochroic grains in central Transantarctic Mountains samples. These grains

are generally irregular in shape. Much stronger colored lemon-grass-green epidote grains occur in

heavy mineral fractions from the Dry Valleys, and some have characteristic blue and yellow

anomolous interference colours. Accessory minerals include apatite, rutile, zircon, tourmaline,

monazite and sphene, which are mechanically and chemically stable (Morton and Hallsworth,

1999). Most of the stable minerals in the Sirius Group are rounded, but subhedral apatite, zircon, monazite and sphene occur in DVDP-11.

Biotite is quite abundant in some intervals of DVDP-11, and is present in the Mount

Feather Diamicton and the Sirius Group at Mount Sirius (Table 5.2). Chlorite also occurs in trace amounts. Olivine was identified optically in the upper part of DVDP-11. Trace amounts of olivine occur in some heavy mineral fractions of the Sirius Group of the central Transantarctic

Mountains. X-ray diffraction analyses confirmed the presence of ferroan forsterite in samples from DVDP-11 and the Dominion Range.

Some non-magnetic opaque grains were encountered, but could not be identified due to oxide coatings or extreme etching. Rock fragments were encountered in minor amounts, but could not be classified based on optical properties. A ternary plot of clinopyroxene, amphibole and stable minerals illustrates that the heavy mineral fractions of Sirius Group sediments on

Mount Fleming and Mount Feather are almost entirely composed of garnet and stable minerals, such as apatite, rutile, and zircon (Figure 5.8).

106 Mount Sirius DVDP-11 3500 21.3-23.8 mbsf 8 0 0 - Pyx 6 0 0 - c 4 0 0 - I Hb/ Cpx ^ cpx 200 - ■ ? . J y\yi A J l . mIL - 9 . . - 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 20 2 0 Meyer Desert DVDP-11 2000 ■ DR 13-4 120.06-120.16 mbsf 1500 - c r ■ 0 Py> g 1000 - § o <3 500 - Aim n yPX L 0 - ■I—1—t-iW 1 1 Ï—r-T 1 1 r fP r-T 1 -TT'i—1X 1 T"i‘ 1 -c -T'l t 1 2. 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 20 2 0 DVDP-11 1000 - 95DMHQ51 156.52-156.62 mbsf 800 ■ Pyx (0 c 600 • o Pyx Cpx O 400 • Cpx 200 - 0 -f 'ixJ 2! 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 20 2 0 DVDP-11 1000 J 1200 1 PNW95-066 311.62-311.72 mbsf 800 - 1000 ■ Pyx

CO 800 - f 600 - c 3 3 600 ■ 400 ■ Cpx c O o U 400 - “ ? l c 200 - 200 - 0 -t y j 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 20 20

Figure 5.6. X-ray dififractioa analyses of fine-sand heavy-mineral fractions from four Sirius Group locations in the central Transantarctic Mountains and from four depths in the DVDP-11 core. C= corundum (standard), M=mica, Pyx=pyroxene, Cpx=clinopyroxene, Pig=pigeonite, Ca=calcite, Alm=almandine, Hb/Tr=homblende/tremolite, Ac=actinolite.

107 (a)

: :

(b)

Figure 5.7. Garnet grains of different origins: (a) rounded garnet grain probably recycled through sedimentary rocks, (b) etched garnet affected by either burial metamorphism or thermal alteration. Both grains are from the Mount Sirius succession from the size fraction 125-250 micrometer, (a) from sample S-9, and (b) from sample S-6.

108 % Amphiboles

DVDP-11

Victoria Land Sirius Group

% Clinopyroxenes % Garnet + stable minerals Mount Sirius

Figure 5.8. Relative abundances of clinopyroxenes, amphiboles, and stable minerais plus gamet for Sirius Group and DVDP-11 sediments.

In contrast. Dominion Range, Bennett Platform and Roberts Massif deposits are dominated by clinopyroxenes. DVDP-11 samples contain more amphiboles and stable minerals than most Sirius

Group samples. At Mount Sirius, in most heavy mineral fractions gamet is dominant, followed by pyroxenes. A few large grains of sillimanite are also present. Similar to the XRD analyses, stratigraphie distributions of heavy minerals (Figures 5.9 and 5.10) are characterized by rapid shifts in abundance at lithological boundaries. However, since the heavy minerals of the fine-sand fraction of the samples were studied, the compositional differences cannot be explained by grain- size effects.

Percentages of heavy minerals in the fine sand fraction vary from <1 to 27% (Table 5.2).

The highest percentages of heavy minerals occur in deposits at Oliver Bluffs, Bennett Platform

109 wl. % heavy % colorless Oliver brown Cpx % ortho- % amphiboles % epidole % garnet %rock minerals % opaque % oxidlxed Bluffs Cpx pyroxene fragments 30 0 30 0 40 0 to 0 to 0 20 0 30 0

10 -

20 -

30 -

40 -

50 - È 60 -

Bennett **™t Platform

30 30 20 0 30 0 40-

60-

70-

80-

90-

100

Figure 5.9. Stratigraphie distribution of the dominant heavy minerals in Oliver Bluffs and Bennett Platform suceessions of the Sirius Group, eentral Transantarctie Mountains. wl.% heavy % brown Cpx % green Cpx % ortho- % amphiboles % epidote % blotlle % opaque % rock % oxidized DVDP-11 minerals pyroxene fragments

30 0

100 -

200 -

300 -

Figure 5.10. Stratigraphie distribution of dominant heavy minerals in DVDP-11, Dry Valleys. Cpx=clinopyroxene. and Prospect Mesa; the lowest percentages at Mount Sirius, Mount Feather and Mount Fleming.

A clear relationship exists between the gamet content of the heavy mineral fraction and

the weight percentage of heavy minerals in a sample. All samples with >10 % gamet have <5 wt.

% of heavy minerals in the fine sand fraction (Table 5.2). Percentages of magnetic minerals are

generally < 1%, except for Bennett Platform deposits, where some samples contain up to > 2 %

magnetic minerals.

5.3.3 Thin section description and detrital modes

Framework grains in the Sirius Group generally show a poorly-sorted size distribution

(e.g. Figure 5. II). More than 90 % of the grains counted occur in the size range of 60-250 pm.

The percentage matrix (< 60 p.m) material varies from 44 to 71 %. Modal analysis of the thin sections shows that quartz is generally angular and forms the major constituent (Table 5.3).

Monocrystalline quartz is abundant, but polycrystalline quartz also occurs. Some of the quartz is strained. K-feldspars include small amounts of microcline with characteristic twinning, but most of the grains are altered. Plagioclase occurs both as altered grains and as relatively fresh, small, angular grains. Oliver Bluffs and Bennett Platform sediments contain major amounts of plagioclase and clinopyroxene in equal proportions. Opaque grains consist of detrital coal and minor oxides. As with the bulk XRD results, the detrital modes for Table Mountain form an outlier in that more than 90 % of the grains are quartz, and most of them are rounded, in contrast to the other Sirius Group deposits (Figure 5.12a).

The petrology of granules and pebbles is quite uniform. Dolente is the main clast type, but basalt is also recognized in deposits from Mount Sirius, the Dominion Range (section 13) and

Roberts Massif. Sediments from Mount Sirius, the Dominion Range (section 13) and Roberts

112 Sample Qa K-fsp Flag Qr Cpx Opx Amph Gt Mica L-bas L-gran L-met L-vol L-sed Opaq Carb Other (%) (%) (%) (%) (%) (%) (%)(%) (%) (%) (%) (%) (%) (%) (%) (%)(%)

S-3 54 < 1 9 7 2 0 < 1 < 1 < 1 2 0 0 5 10 8 < 1 4 S-9 40 3 10 7 < 1 < 1 < 1 < 1 4 5 0 3 6 13 4 0 4 13-2 43 3 8 7 7 1 0 < 1 < 1 5 0 < 1 13 8 < 1 1 3 13-8 35 5 9 13 8 0 0 < 1 0 7 0 < 1 11 9 0 2 1 5-10 17 < 1 5 11 10 < 1 0 0 < 1 15 0 2 17 4 11 7 2 95DMH143 50 0 14 6 13 0 < 1 0 < 1 2 0 < 1 6 4 4 0 1 PNW95-003 39 18 6 6 10 0 0 0 0 3 0 0 14 2 2 < 1 1 PNW95-006 43 10 11 4 8 1 0 0 < 1 1 0 0 16 5 0 0 1 PNW95-0I3 44 2 11 7 17 1 < 1 0 < 1 5 0 < 1 7 3 < 1 0 3 PNW95-014 41 < 1 11 8 20 0 0 0 0 4 0 1 8 3 1 0 0 PNW95-021 32 11 16 10 11 0 0 0 < 1 3 < 1 < 1 12 3 1 0 1 PNW95-037 37 2 16 3 12 0 < 1 0 0 13 0 < 1 11 3 0 0 1 PNW95-055 27 2 3 22 22 0 0 0 0 14 0 < 1 6 3 1 0 0 PNW95-061 37 0 7 23 14 < 1 < 1 0 0 9 0 0 5 2 1 0 2 PNW95-067 21 3 14 16 16 1 0 0 0 9 0 0 8 10 1 < 1 1 PNW95-073 52 0 9 7 7 < 1 0 0 0 4 0 0 7 8 4 < 1 1 TM-7a 23 67 < 1 2 4 < 1 0 0 0 < 1 0 0 < 1 < 1 < 1 0 4 MtFlem#2 48 8 13 8 0 0 0 0 5 0 < 1 < 1 7 8 1 0 2 94QH27 39 2 25 13 0 0 2 0 14 0 4 < 1 < 1 < 1 < 1 0 2 94QH42 46 2 8 11 0 0 3 0 29 0 < 1 0 < 1 0 < 1 0 2 94TS14 68 < 1 10 4 0 0 0 0 7 0 < 1 4 3 2 0 < 1 2 94TS27 73 2 6 3 0 0 0 0 5 0 < 1 4 < 1 2 3 0 1

Qa = angular quartz Cpx = clinopyroxene L-bas = basic igenous rock fragments L- vol = volcanic rock fragments Qr = rounded quartz Opx = orthopyroxene L-gran = granitic rock fragments Opaq = opaque grains K-fsp = K-feldspar Amph = amphibole L- met = metamorphic rock fragments Carb = carbonate rock fragments Plag= plagioclase Gt = garnet L- sed = sedimentary rock fragments

Table 5.3. Detrital inodes in 22 thin sections of the Sirius Group Fsp «J? ^ 1

- ^ 7‘ -\g)lcaiiic: ; V, -V

Figure 5.11. Thin section micrographs of sample DR5-10 from Oliver Bluffs section 5. (a) In plain light, (b) with crossed polarizers. Note numerous coal fragments, volcanic rock fragments and dolerite fragments. Scale in lower right comer is 50 micrometers.

114 i

Figure 5.12. Thinsection micrographs of poorly sorted sandstone from Table Mountain, sample TM-7a (a) and the Upper Fleming Till, sample MtFlem#2 (b). The Upper Fleming Till has a high abundance of phyllosilicates, which show no preferred orientation. Note the difference in grain shape and quartz composition between (a) and (b): (a) contains more rounded quartz of various varieties, whereas mineral grains in (b) are mainly angular monocrystalline quartz and feldspar. Qm=monocrystalline quartz, Qp=polycrystalline quartz, Qs=strained quartz, Fsp=feldpar. Scale bar is 100 micrometers.

115 Massif also contain volcaniclastic detritus, such as vitric tuff. Detrital carbonate clasts are present

in Dominion Range and Mount Sirius diamictons. Fine-grained metasediment clasts and

significant amounts of mica occur in Mount Sirius deposits. Patchy carbonate cement occurs in a

diamict from the upper part of the Mount Sirius succession. The matrix of Mount Sirius and

Mount Heming diamicts contains abundant fine-grained biréfringent material with random

orientation (Figure 5.12b). Oriented bands of biréfringent material are visible in a sample from

the lower part of the Meyer Desert succession at the Dominion Range and in a sample from

Roberts Massif. Some framework grains have thick rims of biréfringent material and carbonate

appears to be replacing some of the rock fragments. A thin section of a brown diamict exposed

within the dolerite erosion surface at Roberts Massif contains much authigenic material, probably

also carbonate, whereas a thin section of a gray diamict does not indicate any obvious secondary

alteration.

Argillaceous cutans within diamicts at Bennett Platform are evidence of post-depositional

relocation of clay particles, but most diamicts appear to be very fresh and chemically unaltered.

Carbonate cement was identified in thin section around granules of volcanic origin and along the

walls of micro-cracks within diamicts in the lower part of the section on Bennett Platform. Patchy

carbonate cement and carbonate-filled microcracks also occur in the Oliver Bluffs section.

5.4 Source rocks

The XRD analyses show that the bulk composition of the sediments is derived from three main sources: a mafic igneous source and a felsic igneous source, including its weathered and recycled products (Figure 5.13). The most likely mafic igneous source is the Ferrar Group, which

116 Feldspar

o, A' • Oliver Bluffs

o Bennett Platform

• Meyer Desert Section 13

° Mount Sirius

X Roberts Massif

■ Dry Valleys Sirius Group

+ DVDP-11 and Prospect Mesa

o Quartz Hills

° Tillite Spur

Table Mountain

Felsic igneous and Mafic igneous Q u a r t z sedimentary rocks rocks Pyroxene

Figure 5.13. Relative abundances of quartz, feldspar, and pyroxene in Sirius Group and DVDP-11 sediments as determined by XRD analysis, and provenance interpretation. Quartz=4.25 X; Pyroxene and Feldspar abundances are based on averages of the peaks in Table 5.1. forms thick dolente sills and basalt flows in the Transantarctic Mountains. This is confirmed by

the abundance of colorless clinopyroxenes with cleavage in the heavy mineral fractions, and the

identification of pigeonite in the X-ray analyses of the heavy mineral fractions (Figure 5.6).

DVDP-11 and Prospect Mesa may contain minor amounts of pyroxenes derived from the

McMurdo Volcanics Group or metamorphic basement, as is indicated by the presence of

titanaugite and green clinopyroxenes. The Beacon Supergroup, a stratigraphie sequence of

Paleozo ic-Mesozoic sedimentary rocks, is likely the main source for the deposits with higher

feldspar and quartz contents.

5.4.1 Sirius Group of Victoria Land

The Sirius Group at Table Mountain has a very high quartz content, as determined by

XRD and modal analysis (90 % quartz). The greater proportion of the quartz is rounded, in

contrast to the dominance of angular quartz in the other Sirius Group deposits. The sediment from

Table Mountain investigated here is a poorly sorted sandstone, which may have be a glaciofluvial

deposit, so that grain-size sorting may have had an effect on the composition. However, the Sirius

Group at Table Mountain overlies the Devonian part of the Beacon Supergroup (Taylor Group),

which is dominated by sorted quartz arenites with well-rounded grains (Korsch, 1974). Therefore, the deposit may also represent local recycling of Beacon Supergroup sandstone, as was suggested previously by Barrett and Powell (1982).

The low abundance of heavy minerals in the Sirius Group at Mount Fleming and Mount

Feather suggests that sedimentary rocks were the source for the sand-sized components.

Chemically and mechanically stable minerals, such as garnet and apatite, are dominant heavy minerals, but the heavy mineral fraction on Mount Feather also contains epidote and zircon.

118 Possible sources for these heavy minerals include sedimentary rocks, such as the Beacon

Supergroup, the Precambrium to Lower Paleozoic basement and/or the Granite Harbour

Intrusives (Skinner and Ricker, 1968). Brady and McKelvey (1979) reported the presence of

Ferrar Dolerite, Beacon Supergroup and pre-Beacon metasediment clasts in the Sirius Group at

Mount Feather. The color of the matrix of the Feather Diamicton suggested to Brady and

McKelvey (1983) that it was reworked from the Triassic Lashly Formation of the Beacon

Supergroup. The Feather Diamicton contains Triassic pollen (Askin, 1998), which confirms the contribution of the Triassic Beacon Supergroup to the matrix of the sediments.

The difference in heavy mineral composition of the Dry Valleys Sirius Group and diamicts from the nearby Prince Albert Mountains, between David and Mawson Glaciers, is striking. The heavy mineral fractions of diamicts in the Prince Albert Mountains consist almost entirely of pyroxenes derived from the Ferrar Group (Krook, 1994). Trace amounts of garnet, tourmaline, zircon and rutile also occur, but mainly in the silt fraction.

5.4.2 DVDP-11 and Prospect Mesa

In the DVDP-11 core, a sharp change in heavy mineral composition takes place at around

300 mbsf, where abundances of biotite, sphene, oxidized grains and rock fragments rapidly increase uphole, whereas orthopyroxenes and apatite diminish (Figure 5.10). Above 300 mbsf the composition changes gradually uphole from one dominated by metamorphic source rocks to a higher contribution from volcanic sources. The lower part of unit 8 (> 5.9 Ma; Ishman and Rieck,

1992), is dominated by colorless clinopyroxenes, mainly pigeonite (Figure 5.6), green pyroxenes, green amphiboles, some garnet and apatite. The clinopyroxenes and orthopyroxenes are likely derived from Ferrar Dolerite sills, which are exposed at the head of Taylor Valley. The Basement

119 Sill (Gunn, 1962), which extends from Mackay Glacier to the head of Ferrar Glacier, is a

hypersthene tholeiite and is the most likely source for the high abundance of orthopyroxenes. The

presence of pigeonite is also consistent with derivation from tholeiitic mafic igneous rocks. The

heavy mineral composition of the lower part of unit 8 (upper Miocene) may suggest active erosion of Ferrar Group rocks at the head of Taylor Valley at that time, perhaps by an outlet glacier of the East Antarctic Ice Sheet.

The change in composition at 300 mbsf indicates that low-grade metamorphic rocks become more important sources during deposition of the upper part of lithostratigraphic unit 8.

Biotite schists of the Skelton Group are the most likely source of these metamorphic minerals, since it forms the base of Taylor Valley and was drilled at DVDP-12 farther upvalley (Lopatin,

1972; Mudrey, 1975). Heavy mineral distributions in, and XRD analyses of, DVDP-11 show an upsection increase of brown titanium-rich pyroxenes and opaques, and a decrease in green pyroxenes and amphiboles (Figure 5.10), which signals an increasing contribution from the

McMurdo Volcanic Group towards the top of the succession. The low pyroxene content at 156 mbsf may be due to chemical or mechanical instability, perhaps augmented by the cleavage of clinopyroxenes, since many abraded and altered grains are present at this level.

The heavy mineral composition of the DVDP-11 core generally follows clast composition (McKelvey, 1982) and lithic analysis (Porter and Beget, 1981), except for the major change in composition at 300 mbsf within Unit 8, which was not previously reported.

Lithostratigraphic Units 1,2, 3, 5 and 6 represent a Ross Sea provenance and Units 4, 7 and 8 a

Taylor Valley provenance. In general, the units with a Taylor Valley provenance are characterized by higher abundances of epidote and biotite, typical indicators of metamorphic source rocks, such as basement rocks of the Transantarctic Mountains. Units 3 and 5, with a Ross

120 Sea provenance, contain more opaque minerals, which generally indicate a volcanic source, in

this case the McMurdo Volcanic Group.

The heavy mineral fraction of the Prospect Mesa lower diamicton (late Miocene-early

Pliocene) in Wright Valley consists almost entirely of colorless clinopyroxenes and colorless

amphiboles, which suggests significant erosion of Ferrar Group and metamorphic basement

source rocks. Ferrar Dolerite outcrops are more abundant to the west than to the east of Prospect

Mesa. In addition, the general absence of much granitic material with sources east of Prospect

Mesa also suggests that the lower diamicton of Prospect Mesa was deposited by ice flowing from

the west, most likely the East Antarctic Ice Sheet (Prentice et al., 1993).

5.4.3 Mount Sirius

The Sirius Group at Mount Sirius has a low abundance of heavy minerals, suggesting that it may have originated from sedimentary rocks (Table 5.2). Monocrystalline quartz, polycrystalline quartz, altered K-feldspar and plagioclase, all found in the Sirius Group at Mount

Sirius (Table 5.3), are major components of the Permian and Triassic arkoses and arenites of the

Beacon Supergroup (Barrett, 1969). Pink and colorless almandine garnet could be recycled accessory heavy minerals from the Beacon Supergroup (Barrett, 1966). Most of the pyroxenes originated from the Ferrar Group.

The presence of large sillimanite grains, fine-grained metasediment clasts with minor foliation, and significant amounts of mica indicates an additional source, besides the Beacon

Supergroup and the Ferrar Group, for the Mount Sirius deposits. Fine- to medium-grained mica schists of the Nimrod Group contain micas and almandine garnet (Gunner, 1969). Large (up to 1 mm) grains of sillimanite occur in metaquartzites of the Nimrod Group. Outcrops of interbedded

121 mica schists and metaquartzites of the Nimrod Group are known from the Miller Range southwest

of Mount Sirius (Gunner, 1969). Therefore, some of the material in the Sirius Group at Mount

Sirius may be derived from the Miller Range or a subglacial extension of this part of the Nimrod

Group.

Detritai carbonate in the Mount Sirius sediments suggests that Lower Paleozoic basement

rocks also contributed material. Limestone-bearing diamictons of the Sirius Group have also been

reported by Prentice et al. (1986) from the Queen Alexandra Range (3,490-3,825 m elevation),

and the highest surface of the northern Dominion Range (> 2400 m elevation). The diamicts are

yellow to gray, massive, and contain striated gravel. The limestone clasts most likely were

derived from the Shackleton Limestone, which crops out in the upstream Beardmore region

(Young and Rybum, 1968).

The stratigraphie distributions of volcanic and basaltic rock fragments, heavy minerals,

clay mineral content, and sand percentage suggest that the lower part of the Mount Sirius succession was more influenced by erosion of weathered Kirkpatrick Basalt rocks (Elliot, 1972),

whereas the upper part experienced a greater supply of Beacon Supergroup detritus, specifically the volcaniclastic formations in the Triassic part of the Beacon Supergroup. Additional support for the Beacon Supergroup provenance comes from detritai Triassic pollen assemblages recovered from the Sirius Group at Mount Sirius (Askin, pers. comm., 1999). The Kirkpatrick

Basalt is the highest stratigraphie unit in the Transantarctic Mountains, and remnants remain today at high elevation at Otway Massif and in the Queen Alexandra Range (McGregor, 1965;

Elliot, 1970). The Kirkpatrick Basalt overlies the volcaniclastic sandstones of the Prebble and

Falla Formations of the Beacon Supergroup (Barrett, 1972; Barrett and Elliott, 1972). Therefore, the Mount Sirius deposits show evidence predominantly of erosion of the upper section of the stratigraphie sequence of the Transantarctic Mountains.

122 5.4.4 Beardmore and Shackleton Glacier areas

Similar to the Sirius Group at Mount Siri us, diamictons firom Meyer Desert section 13 and Roberts Massif also contain basaltic rock fragments and relatively low heavy mineral percentages. The outcrop of Kirkpatrick Basalt preserved at Otway Massif (Elliot et al., 1996) is directly to the east of Meyer Desert, but the basalfts may have been more extensive when the

Sirius Group was deposited at this location.

At Bennett Platform, the feldspar/quartz 5CRD ratios show a distribution similar to the sand content (Figure 5.3), suggesting that the sand content is mainly controlled by the addition of feldspar to the sand fraction of the diamictons. M ost of the feldspar (mainly plagioclase) and the colorless clinopyroxenes (including pigeonite) aree derived from the Ferrar Group. The amphiboles and stable minerals are derived from tdie Beacon Supergroup (La Prade, 1982), and from the metamorphic basement rocks of the Beardmore Group.

Olivine and abundant brown pyroxenes weere encountered in the heavy mineral fraction of one sample at the base of the Bennett Platform section (sample PNW95-068). Olivine is also present in trace amounts in deposits from Oliver Bluffs (Table 5.2), the higher areas of Meyer

Desert, and one thin section from a sample at Roberts Massif. Brown, Ti-rich clinopyroxenes and olivine are characteristic minerals of rift-type volcuinism (Nechaev and Isphording, 1993). Early

Miocene alkalic volcanism associated with olivine: basalts occurred in the southern Transantarctic

Mountains (Stump et al., 1980), but dolerite sills o«f the Ferrar Group at Nilsen Plateau (between the heads of Amundsen and Scott Glaciers farther rto the southeast; Elliot et al., 1996), at Mt.

Darwin (at the head of the Beardmore Glacier), ancd on the west side of Bowden Névé (Grindley,

1963) are also reported to contain significant amoinnts of olivine.

123; The Oliver Bluffs succession is characterized by rapid changes in composition. The stratigraphie distribution of heavy minerals at Oliver Bluffs corresponds to lithological changes

(Figure 5.9). The upper part of unit 3 has a different composition than the basal conglomerate: more garnet, stable minerals (mainly apatite) and amphiboles, but also very high clinopyroxene contents. The total heavy mineral content is high in the upper part of Unit 3, which also has a high number of rock fragments in the heavy mineral fraction. The upper part of Unit 4 is compositionally different from the lower part of Unit 4, which has a higher total clay content

(XRD) and lower heavy mineral, calcite, chlorite and pyroxene contents. The change in composition is likely caused by a shift in depositional environments as indicated by the lithological change. However, the shifts in heavy mineral composition of the fine-sand fractions also suggest a change in the relative contribution of sediment sources.

The distributions of brown clinopyroxenes and orthopyroxenes throughout the Oliver

Bluffs succession suggest that they have a similar source, most likely hypersthene tholeiite of the

Ferrar Group. The distribution of pyroxenes is probably a result of erosion of Ferrar Group rocks of variable composition, since olivine is also encountered in the heavy mineral fraction of one sample. Amphiboles are colorless and are probably derived from metamorphic basement rocks of the Beardmore Group. The detritai carbonate clasts identified in thin section suggest that the

Shackleton Limestone formed an additional source rock for the deposits at Oliver Bluffs.

The Sirius Group deposits of Oliver Bluffs and Bennett Platform show similarities in composition: high abundances of heavy minerals and a dominance of clinopyroxenes. However, the higher amount of magnetic minerals and the poor correlation between heavy mineral and magnetic mineral contents in the Bennett Platform sediments suggests that several mafic igneous or metamorphic basement sources supplied heavy minerals and magnetic minerals to the Sirius

Group at Bennett Platform. The Oliver Bluffs heavy mineral fractions exhibit a good correlation

124 between the wt. % of heavy minerals and magnetic minerals, suggesting that the magnetic

minerals mainly originate from the Ferrar Group, which supplied most of the heavy minerals.

5.4.5 Reedy Glacier area

The bedrock geology in the Reedy Glacier area is different from that in the rest of the

Transantarctic Mountains, and this difference is reflected in the composition of the Sirius Group.

High feldspar contents, and the presence of granite rock fragments, amphibole, and mica in

Quartz Hills sediments point to sources composed of granitic and metamorphic rocks. Granite and

gneisses, which form a large part of the basement in the , are the most likely

sources of the glacial sediments at Quartz Hills (Minshew, 1966).

The Tillite Spur section is about 35 m thick and crops out in a paleovalley within the

Wisconsin Plateau, overlying Permian strata of the Beacon Supergroup. The sediments are

quartz-rich, as is indicated by the XRD analysis and in thin sections, but micas and metasediment

fragments are also present. The most likely sources for the Sirius Group at Tillite Spur are

phyllites, schists and slates of the LaGorce Formation, which crops out in the foothills along the

lower Reedy Glacier area, but also as isolated patches to the south. Some of the quartz may have

been derived from locally reworked Beacon Supergroup. The cyclicity in the abundance of mica and feldspar may be related to glacial versus interglacial depositional environments and subsequent changes in sediment sources (Wilson et al., 1998). Units 8 and 9 are more quartz-rich and may include a larger component of recycled sedimentary rocks, e.g. from the Beacon

Supergroup.

125 Reedy Shackleton Beard- Nimrod Glacier Glacier Dry Valleys S : : e r David i i I Glacier oiac MS ^ i fl TS Q. QH BP MFe RM |o ._Q_ÆL_. If TM ta ^ PM cj o) 3 ëè Scale: ^ 400 km ^ C $ 4 0 m DVDP-11

Bulk X-ray diffraction Pyroxene

Quartz

80 • 60 - % silt 40 ■ % sand 20 ■ % clay

2 0 - W t 15 ■ % Heavy minerals

10 •

60

40 - % Clinopyroxene

80 60 % Garnet + stable minerals 40 -

20 •

0 X- 100 80 % Biotite

TS QH BP MDOB MS MFe TM MFI PMRM

Figure 5.14. Distribution of minerals and average grain-sizes along the Transantarctic Mountain front and relation to géomorphologie setting of stratigraphie sections. Heavy minerals are given as percentages of the heavy mineral fraction. Values o f Reedy Glacier Sirius Group are calculated from detritai modes.

1 2 6 5.5 Provenance and paleodrainage of the Sirius Group

The main controls on ice flow are the orientation of ice-surface gradients and the subglacial topography. Ice-surface gradients are determined by th e location of the main - accumulation areas. In tectonically active regions, the subglacial topography may be largely structurally controlled: surface uplift influences the flow direction of local glaciers and ice caps and the development of structural grabens in areas with weak crust may result in channeling of ice-sheet drainage systems (e.g. Wilson, 1999; Salvini and Storti^ 1999). In this study, sediment provenance of the Sirius Group and paleo-ice flow indicators, such as striated pavements and clast-orientation fabrics in diamictons, are used for paleo-drainage reconstructions, and to constrain ice-extent and tectonic setting during deposition.

Figure 5.14 illustrates the relation between the mineralogy of the Sirius Group and the geomorphological setting of Sirius Group outcrops. The outcrops studied from high plateaus are systematically enriched in quartz as compared to the valley sections. Abundances of clinopyroxene and stable minerals (including garnet) in the heavy mineral fractions show opposite distributions in relation to the geomorphological setting of the deposits. The compositional differences between the Sirius Group from the D ry Valleys and in the central

Transantarctic Mountains can be explained by a lack of preservation of valley sections in the Dry

Valleys. The valley sections from the central Transantarctic Mountains have compositions similar to those of deposits from DVDP-11 and Prospect Mesa. The heavy mineral percentages indicate that the Sirius Group in the Reedy Glacier area has a completely different source rock composition than the other Sirius Group deposits discussed here. However, diamictons from the valley section at Quartz Hills do contain more heavy minerals than the diamictons from the plateau section at Tillite Spur, which is similar to relationships observed elsewhere.

127 175"E I Ross N Ice Shelf 20 km

85° S

Otway Dominion Range 1 ) ^

170° E 175°E

Basement complex Kirkpatrick Basalt

Beacon Supergroup Paleo-ice flow indicators

Figure 5.15. Simplified geologic map of the central Transantarctic Mountains. After Elliot et al. (1996), and Barrett (1972). Ice-flow indicators from: Mayewski (1975); Mayewski & Goldthwait (1985); McKelvey et al. (1991).

1 2 8 5.5.1 Central Transantarctic Mountains

At Mount Sirius the stratigraphically and topographically highest regions of the

Transantarctic Mountains formed the source rocks for the Sirius Group. The provenance of the

Mount Sirius glacial sediments suggests that ice was flowing at a higher base level within the

Transantarctic Mountains and that the deposits originated from a different source than that of the

Sirius Group firom the nearby Dominion Range. The diamictons with Shackleton Limestone clasts firom the Queen Alexandra Range are associated with northeast-trending bedrock striations

(Prentice et al., 1986). Evidence for a northeast ice-flow direction was also found at Mount Sirius, where the Nimrod Group components suggest derivation from an area to the southwest (Figure

5.15). At present, the Bowden Névé is not a major drainage corridor for the East Antarctic Ice

Sheet. However, the paleo-ice flow direction indicators (striae) and the compositions of the deposits on Mount Sirius, the Queen Alexandra Range, and the Queen Elizabeth Range point to a uniform source and flow direction from the southwest through the Bowden Névé basin. Perhaps the Queen Alexandra and Queen Elizabeth Ranges were overridden by the ice during deposition of the Sirius Group at Mount Sirius, which may have formed the lower part of a broad (> 30 km wide) paleo-valley system, or the diamicts in the Queen Alexandra and Queen Elizabeth Ranges represent an even older glacial phase than the Sirius Group at Mount Sirius. The lower age boundary of Sirius Group diamictons at Mount Sirius is provided by forams and diatoms of

Cretaceous through late Oligocene/early Miocene age (Webb and Harwood, 1991).

Fabric data suggest that ice-flow directions at the Meyer Desert were mainly N-S (McKelvey et al., 1991). At Roberts Massif, paleo-ice flow directions from glacial grooves on a dolerite erosional surface and fabric data are also N-S (Figure 5.15). The N-S paleo-ice flow direction is slightly oblique to the Mount Sirius-Queen Alexandra Range system, but its orientation is parallel

129 to, but at a higher base level than, the present Beardmore and Shackleton glacial troughs, which

are probably fault-controlled. The presence of Kirkpatrick Basalt rock fragments, olivine and

brown pyroxenes in the Sirius Group deposits at Meyer Desert and Roberts Massif, and olivine

and brown pyroxenes at Bennett Platform (Lithostratigraphic Unit 1), suggest that the Sirius

Group at these locations may have had a common source. So far, no age-diagnostic biogenic

components have been described from the Sirius Group successions at the Meyer Desert (section

13) and Roberts Massif.

The Oliver Bluffs and Cloudmaker Formations (Figure 5.15) occur within a valley

incised into the Meyer Desert base level, and therefore represent a younger drainage system (c^

McKelvey et al., 1991). The partly marine Cloudmaker Formation is correlated to the lower part

of the Oliver Bluffs succession, which has an early Pliocene minimum age based on diatom

assemblages in silt-sized biogenic clasts (Webb et al., 1994; Harwood and Webb, 1998).

5.5.2 Victoria Land

In the Dry Valleys, ice-flow indicators occur in two different orientations: NW-SE

(Brady and McKelvey, 1983) and NNE-SSW (Denton et al., 1984). The NW-SE orientation is derived from clast-fabric data from the Feather Diamicton and from striated pavements at only two locations in Victoria Land at ca. 2700 m. above sea level (Figure 5.16). The NNE-SSW paleo-ice flow direction is more common and occurs on mountain summits and plateaus of the higher ranges of the Dry Valleys > 2400 m above sea level. The latter orientation was also found in clast fabrics of the Upper Fleming Till (Stroeven and Prentice, 1997). Both paleo-ice flow directions are at an angle to the present orientation of the glacial troughs.

ISO R oss S ea 77 S 77 S M ackayG\0^^ % % Prospect M esa Ross Island Mt.Fleming DVDP 11

Mt. Feather 2 0 k m 78 S Table Mtn 165 E

Basement complex McMurdo Volcanics Group

Beacon Supergroup Dry Valleys

Figure 5.16. Simplified geological map of the south Victoria Land (After: Lopatin, 1972). Paleo- ice flow indicators are from: Denton et al. (1984); Brady & McKelvey (1979) and (1983); Stroeven & Prentice (1997).

The Mount Feather Diamicton and the Sirius Group at Table Mountain rest within depressions cut by the main glacial troughs of Taylor and Wright Valley. This geomorphological relationship suggested to Brady and McKelvey (1983) and Barrett and Powell (1982) that the deposits predate the carving of Taylor Valley, before the middle Miocene. Meta-sediment basement clasts in the Feather Diamicton are probably derived from the Mawson Formation, which crops out to the north in Central Victoria Land, but may extend subglacially (Brady and

McKelvey, 1983). Meta-sediment clasts are also reported from the Mount Fleming Upper Till

(Stroeven and Prentice, 1997), and may have a similar origin, although the heavy mineral compositions point to slight differences in source rock lithologies. The provenance and fabric data suggest that the Feather Diamicton was deposited by ice flowing along a south-east to north­

131 west oriented flow-line. The fabric data and some rat-taiis point to a more N-S oriented paleo-ice

flow for the Upper Fleming Till. The orientation of the ice-flow and the composition of the

sediments suggest that perhaps the Feather Diamicton and the Upper Fleming Till did not

originate by East Antarctic Ice Sheet deposition, but are derived from local ice domes located

close to or within the ancestral Transantarctic Mountains. Unfortunately, age constraints are poor

for these glacial deposits. A late Pliocene age proposed by Harwood (1986) for the Mount Feather

Sirius Group could not be reproduced when analyzing a core from the Feather Diamicton

(Harwood and Rose, 1998). Some of the diatoms recovered from Mount Feather and Mount

Fleming are thought to be wind-blown (Barrett et al., 1997; Stroeven et al., 1998) and cannot be

used for age constraints on the depositional age of the diamictons.

The heavy mineral assemblages of deposits from the glacial troughs indicate that outlet glaciers of the East Antarctic Ice Sheet were flowing through the Dry Valleys from the west in the late Miocene to early Pliocene. The heavy mineral assemblage of the lowermost diamicton in

DVDP-11 suggests that overdeepening and headward erosion into Ferrar Group rocks of Taylor

Valley continued until the late Miocene. A general upward increase in volcanogenic detritus is apparent from the early Pliocene onward in DVDP-11. This is consistent with the early Pliocene heavy mineral record from CIROS-2 beneath the Ferrar Glacier (Ehrmann and Polozek, 1998) and indicates an increasing influence of local alpine glaciers and the after the early

Pliocene. Prentice et al. (1993) also suggested that in Wright Valley there is no evidence for major glaciation from the west after 3.9 Ma.

132 5.5.3 Reedy Glacier area

Striations on the valley floor at Tillite Spur and the clast fabric of one gray conglomeratic diamictite suggest that the gray diamictites were deposited by local ice flowing westward from the Wisconsin plateau (Wilson et al., 1998). The gray conglomeratic diamictites are rich in micas, which supports this interpretation, since phyllites are an important basement lithology on the

Wisconsin plateau (Minshew, 1966). According to fabric data (Wilson et al. 1998), brown-olive diamictites were deposited by northward flowing ice, most likely the paleo-Reedy Glacier. The lower abundance of mica supports a change in source area between the two lithologies. Hanvood

(1986) and Webb & Harwood (1984) recorded relatively rich Eocene to Pliocene assemblages of marine microfossils in the middle and upper part of unit 4 (Mercer, 1968). A few diatoms were also recovered from the brown/olive-brown fine-grained beds with low mica content in the lower part of the section (Wilson et al., 1998).

The glacio-Iacustrine and marine deposits at Quartz Hills indicate that glaciers eroded the lowest stratigraphie units in the Transantarctic Mountains, here consisting of granitic and metamorphic basement. The stratigraphie relationship of the Tillite Spur and Quartz Hills exposures is unknown (Wilson et al., 1998). The Tillite Spur succession probably reflects alternations of local and East Antarctic Ice Sheet sources, which are recorded in the bulk mineralogy (Figure 5.3), clast lithology and clast fabric data of the sediments (Wilson et al.,

1998). Provenance analysis based on Rb-Sr dating of feldspars suggested that the Tillite Spur sediments could have been deposited by a local ice-cap, whereas feldspars from the Quartz Hills sediments contain a Precambrian component, indicating a source in East Antarctica (Faure et al.,

1983). Therefore, similar to other areas, the Sirius Group from the glacial troughs (Quartz Hills) was deposited when outlet glaciers were flowing through, and actively eroding the rocks exposed

133 within the glacial troughs, whereas higher elevation successions (Tillite Spur) represent either

local glaciation or overriding of the Transantarctic Mountains by the East Antarctic Ice Sheet.

5.6 Weathering and diagenesis

Chabazite was identified in Sirius Group samples by XRD and microscopic analysis.

Chabazite is a Ca-rich zeolite, and Ca can be supplied through chemical breakdown of

plagioglase. Chabazite was previously reported from modem Antarctic soils at Prospect Mesa in

Wright Valley (Gibson et al., 1983) and occurs in the Sirius Group at Table Mountain (Dickinson

& Grapes, 1997). It is believed to have precipitated at the boundary between ice-free and ice- cemented permafrost horizons in the active layer of the polar soil. The occurrence of chabazite is related to the presence of patterned ground, indicating freeze-thaw cycles. Dickinson & Grapes

(1997) argued that the patterned ground at Table Mountain resulted from a period that was warmer and wetter than at present, but Gibson et al. (1983) state that diurnal freeze-thaw cycles do occur on clear summer days in the lower regions of the Dry Valleys. In a diamicton from

Roberts Massif (samples PNW95-012 through 014), chabazite occurs in sediments with low bulk density, high porosity and a rotational fabric, all features consistent with peri-glacial reworking

(Chapter 3).

In XRD patterns of samples from the Oliver Bluffs and Bennett Platform successions, the occurrence of chabazite, calcite and the maximum abundance of total clays coincide with lower abundances of feldspar (mainly plagioclase) and pyroxenes. The 4.47 - “total clays” peak in the

XRD patterns is dominated by illite-montmorillonite, an authigenic soil mica, which is the principal clay mineral formed in glacial sediments upon exposure to the atmosphere (Andersen et

134 Calc/Q lllite-Mont/Q Chab/Q Oliver Bluffs 0.0 2.0 0.0 0.5 0.0 2.0

40 -

CO

Calc/Q llüte-Mont/Q Chab/Q Bennett Platform 0.0 1.0 0.0 1.0 0.0 1.0

o

55 -

95 -

Figure 5.17. Stratigraphie distribution of Calcite, Illite-Montmorillonite and Chabazite from XRD bulk mineralogy, expressed as mineral/quartz ratios.

135 al., 1997). niite-montmorillonite is found in many Antarctic soils today (Ugolini & Jackson,

1982) and throughout the Upper Neogene section of the DVDP-11 drillcore (Ugolini et al., 1981).

Approximately 60 m below the top of the outcrop at section 5 of Oliver Bluffs (Figure 5.17) calcite and chabazite abundances reach a maximum; this level is several metres below the peak abundance of total clay and lower feldspar and pyroxene abundances. This pattern suggests the presence of a polar paleosol at this level. Chemical weathering of unstable minerals may have resulted in the formation of calcite and chabazite at the top of an ice-cemented horizon, several metres below the surface.

At Bennett Platform, two subaerial exposure horizons can be recognized. One coincides with stratified sediments at ca. 55 m below the top of the outcrop, in lithostratigraphic unit 6 , the other at the top of lithostratigraphic unit 3, which consists of stratified diamictons (Figure 5.17).

Chabazite occurs at both levels and is associated with a higher relative abundance of clay minerals. More calcite is also detected at the top of unit 3. Despite a careful search, fossil plants or pollen were not found at Bennett Platform (Webb et al., I996d).

Gibson et al. (1983) stress the importance of chemical weathering in Antarctica.

Although weathering rates are low, considerable amounts of montmorillonite were found in soils in the Shackleton Glacier area (Claridge and Campbell, 1968) and Ugolini and Jackson (1982) demonstrated that feldspars are transformed to authigenic soil mica and vermiculite montmorillonite under present climate conditions in Wright Valley. In DVDP-11, clay minerals consist of a complex association of interstratified montmorillonite-vermiculite-chlorite and mica, as well as non-stratified vermiculite (Ugolini et al., 1981). The clay minerals in DVDP-11 are detritai and were derived from soils. The total clay content indicated by XRD data in this study

(Figure 5.5) is higher for the lower part of the hole (> 200 mbsf), which is either related to variability in grain-size of the sediments or signifies an increased supply of clay minerals from

136 soils. In DVDP-11 calcite is observed in Lithostratigraphic Units 4,5 and 8 , but is absent in the

lower Pliocene interval perhaps due to a slightly different carbonate chemistry. Despite modest

warming indicated by foraminiferal and isotope data from DVDP-11 (Ishman & Rieck, 1992;

Prentice et al., 1999), chemical analyses suggested that weathering rates remained low throughout

the early Pliocene (Ugolini et al., 1981).

5.7 Discussion

At most Sirius Group localities, heavy mineral assemblages reflect contributions from two main sources: mafic igneous rocks of the Ferrar Group and sedimentary rocks of the Beacon

Supergroup. The relative proportions of minerals associated with these two sources vary considerably between localities. Similar patterns are recognized in the X-ray mineralogy and the detritai modes. In general, the composition of glacial sediments is controlled by 1 ) source rock composition; 2) hydraulic sorting; 3) mechanical abrasion; and 4) dissolution (Johnsson, 1993).

The differences in composition could be related to differences in the relative contributions from the two sources to sediments at different localities. A difference in source rock contribution is not readily explained if deposition occurred during one episode of glacial expansion, since the basement geology of the Transantarctic Mountains is quite uniform.

However, if the Sirius Group deposits represent different stages of unroofing of the Transantarctic

Mountains, then the variable compositions could be explained by a difference in the proportion of source rocks exposed to glacial erosion at different times. This would require that the deposits are of different ages, a suggestion that is supported by the diatom ages provided for some Sirius

Group deposits and by the paleo-ice flow data.

137 Alternatively, the differences in heavy minerai composition couid be controlled by

environmental factors. In an ice-contact environment, hydraulic sorting is not a major factor;

however, mechanical abrasion during transport may have affected the heavy mineral assemblages

in the Sirius Group. The cleavage of the clinopyroxenes may have reduced their mechanical

stability; but since there appear to be no major differences in depositional environment between

Sirius Group localities, this cannot account for differences in composition between localities.

Weathering of the parent rocks may have removed certain unstable minerals, such as pyroxenes,

from the assemblage before it was eroded and transported. At present, chemical weathering in

Antarctica only affects small pigeonite crystals (Claridge & Campbell, 1987). However, since the weathering regime prior to deposition of each Sirius Group succession is unknown, removal of pyroxenes by chemical weathering prior to glacial erosion has to be taken into account.

The subaerial exposure indicated by chabazite and calcite horizons at Oliver Bluffs is consistent with the presence of in situ Nothofagiis wood, roots and leaf mats at the same level in the succession. The plant fossils have been interpreted as remains of shrubs, a vegetation type that is consistent with the presence of permafrost and active layers several meters-thick.

Other possible paleosols occur in the upper part of lithostratigraphic unit 4. These paleosols were previously described by Retallack & ICrull (1996) and indicate warmer and wetter conditions than at present, compatible with the tundra climate indicated by the fossil flora in the lower part of the succession (Askin & Markgraf, 1986; Hill et al., 1996; Francis & Hill, 1996). A succession of waterlaid tills recovered from the Cloudmaker Formation, which occurs downvalley from Oliver

Bluffs, suggests that the Beardmore Valley was once a paleoQord (Webb et al., 1996a). The

Cloudmaker Formation now occurs at ca. 1100 m a.s.l. and its upper part is correlated to the lower part of the Oliver Bluffs succession. This correlation implies that the fossil plants recovered from Oliver Bluffs were growing near sea level. The difference between the present climate at

138 Oliver Bluffs and the paleoclimate reconstructed from the fossil plants and soils may be caused in part by the difference in altitude at the time of deposition (several metres a.s.l.) and the present altitude of the outcrop (ca. 1 1 0 0 m a.s.l.).

The presence of chabazite in proglacial facies at Oliver Bluffs and Bennett Platform suggests that permafrost conditions existed during part of the deglaciation stage. The sediments were deposited in an ice-proximal setting recording several advances and retreats. The polar paleosols may represent several thousands of years of exposure to the atmosphere (cf. Retallack and Krull, 1996; Claridge and Campbell, 1968; Ugolini and Jackson, 1982). This time frame suggests that the Sirius Group successions may represent 10,000-100,000 year glacial-interglacial cycles. A similar time frame was proposed previously by Wilson et al. (1998). However, there is no direct time control to support these estimates.

5.8 Summary and concluding remarks

Mineralogical analysis of the Sirius Group is successful in identifying differences in provenance of individual Sirius Group successions. The sedimentary record included in the Sirius

Group shows a large range of mineralogical compositions. The accessory minerals are related to source rocks that crop out in the Transantarctic Mountains. The successions can roughly be subdivided into two major lithological groups: 1 ) those that have a high heavy mineral content and high abundance of pyroxenes; and 2 ) those that have a low heavy mineral content and a relatively high abundance of garnet and other mechanically and chemically stable minerals. The pyroxenes are mainly derived from the Jurassic Ferrar Group. The garnet and other stable minerals are probably derived from Paleozoic sedimentary rocks of the Beacon Supergroup,

Proterozoic metasediments, and perhaps Cenozoic sedimentary rocks. Most of the deposits in

139 Group 1 occur within the present glacial troughs, whereas the deposits of Group 2 are situated on

high mountain plateaus. The difference in mineralogy of the two groups can be explained by a

difference in their relation to the tectonic history of the Transantarctic Mountains and/or a

difference in pre-erosional weathering regime. Both explanations require different ages for

individual Sirius Group successions, which is in agreement with the paleontological evidence and

paleo-ice flow data.

Horizons enriched in chabazite, calcite and authigenic soil mica indicate exposure to the

atmosphere at several levels in the Oliver Bluffs and Bennett Platform successions. The nature of

the subaerial alteration suggests that climate conditions were cold and dry, similar to present-day

conditions in the lower regions of the Dry Valleys. This paleoenvironmental interpretation is consistent with interpretations based on paleosols (Retallack & Krull, 1996) and vegetation

(Francis & Hill, 1996).

140 CHAPTER 6

CHEMICAL COMPOSITION OF THE SIRIUS GROUP

6.1 Introduction

The aim of this chapter is to evaluate the provenance and weathering history of Upper

Cenozoic sediments of the Sirius Group. Whole rock chemical analyses can be used to evaluate

the mineralogical composition of sediment, which, in the case of glacial sediments, is mainly

controlled by the composition of the source rocks in the catchment area of the

system, and by the degree of chemical weathering upon exposure to the atmosphere. Bulk

chemical analyses have been used before to evaluate paleoenvironmental and provenance changes

in the CIROS-I and CRP-1 drillcores in McMurdo Sound (Roser & Pyne, 1989; Bellanca et al.,

1997; Krissek & Kyle, 1998). This chapter provides an interpretation of the bulk chemical composition of Upper Cenozoic terrestrial glacial sediments from Antarctica.

Bulk chemical analysis has the advantage that all size fractions of the sediment matrix are investigated in a quantitative way. In sand provenance studies, only the coarse fraction is considered, and bulk XRD data are only suitable for semi-quantitative interpretations. However, the interpretation of bulk chemical data involves making assumptions about the mineralogy of the sediments. An integrated approach using both mineralogical and chemical data helps to reduce these uncertainties.

141 North Victoria

MFlem-83 MFeath-8 TM-5 TM-6

Ice Shelf

- PNW95- 053.055 057,059 061,063 DR5-4 -1 DR5-9 DR5-11

DR13-4 PNW95 DR13-8 002.012 DR13 94QH13 94QH39 94TS02 94TS18

Figure 6.1. Locations of Sirius Group samples selected for ICP-OES chemical analysis are indicated by a black dot. Small black squares indicate other important Sirius Group localities (after Mayewski, 1975 and Stroeven, 1997).

142 The bulk chemical composition of clastic sediments is determined by the following

factors: 1 ) the composition of the source rocks, 2 ) the depositional environment (mechanical

abrasion and hydraulic sorting), 3) pre-depositional and post-depositional weathering, and 4)

diagenesis and metamorphism. Tills are quite suitable for provenance studies because they

originate by mechanical erosion and the depositional processes preserve all sediment components.

In tills, most minerals are detrital, and reflect the source rock mineralogy and weathering

conditions prior to deposition of the sediments. In proglacial mudrocks, both hydraulic sorting

and post-depositional alteration may have influenced the mineralogy. Because weathering

products dominantly occur in the finer grain-size fractions, proglacial mudrocks can be used to

assess the paleoclimate conditions during deglaciation.

6.2 Methods

A total of 32 Sirius Group diamicts and mud(stone)s were analyzed. Samples from a large geographical area were selected for analyses, so as to cover as many Sirius Group outcrops as possible (Figure 6.1). Dried samples were ground in a ball mill and sent to XRAL Laboratories in Toronto, , for Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES) analysis. At XRAL, the samples were prepared by LiBOn fusion and 10 major elements were analyzed: Si, Ti, Al, Fe, Mg, Mn, Na, K, Ca, and P, as well as 6 trace elements: Ba, Cr, Nb, Sr, Y, and Zr. Major element abundances were converted to oxides. Loss on ignition (LOI) was not determined. Because sand provenance analysis revealed the presence of detrital carbonate

(Chapter 5), the percentage calcium carbonate was also determined. The data and further details about laboratory methods and accuracy are listed and discussed in Appendix D.

143 6.3 Results

The results of the major and trace element analyses are listed in Table 6.1 and Table 6.2, respectively. Plots of major elements versus SiOi show consistent trends (Figure 6 .2 ). In general,

Fe^Os, MgO, CaO, and MnO are negatively correlated with SiOi- TiO? and P 2O5 concentrations are constant with respect to variations in SiO% concentration. AI 2O3 values are constant with varying Si 0 2 concentrations. Both K 2O and Na 2Û concentrations slightly increase with increasing

SiÛ2. However a few outliers can be identified. The Table Mountain Sirius Group has extremely high concentrations of Si 0 2 (Table 6.1) and does not follow the general trends for the other major elements. The Reedy Glacier Sirius Group is also distinctly different in many aspects, especially considering elements such as Ti, P, Na and K.

Sr and Cr show a decline with increasing Si 0 2 concentration, although the data are highly scattered (Figure 6.3). Zr and Ba show a positive relation to Si 0 2 concentrations. Y has a constant value with increasing Si 0 2 concentrations. A plot of Nb versus S1 0 2 is not shown here because too many values of Nb were below or near detection limits (Table 6.2). Again, Sirius Group samples from Table Mountain and the Reedy Glacier area are outliers to the general patterns.

Calcium carbonate concentrations in the Sirius Group are generally low (<1.0 wt. %), with a few exceptions (Table 6.3). Sediments from Bennett Platform, Oliver Bluffs, Meyer Desert section 13, and Mount Sirius contain up to 2.6 wt. % calcium carbonate.

144 Sample SÎO2 TiOz AI2 O3 Fe^Os MgO MnO CaO NazO KzO PzOs wt % wt % wt % wt % wt % wt % wt% wt % wt % wt %

PNW95-002 64.82 0.63 15.04 4.66 2.17 0.06 2.24 1.71 2.4 0.07

PNW95-012 65.89 0 . 6 8 14.76 5.6 2.37 0.09 3.51 1.9 2.18 0.07 PNW95-053 65.46 0.7 13.42 5.98 2.95 0 . 1 4.87 2.25 1.76 0.09 PNW95-055 65.04 0.78 13.09 6.33 2.7 0 . 1 4.55 2.25 1.72 0.09 PNW95-057 65.46 0.67 13.28 4.89 2 . 1 1 0.09 3.12 1.87 1.96 0.07 PNW95-059 6 6 . 1 0.67 12.87 4.95 2.24 0.09 3.47 1.89 1.84 0.07 PNW95-061 61.61 0.73 13.81 6.31 2.67 0 . 1 4.24 1.83 1 . 8 8 0.09 PNW95-063 61.4 0.72 13.3 7.26 3.78 0 . 1 2 6.14 1.9 1.53 0.09 PNW95-066 67.82 0.72 1 1 . 6 4.36 1.94 0.08 3.26 2.18 1.7 0.09 PNW95-068 62.9 0.63 12.79 4.93 2.35 0.08 3.57 1.89 1.84 0.09 95DMH051 59.26 0.62 12.17 5.32 3.32 0.09 6 . 0 2 1.95 1.47 0.09 DR8-Unit2 59.05 0.62 12.47 4.92 2 . 8 0.09 5.3 1.91 1.54 0 . 1 1 DR5-4 56.48 0.55 13.32 6.43 4.49 0 . 1 6.91 1.7 1.48 0.09 DR5-9 59.26 0.58 13.04 5.16 2.44 0.08 3.99 1.74 1.82 0.07 DR5-1I 58.83 0.58 13.57 4.65 2.3 0.08 4 1.82 1.96 0.09 DR5-16 59.47 0 . 6 13.17 4.36 2.19 0.09 4.3 1.99 1.67 0.07 DRI3-4 61.61 0.58 12.89 5.02 2.44 0.08 3.72 1.74 1.79 0.09 DR13-8 62.9 0.55 15.4 4.55 1.46 0.08 2.49 1.95 2.35 0.07 DR13-9 68.24 0.65 16.04 5.58 2.19 0.09 3.48 2.36 2.23 0.09 S-4 68.03 0.53 11.96 3.76 1.41 0.05 1.89 1 . 6 8 2.13 0.07

S- 6 67.39 0.55 1 1 . 6 4.19 1.79 0.06 2.45 1 . 6 6 2.16 0 . 1 1 S-9 63.97 0.62 13.43 4.06 1.54 0.05 1.62 1.44 2.32 0 . 1 1 S-12 60.76 0.63 13.06 4.93 1.59 0.05 2.62 1.36 2 . 1 1 0.07 TM-5 78.3 0.25 4.76 2.77 2.07 0.04 3.06 0.61 0.47 0 . 0 2

TM - 6 76.16 0.28 4.63 3.52 2.42 0.05 3.06 0.59 0.43 0.05 MFeath-S4 73.59 0.58 12.17 2.46 0 . 6 8 0.03 0.87 1 . 8 6 2.14 0.07 MFIem-83 69.1 0.52 13.68 2.63 0.71 0.03 0.98 2.14 2.57 0.07 94TS02 71.24 0.83 11.87 3.8 0.9 0.06 0.46 1 . 0 2 4.1 0.09 94TS18 69.96 0.55 13.93 3.35 0 . 8 0.04 0.39 1.09 4.46 0 . 1 1 94TS23 71.88 0.58 11.77 3.1 0 . 8 0.03 0.24 1 . 1 1 2.69 0.09 94QH13 59.9 0 . 8 8 15.12 5.96 2.37 0 . 1 2 . 6 6 2.08 4.19 0.27 94QH39 58.83 0.92 14.74 5.98 2.55 0 . 1 2.69 1.89 4.1 0.32

Table 6.1. Major element data for the Sirius Group as determined by ICP-OES.

145 Sample Ba Cr Nb Sr Y Zr ppm ppm ppm ppm ppm ppm

PNW95-002 483 41 1 1 160 2 0 186 PNW95-0I2 479 76 1 1 208 2 1 159 PNW95-053 420 65 - 190 2 1 144 PNW95-055 391 47 1 1 158 2 1 156 PNW95-057 443 52 - 168 2 1 166 PNW95-059 420 53 1 0 179 2 0 182 PNW95-06I 413 43 1 1 169 2 1 162 PNW95-063 367 55 13 172 2 1 138 PNW95-066 443 44 1 2 214 19 158 PNW95-068 411 55 15 163 19 141 95DMH05I 389 78 - 237 17 147 DR8-Unit2 431 6 6 1 0 273 18 136 DR5-4 355 1 1 0 - 219 16 109 DR5-9 449 49 - 226 18 144 DR5-LI 458 46 1 2 168 2 0 149 DR5-16 458 53 - 230 18 141 DR13-4 408 53 - 188 18 171 DR13-S 500 32 - 199 23 174 DRI3-9 469 70 - 2 1 2 2 0 180 S-4 454 39 - 162 18 207

S- 6 505 52 - 167 17 184 S-9 500 46 - 130 2 0 173 S-12 418 32 - 191 2 1 170 TM-5 114 42 - 47 0 124

TM - 6 99 48 - 41 0 114 MFeath-84 464 40 - 158 19 218 MFlem-83 602 33 1 1 179 19 168 94TS02 453 25 13 71 41 432 94TS1S 480 18 - 6 8 38 197 94TS23 353 43 1 2 55 2 1 229 94QH13 365 35 24 216 38 243 94QH39 373 43 24 155 54 283

Table 6.2. Trace element data for the Sirius Group, determined b-y ICP-OES.

146 1.0 0.2

0.8 ■ g A A ♦ ■g 0.4 ■

0.2 ■

0.0 0.0 50 55 65 7060 75 80 50 55 60 65 70 75 80 w t % SiOj w t % SiO, 8 7 15 • 6 5 < m ♦ ü 10 ■ SS 4 5 S 3 2 1 0 50 55 6560 70 75 80 5055 60 65 70 75 80 Wt % SiO, w t % SiO, 8 2.5 7 ♦♦ 6 2.0 - °- £ 5 Z 4 3 5 2 1 0.5 - 0 0.0 50 55 60 65 70 75 80 5055 60 65 70 75 80 Wt % SiO, w t % SiO , 5 5

4 4

3 3 s- 2 2 ♦ ♦♦ 1 1 0 0 50 55 6065 70 75 80 50 6065 70 75 8055 Wt % SiO, Wt % SiO, 0.6 ° Tillite Spur 0.5 - 4. Quartz Hiiis 0“ 0.4 - oT * Tabie Mountain -g 0.3 ■ ♦ Other Sirius Group

0.0 50 55 60 65 70 8075 Wt % SiO, Figure 6.2. Plots of concentrations of major element versus SiO, Regression lines are based on “other Sirius Group” data only. TiUite Spur, Quartz Hills and Table Mountain data are excluded, because their compositions form outliers to the general pattern.

147 120 m

80 ■ E 300 î & 250 u

50 55 60 6570 75 60 60 65 70 % SI02 % S102

300 700 1

250 600 ■ 500 ♦ ♦ „ 200 • ? 400 • a 150 • & mn 300 100 ■ cc 200 0045. 100

50 55 60 6570 75 80 50 55 60 70 7565 80 % SI02 % SI02

n Tillite Spur 50 ■ A Q uartz Hills 40 • X Table Mountain ?a a 30 • ♦ Other SIrlus Group

50 55 60 65 70 75 80 % SI02

Figure 6.3. Plots of concentrations of trace elements versus SiO^ Linacr regression lines are based on “other Sirius Group” data only, because Tillite Spur, Quartz Hills and Table Mountain data arc outliers to the general pattern. Sample CIA AIzOzTTlOz wt. % CaCO;

PNW95-002 62 23.73 0.2 PNW95-0I2 56 21.58 0.1 PNW95-053 49 19.15 0.4 PNW95-055 49 16.7 0.1 PNW95-057 58 19.91 1.3 PNW95-059 55 19.29 0.8 PNW95-061 52 18.82 0.2 PNW95-063 48 18.55 1.4 PNW95-066 52 16.17 0.8 PNW95-068 54 20.18 0.7 95DMH05I 47 19.72 1.8 DR8-Unit2 51 20.21 2.6 DR5-4 46 24.2 1.4 DR5-9 52 22.33 0.2 DR5-11 53 23.24 0.3 DR5-I6 52 21.93 0.6 DR13-4 55 22.07 1 DR13-8 63 27.98 1.2 DR13-9 58 24.66 1 S-4 59 22.41 0.1 S-6 58 21.08 1.2 S-9 65 21.77 0.7 S-12 61 20.6 0.8 TM-5 43 19.03 0.9 TM-6 40 16.32 0.1 MFeath-84 66 20.84 0.6 MFlem-83 63 26.46 0 94TS02 63 14.23 0 94TSI8 65 25.3 0 94TS23 70 20.16 0 94QH13 55 17.1 0.2 94QH39 55 16.06 0.2

Table 6.3. Chemical index of alteration (CIA), ALOs/TiOz ratio and calcium carbonate concentration of Sirius Group sediments. For the definition of the CIA see section 6.5.

149 6.4. Provenance

Provenance analysis of clastic sedimentary rocks using chemical data requires that we consider only immobile elements. Elemental abundances are modified during weathering, erosion, transport and deposition of source rocks and their products, and after deposition due to diagenesis, metamorphism and hydrothermal alteration (Frakes and Kronberg, 1997). The challenge is to demonstrate which elements were immobile under both surface and subsurface conditions, in order to correctly identify the source area of the sediments. In this case hydraulic sorting and diagenesis are not major factors, but weathering of source rocks prior to erosion and transport cannot be entirely excluded.

The mineralogical analysis (Chapter 5) suggested that the Sirius Group is partly derived from sources in the Transantarctic Mountains: the Ferrar Group dolerites and basalts, the Beacon

Supergroup sandstones and mudstones. Paleozoic metasediments and plutonic rocks, and

Precambrian metamorphic basement. However, especially the finer fractions of the Sirius Group may contain components derived from distant sources now covered by the East Antarctic Ice

Sheet, e.g. crystalline terranes of various compositions (e.g. Faure et al., 1995), and intracratonic sedimentary sequences (e.g. Webb, 1994).

6.4.1. Grain-size effects

The positive relation of Zr and Ba with SiOz suggests that these elements are enriched in the coarse quartz-rich fraction. Since Y and Al do not show a relation to SiOz these elements are probably carried mostly in the fine fractions of the diamicts, which make up the bulk of the

150 18 6 17 ■ 5 16 - 4 S 15 - OO) < 14 -» 3 2 1 10 0 0 10 20 30 0 2010 30 % clay % clay 2.4

2.5 - 2.2 - 2 -• z

0.5 - 1.2 -

0 10 20 30 0 10 20 30 % Clay % clay 120 250

100 - 200 - 80 - 150 - 9 r GO - 100 O 40 - N 50 • 20 -

0 10 20 30 0 10 20 30 % Clay % clay 25 25

20 - 20 -

^ 15 - ^ 15 - E I a. & 10 - & 10 • >- >-

0 10 20 30 40 50 60 70 0 10 20 30 % sand % clay

Figure 6.4. Plots of selected elements versus wt. % in grain-size fractions. The outlier (Cr and Y) is a sample from a siltstone from Oliver Bluffs (sample 5-16). Grain-size data from Chapter 3 (only central Transantarctic Mountains).

151 sediment. Plots of major and trace elements with size-ffaction data (Figure 6.4) show that clay- rich tills are enriched in KiO, Zr and Y, and depleted in MgO, NaiO, and Cr. The clay fraction of the sediments may originate from an acidic source, whereas the sand and silt fraction are derived from a basic igneous source. High clay contents in some Sirius Group sediments may result from a large contribution of acidic source rocks containing abundant clay minerals. Therefore, the range in compositions and grain-sizes of Sirius Group diamicts may be a result of differences in pre-glacial weathering and/or a difference in the contribution of different source rocks: MgO,

Na%0, and Cr-rich, but clay-poor diamicts mainly derived from unweathered basic igneous rocks, versus clay-rich, and K^O, Zr and Y-rich diamicts mainly derived from acidic igneous rocks or mudrocks (Table 6.4).

152 M ajor elem ents:

SiOz TiOz AI2 O3 FeO MgO MnO CaO N a,0 KzO P 2 O5 wt % wt % wt % wt % wt % wt % wt % wt % wt % wt %

Dolente^ 50.4 0.44 15.51 8.72 0.17 1 0 . 6 10.87 1.42 0.37 0.08 Dolente^ 52.91 0.64 14.89 9.09 0.15 6 . 2 2 10.84 1.58 0.51 0.08 Dolente* 53.75 0.7 14.33 9.65 0.18 6.64 1 0 . 6 1.83 0.81 0.18 Dolerite* 55.65 1.03 13.95 10.3 0.17 4.5 8.51 2.5 1.45 0.23 Basalt 54.43 0.7 15.03 9.5 0.18 6.76 10.67 1.99 0 . 6 8 0.096 Basalt’ 56.6 1.28 12.92 11.57 0 . 2 2 3.44 7.91 2.29 1.29 0.16 Pagoda Pm'* 63.2 0.79 16.4 6.19 607* 2.99 0.76 1.18 3.55 0.19 Mackellar Fm"* 60.4 0 . 8 17.4 6.56 501.6* 3.12 0.75 1.23 3.65 0.23 Fairchild Fra“ 63.9 0.85 16 5.91 496.7* 2.37 0.32 0 . 2 1 0.4 0.03 Buckley Fm"* 56.1 0.7 15.6 2.25 191.7* 1 . 1 0 . 6 0.9 3.2 0 . 1 Lower Beacon^ 89.73 0.24 5.7 1.24 0 . 0 2 0.58 0 . 6 8 0 . 2 1.44 0 . 0 2 Lashly Fm^ 77.86 0.53 12.7 2.71 0.03 0.7 1.06 1.72 2.28 0.03 McM Volcanic ^ 42.69 3.82 13.83 11.78 0 . 0 2 9.71 10.78 3.53 1.49 0 . 8 6 McM Volcanic ^ 52.76 1.73 19.1 6.74 0 . 2 2.03 5.06 7.23 3.86 0.58 Granitoid^ 65.25 0.72 16.14 4.39 0.07 1.56 3.88 3.63 3.69 0.17 • in ppm

Trace elements;

Ba Cr Nb Sr Y Zr AI2 O3 / ppm ppm ppm ppm ppm ppm TiOz

Dolerite* 157 352 1 0 0 53 n=l 35.25 Dolerite* 126 132 25 6 6 n=4 23.27 Dolerite* 232 142 126 S3 n=5 20.47 Dolerite' 376 59 138 157 n=4 13.54 Basalt^ 204 141 5 124 26 104 n=18 21.47 Basalt^ 128.8 n= 1 2 10.09 Pagoda Fm"* 101.4 n= 1 0 20.76 Mackellar Fm'’ 114.8 n=39 21.75 Fairchild Fm^ 104.8 n= 6 18.82 Buckley Fm^ 90.4 n=36 27.29 Lower Beacon^ 309 18 5 47 1 2 142 n=42 23.75 Lashly Fm^ 513 37 1 2 147 24 234 n=17 23.96 McM Volcanic ^ 346 377 114 970 32 324 n=26 3.62 McM Volcanic ^ 733 15 204 951 43 582 n=26 11.04 Granitoid^ 913 15 13 516 29 2 2 2 n= 8 22.41

‘ Kyle, 1980 ^ Siders & Elliot, 1985 ^ Faure et al., 1974 ■* Homer, 1992 ^ Roser & Pyne, 1989

Table 6.4. Chemical composition of Sirius Group source rocks.

153 4.5 o Sirius Group 4.0 - 3.5 - ■ Ferrar Group

O 3.0 - o Beacon Supergroup mudrocks 2.5 - X Lower Beacon è 2 .0 - X Lashly Formation

A. McMurdo Voicanics 0.5 - + Granitoid basement 0.0 0 105 15 20 25

W t % ALO2^3

1.4 o Sirius Group

■ Ferrar Group

0.8 - ♦ ♦ ♦ Beacon Supergroup mudrocks

5 0.6 - X Lower Beacon 0.4 -

0.2 - X Lashly Formation

0.0 + Granitoid Basement 0 5 10 15 20

W t % A I2O 3

Figure 6.5. Ti-Al ratios of the Sirius Group compared to those of source rocks in the Transantarctic Mountains (data from Faure et al., 1974; Siders & Elliot, 1985; Kyle, 1980; Homer, 1992; Roser & Pyne, 1989). (a) source rocks including McMurdo Voicanics, (b) source rocks excluding McMurdo Voicanics.

1 5 4 6.4.1 Al:Ti ratios of diamicts

AJ:Ti ratios in glacial diamictons generally reflect the composition of the source rocks,

because of the absence of hydraulic sorting and the low weathering rates in glacial environments

(Young & Nesbitt, 1998). Ti and Al are relatively immobile under most weathering regimes and

the ratio of Al to Ti varies greatly in igneous source rock types, with basic igneous rocks having

low Al concentrations relative to Ti and acidic igneous rocks having high Al concentrations

relative to Ti. The Al/Ti ratios of the Sirius Group are quite uniform (Table 6.3) and are

consistent with a provenance of Ferrar Group, Beacon Supergroup and crystalline basement rocks

(Table 6.4), with the exception of some low values in the Sirius Group at the Reedy glacier (TS

and QH samples). Low values in the Sirius Group from Quartz Hills may be partly caused by

hydraulic sorting, which concentrated Ti-rich biotite in the finer fraction, since the sediments are

muddy and were deposited in a subaquatic environment (Mercer, 1968; Wilson et al., 1998a).

Plots of Al vs. Ti further explore the relation of the Sirius Group to its source rocks in the

Transantarctic Mountains (Figure 6.5). Most of the Sirius Group samples are slightly depleted in both Al and Ti compared to the igneous source rocks, which suggests that recycling of sedimentary rocks occurred, e.g. reworking of the Beacon Supergroup. Al and Ti are generally enriched in the finer fractions of sediments, which means that especially incorporation of coarse­ grained sedimentary components, such as from the coarse sandstones in the Lower Beacon would be especially effective for reducing both Al and Ti concentrations in the diamict (Young &

Nesbitt, 1998).

The Ti and Al concentrations of the Sirius Group are similar to those of basal tills from the upper part of CEROS-I from McMurdo Sound (Figure 6.6; Roser & Pyne, 1989) and diamicts from CRP-1 (not shown here; Krissek & Kyle, 1998). Some of the Sirius Group sediments are

155 ■ CIROS-1 above 366 mbsf A CIROS-1 below 366 mbsf F 0.8 o Sirius Group + Average Upper Crust

%

5 10 15 20 wt % AI2O3

1.0

K Catfish Creek till

■ Port Stanley till 0.7 -

M 0 . 6 - & NWT till on Mesoproterozoic O *- 0.5 - ♦ Matachewan till on Archean 0.4 -

0 Sirius Group

0.2 - + Average upper crust

0.0 0 5 10 15 20 wt % ALO,

Figure 6.6. Plot o f wt. % Al^O, versus wt. % TiC, to compare provenance o f the Sirius Group to diamictons from CIROS-1 (McMurdo Sound) and to Canadian Pleistocene tills. Values of CIROS-1 diamictons from Roser & Pyne (1989), values for Canadian tills and Average Upper Crust from Young & Nesbitt (1998) and Nesbitt & Young (1984).

1 5 6 500 700 450 □ 600 ♦Mt. Fleming 400

350 500 • E ~ 300 A â 400 S 250 n A m 130 300 N 200 ^ □

150 200 ■ 100 ji ♦ 100 • 50 0 0 ■ 10 20 30 40 50 60 0.0 1.0 2.0 3.0 4.0 5.0

Y (p p m ) W t % KzO

500 60 450 50 400

350 40 □ A E 300 Q. A 3 :2 5 0 - S 30 ^ 200 • 20 150 ■ Mt. Fleming 100 ■ 50 - 0 - 5 10 15 20 200 4 0 0 600 8 0 0 0 Ba (ppm) W t % ALO,

° Tillite Spur 6 Quartz Hills X Table Mountain ♦ Other Sirius Group

Figure 6.7. Immobile elements in the Sirius Group. Note that the Upper Fleming Tilhs more enriched in Ba and IQO.

157 more enriched in Al, however, which suggests a higher contribution from an acidic igneous

source to the Sirius Group than to the sediments in CIROS-1, or the incorporation of muddy

sediments derived from an acidic source rock, such as Permian mudrocks of the Beacon

Supergroup (Homer, 1992).

In comparison to Pleistocene tills from the Canadian Shield (Young & Nesbitt, 1998), the

Sirius Group is more enriched in both Ti and Al. The Ti and Al concentrations show a closer

resemblance to the average composition of the upper crust. An explanation for this difference

may be the presence of a higher amount of clay in the Sirius Group relative to Pleistocene

deposits from the Canadian Shield (See Chapter 3), since Ti and Al are generally enriched in

phyllosilicates, which make up the fine fraction of sediments.

6.4.2 Recycling of sedimentary rocks

Plots of Zr to Y, Ba to K20, Zr to Ba and Y to Al (Figure 6.7) show that the Sirius Group

deposits from the central Transantarctic Mountains and most deposits from the Dry Valleys have

similar provenance, whereas the Sirius Group samples from the Table Mountain and Reedy

Glacier form consistent outliers. With the exception of the sediments from the Reedy Glacier

area, the concentrations of Y in the Sirius Group are low compared to source rocks within the

Transantarctic Mountains (Tables 6.2 and 6.4). A plot of Ba/Zr ratios vs. Y/Al ratios illustrates

the effect of mixing of materials from different sources. Zr is enriched in the sandy diamicts of

Mount Sirius and Mount Feather. Zr enrichment in sediments is an indicator of recycling of

sedimentary rocks (McLennan et al., 1992). Sandy tills are characterized by low Al/Ba ratios and

high Zr/AI ratios, whereas silty tills show an opposite pattern (Figure 6.7.). Sediments from

Oliver Bluffs and Mount Fleming clearly reflect a higher contribution of granitoid basement

158 40 14 Lower 35 Beacon». Mount Sirius 12 - □ F eather 30 g 10 - Lashly"# ♦Mt. Fleming 5 25 JO 8 - 2 20 ^ ♦♦♦ Granitoid I 6 - N 15 □A N 4 ■ « san d y F errar silty tills 2 - G roup tills 0 0.02 0.03 0.04 0.05 2 3 ALO,/Ba ratio Ba/Zr ratio

4.5

4 Granitoid

3.5 - ^Dolerite (Ferrar Gp) Q Tillite Spur o 3 Lashly Fm A ra 2 .5 Quartz Hills y . X. Table Mountain 2 Lower Beacon m Basalt ♦ Other Sirius Group 1.5 (Ferrar Gp) ° A Source rocks

0 .5 ■

0 -

Y/Al^Gj ratio

Figure 6.8. Major and trace element ratios indicating grain-size effects and provenance variability in the Sirius Group. Source rock data from Roser & Pyne (1989).

159 rocks or incorporation of sediments derived from such sources.

Sedimentary rocks are characterized by high Zr/Y ratios, granitoid basement by high

Ba/Zr ratios and the Ferrar Group by both low Zr/Y and Ba/Zr ratios. A plot of the Ba/Zr ratios

vs. the Zr/Y ratios illustrates that except for the Reedy Glacier Sirius Group, all other Sirius

Group deposits appear to be mixtures of three main source rocks: sedimentary rocks, such as the

Beacon Supergroup, the Ferrar Group and granitoid basement (Figure 6.8). Sirius Group deposits

from the Dominion Range, Bennett Platform, Roberts Massif and the lower part of Mount Sirius

are most enriched in Ferrar Group and granitoid components. The upper part of the Mount Sirius

succession. Mount Feather, Mount Fleming, and the Tillite Spur Sirius Group have a different

provenance (Figure 6.8). The Mount Sirius, Mount Feather and Tillite Spur deposits have high

Zr/Y ratios, suggesting that recycled sedimentary rocks are important components. The Triassic

Beacon Supergroup has ratios similar to those of Mount Sirius and Mount Feather deposits,

suggesting that the latter were derived from source rocks similar to the ones that supplied the

Beacon Supergroup, or that the Sirius Group at these locations was partly derived from the

Triassic Beacon Supergroup. Other sedimentary rocks, such as Cambrian (?) sandstones,

equivalents of the Beacon Supergroup (Minshew, 1966) or perhaps Cenozoic sedimentary

sequences that are now hidden by the ice or are no longer preserved, may have contributed to the

Tillite Spur diamicts.

The Mount Fleming Upper Till has relatively high Ba and K :0 concentrations (Figure

6.7), indicating a larger contribution from an immature acidic source, such as granitoid basement or sediments derived from a granitoid source terrain. Sediments of the Lashly Formation of the

Beacon Supergroup are petrographically immature and contain higher percentages of K-feldspar and plagioclase than the underlying Beacon formations (Korsch, 1974). The Fleming Upper Till overlies Fleming Formation sandstones, which lie directly below the Lashly Formation (Stroeven

1 6 0 Meyer Desert 13, paleosol n=1 PALEOSOL Pleistocene glacial clays (Nesbitt & Young, 1982) n=20 Gowganda argillites (Nesbitt & Young, 1982) n=78 Quartz Hills mudstone n=2 GLACIAL CLAYS Oliver Bluffs stratified mudstone/diamict, unit 4 n=3 Bennett Platform mudstone, unit 5 and 6 n=2 Pleistocene tills (Nesbitt & Young, 1982) n=9 Gowganda Fm diamictites (Nesbitt & Young. 1982) n=77 Tillite Spur diamict n=2 Dry Valleys Sirius Group diamicts n=2 Mount Sirius upper diamict, unit 3 n=2 Mount Sirius lower diamict, units 1 and 2 n=2 DIAMICT Roberts Massif diamict polar plateau side in=1 Roberts Massif diamict on glacial grooves n=1 Meyer Desert 13, diamict n=2 Oliver Bluffs diamict, units 1 and 3 n=: Bennett Platform diamict lower, units 1 througti 4

Bennett Platform diamict upper, units 7 and 8 n= Arcfiean Shield (Nesbitt & Young, 1982) n= BASEMENT ROCK I 1----- —I— 40 50 60 70 80

CIA

Figure 6.9. Chemical Index o f Alteration (CIA) of Sirius Group deposits compared to Pleistocene and Proterozoic glacial and interglacial sediments of the Canadian Shield (data from Nesbitt & Young, 1984).

1 6 1 & Prentice, 1997). However, the available chemical data indicate that the Lashly Formation has

lower Ba concentrations than the Mount Fleming Upper Till. Therefore, reworking of other

sedimentary rocks derived from weathered and/or acidic basement sources cannot be ruled out.

The chemical composition of the Sirius Group till on Table Mountain is consistent with

very low K-feldspar concentrations and very high quartz contents (see modal analysis. Chapter

5). As suggested by the roundness of the quartz grains and the coarse grain-size of the sediment,

this deposit probably originated from local recycling of Beacon Supergroup sandstones or fluvio-

glacial reworking after deposition of the till (Chapter 5).

In addition to high Y, Al and K 2O concentrations (Figure 6.7) the Sirius Group from

Quartz Hills also has low SiOz, high TiO: and P 2O5 concentrations, which points to combined

inputs from granitic and metamorphic basement sources. The abundance of micas (see Chapter 5)

is reflected in the high Ti concentration (biotite), and suggests that phyllite-schists contributed to these sediments (Minshew, 1966).

6.5. Weathering

During chemical weathering of the upper crust the dominant process is the degradation of feldspars and other labile components and the formation of clay minerals. A measure of the degree of chemical weathering can be obtained through calculation of the Chemical Index of

Alteration (CIA) from bulk inorganic chemical analyses:

CIA = [Al2 0 3 /(Al2 0 3 + CaO^+NazO+KzO)] x 100 where components are expressed as molar proportions, and CaO* represents the amount of CaO fixed in silicate minerals (Nesbitt & Young, 1982). Because of the presence of detrital carbonate in the Sirius Group, CaO values were corrected to represent only silicate minerals. The following

162 CIA Ti/AI Ba/Sr

0.05 0.10 1.5 2.0 2.5 3.0 60 040 Diamicton \ 40- (m) Stratified diamicton Breccia 60- Gravel/ Conglomerate 70- Sand and gravel

80- Sand CP Mudstone 90- Paleosol

100 Gravel lag (a) Bennett Platform

10 m - 450 m Section 5 0 m - CIA TUAI B a /S r 40 45 50 55 60 0.00 0.05 0.10 0

Section 8

Unit

O Unit 4

Unit 3b Unit s Unit 3a T I 3 ; 3a Unit 2 Unit1 Oliver Bluffs

Figure 6.10. Changes in weathering indices and Ti/Al ratios along stratigraphie sections at Bennett Platform (a) and Oliver Bluffs (b).

1 6 3 relation was used, assuming that all carbonate CaO was present as calcite: CaO* = mol CaO -

mol CO: (calcite).

The CIA index was originally developed for lutites (Nesbitt & Young, 1982), because

hydraulic sorting during sediment transport affects the CIA values. Glacial transport does not

modify the composition of the sediment source through processes such as sorting, so the CIA can

be used to evaluate weathering of source rocks prior to glacial erosion (Nesbitt & Young, 1982;

Krissek & Kyle, 1998). Chemical studies of both the fine fraction (< 63 |um) and bulk samples

from diamicts in CRP-I show that the presence of a sand fraction has an effect on the chemical

composition of the sediments, but it has an insignificant effect on the CIA (Krissek & Kyle,

1998).

The CIA values of the Sirius Group vary between 47 and 70 (Table 6.4). The Table

Mountain CIAs were omitted, because those sediments are so coarse that their CIAs cannot be compared to those of the other sediments. In general, Sirius Group mudstones have low CIAs

(52-58) compared to Pleistocene glacial clays from the Canadian Shield (Figure 6.9) and

Proterozoic Gowganda Formation argillites (CIAs of 60-65; Nesbitt & Young, 1982). However, the diamicts associated with these Sirius Group mudstones (Figure 6.10) have even lower CIAs

(47-53). These values are also lower than the CIAs of Pleistocene diamicts from the Canadian

Shield (average of 52) and the Gowganda Formation diamictites (average of 56; Nesbitt &

Young, 1982). Other diamicts of the Sirius Group have significantly higher CIAs (56-62).

Comparison of the chemical compositions of the Sirius Group to predicted weathering trends of source rocks within the Transantarctic Mountains shows that the relations are inconsistent with weathering being the sole cause of compositional differences within the Sirius Group (Figure

6 . 11).

1 6 4 Sirius Group of the Transantarctic Mountains

» Shackleton 100 - Kaol. ° Beardmore • Mount Sirius 8 0 - mite illite Mu. ■ Dry Valleys Mu. R e ed y O 60 Bio. K-fsp.

Cpx. Hb. Bio. CN K CNK FM

Source rocks within the Transantarctic Mountains

A A • Ferrar Group Kaol. ° Beacon Supergroup Illite * M cM urdoV olcanics iiiite. + Granitoid basement

Cpx. Hb. Bio. CN K CNK FM

Figure 6.11. Ternary plots of Ca0*-i-Na,0 - ALO;- K.O (CN-A-K) and CaQ*-rNa,QTK20 - ALO; - FeO-rMgO (CNK-A-FM) systems. In the upper diagrams the Sirius Group is plotted, in the lower diagrams source rocks within the Transantarctic Mountains are plotted with theoretical weathering trends (Nesbitt & Young, 1984; 1989). The linear trends in the upper diagrams are inconsistent with weathering being the sole control of the variability in the chemical composition of the Sirius Group. Recycling of sedimentary rocks with a more evolved chemical composition is a possibility.

1 6 5 6.6. Discussion

The low CIAs (47-53) of diamicts from Oliver Bluffs and Bennett Platform suggest that

the source rocks had experienced little or no weathering. Proglacial clays from these locations

also have low CIAs, suggesting that climates were cool-temperate, but colder than Northern

Hemisphere mid-latitude glacial climates (Figure 6.10). The deposits with low CIA values all

occur in the valley sections of the Sirius Group. The diamicts with high CIAs are erosional

remnants in the higher regions of the Transantarctic Mountains, overlying horizontal rock

plateaus and scoured surfaces. The composition of high CIA Sirius Group deposits can be explained by three processes. First, the sediments may have been reworked from older sedimentary rocks deposited under warmer climatic conditions. Second, the tills may have been derived from weathered crystalline source rocks. Third, the sediments may have been affected by post-depositional weathering.

A paleosol at Meyer Desert section 13 has a CIA (63) that is several points higher than the CIAs of the glacial sediments immediately below it and in other parts of the succession (55-

58). This fact suggests that the contemporaneous weathering regime during deposition of the

Sirius Group at Meyer Desert section 13 had a significant impact on the composition of the sediments in the paleosol. The proglacial muds in the Bennett Platform section also have a higher

CIA than the diamictons (Figure 6.10). However, all the other samples discussed here are from the centers of diamict units, which makes it unlikely that post-depositional weathering during subaerial exposure affected their CIAs. The CIAs of these diamicts appear to reflect the weathering state of bedrock surfaces eroded by the ice, or the reworking of sedimentary rocks deposited under specific climate-related weathering conditions.

166 CIA 30 40 50 60 70 80

CRP-1 -2 3 Ma

o â

CIROS-1

c

Figure 6.12. Decreasing CIA with time in sediments from the Victoria Land basin. Data from Roser & Pyne (1989) and Krissek & Kyle (1998). Figure after Krissek & Kyle (1998).

CIAs of the Sirius Group are comparable to those of Cenozoic sediments from drillcores in the Victoria Land basin (Krissek & Kyle, 1998). The CIA record of the Victoria Land basin is presented in Figure 6.12. However, it should be noted that these CIAs have not been corrected for the presence of high-Ca carbonate, and they should therefore be regarded as minimum CIAs.

Carbonate analyses of sediments from CRP-1 shows that carbonate concentration is generally <

2% with a few exceptions. However, the carbonate concentrations in some of the lithological units sampled for chemical analysis are as high as 6.3 wt. % (Dietrich & Klosa, 1998), which is high enough to have a significant effect on the CIA. Carbonate concentrations in the CIROS-1 core were not analyzed, but it is likely that some low CIA values are also caused by higher carbonate concentrations in the sediments. Another factor that should be noted is that sediment

167 provenance also affects the CIA, so that a provenance change to more basic source lithologies

produces an apparent decrease in weathering intensity.

The CIAs in a combined record from the CRP-1 and CIROS-I drillcores in McMurdo

Sound show a decreasing trend through time, from values of 50-60 in the Late Eocene to 40-45 in

the Miocene (Figure 6.12). This trend can be partly attributed to a change in weathering regime,

from more temperate and humid in the Eocene to colder in the Oligocene and Miocene. A

transition to more glacially influenced lithologies at the Eocene-OIigocene boundary in CIROS-1

supports this interpretation (Hambrey et al., 1989a).

The Sirius Group records a similar range of CIA values (Figure 6.13), possibly also

caused by differences in contemporaneous weathering regimes. Variability in the contribution of

Ferrar Group source rocks possibly also had an effect on the CIAs, since the Sirius Group

successions with the lowest CIAs appear to be most enriched in clinopyroxenes (See Chapter 5).

However, it is unlikely that this would entirely explain the variability in the weathering ratios.

Because the Victoria Land basin CIA record (Figure 6.12) is not corrected for the

presence of carbonate, and because there may be differences in provenance, it is not possible to

directly compare the CIAs of the Victoria land basin with those of the Sirius Group. However, the

trend towards a less active weathering regime through time in the Victoria Land basin suggests

that the variability in the Sirius Group CIAs may be related to differences in ages of source rocks

and weathering products that were stripped off the landscape by the ice. Figure 6.13 also lists the

recycled marine microfossil groups that were recovered from the Sirius Group. It becomes

apparent that the Sirius Group sediments with the highest CIAs contain the oldest microfossil

faunas and floras. Both the Tillite Spur and the Mount Feather diamictons contain Eocene calcareous foraminifera (Webb & Harwood, 1984; Harwood, 1986). The CIAs of these sediments are 63-70 (Table 6.3) and high Zr/Y ratios indicate recycling of sedimentary rocks (Figure 6.8).

168 CIA

Marine microfossils Oliver Bluffs lower

Bennett Platform upper Bennett Platform lower Plio. Oliver Bluffs upper ami) Plio. Quartz Hills anfflD Plio. Roberts Massif polar plateau side Bennett Platform units 5 and 6 Meyer Desert 13

ami) Mount Sirius upper . ■ Olig. Roberts Massif brown diamict on glacial grooves Mount Sirius lower - n=2

Meyer Desert 13 n=1

Dry Valleys n=2 amD © Plio. Eoc. Tillite Spur n=2 amD ® Plio. Eoc.

EH diamict ® indurated sediment clasts H mudstone <®D diatoms H paieosoi © calcareous foraminifera

Figure 6.13. Range of CIAs within the Sirius Group. The spread in CIAs suggests that significant differences in degree of weathering occurred in the parent rocks of the sediments. The presence of reworked sediment clasts and Eocene foraminifera suggest that the highest CIAs are partially caused by reworking of Eocene marine sediments. Slightly lower CIAs occur in sediments of Mount Sirius where Oligocene marine micro fossils were found. Only Neogene microfossils occur in the Sirius Group sediments with the lowest CIAs. Microfossil data from Webb et al. (1984) and Harwood (1986).

169 Harwood (1986, p. 145) stated that Pliocene diatoms are dominant in these sediments, but that

this may be an artifact of the laboratory procedures; sediment clasts that contain most of the older

floras were not disaggregated with acid, in order to preserve calcareous and organic-walled

microfossils. Therefore, it seems likely that the Sirius Group formations at Tillite Spur and Mount

Feather contain a large contribution of weathered materials of Eocene age. Late Oligocene to

early Miocene diatom floras were recovered from the upper part of the succession at Mount

Sirius, where CIAs are 58-59. Only Neogene diatom floras and other siliceous microfossils were

recovered firom the other Sirius Group deposits, which have CIAs < 55. The decreasing

weathering ratios in combination with the decreasing ages of the oldest microfossil assemblages,

suggest that the weathering history of East Antarctica is reflected in the composition of the Sirius

Group through recycling of Cenozoic sedimentary rocks and weathering products.

6.7 Summary and concluding remarks

Bulk chemical analyses of the Sirius Group prove to be very effective for distinguishing

different provenance and weathering histories of parent rocks for individual Sirius Group

successions. Except for the Sirius Group at Reedy Glacier, the composition of most Sirius Group

deposits reflects contributions from basic igneous Ferrar Group rocks, (recycled) granitoid

basement, or sedimentary source rocks. The successions at Oliver Bluffs, Bennett Platform,

Roberts Massif and the lower part of the Mount Sirius succession contain the highest contributions of Ferrar Group rocks. The upper part of the Mount Sirius succession, and sediments from Mount Feather and Tillite Spur show much evidence of recycling of sedimentary rocks. These sedimentary rocks include the Beacon Supergroup and Cenozoic sedimentary rocks, as confirmed by the presence of Lower Cenozoic microfossil assemblages in these deposits. The

170 Sirius Group at Mount Fleming has a more granitoid provenance and possibly indicates erosion of

granitoid basement source rocks and/or immature sandstones of the Lashly Formation.

The CIAs of Sirius Group diamicts show a wide range of paleoweathering conditions,

which is explained by the recycling of sedimentary source rocks of different ages and with

different weathering histories. Diamicts from highly elevated flat mountain surfaces show the

highest CIAs, and these diamicts contain recycled marine microfossils including Lower Cenozoic

assemblages. The latter suggests that the high CIAs in these deposits can be attributed to

recycling of previously weathered Lower Cenozoic sequences from East Antarctica. The Sirius

Group deposits with lower CIAs show a lower contribution of sedimentary source rocks and

evidence of recycling only Neogene marine microfossil assemblages. The CIAs of these diamicts

and proglacial mudrocks are lower than those of Pleistocene mid-latitude glacial climates, and

suggest that conditions were cool-temperate during deposition of the sediments at Oliver Bluffs and Bennett Platform.

171 CHAPTER?

CHEMICAL COMPOSITION OF THE UPPER PLIOCENE BARDIN BLUFFS FORMATION

OF THE PAGODROMA GROUP, LAMBERT GRABEN, EAST-ANTARCTICA

7.1 Introduction

The lower Miocene to Pliocene Pagodroma Group is a succession of glacigenic Qordal

strata cropping out in the Prince Charles Mountains along the western margin of the Lambert

Graben (Bardin, 1982; McKelvey, 1994; Hambrey & McKelvey, 2000). The sediments have a

cumulative thickness of > 800 m and provide evidence for major shifts in the grounding-line of

the Amery Ice Shelf- drainage system throughout the Neogene. The Pagodroma

Group can be regarded as the East Antarctic equivalent of the Sirius Group in the Transantarctic

Mountains, with both systems recording Neogene ice-proximal conditions on the Antarctic continent (McKelvey, 1994). The Pagodroma Group was deposited in a marine environment, which has the advantage that ages can be derived from in situ fossil assemblages, in contrast to the mainly terrestrial Sirius Group, where all microfossil assemblages are reworked.

172 West Ice Shett

Prydz Bay

C',-- Amety I ,'Ice Shelf

y

Gamburtsev Subgiacial Mountains • 70'-e\# 60°E

,50'E Ice flow path

-- Ice surface 3 5 0 0 elevation Rock outcrop 500 I (Prirxte Charles km Mountains)

Figure 7.1. Location of the Pagodroma Gorge in the Prince Charles Mountains.

173 The Pagodroma Group is divided into four geographically separate stratigraphie units,

classified as formations (Table 7.1). Each formation fills a high-relief glacially moulded

paleovalley, with baselevels extending from sea level to 1500 m above sea level. The Neogene

age of the deposits suggests that significant uplift has occurred since the successions were

deposited. The dominant lithofacies are diamicts, boulder-gravels, and sandy gravels deposited in a Qordal environment with calving glaciers (McKelvey, 1994; Hambrey & McKelvey, 2000).

Form ation Altitude (m) Substrate Age Dominant Facies

Mt. Johnston 1268 to 1483 Proterozoic gneiss, > early diamict, boulder gravel metavolcanics Miocene and gravelly sand

Fisher Bench 270 to 560+ Proterozoic meta­ middle massive diamict, <90 to 190 volcanics Miocene boulder gravel

Battye Glacier 316 to 486 Proterozoic ?middle massive diamict, boulder granulites Miocene gravel and gravelly sand

Bardin Bluffs 0 to 70+ Permo-Triassic Pliocene massive/weakly stratified sandstones diamict, massive gravelly sands

Table 7.1. Stratigraphy of the Pagodroma Group, northern Prince Charles Mountains (After: Hambrey & McKelvey, 2000).

This study focuses on the Pliocene Bardin Bluffs Formation, because of its significance to the debate about Pliocene dynamic versus stable ice-sheets (Harwood & Webb, 1998; Stroeven et al.,

1998). The Bardin Bluffs Formation is best exposed in Pagodroma Gorge in the Amery Oasis

(Figure 7.1). The sediments blanket the basement surface and fill a channelized paleotopography

(Whitehead & McKelvey, in prep.). Samples were collected from fine- grained lithologies for siliceous microfossils by Jason Whitehead, University of Tasmania, during an AN ARE expedition in austral summer 1997/1998. These mudrocks can also be used to determine the

174 1 '|

Radok Lake

Bardin Bluffs Fm Hiatella Location

I— I Basement Quike Gully Recent alluvium Quike Gully I I Snow & lake ice Bardins Knoll Scarp Bardin Bluffs

True Adamson Spur North Glossopteris Gully t1 km Southern Side of Pagodroma Gorge

Figure 7.2. Map of Pagodroma Gorge with sample localities. Samples in this study -were taken from locations 1 through 6. See Figure 7.3 for detailed stratigraphie sections of sarmple localities. From: Whitehead & McKelvey, in prep.

175 weathering history on land and sediment provenance from their chemical compositions, which is

the major objective of this chapter.

The Bardin Bluffs Formation was deposited in the late Pliocene (< 3.1 Ma) to early

Pleistocene (Whitehead & McKelvey, in prep.). The sediments unconformably overlie the Permo-

Triassic Amery Group, which is chemically weathered to a depth of 2 m. The type section of the

Bardin Bluffs Formation is at Bardin Bluffs in Pagodroma Gorge (Figure 7.2). The Bardin Bluffs

Formation is divided into two members (Bardin, 1982; Whitehead & McKelvey, in prep).

Member 1 has a gray color and consists of glaciomarine silts and sands with minor cyclopels-

cyclopsams and slope breccias. Member 2 is brown in color and consists of waterlaid tills with

sands and gravels perhaps released from proximal melt-out streams. Based on diatom

biostratigraphy. Member 1 was deposited between 3.0 and 3.1 Ma and Member 2 was deposited <

3 Ma. Planktonic foraminifers in the upper part of Member 1 at Bardin Bluffs indicate a Pliocene

age and open marine conditions with normal salinity.

The lithological and mineralogical composition of the Bardin Bluffs Formation is summarized in

Bardin (1982). His conclusions based on studies of heavy minerals, clay minerals, authigenic

minerals, and gross chemical composition were that Member 1 (gray moraine, unit Me of Bardin,

1982), showed evidence of conditions slightly warmer than present, whereas the overlying brown

Member 2 (Mk of Bardin, 1982), separated from Member 1 by gravels, was considerably less

weathered. However, Bardin (1982) did not recognize the largely marine character of the

Pagodroma Group (McKelvey, 1984; Hambrey & McKelvey, 2000) and interpreted the abundance of secondary minerals in terms of their formation under subaerial conditions. The revised depositional setting requires a réévaluation of the compositional data. The present study investigates whether Member 1 was deposited under warmer conditions than Member 2 using

176 © 1 3 o IS « >• kill j l! g E l O I « 1 S 1 I I t © . I I s I 5 3 î Ü ÿ i a 9 c M Scale 10m a S. a. © . I (0 Member 1 Î

© L o i w

(D

■s Member 2 Member i CO a Unit 3 Unit 2 Unit 1 0 3 O. cn c o E eg Member 2 em ber i <

Figure 7.3. Detailed stratigraphie logs of sample locations within the Bardin Bluffs Formation. From Whitehead & McKelvey, in prep.

177 bulk chemical data, carbonate content, and a calculated weathering index, the chemical index of

alteration (CIA; Nesbitt & Young, 1982).

7.2 Methods

A total of 20 mudstone and muddy fine sandstone samples of the Bardin Bluffs

Formation were analyzed. The samples are from 6 different localities in Pagodroma Gorge

(Figure 7.3) and from the two members described above: 13 samples originated from Member 1,

and 7 samples from Member 2. The procedures of the chemical analysis are similar to those

described in section 6.2 of Chapter 6. The data and further details about the procedures are listed

in Appendix D.

7.3 Results

Bulk chemical compositions of individual samples from the six localities in Figure 7.2 differ significantly (Table 7.2). Part of the variability may be explained by differences in the percentage of sand of the mudrocks and fine sandstones (Figure 7.4). The coarsest lithologies are more Si-rich, but the concentrations of Fe, Mg, P, Ti and AI have slightly higher values in the finer lithologies than in the sandy facies. Although Ti and A1 concentrations decline gradually with increasing Si concentration, the concentrations of Fe, Mg and P in the coarser lithologies are scattered. Some high values occur in the CaO concentrations, which correspond with high calcium carbonate concentrations.

Plots of trace elements vs. SiOi show that Cr follows the same pattern as Fe, Mg and P

(Figure 7.5). Zr has a positive relation to SiOa, with the coarser sediments being enriched in Zr.

Strontium, Y and Ba have relatively constant concentrations in all samples (with a few exceptions for Sr) and no grain-size effects are found. Trace element data are listed in Table 7.3.

178 Sample SiOz TiOz AI2 O3 FezO] MgO MnO CaO NazO KzO P2 O5 wt % wt % wt % wt % wt % wt % wt % wt % wt % wt %

Hiatella Location (I)

1-006 51.56 0.83 17.63 5.69 1.51 0.05 2.34 1.52 3.81 0.14 1-009 53.27 0.93 19.65 5.56 1.53 0.04 0.32 1.02 3.87 0.16 1-011 57.98 0.58 13.66 4.02 1.04 0.05 2.36 1.04 3.57 0.09 1-016 51.13 0.80 16.50 5.69 1.31 0.05 2.46 0.98 3.65 0.11 1-020 52.84 0.87 17.65 6.56 1.46 0.08 4.95 0.98 3.81 0.14 1-025 55.62 0.77 15.13 6.55 1.16 0.43 0.49 0.82 3.20 0.14

Lower QuikeGully (2)

7-001 66.53 0.82 15.49 5.05 1.53 0.05 0.64 4.58 3.78 0.11

Upper QuikeGully (3)

6-004 58.62 0.87 16.50 5.19 1.49 0.04 0.57 1.25 3.63 0.14 6-014 60.11 0.73 13.70 5.49 1.19 0.05 0.42 1.02 3.14 0.11 6-033 58.83 0.87 17.29 5.33 1.49 0.05 0.46 1.21 3.78 0.14

Bardins Knoll (4)

8-001 67.39 0.67 12.19 3.90 0.95 0.08 0.64 1.01 3.46 0.11

Bardin Bluffs (5)

3AA-026 63.11 0.75 14.70 1.56 0.53 -0.01 0.43 0.40 3.76 0.07 3AA-027 65.89 0.77 14.64 2.12 0.66 -0.01 0.31 0.46 3.78 0.07 3AB-025 71.88 0.62 12.87 2.06 0.80 0.01 0.25 0.55 3.60 0.05 3AB-12 61.18 0.93 20.22 5.99 1.53 0.08 0.59 0.65 4.22 0.14 3AB-18 51.99 0.87 18.74 5.10 1.53 0.05 0.43 0.54 3.63 0.11 3AD-4 62.47 0.87 16.17 5.15 1.29 0.06 1.55 0.66 3.85 0.16

Adamson Spur (6) lOB-001 57.55 0.78 14.57 3.69 1.11 0.04 0.49 0.57 3.38 0.11 lOB-002 61.83 0.93 17.46 4.49 1.46 0.05 0.50 0.67 3.87 0.11 9-001 59.90 0.88 15.06 4.93 1.09 0.06 0.60 0.67 3.90 0.14

Table 7.2. Major element data as determined by ICP-OES analysis for the Bardin Bluffs Formation, Prince Charles Mountains.

179 Sample Ba Cr Nb Sr Y Zr ppm ppm ppm ppm ppm ppm

Hiatella Location (I)

1-006 695 97 17 275 32 197 1-009 663 94 16 159 32 197 1-011 624 72 - 159 23 141 1-016 652 79 34 209 32 232 1-020 798 104 15 412 33 243 1-025 752 77 - 154 29 207

Lower Quike Gully (2)

7-001 659 59 - 104 35 219

Upper Quike Gully (3)

6-004 829 89 19 116 41 235 6-014 534 79 21 84 32 217 6-033 682 89 - 104 32 188

Bardins Knoll (4)

8-001 688 57 17 118 36 311

Bardin Bluffs (5)

3AA-026 795 90 15 125 26 261 3AA-027 778 71 32 124 31 247 3AB-025 704 50 - 132 33 212 3AB-12 1130 96 20 125 45 186 3AB-18 717 101 31 105 33 149 3AD-4 816 76 19 209 39 260

Adamson Spur (6) lOB-001 735 85 29 139 28 279 lOB-002 852 84 0 162 36 319 9-001 775 87 13 120 37 317

Table 7.3. Trace elements as determined by ICP-OES in the Bardin Bluffs Formation, Prince Charles Mountains.

180 Sample Lithologie description CIA. A12 0 3 /^ 1 0 2 CaCOs wt. % Hiatella Location ( 1)

1-006 gray-brown mudstone with shell fragments 69 21.1 2.7 1-009 brown mudstone 76 21.0 0.2 1-011 gray mudstone with shell fragments 61 23.4 1.2 1-016 gray mudstone with shell fragments 68 20.6 2.1 1-020 brown mudstone with shell fragments 69 20.3 6.7 1-025 gray-brown mudstone 76 19.7 1.0

Lower Quike Gully (2)

7-001 brown muddy sandstone 57 19.0 1.0

Upper Quike Gully (3)

6-004 reddish-brown clayey mudstone 71 19.0 0.2 6-014 reddish brown mudstone 71 18.7 0.2 6-033 reddish-brown mudstone 73 19.9 0.4

Bardins Knoll (4)

8-001 gray-brown sandy mudstone/very fine sandstone 66 18.3 0.2

Bardin Bluffs (5)

3 AA-026 gray muddy very fine sandstone 74 19.6 0.2 3 AA-027 gray sandy mudstone 73 19.1 0.0 3AB-025 gray sandy mudstone 72 20.8 0.2 3AB-12 gray-brown mudstone 77 21.6 0.7 3AB-18 gray-brown mudstone 79 21.6 0.5 3AD-4 gray-brown mudstone 72 18.6 1.9

Adamson Spur (6) lOB-001 gray sandy mudstone 75 18.6 0.7 1 OB-002 gray-brown mudstone 76 18.7 0.7 9-001 brown very fine sandstone, laminite 72 17.0 0.5

Table 7.4. Carbonate concentration, chemical index of alteration (CIA), and Al 2 0 3 /T i 0 2 ratio for the Bardin Bluffs Formation, Prince Charles Mountains.

181 1 0.5 □ • 0.8 0.4 - o p 0.6 - 2 0 .3 - 5? % 0.4 ■ $ 0.2 - • mudstones 0.2 ■ o sandy mudstones and sandstones

0 - 50 55 60 65 70 75 50 55 60 65 70 75 w t % SiO, wt % SiO, 25 6

20 ■ 5 O 4 f 15- Ü 3 10 - 2 1 0 50 55 60 65 70 75 50 55 60 65 70 75 w t % SiO, wt % SiO;

7 5 6 4 5 £ 4 3 3 2 2 1 1 □ •••, 0 0 50 55 60 65 70 75 50 55 60 65 70 75 wt % SiO , wt % SiO, 2 5 4 .5 oo 3 1 O 2 0.5 1 0 0 50 55 60 65 70 75 5560 6550 70 75

wt % SiO , wt % SiO, 0.2

0.15 - m o* • • □ □

^ 0.05 ■

50 55 60 65 70 75

w t % SiO ,

Figure 7.4. Plots of major elements versus wt. % SiO; for the Bardin Bluffs Formation. Note grain-size effect on composition.

1 8 2 350 a • □ • • 300 • • • □ ^ 80 a - ° 250 E . c Q.E 200 *• • • â 60 Q- a 150 * # 100 ■ 20 50 ■ 0 ------a------1------0 •------,------g ------60 65 60 65 Wt % SiO, wt%SiO, 1200 1000 800 E 300 CLh 600 U. 2 0 0 • ca CD 400 200 - 0 60 65 60 65 Wt % SiO, wt % SiO,

• mudstones □ sandy mudstones and sandstones

wt% SiO

Figure 7.5. Plots of trace elements versus wt % SiO; for the Bardin Bluffs Formation. Note grain-size effect on trace element composition.

183 Carbonate concentrations range from 0 to 6.7 % and are highest at the Hiatella Location

in Pagodroma Gorge, where numerous shell fragments occur in the mudrocks. Other relatively

high carbonate concentrations are found at the top of Member I in the Bardin Bluffs section,

where foraminifera and molluscs are reported (McKelvey et al., in prep.). A plot of calcium

carbonate concentration versus Sr shows that excess Sr resides in the carbonate (Figure 7.6).

250 - w 200 = 0.909 150 -

100 -

0.00 2.00 4.00 6.00 8.00 wt. % CaCO

Figure 7.6. Plot of wt. % CaCOa vs. Sr in ppm. The good correlation suggests that Sr resides in the Ca-carbonate.

7.4 Discussion

7.4.1 Provenance

The Si-Ti and Si-Al plots (Figure 7.4) show that Ti and Al behave in a similar fashion with respect to hydraulic sorting and the elements appear to be relatively immobile because they plot on a straight line (cf. Fralick and Kronberg, 1997). Ti and Al are considered to be relatively immobile in most weathering regimes and are useful indicators of provenance, because their

184 initial ratios vary greatly in primary (igneous) source rocks (Young and Nesbitt, 1998). The

Bardin Bluffs Formation has intermediate Ti and Al values (Figure 7.7a). Crystalline basement

rocks from the northern Prince Charles Mountains are also plotted, and the pellitic gneisses from

the northern Prince Charles Mountains show some overlap with the Bardin Bluffs Formation.

Cr also shows evidence of limited mobility, because samples plot on a straight line relative to SiO: concentrations (Figure 7.5). A plot of Cr versus TiOz further explores the relationship of the Bardin Bluffs Formation with its source rocks. The Cr values are also nearest to those of politic gneisses (Figure 7.8a). When all three elements Cr, Al and Ti, are considered, the Bardin Bluffs Formation samples plot in a tight cluster, suggesting that these elements have a common source. The closest in terms of composition is the pellitic gneiss (Figure 7.8b).

The variability in the Ti and Al values of the Bardin Bluffs Formation can be attributed to differences in grain-size (Figure 7.7b). Sandy mudstones and sandstones have lower Ti and Al concentrations than the mudstones. The mudstones are enriched in Al with respect to known basement rocks from the area, which suggests that the source rocks were weathered before erosion occurred, and that the weathering products (clay minerals) were concentrated in the muds by hydraulic sorting (cf. Young & Nesbitt, 1998). Zr is also relatively immobile judging from the

Si-Zr plot (Figure 7.5). In contrast to Cr, Al and Ti, which concentrate in the mud fraction, Zr- bearing minerals, such as zircon, concentrate in the sand fraction due to their high mechanical and chemical stability. A plot of TiOz/Zr ratio versus Zr/AlzOz ratio illustrates this (Figure 7.9): the sandy mudstones and sandstones are relatively enriched in Zr. Because Zr is also immobile, the fact that all Bardin Bluffs samples plot on a line suggests that they have a uniform provenance

(cf. Fralick & Kronberg, 1997).

185 1.8

1.6 ■ @ quartz monzonite

1.4 average vCharnockite 1.2 - pellitic g n e i s s ^ 1

0.8 ■

0.6 -

0.4 ■ mafic granulites granitoid 0.2 -

0 10 15 20 25 wt. % AIjO,

• rrudstones □ sandy midstones and sandstones

O F 0.8 - 5? weathering and sorting of muds ^ 0.6 -

0.4 -

weathering and 0.2 - sorting of sands

0 5 10 15 20 25 wt. % AIjO,

Figure 7.7. Plot illustrating the uniform AJ/Ti ratios. Minor variability is mainly caused by grain-size effects (cf. Young and Nesbitt, 1998). Solid circles are mudstones, open squares are sandy mudstones and sandstones. A) shows the relationship between Bardin Bluffs Fm Ti-Al contents and Ti-Al contents of basement rocks from the northern Prince Charles Mountains (data from Munksgaard et al, 1992; Sheraton et al., 1995; Kinny et al., 1997; Zhao et al., 1997.). B) illustrates the effects of weathering and hydraulic sorting on the Ti-Al contents of the sediments. Samples plotting to the right of the dashed line are enriched in Al with respect to the source rocks due to weathering, and concentration of weathering products in the mudstones by hydraulic sorting.

186 10000

mafic granulites

1000 pellitic gneiss I

1 100 - quartz monzonite/ ô average chamockite

10 granitoid

1.5

w t. % n o

0.12

quartz monzonite

0.08 - @ average ctiamockite

pellitic gneiss O 0.06 - _ □

0.04 - mafic igranulites 0.02 granitoid

0 0.05 0.1 0.15 nO/Crx 10.000

Figure 7.8. Plots illustrating the relation between immobile elements Ti, AJ and Cr. A) Cr-Ti plot of Bardin Bluffs Fm and basement rocks from the northern Prince Charles Mountains. Note that the vertical scale is logarithmic. B) The Bardin Bluffs Fm forms a tight cluster on a 3 variable plot, suggesting that Ti, Al, and Cr are derived from the same source in all sediments. Note the similarity in immobile element composition of the Bardin Bluffs Fm and pellitic gneiss. (Basement chemical data from: Munksgaard et al, 1992; Sheraton et al., 1995; Kinny et al.,1997; Zhao et al., 1997.)

187 quartz monzonite/ chamockite

politic gneiss

mane granulites

TiOj/Zr

Figure 7.9. Plot illustrating concentration of Zr in the sand fraction and concentration of Ti and Al in the finer fractions.

188 7.4.2 Weathering

To estimate the weathering intensities in the source rocks of the Bardin Bluffs Formation

the chemical index of alteration (CIA) was calculated from the chemical analyses. For a

definition of the index see Chapter 6, section 6.5.

Most of the CIAs of the Bardin Bluffs Formation fall within the range of average shales

(Figure 7.10). The characteristics of the TizAl plot (Figure 7.7b) suggest that some chemical

weathering took place. In the absence of chemical weathering, the samples would plot on an

almost vertical line (indicated by the dashed line in Figure 7.7), with the muds plotting at the top

of the trend and the sands at the bottom (Young and Nesbitt, 1998). Instead, the samples plot on a

line with a much lower slope, which can be interpreted by loss of Al-rich clay mineral phases from soils during weathering and concentration of these phases in basin muds by hydraulic sorting.

No systematic difference in weathering index was found between the two members

(Figures 7.10 and 7.11). Although grain-size differences affect the composition of the sediments

(e.g. Figure 7.7), there is no systematic difference in the CIAs of coarser and finer lithologies in this study (Figure 7.1 lb). A comparison of the composition of the Bardin Bluffs Formation with predicted weathering products of possible crystalline source rocks (Figure 7.12) shows that the sediments may represent weathered acidic to intermediate crystalline basement rocks (or sedimentary rocks derived from these sources). The three sandy samples, which deviate from the general trend in the CNK-A-FM plot, are from unit 2 of Member 1 at the Bardin Bluffs locality

(Figure 7.3). Bardin (1982) found considerable enrichment of authigenic minerals in these sediments. Almandine garnet is the main heavy mineral, whereas hornblende and pyroxenes are

189 56-60 61-65 66-70 „ .,5 76-80 >go

Figure 7.10. Comparison of CIAs of Member 1 (dark gray) and Member 2 (light gray). Most CIAs fall between 71 and 75. There appears to be no significant difference between members, although the number of samples is too low to draw any definate conclusions.

25 25 -

20 cn • • • 20 - ♦ o ir 15 - o O — ~10 10 - ♦ Merrfcer 1 <

5 - X Member 2 5 - • mudstones □ sandy mudstones and sandstones 0 0 - 50 60 70 80 50 60 70 80 CIA CIA

Figure 7.11. Plots of CIAs v.y. Al^Oj/TiOj ratios show the variability in CIAs with relatively constant Al^Oj/TiO, ratios.

190 insignificant. This fact is in agreement with the low Fe, Mg, Mn, and Na concentrations (Table

7.2), which suggest that plagioclase and unstable mafic heavy minerals have been removed by

leaching.

Bardin (1982) argued that the leaching and authigenic precipitation resulted from

chemical weathering upon exposure to considerably "warm" and "moist" climates and that there

is a difference in weathering state between Member 1 and Member 2. However, since it was

established that the sediments of Member I were deposited in a marine environment, the

significance of the secondary minerals changes. Secondary minerals in unit 2 of Member 1 at

Bardin Bluffs include: limonite, calcite, marcasite, and iron-rich carbonates which together form

up to 31% of the rock (Bardin, 1982). In a marine environment, the abundance of these types of

authigenic minerals is controlled by many factors, such as porosity of the sediments, water depth,

pore water chemistry, and temperature, and the abundance of secondary minerals is not a

significant paleoclimatic indicator. The high abundance of authigenic minerals in the lower part of Member 1 corresponds to the occurrence of sandstones, which have higher porosity than the overlying mudstones.

The leached nature of the sediments in Member 1 may suggest that chemical weathering under warmer conditions than today occurred prior to erosion and transport of the materials in unit 2 of Member 1. The base of Member 1 at Bardin Bluffs overlies a 2-m thick zone of weathered sediments of the Amery Group, and it is possible that the leached sediments were derived from the weathered zone. The high abundance of detrital coal in the sediment matrix supports an Amery Group provenance (Bardin, 1982). Alternatively, the leaching may have occurred by alkaline pore-waters during advanced diagenesis in a marine environment (cf. Bridle

& Robinson, 1989). The low CIAs of Member 1 sandstones from Bardin’s Knoll (66) are in the range reported for Pleistocene proglacial sediments of mid-latitude regions (Nesbitt & Young,

191 Upper Pliocene Bardin Bluffs Formation, East Antarctica

• mudstones 100 - Kaol. o sandy mudstones and sandstones lllite lllite Mu. Mu

O 60 - Bio. K-fsp.

Cpx. Hb. Bio. ON K CNK FM

♦ granite/granodiorite o gneiss Kaol. - pellitic gneiss o chamockite/ lllite lllite quartz monzonite Mu. Mu. • mafic granulites

Bio. K-fsp.

Cpx. Hb. Bio. CN K CNK FM

Figure 7.12. Ternary plots o f CaO*-i-Na,0 - AljOj- K^O (CN-A-K) and CaO^+Na^O+BLO - AljOj - FeO+MgO (CNK-A-FM) systems. In the upper diagrams the Bardin Bluffs Fm is plotted, in the lower diagrams source rocks within the northern Prince Charles Mountains are plotted with estimated weathering trends (Nesbitt & Young, 1984; 1989).

192 1982). The significance of the low CIA is that it indicates the maximum level of chemical

weathering that could have occurred during the depositional cycle of Member I. The low CIA

suggests that conditions were not warmer than cold-temperate. The presence of ice-rafted debris

(IRD) suggests that the sediments were deposited in a distal glacial environment, which means

that the CIAs could represent weathering rates during the time of deposition.

The presence of cyclopels and cyclopsams (rhythmic ice-proximal glaciomarine deposits)

in Member 1 of the Adamson Spur location requires ice-proximal depositional conditions (Figure

7.3). The CIA (72) of a very fine sandstone in this unit suggests moderate chemical weathering of

the materials before deposition, which is incompatible with ice-proximal glacial conditions.

Therefore, the weathering probably occurred within the source rocks a considerable time before

erosion of the materials by the glacier, when the area was subject to a "warmer" climate regime.

The problem is to find out when the chemical weathering took place, because if the chemically

leached materials were Permian paleosols within the Amery Group, the data would have no

bearing on the Late Neogene climate of the area.

Member 2 sediments from Lower Quike Gully have the lowest CIA (57), which suggests

that unweathered crystalline rocks or immature sedimentary rocks contributed most of the

sediment. Such low CIAs require that climate conditions were cold enough to limit chemical

weathering of most minerals, and are in agreement with the glacial nature of the deposits

(diamictons and mudstones with IRD). Higher CIAs were found for Member 2 at Adamson Spur

and at Upper Quike Gully. For these localities recycling of weathered sedimentary rocks cannot

be ruled out.

The lowest CIAs for Member 1 (66) is less than the lowest CLA for Member 2 (57). If these CIAs represent the respective contemporaneous weathering regimes, then Member 1 appears to have been deposited under slightly warmer conditions than Member 2. The presence of

193 in situ planktonic foraminifera in combination with IRD in the upper part of Member I at the type

locality (Figure 7.3) supports open marine conditions with glaciers at sea level at some distance

from the site of deposition. The upward increase in calcium carbonate in Member 1 indicates an

upsection increase in productivity. The transition from sandstones in unit 2, to mudstones in unit

3 at Bardin Bluffs (Figure 7.3) indicates a relative sea level rise and/or a receding ice margin. The

Member I sequence is truncated at the top by an unconformity, suggesting that part of the record

was not preserved, so that it is not known whether unit 3 represents peak interglacial conditions

for Member 1. Member 2 is dominated by diamict and probably represents ice-proximal

deposition.

The transition towards colder conditions from the mid-Pliocene (Member I) to the late

Pliocene (Member 2) is observed in many other records, including oxygen isotope records from

the deep sea (Shackleton et al., 1995). Mid-Pliocene records from around the world show

evidence of considerable global warming during this time (e.g. Dowsett et al., 1996). Evidence from the Bardin Bluffs Formation suggests that, in the Antarctic, cold climatic conditions prevailed during most of the mid-Pliocene, but it cannot be ruled out that conditions were slightly warmer and more humid than in the late Pliocene and Pleistocene. The large amounts of weathered material present in sediments indicates significant weathering prior to erosion, transport and deposition at Bardin Bluffs. However at this time it is not possible to conclude whether the material was derived from Mesozoic paleosols or from Upper Cenozoic weathered horizons.

194 CHAPTERS

MINERALOGY AND CHEMICAL COMPOSITION OF EOCENE AND PLIOCENE

ERRATICS, MOUNT DISCOVERY AREA, MCMURDO SOUND

8.1. Introduction

Cenozoic rocks in the Ross Sea area are not well exposed at the surface and only a few

drillholes exist to study these strata. Upper Eocene to Recent deposits have been encountered in

sitti in the Ross Sea region in the CIROS-1 (Wilson et al., 1998b), the MSSTS-1, and the CRP

drillholes, however drillholes provide only limited geographic and paleoenvironmental

information. Fortunately, hundreds of erratics of fossiliferous Eocene to Pliocene sediments have

been collected and described from moraines in the McMurdo Sound area (Harwood & Levy,

2000). The erratics were probably derived from outcrops of Eocene or younger rocks covered by the ice-sheet, and have become a valuable additional source of paleo-environmental information.

195 M a c k a y G i .

M c M u r d o S o u n d T a y l o r isrand G l a c i e r R o s s I c e S h e l f W P ^ v e r y # ^ Island

Figure 8.1. Location of the Mount Discovery area.

The samples studied here were collected in December 1997 during one day of field work in the Mount Discovery area with David Harwood (Figure 8.1). The purpose of the trip was to collect more fragments of a mudstone, informally named D-1 (Harwood & Levy, 2000), which is part of a middle to upper Eocene suite of erratics representing deposition in coastal marine environments (Levy and Harwood, 2000). The suite of erratics represents a variety of facies, including sandstone, sandy mudstone, conglomerate, mudstone, and diamictite. During the field work on the Mount Discovery coastal moraine, three lithofacies of erratics were encountered: 1) an olive-green volcaniclastic diamictite with rounded clasts, mainly of basalt lithology, 2) a yellowish gray fossiliferous quartz-rich coarse-grained sandstone, and 3) an olive-brown fissile mudstone (D-1).

The sandstone contains a marine macrofauna dominated by bivalves (Figure 8.2) and probably represents a much shallower depositional environment than the sediments encountered in McMurdo Sound drillcores (cf. Levy & Harwood, 2000).

196 Figure 8.2. Eocene sandstone erratic from the Mount Discovery coastal moraine. Note mollusc fauna.

The mudstone D-1 is one of the key lithologies used to characterize Eocene paleo- environments on the basis of erratics (Harwood & Levy, 2000). D-1 contains marine palynomorphs, terrestrial palynomorphs, and siliceous microfossils, and its age is constrained to the middle to late Eocene based on dinoflagellate cysts and siliceous microfossils. The well-dated

Eocene mudstone provides an opportunity to expand our knowledge of Eocene depositional environments and paleoclimates in the Ross Sea area through compositional studies.

The volcaniclastic diamictite may represent erratics of the Scallop Hill Formation

(Eggers, 1979), which crops out in situ on Brown Peninsula, north of Mount Discovery (Figure

8.1). Volcaniclastic rocks similar to the erratics discussed here have been found in the saddle west of Scallop Hill (concrete facies, Leckie & Webb, 1979). The Scallop Hill Formation was dated to late Pliocene based on K-Ar dating (Webb & Andreasen, 1986) and was deposited in a shallow marine environment (Eggers, 1979).

197 The Eocene mudstone and the matrix of the diamictite were analyzed by X-ray

diffraction (XRD) at McMurdo Station to characterize the mineralogical compositions. The bulk

geochemistry of the Eocene mudstone was determined to assess the paleoweathering conditions

on the continent during the time of its deposition. The chemical results were compared to bulk

chemical data from the lower part of CIROS-1 (Roser & Pyne, 1989; Krissek & Kyle, 1998). The

weathering history of middle to upper Eocene sediments can be regarded as a baseline for

Tertiary weathering conditions on the Antarctic continent, since the middle to upper Eocene

weathering history reflects the climatic conditions on the continent before the initiation of cooling

and the development of large ice-sheets.

8.2 Methods

8.2.1 X-ray diffraction

For XRD analysis, a sample of the fissile mudstone (D-1) and two samples o f erratics of the volcaniclastic diamictite (Scallop Hill Formation) were dried at 60°C and powdered in a ball mill for 5 minutes. Procedures of the XRD analyses are as described in Chapter 5, section 5.2.

8.2.2 Chemical composition

Some of the ground mudstone sample (D-1) was sent to XRAL Laboratories in Toronto,

Canada, for Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES) analysis. At

XRAL the sample was prepared by LiBO? fusion. Ten major elements and six trace elements were analyzed: Si, Ti, Al, Fe, Mg, Mn, Na, K, Ca, and P, and Ba, Cr, Nb, Sr, Y, and Z r. Major element abundances were converted to wt. % of the respective oxides. The percentage calcium

198 carbonate was also determined. Further details about laboratory methods and accuracy, are listed

in Appendix D.

8.3 Results

The matrix of the volcaniclastic diamictites is primarily composed of plagioclase and K-

feldspar with minor additions of siderite, phillipsite and pyroxenes. Much amorphous material is

also present, presumably volcanic glass. The two erratics were recovered at some distance from

each other on the coastal moraine at Mount Dicovery, but have an almost identical matrix composition.

Felctepars 500-

Siderite 250-

Phillipsite Plagioclase süeri& LAiifL,L^ 40 50 60 20

Figure 8.3. Bulk X-ray diffractogram of a volcaniclastic diamict erratic of Pliocene age. Unprocessed diffractogram in upper right comer.

199 3000 1 Q uartz (Q) Eocene erratic 2500 - D-1

1000 - Quartz

500 - Aragonite (A) Fsi Clay minerais Ce

313 23 33 43 53 63 2 thêta

Figure 8.4. Bulk X-ray diffractogram of a fissile mudstone of Eocene age (erratic D-1). Fsp=feldspars.

The X-ray mineralogy shows that D-1 is a calcareous mudstone containing calcium carbonate and aragonite (Figure 8.4). The d-spacing of the carbonate peak is slightly smaller than that of pure calcium carbonate (3.027 Â vs. 3.035 Â), suggesting that minor substitution of magnesium has taken place. The major detrital mineral component is quartz. Feldspars are only a minor component of the rock and some clay minerals are present. The chemical composition reflects the high carbonate and low Mg concentration.

200 Major elements: SiOz TiOz AlzOj FezOz MgO MnO CaO KzO NazO PzOs wt. % 48.99 0.28 5.54 3.15 1.53 0.39 20.57 1.13 0.63 0.14 Trace elements: Carbonate analysis: Ba Cr Nb Sr Y Zr CaCOs ppm 323 19 - 216 14 168 wt. % 35.3 Chemical index of alteration (CIA.): 60

Table 8.1. Chemical composition of middle to late Eocene mudstone erratic D-1.

8.4 Discussion

8.4.1 Volcaniclastic erratics

The almost identical composition of the two volcaniclastic erratics suggests that they

originate from the same sedimentary unit within the Scallop Hill Formation. The dominance of

feldspars and the presence of amorphous material (glass) supports the conclusion of Eggers

(1979) that the rocks that occur in situ at Scallop Hill formed in a shallow marine environment near an ice-shelf, in combination with active volcanism. The carbonate cement probably consists of siderite, which gives the rock an indurated appearance. Both erratics contain phillipsite, a zeolite often encountered in basaltic volcaniclastic sediments. Phillipsite was found previously in two former lakes in Taylor Valley where it formed from vitric ash (LinkJetter, 1974) that was washed into the lakes and zeolitized under alkaline and saline conditions.

201 The lithologie character of the erratics resembles the concrete facies reported by Leckie

& Webb (1979) from the coast of Black Island, but this facies has not been found at Scallop Hill

(Eggers, 1979). The type locality of the Scallop Hill Formation, however, does represent only a portion of a more extensive (glacio-)marine volcaniclastic succession that was deposited in the coastal areas of the volcanic islands in McMurdo Sound (Leckie & Webb, 1979; Webb &

Andreasen, 1986).

8.4.2 Eocene mudstone erratic

The Ti/Al ratio of the Eocene mudstone (D-1) erratic resembles those of Upper Eocene mudstones in CIROS-1 and is consistent with a provenance from a mixture of sources, including:

Ferrar Group, Beacon Supergroup, and basement rocks (Krissek & Kyle, 1998).

The chemical index of alteration (CIA) of the mudstone erratic is higher than the average

CIA of Upper Eocene mudstones in CIROS-1. However, CIA values for CIROS-1 are quite variable, and the CIA values from CIROS-1 were not corrected for the presence of calcium carbonate. Some of the CIA values in CIROS-1 are slightly higher than the CIA of the Eocene erratic.

The CIA value of 60 for the mudstone erratic suggests a weathering history similar to that experienced by Pleistocene glacial clays from the Canadian Shield (Nesbitt & Young, 1982), suggesting that climates on the continent were cool-temperate, perhaps with glaciers in higher elevated regions. This interpretation is in agreement with the paleoenvironment deduced from fossil wood (Francis, 2000) and clay mineralogy (Holmes, 2000).

The high carbonate content of the mudstone D-1 may be a result of the presence of detrital carbonate or calcareous skeletal debris. Besides its high carbonate content, the

202 phosphorous content of the mudstone erratic D-1 is also higher than average values for Eocene

mudstones in CIROS-1. These compositional differences could either suggest that marine

productivity had a stronger influence on the mudstone erratic D-1, or that volcanic rocks were a

more important source to D-1.

The depositional environment of the Eocene mudstones in CIROS-1 has been debated, but was probably below wave-base, perhaps even in a rather deep-water bathyal environment

(Hambrey et al., 1989; Fielding et al., 1997; Claps et al., 1997). The high carbonate content of the

Eocene erratic D-1 may require a different depositional environment for the D-1 mudstone than the environment in which the CIROS-1 mudstones were deposited. Although the relationship between depositional environments and composition of mudrocks has not been well-studied, calcareous mudrocks are generally interpreted forming in a shallower depositional environment than non-calcareous mudrocks (Blatt, 1992).

203 CHAPTER 9

SEDIMENT DEFORMATION, CAPE ROBERTS DRILLING PROJECT, MCMURDO

SOUND, ROSS SEA AREA

9.1 INTRODUCTION

One of the objectives of the Cape Roberts Drilling Project is to reconstruct the glacial and climatic history of the Victoria Land basin. In that perspective, the record of grounded ice within the cores is important because it records times of major glacial expansion onto the Ross Sea continental shelf. This chapter consists of studies on two cores, CRP-1 and CRP-2/2A drilled in

1997 and 1998 at Cape Roberts.

Sedimentological properties that are used by the Cape Roberts Science Team to identify the record o f grounded ice are:

-Clast orientation fabrics (Atkins, in press; Cape Roberts Science Team, 1998; 1999)

-Microfabrics of diamicts (Van der Meer & Hiemstra, 1998; Van der Meer, in press)

-Vertical facies associations and comparison to modem depositional environments (Powell et al.,

1998; in press)

-Macroscopic evidence of glaciotectonic deformation (this study)

204 The term "glaciotectonic" refers to structures formed by glacially induced deformation of

pre-existing bedrock or sediments. Glaciotectonism has been recognized as an important process

in the terrestrial subglacial environment for a long time (Aber, 1988). In marine environments

accounts of glaciotectonic deformation have been limited due to problems recovering diagnostic

lithological records with drilling from ship platforms. However, combinations of seismic data and

core descriptions have established that the Barents Sea shelf was subject to glaciotectonic

deformation by grounded ice during the late Quaternary (Gattaulin et al., 1993; Saettem, 1994).

The drilling at Cape Roberts was based on a fast-ice platform, which resulted in

exceptional high recovery (ca. 98 %) and preservation of structural features, such as brecciation

and soft-sediment deformation in the cores. In this chapter the record of sediment deformation is

interpreted in terms of depositional envirorunents, with emphasis on the recognition of

glaciotectonic deformation. The results are compared to other methods used to reconstruct the

record of grounded ice.

Sections 9.2 and 9.3 have been previously published as two separate papers (Passchier et al., 1998; Passchier, in press) in the Cape Roberts Project Scientific Results volumes (Terra

Antarctica). The CRP-1 paper was written with T.J. Wilson and T.S. Paulsen, with ca. 70 % of the data analysis and writing done by Passchier, ca. 20 % by Wilson, and ca. 10 % by Paulsen.

205 9.2. ORIGIN OF BRECCIAS IN THE CRP-1 CORE (with T.J. Wilson and T.S. Paulsen)

9.2.1 INTRODUCTION

The CRP-1 core was drilled on a bathymetric high, Roberts Ridge, 10-15 km off Cape

Roberts, in northern McMurdo Sound, western Ross Sea (Cape Roberts Science Team, 1998a).

Drilling sampled a gently tilted sequence on the margin of the Victoria Land Basin, one of four major extensional basins within the Ross Sea continental shelf. Drilling was carried out from a fast sea-ice platform in a water depth of approximately 150 m. The drill-site is located on the western side of Roberts Ridge, which rises from a depth of 500 m below sea level in the west to less than a 100 m near the drillsite. Slopes on the northern and western sides of Roberts Ridge are steep, probably due to structurally controlled glacial erosion of the 900 m deep MacKay Sea

Valley (Cape Roberts Science Team, 1998a). The regional area has been glaciated during the time span represented by the core material, and the drillsite location may have been overridden by glaciers repeatedly (Barrett, 1986, 1989).

The CRP-1 core consists of a Quaternary glacigenic interval down to 43.55 meters below the sea floor (mbsf) and an early Miocene glacigenic interval down to the base of the core at

147.69 mbsf. Diamictites, sandstones, and siltstones are the main lithologies. The Quaternary section is composed of unconsolidated strata, the upper part of the Miocene section is composed of semi-consolidated strata, and the lower part of the Miocene section is largely composed of lithified strata. Numerous enigmatic brecciated intervals occur throughout the CRP-1 core, but are dominant between -40 and -85 mbsf in the lower Quaternary and the upper Miocene sections.

Breccias and soft-sediment deformation features were encountered in the CIROS-I and the MSSTS-1 drill cores from further south in McMurdo Sound. In the MSSTS-1 core the

206 deformation features were interpreted as the result of subglacial deformation (Barrett &

McKelvey, 1986); in the CIROS-I core they were attributed to mass-gravity flows (Hambrey et

al., 1989a). Upon the completion of drilling at Cape Roberts considerable debate remained about

the origin of breccias that occur within the CRP-1 core. Did the breccias originate by e.g.

tectonic, glacio-tectonic, mass-gravity or periglacial processes (cf. Laznicka, 1988)? We used

scanned images of the slabbed core to characterize breccia textures, with the purpose of providing

information on the physical processes of deformation. We conclude that brecciation occurred by

in situ fracturing and by horizontal planar shearing in subglacial and mass-movement settings

(Dreimanis, 1993, Hart & Roberts, 1994).

9.2.2 METHODOLOGY

After splitting of the CRP-1 core at the drillsite, the slabbed face of the working half of

the core was scanned at a resolution of 5 pixels/mm using CoreScan® equipment leased from

DMT, Germany. CoreScan images were inspected at The Ohio State University to describe breccia textures, other deformation features, and the relations between breccia occurrences, unconformities and lithological boundaries. The depths used in this paper are as initially determined at the drill site and as reported on the CoreScan images. Depths below 45.28 mbsf are the same as the ones used on the 1:20 Core Logs published in the Initial Report (Cape Roberts

Science Team, 1998c). Sediment-filled fractures were examined in two thin sections, taken from

-44 and -46 mbsf.

207 9.2.3 BRECCIA CHARACTERISTICS

Textures

Brecciated sediments consist of multiple stacked zones of alternating textures. The

textures range in thickness from several centimeters up to several decimeters, but never exceed a

meter. We classified breccias based on textural criteria, generally following Laznicka (1988).

Five breccia types were identified and include:

1) Crackle breccias, which are defined as fracturing of the rock, in which no displacement of the fragments has occurred. The fragments are angular to subangular pebbles

(Hambrey et al., 1997), and fit together in a jigsaw puzzle pattern (Figure 9.1c).

2) Mosaic breccias are similar to crackle breccias, but differ in that the fractures have been filled with sediment matrix.

3) Chaotic breccias are defined as a mixture of large fractured pebbles in a grouadmass of smaller fragments (Figure 9.1b).

4) Rubble breccias are defined as a mixture of subrounded, small pebbles, with irregular intergranular spaces (Figure 9.1a).

5) Matrix-supported breccias consist of pebbles surrounded by a fine-grained matrix

(Figure 9.Id).

Boundaries

Boundaries between breccias and non-brecciated zones within a breccia interval are either sharp or gradational. Sharp and gradational boundaries are commonly observed in the same brecciated interval (Figure 9.2c). Most boundaries between different breccia types within brecciated intervals are horizontal and planar. Irregular and inclined boundaries are rare (Figure

208 9.2a+b). Sharp and planar boundaries typically separate rubble breccias from adjacent crackle breccias. Lower boundaries of brecciated intervals are as commonly irregular as they are planar.

Thickness and internal arrangement of brecciated zones

The thickness of brecciated intervals ranges from 8 to 635 cm. In two cases, brecciated intervals reach ~6 m thickness and consist of multiple zones of alternating breccia types (Figure 9.2c;

Figure 9.3). Brecciated zones within breccia intervals appear to be horizontal tabular bodies, based on their subhorizontal upper and lower boundaries. In the upper 85 m of the core, the lower sections of brecciated intervals consist of crackle breccias. Crackle breccia zones also occur as layers within intervals dominated by rubble and matrix-supported breccias, e.g. in Figure 9.3 between -49 and -50 mbsf and in Figure 9.4 at 69.55 mbsf. Mosaic breccias typically grade into crackle breccias. Crackle breccias commonly grade into chaotic, rubble and matiix-supported breccias.

209 a '

Figure 9.1. Breccia types in the CRP-1 core, (a) Rubble breccia; characterized by small, subrounded fragments of a variety of grain sizes. Note relatively abrupt change to smaller grain size of rubble breccia across planar boundaries with chaotic breccia above and below; (b) Chaotic breccia ; with large, fractured clasts intermixed with fine-grained brecciated material; (c) Crackle breccia; characterized by network of cracks separating angular fragments that have not been displaced; (d) Matrix-supported breccia, with pebble-sized fragments surrounded by a fine­ grained matrix. Core diameter is 61 mm.

210 Sharp

Figure 9.2. Boundaries between breccia types in the CRP-1 core, (a) Inclined boundary (white dashed line); (b) Irregular boundary (white dashed line); (c) Gradational and Sharp boundaries. Most boundaries between breccia textures are sharp and planar. Note stacked nature of breccia types in 2c, where crackle breccia (cr) at base of interval passes upward into chaotic (ch) and rubble (r) breccias. Core diameter is 61 mm.

211 BRECCIA TEXTURES BRECQA TEXTURES C»irSmsanoar»«t -, 9« S«nd Qwvtl , Mlfr3 ( :#niR

43.0- 43.0- 47.0 47.0 m # I

44.0- 48.0- ,0 ■ c h 48.0-

\L=.

45.0- 45.0- 49.0- 49.0

46.0- 5 0 4 -

47.0- 47.0- 51.0-

Figure 9.3. Detailed 1:20 log of breccia 2, below the Quatemary-Miocene boundary in CRP-I. (Key: cr=crackJe, mo=mosaic, ch=chaotic, and r= rubble breccia.) Two types of associations are particularly common: 1) crackle breccias adjacent to rubble breccias with a sharp boundary in between the types; or, 2) crackle breccias grading upwards or downwards into chaotic textures with large fractured clasts in a rubbly groundmass, which in turn pass into rubble breccias. No changes in lithology occur at the boundaries between textures. The crackle breccias clearly represent in situ brecciation with minimal movement of clasts. The smaller and more rounded fragments in the rubble breccias point to grain size reduction and abrasion of clasts, suggesting some degree of transport.

212 BRECCIA TEXTURES LITHOLOGY Qay SIR Sand Gnvel Depth Oepth a t y SIR Sand Gravel (mbsO (mbsO

67.0- 67.0 j - A

.6 - Jt

68.0

A -

A - A - PS, A - 69.0- 69.0

.« -

A -

70.0 -

X -

.6 - J - 71.0 -

^ = soft sedim ent deform ation = microfauits

Figure 9.4. Detailed 1:20 log of breccia 4, CRP-1. (Key: cr=crackle, mo=mosaic, ch=chaotic, and r= rubble breccia.) Crackle breccias, indicative ofin situ brecciation, are of minor importance. Local crackle breccias pass upwards into rubble and chaotic breccias, which grade into soft- sediment deformation features. Boundaries between breccia types in this interval are mainly gradational.

213 Composition of breccia fragments, sediment matrix, and sediment-fîlled cracks

The breccias occur in lithologies ranging from, clayey siltstone to silty fine sandstone,

except for the breccia in Quaternary strata, which occurs in diamicton. In all cases the composition of the breccia clasts is the same as the host rock, indicating that the breccias are intraformational. The sediment matrix surrounding the breccia clasts is generally lighter in colour, but of the same lithological composition (Figure 9. Id). Sediment-filled cracks of mosaic breccias are abundant in the upper part of the Miocene section, commonly directly associated with crackle breccias. Thin section observations indicate that, locally, the sediment fill in cracks is graded, with a wall-parallel stratification of silt and clay (Figure 9.5).

Figure 9.5. Thin section photomicrograph of sediment-filled crack at 44.49 - 44.64 mbsf. The sediment-filled crack is ~I mm across. Note wall-parallel stratification and change in grain size of clastic fill along the length of the crack.

The clay particles have a common extinction angle under crossed polarizers, suggesting that clay particles are aligned parallel to the crack wall. Carbonate cement and microconcretions

2 1 4 are also present in intergranular spaces and cracks (see Baker & Fielding, 1998). In addition to the sediment-filled cracks two clastic dikes occur in the CRP-1 core: one at 133.52 to 133.77 m, and one at 139.05 to 139.31 m. The dikes are approximately 1.5 cm wide and have steep dips.

Association with other deformation features

Planar fractures that cross-cut the core margins are abundant in the uppermost Miocene strata where brecciation is extensive, but elsewhere in the core there is no systematic correlation between high fracture abundance and the presence of brecciated intervals (see Figure 2 in Wilson

& Paulsen, 1998). A conjugate set of listric microfauits with normal displacement occurs at -65-

70 mbsf and a reverse fault was observed at -109 mbsf. Below 55 mbsf, the breccias occur associated with soft-sediment folding (Figure 9.6) and chaotic sediment mixing. Soft-sediment deformation is absent in the upper 55 m of the core.

% I

Figure 9.6. Soft-sediment deformation and brecciation at -67 mbsf. Contorted lamination may represent soft-sediment folding, which is overprinted by brecciation (crackle breccia). The upper picture is a positive CoreScan image, the lower picture is a negative image of the same interval, where open cracks appear white, highlighting the breccia texture.

21 5 Breccia depth thick­ lithology dominant boundaries other defor­ sediment- associated interpretation n (mbsf) ness breccia mation filled with breccia (cm) textures cracks

1 40.35-41.79 144 diamicton chaotic/crackle grad./sharp diamicton subglacial shear? 2 44.10-50.23 635 fine sandstone crackle/chaotic sharp/grad. fractures abundant diamictite subglacial shear 3 55.95-58.45 250 coarse siltstone chaotic/rubble gradational soft sed. def. present interstratifred slope failure diamictite 4 67.60-70.90 330 silty sandstone rubble/chaotic sharp/grad. soft sed. def, diamictite subglacial shear? fractures microfauits Î 79.55-85.52 597 silty fine rubble/crackle sharp/grad. soft sed. def. abundant diamictite subglacial shear sandstone w/ microfabric 6 92.44-92.52 8 mudstone chaotic grad./sharp graded beds slope failure 7 110.38-110.46 8 clayey siltstone matrix-supported gradational soft sed. def. present laminated sediments slope failure microfauits 8 116.22-116.48 26 sandy siltstone chaotic grad/sharp soft sed. def. interstratifred slope failure diamictite t—'t o 9 118.19-119.30 111 silty V. fine rubble/ gradational soft sed. def. laminated slope failure 0\ sandstone matrrx-sup. sediments 10 147.32-147.47 15 fine sandstone matrix-sup. sharp bioturbation slope failure

Key; grad.=gradational; soft sed. def.= soft sediment deformation

Table 9.1. Characterization of breccias in CRP-1 9.2.4 BRECCIA SEQUENCE AND STRATIGRAPfflC BOUNDARIES

Downcore sequence of breccias

The breccias are confined to 10 intervals, which are described in Table 9.1. Brecciation is

most abundant between -40 and -85 mbsf, with only thin breccia intervals occurring below -85

m (Figure 9.7). The Quaternary section is poorly indurated and core recovery is low, so only one

brecciated interval, which occurs within a diamicton, could be described with confidence. Crackle and mosaic breccias dominate in the upper part of the Miocene section, above -85 m, and are less common below this depth, where rubble and matrix-supported breccias dominate.

Two -6 m thick breccia intervals occur in the upper part of the Miocene section (breccias

2 and 5). Breccias 2 (Figure 9.3) and 5 contain numerous sharp, horizontal, planar boundaries, and breccia types occur in the following two arrangements: 1) crackle breccias adjacent to rubble breccias with a sharp boundary in between the types; or, 2) crackle breccias grading upwards or downwards into chaotic breccias with large fractured clasts in a rubbly groundmass, which in turn pass into rabble breccias (Figures 9.2c and 9.3).

Breccia 1 is similar to breccia 2, but breccia 1 contains more discontinuous brecciation and is thinner. Clear planar horizons are not characteristic of breccias 3, 4, 6, 7, 8, 9 and 10, which have gradational boundaries and soft-sediment deformation between breccia types.

2 1 7 so —

no _

120 —

40 —

|^ = dinmicliic • = breccia interval : sandstone %. — normal fault : siltstone x = reverse fault = claysionc ex = soft sediment : muddy deformation packstone = sequence : ioncstonc boundaries

Figure 9.7. Downcore distribution of breccias and other deformation features in the CRP-1 core. Numbers denote breccia units discussed in the text and described in Table 9.1. Lithologie log and sequence boundaries after Cape Roberts Science Team (1998). Note: single breccia in Quaternary section; association of thickest breccia units (2, 5) with sequence boundaries overlain by diamictites; and the association of brecciated intervals with other deformation features from 55 mbsf to the base of the core.

218 Association of breccias with unconformities and facies

Some breccias have developed just below sequence boundaries (Figure 9.7), that have

been defined based on rapid shifts in facies and grain-size (Fielding et al., 1998). The sequence

boundaries at -43 m (the Quatemary-Miocene boundary), at -79 m, -92 m, and at -116 mbsf

overlie breccia 2, 5, 6 and 8, respectively. It is notable that the facies associations that form the

sequences in the lower part o f the core are more complete than in the upper 85 m. It is interpreted

that the advance stage of the glacial cycle is missing from the sequences between 40 and 85 mbsf

(Powell et al., 1998), which suggests that glacial erosion took place. A break in sedimentation and

erosion at the top of brecciated intervals may also be present in the lower part of the core. Shifts

in magnetic susceptibility at 92, 110 and 116 m coincide with the presence of breccias 6, 7 and 8

below these levels (Cape Roberts Science Team, 1998a).

Diamictites overlie sequence boundaries above breccias 1, 2, 3, 4, 5, and 8. Breccias 6, 7

and 9 occur above and below graded beds and stratified sediments.

9.2.5 BRECCIATION MECHANISMS

In situ brecciation

Crackle and mosaic breccias are comprised of a framework of angular and subangular

clasts that show little or no evidence for displacement, suggesting that they formed by in situ

fragmentation of the rock (Figure 9.1c). In the core, crackle and mosaic breccias occur in discrete

zones within intervals consisting of rubble textures and, in the lower part of the core, with soft- sediment deformation. At the top of breccia 5 (-79 mbsf), soft-sediment deformation structures point to chaotic mixing of sediment under high pore-fluid pressures. Sediment-filled cracks forming mosaic breccias occiur throughout breccias 2 and 5. In breccia 2, the orientation of the

219 clays and the wall-parallel stratification of the clastic infill of cracks suggest that they developed

due to forceful injection of sediment. Sediment can become overpressured when it is subjected to

a compressional stress regime. Fractures filled with sediment may have formed as tension veins

were injected by clastic material during dewatering of unconsolidated sediment. Conventional

triaxial compression tests in experiments on a variety of isotropic rocks have shown that when a

rock is loaded to failure extension fractures may develop in low strain situations (Hancock, 1985).

The orientation of the extension fractures will be parallel to the maximum stress direction. A

vertical load is consistent with the dominantly high-angle orientation of the sediment-filled veins

in breccias 2 and 5. Collectively, the jigsaw puzzle morphology of the breccia textures, the presence of clastic vein injections, and their direct association with chaotic sediment mixing, suggest that in situ brecciation occurred as a response to release of high fluid pressures induced by a vertical load.

Shear-induced brecciation

The overall decrease in grain-size within tabular zones of chaotic, rubble and matrix- supported breccias suggests grain-size reduction by shear-related cataclasis. The subhorizontal, tabular geometry of the breccia intervals indicates failure of the sediment by horizontal shear along discrete planar zones. In breccia 2, the stacking and alternation of breccia textures indicate partitioning of strain between horizontal layers (Figure 9.3). Two distinct shear-related deformation styles occur in breccias 2 and 4. Breccia 2 contains sharp boundaries between textures and abundant in situ brecciation (Figure 9.3). Breccia 4 contains mostly rubbly textures grading into soft-sediment deformation (Figure 9.4).

The textural characteristics, stacking of textures, tabular breccia geometries and planar boundaries, suggest a brittle shear model for the higher-level CRP-1 breccias. The presence of

2 2 0 soft-sediment deformation features below 55 mbsf suggests that the sediments were not lithified

during deformation. Since brittle shearing does not likely occur under high pore-water pressures

in unlithified sediments, it probably postdates the dewaterin^ of the sediment. Overall it appears

that subhorizontal shear, with localization of strain in discredie zones marked by matrix-supported

and rubble breccias, is in accord with the macroscopic textural evidence in breccia 2.

Present data do not allow us to constrain why soft-sediment deformation occurs in

association with brecciation below 55 mbsf, but is lacking aEDove. The different deformation

styles may be due to differences in applied stress or different mechanical properties of the

sediment, due to factors such as pore-water pressure, the degree of lithification, or the frozen vs^.

unfrozen state of the sediment (Brodzikowski & Van Loon, BL985). Studies of brittle shear failure

of indurated sandstones show that brecciation occurs during ih e initial stages of cataclasis

(Hancock, 1985). However brittle shear bands lacking ductile deformation features also develop

in shear zones under subfreezing conditions (Echelmeyer & W ang, 1987).

9.2.6 MODELS FOR BRECCIA FORMATION

Hypotheses for the origin of the breccias

Breccias are described from a variety of structural settings and depositional environments

(Laznicka, 1988). To determine the origin of the breccias in tihe CRP-1 core, we examined the composition of the fragments and host rock, the geometry of th e brecciated bodies and their boundaries, the presence of other deformation features and th*e association of breccias with stratigraphie boundaries. In the CRP-1 core, the breccias consist mainly of intraclasts (i.e. the breccias are intraformational) and they are commonly developed below sequence boundaries.

Most brecciated zones in the core appear to comprise (sub)hojrizontal tabular bodies. Clastic

2 2 1 injections and soft-sediment mixing indicate the presence of high water- pressure. Cleavage does

not occur in the core. The absence of cleavage and the lack of exotic clasts eliminates a tectonic

origin for the breccias. Cryostatic brecciation is associated with rapid changes in porosity and

permeability of sediments (Brodzikowski & Van Loon, 1985; Menzies, 1990), which is not

characteristic of breccias in the CRP-1 core. Further, the sediments in the core were deposited in a

marine environment, whereas cryostatic brecciation is mainly described from terrestrial settings

(Laznicka, 1988). Cataclasis by brittle shearing, inferred here from textural evidence, is a

mechanism which does not occur by freezing and thawing of sediment. The brecciated intervals

do display a variety of textures, brecciation is discontinuous and associated with clastic

injections, which is typical of syn-sedimentary breccias, such as those formed in subglacial

environments and mass-movement settings (Brodzikowski & Van Loon, 1985; Peryt &

Jasionowski, 1994). The position of brecciated intervals below sequence boundaries suggests that

the brecciation may be related to the development of these boundaries.

Two possible models are thus considered likely for the formation of the breccias in the

CRP-1 core: 1) subglacial deformation or 2) slope failure and resedimentation. Subglacial deformation is a plausible model for the origin of the breccias because deformation features similar to those described in subglacial shear zones occur in the CRP-1 core. Slope failure and resedimentation is also a plausible model because the Cape Roberts drill site is located on a steep slope of the Roberts Ridge. An important aspect in the characterization of the core is to identify ice-grounding events. For this purpose we attempt to distinguish between deformation structures produced in a subglacial environment and a mass-gravity flow setting. Therefore, we examined textural arrangements and boundaries of individual brecciated intervals in the core, in combination with other sedimentological information, such as lithology, fabric, and primary

2 2 2 sedimentary structures, to determine the applicability of these two models for the origin of the breccias.

Subglacial deformation or slope failure and resedimentation?

Alternating textures over short intervals, sharp boundaries between horizons, and horizontal brittle and ductile shearing are features both observed in mass-movement settings and in glaciotectonized sediments. However, unique sedimentary and structural features of the two environments help to constrain models for the origin of the breccias (Brodzikowski & Van Loon,

1985; Van der Meer, 1987; Dreimanis, 1993; Hart & Roberts, 1994; Peryt and Jasionowski, 1994;

Van der Wateren, 1995). Extension fractures filled with sediment are common in glaciotectonized sediments. Sheared sediment lenses are usually more sinuously curved in mass movement settings and more planar in subglacial environments. The association and interbedding of breccias and soft-sediment deformation structures with graded beds and sorted sediments with primary sedimentary structures is considered diagnostic for mass-gravity flow environments. If a basal till overlies deformed sediments, the deformation is likely caused by glacial action.

Breccias formed by subglacial deformation

Where diamictites overlie a brecciated zone (breccias 1, 2, 3, 5, 8), it is possible that the emplacement of the diamictite is related to the development of the breccia. Other evidence from the cored sequence, such as clast-fabric (Cape Roberts Science Team, 1998b) and microfabric studies (Van der Meer, 1998) on diamictites, point to glacial overriding of the site. The Ross Sea continental shelf is part of the glaciated Antarctic continental margin with a thick sediment cover.

Grounded ice at the location of the drillsite would possibly deform the underlying sediments.

Deformation features similar to those described from subglacial shear zones in both temperate

223 and frozen subglacial environments were encountered in the CRP-1 core, hence this seems a likely scenario for breccia formation.

Ice-sheets terminating on continental shelfs overlie sediment beds. The bed may be composed of a thin layer of basal diamict transported by the glacier, overlying sequences of strata deposited in either glacial or non-glacial environments. Studies in modem glacial environments have established the presence of deforming beds beneath many glaciers and ice-sheets (Boulton

& Hindmarsh, 1987, Murray, 1997, Alley et al., 1997). Studies of Pleistocene glaciations in

Europe have shown that ice acting on unconsolidated sediments can produce extensive deformation (Van der Meer, 1987, Van der Wateren, 1995, Boulton, 1996). Most deformation structures are the result of subglacial shear near the ice-margin, which occurs under both temperate and subfreezing conditions (Boulton, 1996; Echelmeyer & Wang, 1987). Besides breccias, soft-sediment folding, high-angle fractures and clastic injections are common in

Pleistocene subglacial shear zones (Dreimanis, 1993; Van der Wateren, 1995). Echelmeyer and

Wang (1987) identified brittle shear bands and slip surfaces within a frozen basal debris layer beneath a modem Chinese glacier.

The two thick brecciated intervals, breccia 2 and 5, below the Quatemary-Miocene boundary and below -79 mbsf, respectively, differ from other brecciated intervals by their thickness ( - 6 m), more in situ brecciation, and the abundance of sediment injections (Table 9.1).

Both breccias have many sharp, intemal boundaries, which may be slip planes. In breccia 5, a transition from crackle and mosaic breccias at the bottom to mbble and matrix-supported breccias at the top, suggest a change from more in situ brecciation at the bottom to horizontal shear- deformation at the top. A zone of soft-sediment mixing between the diamictite and the brecciated interval indicates high pore-water pressure in the sediment below the diamictite. The sediment injections point to dewatering of this sediment, which may have occurred as a response to a

2 2 4 significant vertical load, such as grounded ice. This interpretation is supported by independent

evidence for glacial overriding at -79 m (above breccia 5) where microstructures in thin sections

suggest that the diamictites at these levels have been tectonized (Van der Meer, 1998). The

position of breccia 2 below the Quatemary-Miocene boundary, a major unconformity that

possibly developed due to glacial erosion, further suggests that brecciation is the result of

subglacial deformation.

Considering the presence of high angle fractures, multiple slip planes, the thickness of the

brecciated intervals and the association with unconformities, subglacial shearing was the likely

cause of formation of breccia 2, (~ 43 m) and breccia 5 (- 79 m). These hypotheses should be

tested by more research on thin sections of rubble breccias, to determine whether these zones

show microscopic signs of shear deformation. Breccia 1 occurs within a clast-poor muddy

diamictite in the Quaternary section. The intemal arrangement and boundaries of breccia textures

within the brecciated interval suggests horizontal brittle shearing took place. However, additional

evidence for subglacial shearing is not available (Table 9.1). Breccia 4 occurs 3 m below a

diamictite that contains a distinct clast fabric and microscopic evidence of deformation (Van der

Meer, 1998), suggesting glacial overriding. The presence of soft-sediment folding in breccia 4

suggests horizontal shearing under high pore-water conditions, which is characteristic of both a

subglacial and gravity flow setting (Hart & Roberts, 1994). Thus, it is not clear whether the

deposition of the diamictite is related to the brecciation in this case.

Other diamictites in the lower part of the core (at -105, 123 and 134 mbsf) interpreted by

Van der Meer (1998) to have been deformed, are not directly associated with breccias, which

suggests that shear-induced brecciation took place under restricted conditions. It is possible that

lithology plays an important role in breccia formation, because 9 out of 10 breccias in the CRP-1 core developed in fine sandstones and siltstones. These lithologies are not abundant in the lower

225 part of the Miocene section, where thick breccias are lacking, but are abundant in the upper part of the Miocene section, where most breccias occur. The thickness of a deforming bed depends upon the lithology exposed beneath the ice-sheet (Boulton, 1996). Diamictites typically have low permeabilities because of their broad grain-size distribution. Sandstones and siltstones are more susceptible to brittle failure than diamictites, because the former are more permeable, resulting in lower pore-water pressures in the sediment. If pore-water pressures are high, the effective stress is reduced, and strain may be accounted for by ductile deformation of a relatively thin basal till.

Brittle failure of underlying sandstones or siltstones results in dewatering of the basal till and an increase of the depth of deformation. Microfabrics are most distinct in the diamictites at 123 and

134 mbsf, where breccias are absent, and less pronounced in the upper part of the core where breccias dominate. In the diamictites in the lower part of the core described by Van der Meer

(1998), localization of strain in a thin deforming bed may have occurred, because the underlying lithologies were impermeable enough to maintain high pore-water pressures and to prevent brittle failure of the sediment.

Slope failure, resedimentation

Slope failure occurs as a result of gravity pull on sediments. Sedimentation at a stable grounding line in a sub-Arctic environment tends to produce morainal banks with steep slopes on which mass movement is a common process (Powell & Molnia, 1989). Meltwater release and sediment deposition at the grounding line increase the gravitational instability of the proglacial environment. In the presence of high pore-water pressures, sediment deposition on a moderate slope can be sufficient to overcome the intemal friction of the sediment. Earthquakes also trigger slope failure and are likely to have occurred in the Miocene tectonic setting, associated with rifting in the Ross Sea and uplift of the Roberts Ridge. In situ brecciation and matrix-supported

226 breccias accompanied by soft-sediment deformation are characteristic of local slope failure in a

water-saturated environment (Brodzikowski & Van Loon, 1985; Peryt & Jasionowski, 1994).

The diamictite above breccia 8 contains thin beds of sandstone, and the lithology

overlying breccia 3 is interstratified diamictite, mudstone and sandstone, which suggests that

these diamictites were redeposited. Breccia 6, 7 and 9 in the lower part of the core are associated

with laminated sediments and graded beds (Table 9.1). Howe et al. (1998) argue that distal

gravity flows (turbidites) are present below 115 mbsf and that ice-proximal gravity flows are a

possible explanation for the laminated sediments deposited between 90 and 115 mbsf. This is

consistent with the nature of the breccias in this interval, which are thin chaotic and matrix-

supported breccias with irregular boundaries. Breccia 10 has irregular boundaries and is matrix-

supported, suggesting that it has also been redeposited.

9.2.7 SUMMARY AND CONCLUSIONS

Macroscopic observations of the breccias within the CRP-1 core suggest in situ fracturing, brittle shearing along discrete shear planes, and ductile deformation were associated with breccia formation in the CRP-1 strata. One breccia was formed during the Quaternary, one some time between the early Miocene and the Quaternary, and 8 breccias were formed in the early Miocene. The two types of settings in which brecciation is likely to have occurred are subglacial shearing and slope failure and resedimentation. This is similar to interpretations of breccias observed in other McMurdo Sound cores (Barrett & McKelvey, 1986; Hambrey et al.

1989a).

Breccias 2 and 5 were likely formed in a subglacial environment, based on multiple sediment-filled veins and brittle-shearing textures, the great thickness of the brecciated intervals

22 7 and their association with unconformities and deformed diamictites. Breccias 1 and 4 may also have formed by subglacial shearing, but the textures are not as pronoimced and the breccias do not directly relate to unconformities or deformed diamictites. Breccias 6, 7, 9 and 10 were likely formed in a mass-movement setting based on the characteristics of the textures (e.g. matrix- supported breccias) and their association with graded beds or stratified sediments. Breccias 3 and

8 are associated with interstratified diamictites, which suggests that they may have formed in an ice-proximal glacio-marine environment. Breccias below 55 m are accompanied by soft-sediment folding and mixing, which is absent above 55 mbsf. This change in deformation style may be related to the temperature and pore-water regime or to differences in the state of lithification of the strata during deformation.

228 9.3 SOFT-SEDIMENT DEFORMATION AND BRECCIATION IN CRP-2/2A

9.3.1 INTRODUCTION

The Cape Roberts Project aims at reconstructing the Cenozoic climatic and tectonic

history of the Transantarctic Mountains and the East Antarctic craton. The Cape Roberts CRP-2 and 2A holes are located ca. 15 km east of Cape Roberts (Figure 9.8). CRP-2 terminated at 57 meters below the sea floor (mbsf) and a new hole, CRP-2A, was started at the same site. Drilling took place in a water depth of 178 m with a core recovery of more than 91 % for CRP-2 and 95 % for CRP-2A. The composite record of CRP-2 and 2A (referred to as CRP-2/2A) comprises a lower Oligocene to Quaternary section down to 624 mbsf.

ANTARCTIC CE SHEET

Ross Ice Shelf

McMurdo Sound

75 S CIROS/y CRP ice-free areas drillsites

150'E

Figure 9.8. Map of the Ross Sea area, showing the location of the CRP drillsites, the CIROS-1 drillsite and the ice-ffee Transantarctic Moimtains.

229 The drillholes are located on the north-south trending Roberts Ridge, which forms the western margin of the Victoria Land Basin, a 20 km wide rift (Hamilton et al., 1998).

The Transantarctic Mountains form the rift flank, which is separated from the basin by a major fault at ca. 10 km west of the drillsites. Drilling takes place in a sedimentary wedge extending from the Transantarctic Mountains. The sedimentary wedge is now truncated to the north by the glacially eroded Mackay Sea Valley, reaching water depths of over 1000 m. Micro fossils suggest a predominantly shallow marine setting for the cored interval, although some core intervals are barren (Cape Roberts Science Team, 1999). Repetitive facies associations of diamictites, sandstones and siltstones, are indications of cyclical changes in glacial proximity and water depth. A 150 meter thick upper Oligocene to lower Miocene section bracketed by Ar/Ar dates on volcanic ash spans only ca. 300,000 yrs, suggesting that much of the record is missing at unconformities.

This paper deals with structures, which developed in the early burial history of the sediment.

Therefore, these structures can be related to the depositional and erosional processes, which were responsible for the accumulation and reworking of semi- or unconsolidated sediment. Sediment deformation features were recorded on the cut surface of the CRP-2/2A cores during normal logging procedures and on core-scan images after drilling was completed. Two whole-core thin sections from CRP-1 were also examined. CRP-1 penetrated an interval of similar age to the upper part of CRP-2/2A and was drilled ca. 1 km to the east of CRP-2/2A (Cape Roberts Science

Team, 1998). Coring induced deformation was recognized, which was partly a result of pre­ existing natural fractures in the sediment. Many deformation features occur in now consolidated core, including healed fractured sediment, here called breccia.

Sedimentation rate and glacial proximity are the primary controls on the stability of the depositional environment. Lithofacies and sequence stratigraphie interpretations suggest that the

230 grounding line of glaciers repeatedly approached and abandoned the location of the drillsite

(Cape Roberts Science Team, 1999). Since the reconstruction of Antarctic ice-volume is a main

objective of the project, macroscopic (this study) and microscopic analyzes (Van der Meer, in

press) of sediment deformation features are performed to identify subglacial overriding. Sheared

breccias were already linked to glacial overriding in the Miocene section of CRP-1 (Passchier et

al., 1998). Other deformation features encountered in the CRP-2/2A core, such as contorted

bedding with preserved lamination, clastic dikes and microfaults, also help to characterize ice-

distal depositional environments.

9.3.2 CHARACTERIZATION OF THE DEFORMATION

Both “ductile” and “brittle” deformation was recorded, with the words “ductile” and “brittle”

used in the following sense: “ductile” deformation as continuous flow due to grain-boundary

sliding and Mohr-Coulomb failure of unlithified overpressurized sediment; “brittle” deformation

as grain-boundary sliding of dewatered sediment showing dislocation of sedimentary fabric along discrete planes. Main features observed are: intraformational brecciation, clastic dikes, folding, contorted bedding, reverse and normal microfaulting, shear zones and slip planes. Detailed descriptions of some features can be found in the CRP-2/2A Initial Report (Cape Roberts Science

Team, 1999) and the most prominent features are listed in Table 9.2.

Brecciation in CRP-2/2A is similar to the patterns described in Passchier et al. (1998) from

CRP-1. The breccias consist of fractured sediment or intraformational sediment clasts in a matrix.

Five types of breccias occur in CRP-2/2A: 1) crackle breccias and 2) mosaic breccias (Figure

9.9), which are defined as in situ fractured rock; 3) rubble breccias, which consist of rounded

231 Figure 9.9. Breccias from CRP-2/2A: a) crackle breccia at ca. 56 mbsf; b) mosaic breccia at ca. 505 mbsf; and c) deformed breccia at ca. 45 mbsf. Note the thrusted fabric in (c).

232 Deformed Deformation type Facies association Interpretation interval(mbsf) 12-16 crackle breccia? bottom of diamictite 7 44.91-47.87 chaotic breccias, thrusts above diamictite proglacial ice-push 50-71 breccias, cataclastic shear zones below diamictite subglacial shear 90.67-92.86 soft-sediment deformation below diamictite subglacial shear? 93.85-98.64 crackle breccia between diamictites in situ fractiufng 99 folding within diamictite mass-gravity flow 131-132 crackle breccia below diamictite in situ fracturing 145-151 brecciated sandstone dikes top of clast-poor unit hydrofracturing 176.2-176.3 microfaults, soft-sediment folding within mudstone 7 183.35-185.95 folds, rotated clasts top of sandstone mass-gravity flow 245 vertical fractures below diamictite 7 297.4 clastic dyke through cement, patch above diamictite 7 297-306 folds (D), faults and breccia (sst) within diamictite collapse ? 306 intraclast, boudinage, clast rotation bottom of diamictite subglacial shear 307 fault with large displacement below diamictite 7 307-308 rubble and crackle breccia below diamictite subglacial shear 311-315 matrix-supported breccia within clast-poor unit collapse? 315-326 faults and clastic dikes interbedded lithologies 7 328 clastic dyke, coarse stratified fill below diamictite subglacial shear 329 clastic dyke? below diamictite subglacial shear 331.9 matrix-supported breccia within mudstone mass-gravity flow 332 clastic dyke, calcified and pyritized within mudstone 7 334 clastic dyke, faulted within mudstone 7 337 clastic dyke, folded and faulted within mudstone 7 362.78-363.08 shear zone, folding, faulting between diamictites subglacial shear 363.22-363.63 clastic wedge between diamictites subglacial shear 364 clastic dyke between diamictites subglacial shear 381-384 microfaults, soft sediment folding stratified diamictite mass-gravity flow 447 pyritized clastic dikes within mudstone tectonic? 468.7 fault cross-cuts calcite patch within mudstone tectonic? 501-503 contorted bedding, folding laminated sandstone slumping 504-506 local mosaic breccias interbedded lithologies gas-hydrate formation? 512-516 contorted bedding, folding graded beds slumping 520-523 clastic dikes, microfaulting sandy conglomerate debris flow 524-527 faults, pyritized clastic dikes below congomerate debris flow 525.8 mosaic breccia above graded beds gas-hydrate formation? 535-540 faulted clastic dikes below congomerate debris flow 543-552 faults, pyritized clastic dikes within mudstone tectonic? 577-579 contorted bedding, folding interbedded lithologies slumping 579.02-580.99 anastomosing networks interbedded lithologies slumping 581-585 contorted bedding, folding above conglomerate slumping 587 mosaic breccia above graded beds gas-hydrate formation? 594.97-595.03 mosaic breccia below clast gas-hydrate formation? 608.7 anastomosing dilated laminae below stratified sandst slumping 608-614 contorted bedding, folding within sandstone slumping 612.00-612.07 microfaulting cuts folding above conglomerate slumping 614.0-614.2 anastomosing dilated laminae above conglomerate slumping

Table 9.2. List of prominent deformation features and interpretation in CRP-2/2A.

233 pebbles; 4) chaotic breccias, with large fractured clasts within a matrix; and 5) matrix-supported

breccias.

Similar to CRP-1, breccia fabrics occur in stacked patterns of in situ brecciated intervals

grading into rubbly and chaotic zones. Some unconsolidated brecciated sandstones in the upper

part of the CRP-2/2A core (-45-48 mbsf) have a thrusted fabric (Figure 9.9c). An extremely thick

mudstone breccia is found between 311 and 315 mbsf in CRP-2/2A. The breccia has angular

clasts, which are tightly packed in a sandstone matrix. Similar lithologies as the mudstone clasts

are preserved beneath the breccia.

> ' 1cm

o f mud - h \ •'

Figure 9.10. Clastic dikes from CRP-2/2A: a) wedge-shaped dyke with irregular margins, filled with poorly-sorted sandstone at 363 mbsf; b) pyrite and carbonate cemented sandstone dyke with planar margins at 446 mbsf; c) brecciated medium sandstone dyke at 145 mbsf. The fabric of the brecciated mudstone (c) suggests injection occurred from below.

Three kinds of clastic dikes can be recognized in CRP-2/2A: 1) dikes with irregular boimdaries and coarse infill, with various orientations from near horizontal to vertical; 2) near­ vertical dikes with an apparent homogeneous infill and planar boundaries; and 3) homogenous clastic injections accompanied by brecciated host rock (Figure 9.10).

234 host sediment

(a)

silica& dyke zeolitey

h o s t downlapping laminae

I (b) 1 cm

Figure 9.11. Clastic dikes from CRP-1: a) thin section (left) and interpretation (right) of clastic dyke at 133 mbsf; and b) thin section (left) and interpretation (right) of clastic dyke at 133 mbsf. The dikes are filled with concentrically-zoned carbonate microconcretions of the types encountered in fractures in other parts of the core (Baker and Fielding, 1998). Microconcretions in sediments of the upper dyke are smaller than those in the lower dyke. The presence of a diamictite intraclast with small micro-concretions in the upper dyke (a) suggests that the pebble was derived from a lower level in the stratigraphie column. Downlapping laminae in the lower dyke suggest settling of sediment occurred as water pressures diminished. However, the downlapping laminae are truncated, suggesting that the dyke was reactivated after deposition of the laminae. Rapid fluctuations in water pressures in dyke systems are characteristic of subglacial and proglacial environments (Van der Meer et al., 1994, Von Brunn and Talbot, 1986).

235 Thin sections were available from two dikes ’with planar margins in the early Miocene section

of CRP-1 (Figure 9. II). Similar clastic dikes occmir in the Oligocene section of CRP-2/2A, but

these dikes are cemented by pyrite in the centre amd by calcite along the walls. A clastic dyke

crosscuts a calcite-cemented patch at 297.4 mbsf in CRP-2/2A.

Microfaults comprise both reverse and normal faults, normal faults being more abundant

(Figure 9.12). The microfaults occur both as single slip planes and as sets of parallel fault planes.

Some faults are bounded by folded strata, (e.g. -512 mbsf) or display drag along the fault plane.

Faults in the lower Oligocene part of the section are frequently mineralized (Figure 9.12c). A

fault crosscuts a calcite-cemented patch at 468.7 cnbsf in CRP-2/2A. Some normal faults were

injected by sediment in a later stage and are classified as clastic dikes.

Figure 9.12. Microfaults from CRP-2/2A: a) micro-block faulting in laminated sediments at 319 mbsf; b) reverse microfaulting in subglaclally deformed sediments at 363 mbsf; c) normal microfaulting with carbonate cemented fault plane: at 543 mbsf. Faults a) and c) are compatible with an extensional stress regime, however do not necessarily share the same origin. Fault b) may have formed in a compressional stress-regime associated with glacial overriding.

Shear zones and slip planes can be classified as 1) small-scale cataclastic shear zones

(Figure 9.13a); 2) soft-sediment folding progressing into bedding attenuation (Figure 9.13b); and

3) anastomosing, augen-like, dilated laminae (Figuzre 9.13c). The cataclastic shear zones occur in

2 3 6 association with breccias and consist of intraformational abraded pebbles and granules within

cemented fine sandstones. The second type of shear zone is sometimes associated with

boudinage, diamictite intraclasts, and rotated clasts with pressure shadows. The anastomosing,

augen-like dilated laminae form shear bands up to > 1 m thick.

Figure 9.13. Shear zones from CRP-2/2A: a) small-scale cataclastic shear zone, with an intraclast of a microlaminated facies and other intraformational pebbles and granules, in situ crackle breccia below the shear zone; b) sheared lower part of diamictite at 306 mbsf with boudinage and intraclast at the top; c) anastomosing dilated laminae at 614 mbsf.

Contorted bedding is identified in intervals of core with interstratified fine sandstone and siltstone (Figure 9.14). Axial planes of folds are oriented in all directions including horizontal.

The chaotic folding of sediment is associated with dipping strata (15-20 degrees) and shear zones of the anastomosing network type.

237 Figure 9.14. Contorted bedding from CRP-2/2A at a) 513 mbsf; and b) 612 mbsf.

9.3.3 STRATIGRAPfflC DISTRIBUTION OF SEDIMENT DEFORMATION FEATURES

Intense deformation is confined to three intervals in CRP-2/2A (Figure 9.15). The character

of the deformation changes downcore (Table 9.2). The uppermost deformed interval is

characterized by brecciation, small-scale cataclastic shear zones and minor soft-sediment deformation. Localized in situ brecciation first occurs in the CRP-2/2A core at 12 mbsf, in a

Quaternary diamicton. Brecciation intensifies at 38 mbsf and decreases below 151 mbsf. A

relatively undeformed interval is present between 151 and 297 mbsf in the upper Oligocene and

lowermost Miocene section of the core.

Deformation intensifies at ca. 297 mbsf. Faults, clastic dikes and shear zones dominate in the interval between 297 and 401 mbsf. At 306 mbsf the bottom of a diamict contains a diamict

238 intraclast, overlying boudinaged mudstone and breccia (Figure 9.13b). A thick matrix-supported

mudstone breccia occurs at 311-315 mbsf. A shear zone and a clastic wedge occur below a diamictite at 362-364 mbsf. Faulted dikes occur at 334, 337 and 535-540 mbsf.

The character of the deformation changes markedly below 485 mbsf. The sediments in this interval of core are now strongly lithified, but display deformation caused in an unlithified, perhaps sometimes water-saturated condition, illustrated by clastic dikes with irregular flow margins (e.g. 517.6 mbsf). Contorted bedding is recognized at 501-503, 512-516, 577-579, 581-

585 and 608-614 mbsf. Thin mosaic breccias also occur below ca. 500 mbsf (Figure 9.16b).

Anastomosing (augen-like) dilated shear zones are present at 580, 608 and 614 mbsf. Bedding is inclined > 15 degrees below 606 mbsf.

239 Deprh Af/Ar Micro- Ocstlc Infefpfeîoîion Sea level (m) ages Breccias fculfs dylces (Mo) 2_U ° 0 5 10 •100 100 200

so hiatus - 16

100- 2 1 .4 - 1 8 ‘22.6 ISO. -20 200. -22 2 5 0 .

- 2 4 3 0 0 . •25.0

- 2 6 3 5 0 .

- 2 8

430. - 30

5 0 0 32

55 0 gravrty “ 3 4 flews H c c e ro l. (1987) GOO eusroric se a level - 3 6

U~J Mudstone Mudsnne breccia Oewocked teprva

Figure 9.15. Downcore distribution of ductile and brittle sediment deformation features. Clastic dikes and microfaults are recorded as the number of features per 10 m intervals of core. In the interpretation column d, p, and g refer to the proximity of the grounding line with respect to the drillsite: d=distal, p=proximaI, g=grounded. Lithological column and chronology from Cape Roberts Science Team (1999). Ar/Ar ages are from McIntosh (in press). Eustatic curve from Haq et al. (1987). Mass-gravity flows in the early Oligocene correspond to sea level highstands of the Haq et al. (1987) eustatic curve. The early Oligocene lowstand corresponds to the initiation of abundant diamictite deposition culminating into grounding of ice on the drillsite. The upper Oligocene section does not show any evidence of grounded ice near the drillsite, but a sea level lowstand does occur. A highstand characterizes the early Miocene, but grounded ice occasionally reached the drillsite.

2 4 0 9.3.4 DISCUSSION

Origin of the deformation

On glaciated continental margins both glacial and marine processes affect deposition,

erosion, and deformation of sediment. At present the Cape Roberts drillsite is characterized as a

polar interglacial depositional environment. Other cores from McMurdo Sound and seismic

interpretations suggest that ice grounded on the shelf periodically since at least the Oligocene

(Bartek et al., 1996). However, whether a thick continental ice-sheet grounded on the shelf or local outlet glaciers is unknown.

Glaciers or ice-sheets terminating on continental margins overly unlithified sediment beds. Subglacial shear deformation occurs when subglacial drainage is poor, resulting in high pore-water pressures at the ice-bed interface. Subglacial shear deformation results in remoulding of sediments into what is called a deformation till or a deforming bed. Deforming beds have been found underneath many modem glaciers and ice-sheets (Boulton & Hindmarsh, 1987; Murray,

1997; Alley et al., 1997) and the characteristic features of subglacial shear deformation have been well-documented from outcrop studies of Pleistocene and Recent glacial deposits (Croot, 1988; van der Wateren, 1995). In a marine environment, the typical stratigraphie position for the presence of glaciotectonic features is at the bottom of, or below diamictites, which are deposited during the retreat stage in subglacial and proglacial environments near the grounding line.

Deformation found within the upper part of diamicts or in overlying sediments may originate from slumping (Dreimanis, 1993).

241 Shear zones

Shear zones at 306 and 363 mbsf occur at the bottom of diamictites and are likely caused by glacial overriding (Figure 9.13b). The features are similar to those described from shear zones below soft-bedded Pleistocene ice-sheets (Van der Wateren, 1995, p. 75).

Cataclastic shear zones, such as occur at ca. 67 mbsf in CRP-2/2A (Figure 9.13a), are not described from Pleistocene Northern Hemisphere glacial deposits, but were also encountered in

CRP-1 (Passchier et al., 1998). Deformation of basal sediment at sub-freezing temperatures occurs along well-defined shear planes, demonstrating more brittle failure characteristics of the basal layer (Echelmeyer and Wang, 1987). However, in CRP-1 the sheared breccias show evidence of fluidized sediment remobilization (Passchier et al., 1998), suggesting that basal conditions were, at least periodically, above the pressure melting point. The cataclastic shear zones are confined to the Miocene sections of the CRP-cores and may represent polythermal basal conditions, intermediate between the margins of Pleistocene temperate ice-sheets and sub­ freezing basal conditions. Lower subglacial pore-water pressures result in a more brittle style of deformation as represented by the sheared breccias (Brodzikowksi & Van Loon, 1985). The absence of great numbers of clastic dikes and microfaults in the lower Miocene section may also indicate more stable proglacial environments, suggesting lower meltwater production rates and reduced proglacial sedimentation rates.

Anastomosing network shear zones are confined to the lower part of the core (Table 9.2). No association with glacial facies exists. The great thickness of some of the shear zones suggests that displacement of large masses of sediment occurred. The augen-like shear zones may represent the basal slide planes of slumps, a rotational form of slope failure (Martinsen, 1994). This interpretation is supported by the presence of inclined beds in the lower part of the core.

242 Breccias

Cataclastic shear zones cut through in situ fractured sediment in CRP-2/2A (Figure 9.13a), suggesting that the fracturing preceded horizontal shearing. This in itself demonstrates that the fracturing of sediment is due to a geological process, since horizontal shearing after fracturing is not compatible with stresses induced by drilling. In addition, the cataclastic shear zones and parts of fractured core are recemented. The unconsolidated nature of the core material in the

Quaternary section, makes it difficult to describe the fabrics and to identify drilling induced deformation. Therefore, the Quaternary breccias will not be further discussed here. Three mechanisms act to produce the breccia fabrics: 1) in situ fracturing of the semi-consolidated sediment; 2) horizontal shearing of fractured sediment producing rubble and chaotic breccias by shear-related cataclasis; 3) abrasion and transport of fractured sediment producing matrix- supported breccias with rounded clasts. Microscopic thin section observations of brecciated sediments (Van der Meer, in press) are being performed to evaluate the mechanisms proposed here.

The presence of a sandstone dyke within a brecciated mudstone at 145-151 mbsf in CRP-

2/2A (Figure 9.10c) suggests a role of overpressurized, fluidized sediment in fracturing for some of the breccias. The mudstone was injected by sand from below, based on the breccia fabric and the presence of well-sorted unconsolidated sandstone below the brecciated mudstone.

Hydrofracturing is a process observed and predicted in ice-marginal areas, where aquicludes overly permeable sediments (Boulton & Caban, 1995, Rijsdijket al., 1999). Hydrofracturing may occur both subglacially and proglacially (Figure 9.16b) and in both marine and terrestrial environments (Rijsdijk et al., 1999; Boulton & Caban; 1995, Dionne & Shilts, 1974; Dreimanis &

Rappol, 1997; Larsen & Mangerud, 1992; Von Brunn & Talbot, 1986.) Due to the high pressures and discontinuous nature of the sediment bodies in ice-contact environments, deformation related

2 4 3 to hydrological processes may be quite extensive beneath soft-bedded ice-sheets (cf. Rijsdijk et

al., 1999).

Lower Miocene sheared breccias below diamictites in CRP-1 were also attributed to

grounded ice (Passchier et al., 1998). Carbonate cement with low 5*^0 values in open fractures of

brecciated rock in the Miocene section of CRP-1 (Baker & Fielding, 1998) may have precipitated

from meteoric water. Clastic veins within sheared breccias of CRP-1 (Passchier et al., 1998)

suggested hydrofracturing occurred in a subglacial environment. However, fluidized sediment

may also have been remobilized into existing tensile fractures, formed due to ice loading (Figure

9.16b).

Sheared breccias in the lower Miocene section of CRP-2/2A probably also formed subglacially (Figure 9.13a and 9.16c). Breccias at 45-48 mbsf in CRP-2/2A show an atypical thrusted fabric, which was not recognized in CRP-1. High angle thrusts are characteristic of proglacial ice-push environments (Van der Wateren, 1995, p. 110). The stratigraphie position of the thrusts is in accord with this interpretation (Table 9.2). The thrusted breccia immediately overlies a subglacially deformed interval and a diamictite unit.

Thin matrix-supported breccias in CRP-1 were regarded as redeposited (Passchier et al.,

1998). A similar origin could be proposed for the matrix-supported mudstone breccia at 311-315 mbsf in CRP-2/2A. However, the uniform composition and the presence of angular mudstone clasts, suggest only minor transport occurred. Perhaps the breccia represents a collapsed horizon, because all strata in the interval between 297 and 327 mbsf are deformed and normal and reverse faults are abundant. Both normal and reverse faulting in non-cohesive sands have been found in model studies of depleting reservoirs (Odonne et al., 1998). A significant unconformity must be present between 297 and 306 mbsf, based on sharp terminations in a number of datasets from the core, including sediment deformation (Cape Roberts Science Team, 1999). Evidence of glacial

2 4 4 sntial surface _ diam ict c i - l ■. ■ . 1 sandstone deformation Ëvvvl mudstone ÿÿÿd fractured sediment — meltwater flow

d

subgbcial ptoglaoâ

Figure 9.16. Hypotheses for the formation of meltwater-related glaciotectonic structures described from the Cape Roberts drillcores (partly after Boulton and Caban, 1995, Rijsdijk et al., 1999). (a) Free flow of meltwater, no deformation, (b) Hydrofracturing of subglacial and proglacial sand and silt. Subglacial sediment mobilisation may also occur into existing tensile fractures, formed due to ice loading. In the proglacial area low permeability of proglacial sediments or the presence of permafrost may impede free drainage when water depths are shallow, (c) Subglacial deformation of fractured sediments and cataclastic shearing, (d) Proglacial hydrofracturing and redeposition of mudstone breccia, (e) Subglacial and proglacial hydrofracturing in relation to diamictites. The high tensile strength of diamicites apparently leads to the formation of thicker clastic dikes rather than small hairline fractures as observed in sand and mud (b and d).

245 overriding is present at 306 mbsf, which means that a relation between brecciation and glacial

proximity is likely.

Intraformational breccias are also described from the Miocene sections of the CIROS-1

(Hambrey et al., 1989a) and MSSTS-1 (Barrett & McKelvey, 1986) cores in McMurdo Sound.

The coincidence of brecciated intervals in the upper Oligocene to lower Miocene section of CRP-

2/2A and CIROS-1 is striking (Figure 9.17). The Oligocene to lowermost Miocene breccias in

CRP-2/2A consist of a fractured mudstone injected by sand (Figure 9.10c), and the thick

mudstone breccia (311-315 mbsf) described above. The breccias in CIROS-1 mainly consist of

mudstone clasts in a sandy mud matrix. Both Hambrey et al. (1989a) and Fieldiag et al. (1997)

suggested that the breccias in CIROS-1 were redeposited.

14 -1 (Ma)

16 -

18 - CRP-2/2A 27 (mbsf) 20 -

139 CIROS-1 2 2 - 145 (mbsf)

24 _

311 129

26 -

28 -

Figure 9.17. Simplified stratigraphie distribution of breccias in the Oligocene-Miocene sections of McMurdo Sound cores. Note that near in situ breccias predominate in the late early Miocene, whereas mudstone breccias are more abundant in the earliest Miocene to Oligocene. CRP-1 data are from Passchier et al. (1998), CIROS-1 data are from Hambrey et al. (1989a) and Fielding et al. (1997).

246 The cause of the initial fracturing of the sediment may be hydrofracturing as illustrated by the brecciated dyke in CRP-2/2A. A lower Miocene mudstone breccia is present at a depth between ca. 48 and 64 mbsf below a diamictite in CIROS-1. A possible relation to a nearby grounding line caimot be ruled out, due to the presence of diamictites stratigraphically above the breccias (Figure

9.16d).

Some breccias cannot be related to ice-contact glacial environments, such as the mosaic breccias, which occur below ca. 500 mbsf (Figure 9.16b). The breccias occur in an interval of core with low marine productivity and slope instability inferred from sedimentary facies (Cape

Roberts Science Team, 1999). The presence of an oily overprint on the sediments below ca. 500 mbsf may suggest organic contents are increasing downhole. Bohrmann et al. (1998, p. 649) describe brecciation as a result of the growth of gas hydrates and subsequent fracturing from

Hydrate Ridge on the western North American continental margin. Gas-hydrates are solid compounds built out of water and gas (e.g. methane), which are stable under restricted temperature and pressure conditions (Kvenvolden, 1993). In polar regions gas-hydrates occur within shelves and permafrost areas at depths > 150 m. Dissociation of gas hydrates may cause catastrophic failure of thick sedimentary sequences (Pauli et al., 1996; Maslin et al., 1998). The mosaic breccias, indications of mass-gravity flow, and the increasing organic content of the lower part of CRP-2/2A may be connected with the formation and dissociation of gas hydrate.

Bohrmann et al. (1998) describe the formation of carbonate with a distinct chemical and isotopic composition in relation to hydrate development. Chemical analysis of carbonate cemented breccias in CRP-2/2A could reveal whether their formation is related to gas hydrates.

247 Clastic dikes

Clastic dikes are described from a range of environments including mass-flow (Shanmugam

et al., 1996), tectonic (e.g. Bergman, 1982, Winslow, 1983) and glaciotectonic settings (e.g.

Dionne & Shilts, 1974; Dreimanis & Rappol, 1997; Larsen & Mangerud, 1992; Von Brunn &

Talbot, 1986). A preliminary study on two clastic dikes in a lower Miocene diamictite from CRP-

1 show a pattern of fluid and sediment flow (Figure 9.11) consistent with formation in an ice-

marginal environment. In a glacial environment clastic dikes form perpendicular to the direction

of ice-flow (Figure 9.16e) and can be injected upward, downward or sideways (Van der Meer et

al., 1994; Dreimanis & Rappol, 1997). Wedge-shaped dikes with a diamict infill, such as at 363

mbsf in CRP-2/2A (Figure 9.10a), can be interpreted as injection of till from the base of a glacier

into unconsolidated sediment during glacial overriding (Dionne & Shilts, 1974). Clastic dikes

from the lower part of the core (below ca. 470 mbsf) are not related to diamictites. Some of these

dikes have irregular flow margins, suggesting injection occurred into fluidized sediments,

compatible with formation in mass-movement environments. The fabrics of most dikes with planar margins in the lower part of the core (Figure 9.10b) have been obliterated by post- depositional mineralisation. Some dikes represent injection of sediment into mineralized normal faults. A tectonic origin for these dikes cannot be mled out, but a glacial origin is unlikely considering the general geological context of the features.

Microfaults

The record of microfaults may include brittle features associated with the present tectonic regime of the basin (cf. Wilson & Paulsen, in press). Macroscopically, post-lithification micro- faulting cannot be distinguished from syn-sedimentary deformation. However, the uneven stratigraphie distribution of microfaults suggests that the present tectonic stress regime is not a

248 major cause of microfaulting throughout the section. Even microfaults and clastic dikes cross­

cutting calcite cement at 297.4 and 468.7 mbsf, may have been formed due to syn-sedimentary

deformation, since carbonate cementation may occur close to the sea floor. Other faults have been

mineralized (Figure 9.12c), suggesting development under higher temperature and pressure

conditions than exist near the surface. This is particularly true for microfaults below 444 mbsf.

However, most micro faults above 444 mbsf reflect a contemporaneous extensional stress regime,

which is compatible with subglacial overriding and slope failure (Martinsen, 1994; Van der

Wateren, 1995). Microfaulting of clastic dikes at 334, 337 and 535-540 mbsf suggests that more

than one episode of deformation affected the sediments. However, both in glacial and mass- movement environments overprinting of structures is observed when deformation continues

(Martinsen, 1994; Van der Wateren, 1995).

Contorted bedding

Contorted bedding is an indicator of low-strain deformation of water-saturated stratified sediments and is commonly associated with slumping in a marine environment (Shanmugam et al, 1995; Hampton et al, 1996). Inclined bedding (15-20 degrees) below 606 mbsf suggests a rotational form of slope failure, compatible with slumping. The facies association of interbedded sand- and mudstones with loading structures, and structureless sands, also suggests that slumping, debris flows and grain flows characterize the lower Oligocene section of CRP-2/2A and that sedimentation rates were high. Clastic dikes with irregular margins below 485 mbsf in the CRP-

2/2A core are associated with high pore-water pressures in the sediment as is common in rapidly deposited sequences. Similar to other nearby paleofjords, it is possible that the Mackay valley was a Qord during non-glacial times and that part of the record off Cape Roberts represents deposition of a fjord mouth delta. Rapid deposition on delta fronts at the mouth of Qords results in

2 4 9 accumulation of underconsolidated sediments with high pore-water volumes, and therefore low

shear strength (Syvitski et al., 1987). The seismic interpretation of the area also favours

deposition on a slope connected to the Transantarctic Mountains in pre-Quateraary time, before

carving of the Mackay sea valley (Hamilton et al., 1998).

Mass gravity flows are common in proglacial ice-proximal environments (Visser & Colliston,

1984, Powell & Molnia, 1989), but the deformed sediments in the lower part of the core are

probably not of this type. Some of the folded sediments below 485 mbsf in CRP-2/2A have

preserved lamination and bedforms, and are relatively rich in organic matter, suggesting that

redeposition of more ice-distal facies occurred. Modioloid bivalves recovered from sediments

below 441 mbsf (Taviani et al., in press) are not compatible with ice grounded at sea level near

the drill-site. Other paleoenvironmental data, including pollen (Askin & Raine, in press),

calcareous nannoplankton (Watkins & Villa, in press), and clay mineralogy data (Ehrmann, in

press) show that climate for the interval below 444 mbsf was probably milder than for the interval above this level.

Relatively undisturbed bioturbated sandstones with relatively high marine productivity are present at 444-485 mbsf and at 545-570 mbsf, suggesting that the deformation was episodic and consisted of multiple events separated by relatively stable periods. The dominance of M odiolus sp. in the macrofossil assemblages in the lower part of the core suggests that the paleoenvironment was deprived of oxygen and enriched in hydrogen-sulphide. A distinct marine diagenetic assemblage of gypsum, pyrite and carbonate nodules is present below 445 mbsf (Cape

Roberts Science Team, 1999), and only present in traces above this depth in the core, suggesting that early Oligocene mass-gravity flows took place in a deeper part of the shelf, below wave base.

2 5 0 9.3.5 COMPARISON TO OTHER CENOZOIC PALEOENVIRONMENTAL RECORDS

Extensive mass-movement in an ice-distal environment characterizes the lower Oligocene

section of CRP-2/2A. Gas hydrates may have formed periodically as suggested by the mosaic

breccias below ca. 500 mbsf. Dissociation of gas hydrates may occur as a response to rapid sea

level change and leads to sediment instability and slope failure (Pauli et al., 1996; Maslin et al.,

1998). Haq (1998) points out that ô^^C excursions at the Eocene-Oligocene boundary and in the mid-Oligocene could relate to gas hydrate response to sea level. Although the exact temporal relation between structures, 5'^C excursions and sea level remains ambiguous, the hypothesis of gas-hydrates as a cause of brecciation and slope failure in CRP-2/2A is testable if more chemical data become available.

A major unconformity in the CIROS-1 hole separates Eocene? to lower Oligocene strata from upper Oligocene to lower Miocene strata (Harwood et al., 1989; Wilson et al., 1998). In CRP-

2/2A, the exact position of the mid-Oligocene unconformity is uncertain, but the interval immediately below it is dominated by diamictites with a Transantarctic Mountains provenance

(Cape Roberts Science Team, 1999). Macroscopic evidence of grounded ice occurs down to 364 mbsf within this interval (Figure 9.15). Whether the grounded ice originated as alpine glaciers or an extensive ice-sheet cannot be established from the deformation record. However, the presence of an early Oligocene ice sheet in Antarctica is compatible with early Oligocene eustatic events

(Figure 9.15) and other stratigraphie records from the Antarctic continental margin. By early

Oligocene time the edge of the Lambert Glacier/Amery Ice Shelf complex in the Prydz Bay area was grounded 140 km beyond the edge of the present-day floating ice, suggesting that the East

Antarctic ice sheet was fully developed by that time (Hambrey et al., 1989b).

2 5 1 The upper Oligocene sediments in CRP-2/2A lack macroscopic evidence of grounded ice

(Figure 9.15). Some upper Oligocene diamictites in CIROS-1 were related to grounded ice,

although debate exists as to the position of the grounding events in the core (Hambrey et al.,

1989a, Fielding et al., 1997, Hiemstra, 1999). Upper Oligocene glacial sediments were also

encountered in DSDP site 28 in the Central Ross Sea (Hayes & Frakes, 1975). Perhaps late

Oligocene to earliest Miocene glaciation in the Ross Sea was characterized by small ice caps on

exposed structural highs, as suggested by De Santis et al. (1997), and the rising Transantarctic

Mountains may have prevented East-Antarctic ice from flowing onto the Ross Sea continental

margin (Barrett, 1999). The Ross Sea provenance of upper Oligocene strata in CRP-2/2A is also

in agreement with this interpretation.

Lower Miocene strata from CRP-1, CRP-2/2A and CIROS-1 are brecciated, which could

be related to glacial overriding. An oxygen-isotope record of the Oligocene-Miocene boundary

(Zachos et al., 1997) shows that a late Oligocene through early Miocene warming trend was

terminated by rapid cooling in the middle Miocene. However, the record also suggests that major changes in Antarctic ice-volume periodically interrupted the late Oligocene-early Miocene warming. In the Ross Sea, a middle Miocene shelf-wide unconformity marks the presence of grounded East- and West Antarctic ice on the shelf (Bartek et al., 1996). This unconformity is also present in the CRP-2/2A core: at 27 mbsf Pliocene-Quaternary strata directly overly lower

Miocene sediments (Cape Roberts Science Team, 1999). Lower Miocene seismic stratigraphie sequences, which contain glacial seismic facies extend over large areas of the Ross Sea continental shelf, suggesting that continental glaciers may also have been draining onto the Ross

Sea margin before the middle Miocene (Bartek et al., 1996). The sheared breccias in CRP-2/2A are in agreement with this interpretation.

252 9.3.6 CONCLUSIONS

The macroscopic and microscopic description of penecontemporaneous sediment deformation from CRP-2/2A contributes to the reconstruction of sedimentary environments in the

Late Cenozoic record of the Ross Sea area:

1) The lower Oligocene section of the core is characterized as fjord-mouth delta

sedimentation with periodic slumping, debris flows and grain-flows, based on the

presence of contorted lamination, shear zones, inclined bedding, clastic dikes and

microfaults. Gas-hydrates may have formed periodically, as is suggested by the presence

of thin mosaic breccias.

2) The mid-Oligocene record marks the presence of grounded ice near the drill-site.

Important sediment deformation features in this interval of core: clastic dikes, including a

till wedge, microfaults, shear zones and a mudstone breccia. The presence of a grounded

continental ice-sheet on the Ross Sea shelf is in agreement with other Antarctic

stratigraphie records and a eustatic lowstand.

3) The upper Oligocene to lowest Miocene succession was deposited in a stable depositional

environment in the absence of grounded ice near the drillsite. Deformation is limited to a

few clastic dikes and micro faults. The absence of grounded ice during most of the late

Oligocene and earliest Miocene is in agreement with a general warming trend in the late

Oligocene and early Miocene, as suggested by low-latitude proxy records.

4) The early Miocene section shows evidence of hydrofracturing of sediment and shearing

related to grounded ice on the drillsite. The deformation may be related to short-lived

episodes of glacial expansion during this generally warm climate interval with a

relatively high sea level.

253 9.4 CONCLUDING REMARKS

Factors controlling the nature of the sedimentary record on tectonically active glaciated

continental margins are: climate, sea level and tectonic subsidence.

Milankovitch scale glacial — interglacial cycles and associated sea level events appear to produce

vertical facies associations, with systematic changes in lithology (Powell et al., 1998; Fielding et

al., 1998). However, basin development, margin uplift and tectonic subsidence lead to changes in the sediment supply and water depth during deposition and influence the characteristics of the vertical facies associations. The stratigraphie distribution of the sediment deformation record in

CRP-1 and CRP-2/2A suggests that there is major difference in the abundance of microfaults and clastic dikes between the Oligocene and Miocene sections. A possible explanation may be that

Oligocene sediment accumulation rates were much higher so that Oligocene strata were more prone to sediment failure. Alternatively tectonic subsidence was more active in the Oligocene as compared to the Neogene creating a depositional system with steeper submarine slopes. Further studies, integrating tectonic and sedimentological results on the GRP cores are expected to address these issues.

Initially the breccias in the GRP cores appeared an enigmatic feature uncharacteristic of sedimentary sequences on other glaciated margins. However, it appears that brecciated sediments within or underlying subglacial deposits have been described and studied before. Ehlers &

Stephan (1983) describe a late Quaternary brecciated diamicton , which they interpret as the result of actively moving or melting ice. The fabric of the joint patterns is strongly developed parallel to the regional direction of ice-flow. Feeser (1988) describes the mechanism of joint development in unconsolidated Lauenburg lake clays underlying a surface, which based on other evidence, was overridden by the Scandinavian Pleistocene ice-sheet. The joint patterns in the

254 Lauenburg Clay are very similar to those from the brecciated mudstones and fine sandstones in

CRP-1. Macro- and microfabric studies of the Lauenburg Clays show that the cause of the jointing is extension within a compressive stress field, which is the same interpretation as we arrived at for some of the crackle breccias in the lower Miocene sediments of CRP-1. The results of the studies on the Lauenburg Clay suggest frozen ground at the time of glaciotectonic deformation to be able to explain the confining pressures needed for brittle failure of the sediments. This is also one of the options proposed for the development of crackle breccias in the

Miocene interval of the CRP cores.

CO 0 2 c 0 tc • 0 : > > • <

cc CO 2 o5 0 : s .

Deformation till— Glacitectonite

Figure 9.18. Cartoon explaining the development of unconformities by subglacial erosion and deformation. Left: complete sequence with diamicts at the bottom, which are either directly deposited by ice or by icebergs, fining upward into sandstones and mudstones deposited in progressively more distal depositional environments. Right: characteristics of a sequence truncated by glacial erosion. Note that the diamict overlying the unconformity may have been deposited upon glacial retreat, so that fabric studies on diamicts fail to identify ice-grounding episodes.

255 Crackle breccias predominantly occur in the Neogene intervals of the CRP cores,

whereas evidence of only ductile glaciotectonic deformation is found in Oligocene strata.

Therefore, the brecciated sediments may have a climatic or paleoenvironmental significance,

indicating a change in sea floor or subglacial temperature conditions between the late Oligocene

and the early Miocene.

The sequence stratigraphy of the CRP-1 and CRP-2/2A holes has been established mainly

using changes in grain-size, bedding characteristics, contacts and diagnostics sedimentary

structures. Vertical facies associations are useful in identifying the proximity of the glacier margin and waterdepth (Powell et al., 1998; Fielding et al., 1998). Glacial overriding of the drill site was established using fabric data of diamicts, which formed the lowermost lithological units of each sequence, and the presence of glaciotectonic features discussed in this chapter. Clast fabric data from CRP-2/2A (Cape Roberts Science Team, 1999) do not show any evidence of subglacial deposition, including intervals where grounded ice has been inferred from other methods (Van der Meer, in press), including the glaciotectonic features (this chapter). The data sets are in apparent disagreement, but it has to be kept in mind that fabric studies of diamicts are providing evidence for restricted sections of core in clast-rich facies, so that the results are inconclusive when no grounded ice is indicated (Figure 9.18). It is therefore recommended that soft-sediment deformation features form an integrated part of sequence stratigraphie interpretations, since it provides a continuous record of syn- and post-depositional processes.

2 5 6 CHAPTER 10

SYNTHESIS AND CONCLUDING REMARKS

10.1 Introduction

In recent years Antarctic Cenozoic studies have focused on two major topics:

• the initiation of Antarctic glaciation, and in the Ross Sea sector, its relation to the tectonic

history of the Transantarctic Mountains;

• Antarctic climate and ice-sheet stability or instability after the middle Miocene.

In this research these issues are addressed with sedimentological studies of the ?Neogene

Sirius Group in the Transantarctic Mountains, the Pliocene Bardin Bluffs Formation in the Prince

Charles Mountains of East Antarctica, Eocene-Quaternary sediment cores, and Eocene-Pliocene erratics from McMurdo Sound.

10.2 Initiation of glaciation in the Ross Sea sector

The paleoenvironmental information extracted from erratics from the Victoria Land coast suggests that conditions in the middle-late Eocene were cool-temperate, perhaps with glaciers in the higher elevated regions of the Transantarctic Mountains (Harwood & Levy, 2000). The chemical data indicating moderate chemical weathering, which are presented in Chapter 8, are in

257 B 0 CIROS-1, McMurdo Sound CIROS-1, McMurdo Sound

smectite, integral breadth CIA (crystallinlty) 30 40 50 60 70 1.5 2 2.5 ■23 Ma

too

I5D

200

250

300

350 U|lo 00 ■29 Ma 400 •33 Ma Ui 450 ■34 Ma

5 5 0

600

650

700

Figure 10.1. The Eocene/Oligocene transition in (A) the clay mineralogy and (B) the bulk chemistry record of the CIROS-1 drillcore, Victoria Land basin, McMurdo Sound. (A) from Ehrmann, 1997; (B) data from Krissek & Kyle, 1998. agreement with this interpretation.

The high smectite content, smectite crystallinity and mineralogy of sediments in the

CIROS-1 drillcore in McMurdo Sound suggest that chemical weathering prevailed on the nearby

Antarctic continent in the Eocene (Ehrmann, 1997). A shift to low smectite concentrations, lower

Chemical Index of Alteration (CIA) values, and low rock magnetism at the Eocene/Oligocene

boundary is interpreted as a transition to a decreased supply of chemically weathered components

from mafic igneous rocks of the Ferrar Group (Ehrmann, 1997; Krissek & Kyle, 1998; Sagnotti et

al., 1998). The concentrations of smectite and magnetic minerals, as well as the CIAs are mainly

controlled by source rock lithology and climate-dependent weathering conditions. A higher

contribution from basaltic source rocks would raise the concentrations of smectite and magnetic

minerals under any climate conditions. However, if chemical weathering was insignificant and

the CIA would be entirely controlled by source rock composition, an increased contribution of

mafic igneous rocks would lower the CIAs (Krissek & Kyle, 1998); this hyposthesis is consistent

with high values observed in the Eocene section of CIROS-1 (Figure 10.1). Therefore the shift in

the concentrations of smectite, magnetic minerals, and CIAs across the Eocene/Oligocene

boundary can best be explained by a decrease in chemical weathering of the source rocks.

The smectite concentrations (Ehrmann, 1997) and CIAs (Krissek & Kyle, 1998) in

CIROS-1 show a steady decrease in chemical weathering products derived from the continent

from the late Eocene to the late-Oligocene, when glacial facies began to dominate the

sedimentary successions of the Victoria Land basin (Figure 10.1). The glaciotectonic features in the mid-Oligocene section of CRP-2/2A, in the central Victoria Land basin (Chapter 9) suggest that the shift in sediment composition is followed by the first advance of grounded ice onto the continental shelf at Cape Roberts. The coincidence of major glacial expansion, with a shift to erosion of unweathered bedrock suggests that the Eocene/Oligocene transition in the Victoria

259 Land basin was a time of significant change in the glaciation style and climate of the nearby

Antarctic continent. In the Ross Sea sector, climatic cooling and ice-sheet growth accompanied

by a decrease in weathering intensities probably began in the late Eocene (34-36 Ma), and

ultimately resulted in major glacial expansion onto the continental shelf in the mid-Oligocene

(28-30 Ma).

10.3 Relation of the Sirius Group to glacial and tectonic history

10.3.1 Integration of Sirius Group compositional data

Sedimentological (Chapters 2, 3 and 4), mineralogical (Chapter 5) and chemical data from

the Sirius Group (Chapters 4 and 6) and related Upper Cenozoic deposits firom the Transantarctic

Mountains are integrated here in order to evaluate the relation between provenance and tectonic

history and the paleo-environmental conditions during deposition of the Sirius Group. The

compositional data combined with paleo-ice flow data, suggest that the Sirius Group represents

multiple depositional events at different stages in the uplift of the Transantarctic Mountains, and

that Sirius Group deposits cannot be correlated to only one glacial phase. Although the Oliver

Bluffs succession is an important Pliocene terrestrial archive, it should not be regarded as

representative of the entire Sirius Group.

The results of the compositional studies of the Sirius Group discussed in Chapters 5 and 6 are summarized in Table 10.1. The Sirius Group deposits show a range of compositions, with one end member representing much recycling of weathered sedimentary rocks, and the other end member representing active erosion of unweathered igneous and metamorphic basement rocks.

260 Locality Geoniorph. WoHM % Cpx % G t+ CIA Zr/Y Source rock types Surface stab. ratio

Mt. Fleming flat <1 1 36 63 9 sed. rocks + acidic ign. Mt. Feather flat <1 <1 58 66 11 sedimentary rocks Mt. Sirius flat 2 24 24 61 10 sed. rocks + basic ign. Tillite Spur flat 6 0 0 68 9 sed. rocks + metam.

Meyer Desert vertical 8 41 2 57 9 sed. rocks + basic ign. Roberts Mass. mixed 11 49 <1 59 8 basic ign. + sed. rocks

Oliver Bluffs vertical 15 38 2 50 8 basic igneous rocks Bennett Platf. vertical 18 46 <1 50 8 basic igneous rocks Quartz Hills vertical 16 0 0 55 6 acidic ign. and metam.

HM=Heavy minerals; Cpx=cIinopyroxene; Gt=gamet; stab= stable minerals; CIA= Chemical Index of Alteration (Nesbitt & Young, 1982); sed.=sedimentary; ign.=igneous; metam.= metamorphic.

Table 10.1. Dominant source rock types in the Sirius Group and the relation to géomorphologie setting.

The first end member is characterized by low abundances of heavy minerals, dominance of garnet and stable minerals in the heavy mineral fraction, relatively high CIAs, and high Zr/Y ratios. The second end member has high abundances of heavy minerals, dominance of clinopyroxene in the heavy mineral fraction, low CIAs, and low Zr/Y ratios.

A comparison of the interpreted source rocks with the microfossil content of the Sirius

Group samples confirms the differences between the two end members. Indurated sediment clasts, and in many cases Paleogene marine microfossils (Webb et al., 1984; Harwood, 1986) and

Permo-Triassic spores (Wilson et al., 1998a; Askin, 1998; Askin, pers comm., 1999) were found in Sirius Group sediments with a major contribution from sedimentary source rocks (Table 10.2).

261 Locality Source rock type Sed. Marine microfossils^ Pollen/ clasts^ Forants Diatoms spores^

Mt. Fleming sed. rocks + acid. ign. X U- MioTL. Plio. Mt. Feather sedimentary rocks X Eoc. U. Pliocene Triassic Mt. Sirius sed. rocks + basic ign. X U. OligTL. Mio. Triassic Tillite Spur sed. rocks + metam. X Hoc. Pliocene Permian/?Cenoz.

Meyer Desert sed. rocks + basic ign. Roberts Mass. basic ign. + sed. rocks

Oliver Bluffs basic igneous rocks Mio./Pliocene Cenozoic Bennett Platf. basic igneous rocks fragments Quartz Hills acid. ign. + metamorph. L. Pliocene ?Cenozoic

'Data from Webb et al. (1984), Harwood (1986), Askin & Markgraf (1986), Webb & Harwood (1991), Wilson et al. (1998a), Askin (1998), and Askin, pers. comm. ( 1 9 9 9 )______.

Table 10.2. Relation between dominant source rock types, presence of indurated sediment clasts, and microfossil data (youngest age).

The paleontological and compositional data combined clearly demonstrate recycling of

Permo-Triassic and Lower Cenozoic sedimentary rocks for one end member of the Sirius Group,

whereas the contribution from sedimentary sources is significantly lower for the other end

member. If the diatoms represent glacially transported marine floras, recycling is restricted to

Upper Cenozoic marine sediments for the second end member, and the high percentages of heavy

minerals suggest active erosion of crystalline basement rocks. The Sirius Group deposits dominated by recycled sediments (Mount Feather, Mount Fleming, Mount Sirius, Tillite Spur) are

all from flat mountain summits or paleovalleys cut into plateaus at high elevation, whereas the deposits supplied by active erosion of crystalline basement (Oliver Bluffs, Bennett Platform,

Quartz Hills) occur in major glacial troughs now draining the East Antarctic Ice Sheet

(Beardmore, Shackleton and Reedy Glaciers).

262 10.3.2 Deposition of the Sirius Group: timing, paleotopography and paleodrainage

Two different scenarios have been proposed for the deposition of the Sirius

Group, and will be discussed here in light of the new data presented in this dissertation: the

Webb-Harwood or dynamic ice-sheet hypothesis and the stable ice-sheet hypothesis. For further

reading on the development of the Sirius Group age debate and the dynamic versus stable ice-

sheet controversy, one is referred to Denton et al. (1984), Webb et ai. (1984), Harwood (1986),

Clapperton & Sugden (1990), Webb & Harwood (1991), Wilson (1995), Stroeven et al. (1996;

1998), Miller & Mabin (1998), and Harwood &. Webb (1998). The Webb-Harwood hypothesis

implies a dynamic East Antarctic Ice Sheet with several periods throughout the Cenozoic when

the East Antarctic Ice Sheet was greatly reduced in size (Harwood, 1986, p.69). Although Webb

& Harwood (1991) state that not aU Sirius Group deposits are Pliocene in age, they believe that

Sirius Group deposits that contain Pliocene diatoms were deposited after partial deglaciation of

East Antarctica in the early Pliocene and by re-expansion of the East Antarctic Ice Sheet. In

contrast, the stable ice-sheet hypothesis favors a stable East Antarctic Ice Sheet since the middle

Miocene, mid-Miocene overriding of the Transantarctic Mountains, and a mid-Miocene or older

age for the Sirius Group (e.g. Denton et al., 1993; Sugden et al., 1999). Supporters of the stable

ice-sheet hypothesis insist that diatoms in the Sirius Group deposits are not valid indicators of

depositional age, because of uncertainties about their biostratigraphic ranges, sources and

transport paths (Clapperton & Sugden, 1991; Stroeven et al., 1998).

If the Sirius Group was deposited during one phase of continental glaciation as proposed by both Webb et al. (1984) and Denton et al. (1984), then the compositional differences (Tables

10.1 and 10.2) between the two Sirius Group end members must result from differences in the geographic distribution of source rocks exposed to glacial erosion upstream from the two types of

263 deposits. According to Webb et al. (1984), Harwood (1986), and Webb & Harwood (1991), the

different biostratigraphic ranges of marine microfossils in the Sirius Group indicate recycling of

Cenozoic sedimentary rocks of varying ages. The erosion of the sedimentary rocks occurred during the mid-late Pliocene, based on the youngest diatoms present. In this scenario, chemically weathered Cenozoic and Permo-Triassic source rocks should have occurred predominantly upstream of Mount Feather, Mount Fleming, Mount Sirius and Tillite Spur, whereas the same source rocks had only a limited exposure upstream of the other Sirius Group localities, such as

Oliver Bluffs.

Pensacola Basin / -9 0 W Aurora 90" E - Basin

0 m 1500 m

180'

Figure 10.2. Subglacial topography of Antarctica. Simplified after BEDMAP (British Antarctic Survey, 2000).

However, the problem with this scenario is that it is not very likely that chemically weathered Cenozoic sediments could survive long-term glaciation of a cratonic region. The East

26 4 Antarctic shield has characteristics similar to other areas formerly covered by continental ice-

sheets. Both East Antarctica and cratonic regions in Europe and North-America experienced

multiple glaciations by continental ice-sheets. All three cratonic regions have central basins (cf.

White, 1972): the Gulf of Botnia for the Scandinavian shield, Hudson Bay for the Canadian

shield and the Wilkes and Pensicola basins for the (Figure 10.2). The

sediment covers that were present on both the Canadian Shield and the Scandinavian shield were

stripped by continental glaciation in the Pleistocene (White, 1972). Therefore, assuming that

extensive glaciation of the Antarctic continent began during the late Eocene/early Oligocene, it

would be very unlikely that extensive exposures of Cretaceous through Eocene sediments in East

Antarctic intra-cratonic basins were preserved into the late Neogene. The clay and chemistry

records of the Victoria Land basin also indicate a decrease in the supply of chemically weathered sediments after the early Oligocene (Figure 10.1). Therefore, considering only the geochemical and mineralogical data, it is more likely that the Cretaceous through Eocene pelagic marine sediments containing planktonic foraminifera and considerable amounts of chemical weathering products were incorporated into the Sirius Group of Mount Feather, Mount Fleming and Tillite

Spur (Table 10.2) upon initiation of glaciation of East Antarctica.

One problem with assigning an older age to the Sirius Group on Mount Feather, Mount

Fleming and Tillite Spur based on compositional data is that these deposits contain Pliocene marine diatoms; according to Webb & Harwood (1991) these diatoms originated from Pliocene marine basins in East Antarctica. However, the youngest diatom ages of the deposits with the highest CIAs (Table 10.1), those on high flat summit plateaus such as Mount Feather (Barrett et al., 1997) and Mount Fleming (Stroeven et al., 1996), are disputed and the diatoms have been regarded as eolian in origin. Harwood & Rose (1998) were unable to reproduce the diatom assemblages found by Harwood (1986) in surficial sediments at Mount Feather, when examining

265 a core drilled into the deposit. Considering these uncertainties, a much older age for the Sirius

Group on flat mountain summits and in paleovalleys, deposited during glacial overriding of a

much lower mountain range, becomes a more plausible scenario. The topography of the

Transantarctic Mountains and the East Antarctic basins during the Paleogene was probably quite

different from the present-day situation. Therefore, the location of the marine basins that

contributed the marine pelagic sediments, and the question of whether deposition occurred by a

local or a continental ice sheet, remain speculative. Although an East Antarctic origin is

possible for some of the microfossil assemblages, deposition by a local ice cap is favored by

feldspar provenance studies of Tillite Spur sediments (Faure et al., 1983), and fabric studies of the

Upper Fleming Till (Stroeven & Prentice, 1997).

The old age suggested here for some Sirius Group deposits would seem to support the stable ice-sheet hypothesis. However, “stabilists” claim that diatom ages of Sirius Group deposits in other parts of the Transantarctic Mountains are also invalid (e.g. Stroeven et al., 1998).

Although, eolian transport of diatoms is accepted by both “stabilists” and “dynamicists”,

Harwood &. Webb (1998) argue, based on the presence of diatomaceous clasts ("ooze") and other criteria, that the Pliocene diatoms in the Oliver Bluffs deposit must have been transported and deposited subglacially and therefore can be used to provide age control.

A young age for the Oliver Bluffs deposits is supported by the differences in composition between this locality and the Mount Feather and Mount Fleming Sirius Group, which are claimed to be middle Miocene or older (e.g. Stroeven, 1997; Stroeven & Kleman, 1999). The young end members of the Sirius Group, including deposits at Oliver Bluffs, Bennett Platform and Quartz

Hills, have an igneous and metamorphic basement provenance and low CIAs (Table 10.1 and

10.2), which require a lower intensity of chemical weathering, and a different ice-sheet drainage and landscape topography prior to deposition.

266 The compositional differences of Sirius Group sediments cannot be explained by either

the Webb-Harwood or the stable ice-sheet hypothesis. Therefore a new scenario is proposed here,

which allows for a time difference between the deposition of the Sirius Group at high plateaus

and paleovalleys (e.g. Mount Feather, Mount Fleming, Tillite Spur, Mount Sirius) and deposits in

the major glacial troughs of the Transantarctic Mountains (e.g., Oliver Bluffs, Bennett Platform,

Quartz Hills). Sirius Group deposits from several glaciations may have been preserved at

different levels in the erosional landscape due to uplift of the Transantarctic Mountains, and a

relative lowering of the ice surface due to glacial incision. In this scenario, Paleogene sediments

with marine micro fossils were incorporated into the Sirius Group upon the initiation of glaciation

in the Ross Sea sector. The deposits in the glacial troughs post-date the stripping of Paleogene

sediments from East Antarctica by continental ice-sheets, and significant uplift and glacial

denudation of the Transantarctic Mountains (Figure 10.3).

In the southern and central Transantarctic Mountains, the exact timing of the overriding

episode(s) and deposition of the oldest end member of the Sirius Group is unknown, but

paleontological constraints suggest that it started during or after the Eocene (Table 10.2). The

paleo-drainage pattern associated with the Mount Sirius succession is different than that of today,

now that the principal drainage is concentrated in the Beardmore glacial trough. If diatoms in the

Sirius Group on Mount Sirius are glacially transported (Harwood, 1986), they require that one

overriding episode of the central Transantarctic Mountains occurred during or after the late

Oligocene to early Miocene. Isotopic K/Ar dates of subglacially erupted basalts in the southern

Transantarctic Mountains also suggest that ice overtopped a topography of 900 m in the early

Miocene (Stump et al., 1980). However, it is possible that there was more than one phase of glacial expansion and overriding.

2 6 7 EAST ANTARCTICA TAM ROSS SEA EAST ANTARCTICA TAM ROSS SEA 9 physical chemical manne weathering manne sediments \CRP sediments

Sirius Sirius Group Group recycling recycling

Figure 10.3. (A) First phase of ice-sheet expansion and Sirius Group deposition incorporating open marine micro faunas and floras and chemically weathered sedimentary rocks. (B) represents a later phase of Sirius Group deposition in the Neogene with a higher mountain range, lower contemporary chemical weathering rates, and deposition in glacial troughs. TAM= Transantarctic Mountains.

The Sirius Group successions at Meyer Desert and Oliver Bluffs indicate deposition on different base levels of areal scouring within the Beardmore Valley. The composition of the

Oliver Bluffs succession indicates active erosion of Ferrar Group and basement rocks within the glacial trough after the early Pliocene (Table 10.1 and 10.2).

In the Dry Valleys, paleo-ice flow data again suggest deposition of the Sirius Group on high summit plateaus during an older episode of overriding of the mountains at an angle to the present orientation of the glacial troughs. In contrast. Upper Miocene sections at Prospect Mesa and DVDP-11 show evidence of ice-flow through the glacial troughs from the west, and increasing influence of the Ross Ice Shelf from the east since the early Pliocene. To Denton et al.

(1993), the maximum ages of ca. 9 Ma of valley deposits in Wright Valley (Prentice et al., 1993) and ca. 6 Ma in Taylor Valley (Ishman & Rieck, 1992) indicated that ice overrode both the

268 already existing valleys and the higher ranges of south Victoria Land in the mid-Miocene.

However, the data also allow for a period of overdeepening of the major glacial troughs in the middle Miocene (prior to 6-9 Ma) and exposure of the higher ranges of South Victoria Land, with overriding of the higher ranges and much shallower ancestral Dry Valleys by ice before the middle Miocene. This hypothesis will be tested in the near future with data from the Cape

Roberts Project, which drilled an ?Eocene through Quaternary succession in the Victoria Land basin (Cape Roberts Science Team, 1998, 1999, 2000).

10.3.3 Landscape evolution of the Transantarctic Mountains

Three different views exist considering the emplacement of the Sirius Group relative to the denudation history of the Transantarctic Mountains and the development of the East Antarctic

Ice Sheet:

1) The Sirius Group was deposited on a pre-existing high-relief glacial topography

(Mayewski, 1975; Mayewski & Goldthwait, 1985). The same concept was envisaged by Denton et al. (1984) when they proposed mid-Miocene overriding of the higher ranges in Victoria Land, and by Webb et al. (1984) when they linked the Sirius Formation to glaciation in the mid-late

Pliocene, based on the presence of early Pliocene marine diatoms in most Sirius Group deposits.

2) Some or all of the Sirius Group deposits represent erosional remnants that predate valley incision by the East Antarctic Ice Sheet and represent overriding of the mountains prior to the development of the glacial troughs (Mercer 1972; Brady & McKelvey, 1979, 1983; Barrett &

Powell, 1982; Van der Wateren & Verbers, 1994)

3) Part of the Sirius Group was deposited by local ice caps and not by the East Antarctic Ice

Sheet (e.g., Faure et al., 1983; Stroeven and Prentice, 1997).

269 These different views stem from the fact that the relation between the development of the

present East Antarctic Ice Sheet and the erosional history of the Transantarctic Mountains is not

well understood and the fact that individual outcrops have been taken as representing the entire

Sirius Group. In discussing the timing of formation of the major glacial troughs that cross the

Transantarctic Mountains, Denton et al. (1984; 1993) and Marchant et al. (I993a,b) assume that the troughs already existed in the middle Miocene during the initiation of widespread glaciation in the Dry Valleys. Webb (1994) suggested that the troughs were Cretaceous to Paleogene in age.

These old ages are based on the present very low erosional rates, comparisons of the butte-and mesa morphology of the Dry Valleys with other "ancient" landscapes, and the apparent tectonic stability of South Victoria Land since the mid-Miocene (e.g. Denton et al., 1993; Sugden et al.,

1999).

It has been established that the landscapes of the high ranges of the Transantarctic

Mountains are very old. In the central Transantarctic Mountains the oldest ages for erosion surfaces are > 8 Ma for Bennett Platform (Kurz and Ackert, 1998). In Victoria Land, > 14 Ma

Ar/Ar istopic ages wer& obtained for in situ ashes in the Asgard Range of the Dry Valleys

(Marchant et al., 1993b), a > 10 Ma "^Ne exposure age for Mount Fleming, and a > 5.3 Ma "‘Ne exposure age for Mount Feather (Schafer et al., 1999). Evidence of glacial overriding of mountain summits is apparent in tihe central Transantarctic Mountains (McKelvey et al., 1984; Prentice et al., 1986), and Victoria Land (Brady & McKelvey, 1983; Denton et al., 1984; Van der Wateren et al., 1999). The minimuna exposure ages suggest that no major erosion occurred after the middle to late Miocene, and therefore that no ice overrode the highest ranges after this time.

Several lines of reasoning suggest that the overdeepening of the glacial troughs may be

Neogene in age. Transantarctic Mountain uplift may have been episodic rather than continuous

270 Mid-Oligocene - early Miocene Kirkpatrick Basalt

Beacon Supergroup Ferrar Dolerite sill Crystalline Basement

Late Miocene - early Pliocene Glacial troughs, /rrrT T T ^

sea - level

Present © Sirius Group

< rrrrrJ ; : ;

s e a - ^ ---' " level — Ip s a i 1

Figure 10.4. Model of landscape evolution in the Transantarctic Mountains. A=?Mid-01igocene/ early Miocene overriding of the Transantarctic Mountains. B=Development of overdeepened glacial troughs due to uplift and structural segmentation of the Transantarctic Mountains, deposition in terrestrial glacial environments as well as in Qords. C=Topographic distribution of the Sirius Group in relation to the present morphology of the Transantarctic Mountains.

271 since its initiation, including differential uplift histories for separate crustal blocks (Behrendt &

Cooper, 1991; Van der Wateren & Verbers, 1994; Van der Wateren et al., 1999). It is believed

that the present glacial troughs coincide with major transverse faults, and that the development of

these faults has controlled ice-flow during East Antarctic Ice Sheet expansion (Van der Wateren

et al., 1999; Wilson, 1999; Salvini & Storti, 1999). It has been noted before that the present

glacial troughs follow re-activated Mesozoic fault systems (Mazzarini et al., 1997). The timing of

the rejuvenation of the transverse faults, is probably related to the initiation of Ross Sea rifting

and Cenozoic volcanic activity in the late Oligocene to early Miocene (Tessensohn, 1994). The

coincidence of increased volcanic activity (e.g. Stump et al., 1980; Kyle, 1990), reactivation of

faults and exposure of high-elevated landscapes in the Transantarctic Mountains between the late

Oligocene and the middle Miocene suggests that there may be a relationship between them. The

structural segmentation of the Transantarctic Mountains allowed ice to enter the Ross Sea basin

through narrow corridors with steep gradients. Reduction of the ice surface gradients due to the

development of deeper and wider glacial troughs may have further inhibited ice from overriding

the high mountain ranges after the middle Miocene, as suggested by ice-sheet modeling (Kerr &

Huybrechts, 1999). The compositional data from the Sirius Group support such a scenario (Figure

10.4).

272 10.4 Pliocene paleoclimate deduced from the Sirius Group and the Bardin Bluffs Formation

10.4.1 The Bardin Bluffs Formation

The Bardin Bluffs formation shows evidence of a much receded Lambert Glacier in the

early-mid Pliocene (Hambrey & McKelvey, 2000). The presence of planktonic foraminifera in

Member 1 (>3.1 Ma) suggests open ocean-conditions with normal marine salinities during local

peak interglacial time (McKelvey et al., unpubl. data). The geochemical studies (Chapter 7)

suggest that weathering intensity decreases across the mid-late Pliocene transition, although

recycling of weathered Permian sedimentary rocks complicates the analysis.

10.4.2 The Sirius Group

The deposits of the Sirius Group at Oliver Bluffs contain the best available information

for m id-late Pliocene terrestrial paleo-environmental reconstruction. The maximum age of this

deposit is relatively well-constrained due to the presence of fragments of diatomaceous "ooze",

which excludes a windblown origin for the diatoms and supports a glacial transport path

(Harwood & Webb, 1998). However, the mineralogical and sedimentological studies point out

that the Oliver Bluffs succession is not necessarily representative of all Sirius Group deposits.

Therefore, correlation of Sirius Group successions to address paleoclimatic issues is not

recommended until a better stratigraphie framework for the Sirius Group is developed.

The Sirius Group at Oliver Bluffs was deposited near sea level (Webb et al., 1994;

1996a), presumably after early Pliocene deglaciation at > 3.8 Ma as indicated by recycled diatoms. The terrestrial glacial succession represents deposition during glacial times in the mid- late Pliocene, < 3.8 Ma. The combined presence of Pliocene diatoms and Nothofagus fossil

273 Figure 10.5. Peat marsh vegetation and dwarf willow (Betula nand) covering higher ground near the Greenland ice margin, ca. 67“N (a). Dwarf willow grows on the lee side of topographic features near the ice margin because of harsh cold and dry conditions caused by the katabatic wind (b). N othofagus from the Sirius Group may have grown in a similar enviroment (cf. Francis & Hill, 1996).

2 7 4 material in the Oliver Bluffs Sirius Group requires survival ofNothofagus after geographical and thermal isolation of the Antarctic region in the early Miocene (Lawver et al., 1992; Kennett &

Hodell, 1993). The survival of N othofagus into the Pliocene in East Antarctica is further supported by the discovery of significant quantities of Nothofagus pollen in mid-Pliocene sediments of DSDP Site 274 (Fleming & Barron, 1996).

The preservation of a sedimentary succession ca. 60 m thick in a terrestrial environment at Oliver Bluffs suggests that repeated glacial overriding of the deposit was limited.

Sedimentological characteristics of the Oliver Bluffs succession are similar to those found in sub­ polar ice-marginal environments (Chapter 3). A modem analogue is the area of continuous permafrost along the margin of the ice-sheet in central west Greenland, north of 67°N (Figure

10.5). Dwarf tree growth occurs at mean annual air temperatures of below -5°C in central West

Greenland, and mean annual temperatures of -15 to -22°C have been regarded as the lower temperature limit for the survival of N othofagus “Krumholz” at Oliver Bluffs (Francis & Hill,

1996), based on examinations of fossil wood. Weathering rates in sub-polar ice marginal areas are low (Andersen et al., 1997; 2000), which is in agreement with the low CIAs for the Oliver Bluffs deposits (Chapter 5). Paleosols from Oliver Bluffs were classified as cold polar desert soils

(Retallack & Krull, 1996). Post depositional chemical alteration of the Oliver Bluffs deposits is limited to the formation of chabazite and carbonate, which occur in present-day soils of Wright

Valley (Gibson, 1983) and is related to the presence of permafrost (Dickinson & Grapes, 1997).

10.5 Neogene climates and ice-volume in the Antarctic region

Evidence for early Pliocene warming of the southern high-latitudes is substantial, but the magnitude of the warming and the associated decrease in Antarctic ice-volume is debatable

275 Sea level Pliocene deposits in Antarctica Site 704 d'"0 Falling Rising—► - colder/\warmer- Marine Bardin DVDP-11 Oliver Site 274 Ma Ma Piain Bluffs Bluffs pollen 1 2.5 h 2.5 2.7 2.7 2.9 2.9

3.1 /\ 3.1 3.3 3.3 3.5 3.5 IQ. 3.7 0> 3.7 X) 1 0) 3.9 c 3.9 • c CD 1 4.1 4.1 NJ E -J € o o\ 4.3 u 4.3 I m 4.5 4.5 3 4.7 4.7 / 4.9 G. budoWes t 4.9 5.1 5.1 5.3 I 5.3 5.5 5.5

Figure 10.6. Pliocene deposits of Antarctica (data from Webb & Harwood, 1991; Pickard et al.,1988; Hambrey & McKelvey, 2000; Fleming & Barron, 1996; Ishman & Rieck, 1992; Prentice et al. 1993; 1999) eompared to the oxygen isotope stratigraphy of planktonie foraminifera at Site 704 in the Southern Oeeaii (Hodell & Venz, 1992) and a eustatic sea level model from the Atlantic coastal plain (Krantz, 1991). The bars representing Pliocene Antarctic sediments delimit the possible time range of each deposit and do not indicate continuous deposition. Note that evidence for significant warming and deglaciation and marine incursion is concentrated in the interval between 4.5 and 4.0 Ma. Other evidence of warming and high eustatic sea level from proxy records occurs between 3.9 and 3.2 Ma. (Webb & Harwood, 1991; Kennett & Hodell, 1993). A number of Antarctic continental margin

records show evidence of modest warming and deglaciation in the early Pliocene (Figure 10.6). In

East Antarctica near Prydz Bay, marine deposits are exposed up to 15 m above sea level in a

coastal section at Marine Plain, and are dated between 4.0 and 4.5 Ma by diatom biostratigraphy

(Pickard et al., 1988). The deposits contain a rich fossil flora and fauna, including cetacean

vertebrate fossils. The cetacean fauna includes a species of dolphin, which shows no adaptation to

cold environments suggesting that it lived under conditions now found in the vicinity of the

Antarctic Convergence (Quilty, 1992). The glaciomarine sediments of the Bardin Bluffs

Formation (Pagodroma Group) suggest a much receded Lambert Glacier during this time and

open marine conditions during a brief interval in the early Pliocene (Hambrey and McKelvey,

2000; McKelvey et al., unpublished data).

The diatom record from the Oliver Bluffs section indicates deglaciation prior to 3.8 Ma

(Webb et al., 1994; 1996a; Harwood & Webb, 1998), but other deposits in the Transantarctic

Mountains also indicate modest warming and deglaciation in the early Pliocene. Glaciomarine

deposits at Prospect Mesa in Wright Valley (Prentice et al., 1993) and in DVDP-11 in Taylor

Valley (Ishman & Rieck, 1992; Prentice et al., 1999) suggest that Qords existed in the Dry

Valleys under slightly warmer conditions during the early Pliocene. Unpublished oxygen isotope

data (Prentice et al., 1999) suggest considerable melting of the valley glacier in Taylor Valley

Qord between 4.5 and 4.0 Ma.

Pliocene sediments are poorly preserved on the Antarctic continental shelves, but ice- proximal Pliocene glacial records are found in trough-mouth fans on the Antarctic continental slopes. Seismic sequences in trough-mouth fans on the margin have been correlated to Leg 113 drillcores and interpreted as collapse structures caused by major early Pliocene

277 decrease in East-Antarctic ice-volume, associated isostatic rebound, and relative sea level fall

(Bart et al., 1999).

Evidence for early Pliocene warming and/or a reduction in ice volume is also found in stable isotope records of the Southern Ocean (Hodell & Venz, 1992), the eastern equatorial

Pacific (Shackleton, 1995; Kwiek & Ravelo, 1999) and the equatorial Atlantic at Ceara Rise

(Billups et al., 1998). According to the stable isotope records the most prominent high-latitude warming occinred between 4.7 and 3.5 Ma. These results are in agreement with warming events identified in the Southern Ocean from siliceous microfossils assemblages, with peak warming centered at -4.5, -4.2 and -3.6 Ma (Bohaty & Harwood, 1998). Subantarctic ice-rafted debris

(IRD) records periodically show higher concentrations of IRD in the early Pliocene. However, the high IRD concentrations have been interpreted either as indicating major deglaciation (Breza et al., 1992) or as evidence of extensive continental glaciation (Wamke et al., 1992; Kennett &

Hodell, 1993).

Evidence of significant mid-late Pliocene warming, -3.0 Ma, is widely identified in the

Northern Hemisphere and is present but less pronounced in Southern Ocean marine records

(Dowsett et al., 1996). Upper Pliocene stable isotope records are difficult to interpret in terms of

East Antarctic ice volume , because the initiation of Northern Hemisphere glaciation complicates the ice-volume interpretations. Appropriate ice-proximal records from the Antarctic region are lacking, making it difficult to reconstruct Antarctic paleoclimates and ice-volume for the last 3

Ma.

27 8 10.6 Summary and concluding remarks

Contrasting compositions, geomorphological settings, and paleo-ice flow directions

suggest that the Sirius Group was deposited by several significantly different ice-sheet drainage

systems rather than during one glacial phase in the Neogene. Part of the Sirius Group was

deposited during glacial overriding of the higher ranges of the Transantarctic Mountains, such as

Tillite Spur, Mount Sirius, Mount Fleming, and Mount Feather. Other Sirius Group deposits are

similar in character to those of DVDP-11 and Prospect Mesa, and represent deposition within

glacial troughs. The concept of Sirius Group deposition within significantly different landscape

topographies during several stages of glacial denudation of the Transantarctic Mountains is a

departure from previous ideas of emplacement of the Sirius Group, and requires that the Sirius

Group be subdivided into two or more stratigraphie units:

1. Deposits associated with glacial overriding of the highest ranges of the Transantarctic

Mountains, or ice-flow at higher base levels and with an orientation different than the

principal modem drainage system, with much evidence of recycling chemically weathered

sedimentary rocks; some of these deposits were probably deposited by local ice caps, whereas

others may have been deposited by the East Antarctic Ice Sheet;

2. Deposits within glacial troughs, which show evidence of active erosion of igneous and

metamorphic basement and low contemporaneous rates of chemical weathering.

Although evidence is circumstantial, and the chronology of the terrestrial deposits quite poor,

the proposed link between development of structurally defined glacial troughs and ice-sheet drainage provides much better explanations of the distribution and preservation of terrestrial and shallow-marine glacial deposits in the Transantarctic Mountains, as well as the surface exposure ages of high mountain plateaus.

279 This study shows that mid-late Pliocene proglacial deposits associated with the

Nothofagus flora exhibit chemical weathering conditions that do not differ much from the present

conditions that occur at sea level today in continental Antarctica. However, glacial depositional

environments were characterized by higher seasonal meltwater production, suggesting that mid-

late Pliocene glacial climates near sea level were slightly more temperate/maritime than today.

The sedimentological, chemical and paleontological data indicate that mean annual temperatures

for the Oliver Bluffs succession were colder than -5 °C, and perhaps as cold as -15 to -22°C

(Francis & Hill, 1996). Continuous permafrost was probably established and the mean annual

precipitation was probably less than 150 mm/year. A short growth season with temperatures up to

5°C allowed tundra vegetation to persist (Hill & Truswell, 1992; Francis & Hill, 1996).

The present study does not provide additional evidence for major deglaciation in the

early Pliocene, although some degree of deglaciation in the early Pliocene (most likely within the

interval 4.5-4.0 Ma) is indicated by evidence from both the Sirius Group and other Neogene

deposits from Antarctica.

The Oliver Bluffs succession of the Sirius Group is an important terrestrial archive for

studying late Neogene paleoclimatic conditions in the southern high-latitudes. The present study

shows that it is very likely that the Sirius Group includes deposits with a range of ages, with the

Oliver Bluffs succession representing one of the youngest end members. The dynamic versus stable ice-sheet debate centers around a basic assumption, which clearly needs re-evaluation: the assumption that most Sirius Group deposits are the same age, and that evidence from one Sirius

Group outcrop can be extrapolated to other Sirius Group outcrops (Webb & Harwood, 1991;

Stroeven et al., 1996). This study has shown that this approach is invalid and that the sedimentological and paleontological evidence should be re-evaluated to revise the stratigraphie framework for the Sirius Group.

280 APPENDIX A

GRAINSIZE ANALYSIS

METHODS AND REPRODUCIBILITY

Pretreatment

The matrix (< 2 mm) of the samples was treated with H 2O2 and excess chemicals were removed

by boiling after the reaction had stopped. Demineralized water was added to the beaker. Water

and chemicals were decanted after 20 hours. The samples were then treated with a dispersant,

Na2P20?, and heated. After cooling the samples were wet-sieved over 63 |i.m (4 phi). The fraction

> 63 p.m was dried, weighed, and sieved in 1 phi intervals using a rotap. Pipette analysis of the

fraction < 63 p.m was performed to determine silt (2-63 |J.m) and clay (<2 jj.m) content.

Reproducibility of sieving and pipette analysis

Five subsamples from a bulk sample (PNW95-066) were processed to identify the effect of heterogeneity of the sediments on the analysis of grain-size. Two other samples were run in duplicate, sample PNW95-002 and PNW95-074. The clay fraction of these samples was also determined. The precision of the clay contents was further determined by pipette analysis of three aliquots from sample PNW95-003. The results of sieve and mud test analyses are presented in

Table A I.

2 8 1 Sample: PNW95-066 Sample: PNW95-074/-002 Sample: PNW95-003

Split sand % mud % Split sand % silt % clay % Split clay %

A 35.25 64.75 A 33.25 42.50 24.25 A 7.74 B 37.12 62.88 B 33.76 41.89 24.35 B 7.54 C 36.20 63.80 C 7.50 D 36.97 63.03 A 28.84 55.56 15.61 E 35.90 64.10 B 28.88 56.65 14.67

Table Al.Variability of sand, mud and clay % induced by sample heterogenities and method. Five bulk splits of sample PNW95-066 were processes separately and sieved over 63 pm, two bulk splits of PNW95-074 and PNW95-002 were processed separately, and three aliquots of the < 63 pm fraction of sample PNW95-003 were pipetted.

Coulter Counter analysis

Initially wt. % per phi interval of 5 samples (PNW95-012, 013, 022, 023, 032) was determined by

pipette analysis in August 1997. The characterization of the silt-fraction by pipette analysis turned out to be too time-consuming and an alternative technique was sought. Grain-size distribution in

the size range between 0.5-1000 pm have been shown to be accurately determined by electroresistance particle-size analyzers (Milligan & Kranck, 1993). In electroresistance particle size analyzers the number and volume of particles in an electrolyte suspension is determined. The sediment is diluted and suspended before it passes through an aperture of known size. Two electrodes at both sides of the aperture measure the voltage when a constant current of sample is passing through the aperture. The intensity of voltage pulses can be interpreted as particles of a given size passing through the aperture and replacing its volume of electrolyte. The pulse intensities are calibrated with particles of known diameter and recorded on separate channels. The data generated by the instrument represent the number of particles in 256 channels, which are then recalculated to numbers of particles per 0.5 phi grain-size class.

Mary Davis who is an experienced technician on Coulter Counter instruments at BPRC

2 8 2 carried out the analyses in this study. First, a test was run on equipment at BPRC for Coulter

Counter analysis of the silt fraction of a Sirius Group sample (PNW95-066). A pipette sample of

the fraction < 63 p.m was transferred to a Nalgene bottle for detailed measurement on a Coulter

Multisizer, with a 100 |0.m aperture and a 2-80 p.m range. Because most of the samples in this

study are diamictites, a broad size distribution of the silt fraction was anticipated. The challenge

of electroresistence particle size analysis is to obtain measurements for sufficient numbers of

particles in the large-diameter size classes, but to avoid a condition called “coincidence”, when

two particles pass through the aperture at the same time. Coincidence can be avoided by keeping

the concentration of particles in the electrolyte low. Longer counting times may seem a solution

to this problem, but it is hard to keep particles in suspension for a longer time, especially the

coarse particles. Multiple measurements do improve the precision of the measurement, but these

take time. Experiments were carried out to optimize the measurement conditions and to evaluate

the effects of sieving, splitting and the character of the grain-size distribution on accuracy and

precision.

Figure Ai shows a logarithmic plot of the number of particles versus grain diameter for 4 subsamples of PNW95-066. All 4 subsamples were sieved over 63 p.m separately and three splits of each sample were run three times on the Coulter Counter. The plot is a straight line, because grain-size distributions of diamictites are thought to be fractal as a result of the crushing process of grains at the bottom of a glacier or ice-sheet (LeHooke & Iversen, 1996). The subsamples with the highest concentrations of particles show quite consistent results, but the lower two curves show some distortion at the lower end. A dip in the number of particles occurs at the coarser end of the curve. The reasons for the dip may be;

- the number of coarse particles is always very low in the presence of many fine particles, which reduces the accuracy of the measurement

283 - the number of coarse particles is low because a small number of them become trapped in the

sediment and in the sieve when sieving over 4 phi (63 p.m)

- a true deficit in coarse particles may be present in the sediment

- coarse particles settle out of suspension during the Coulter Counter measurement

1000000

00000

10000

1000

100

44.2 31.3 22.1 15.6 11.1 7.81 5.52 3.91 2.76 1.95 particle diameter (um)

Figure A l. Logaritmic plot of particle diameter (|i.m) versus number of particles in a test run of four subsamples of the same sample.

Because Coulter Counter data were combined with sieve data, numbers of particles were recalculated to wt %, using the following formula:

Wt % = n4/37t(l/2df5

Where n is the number of particles, d is the particle size, and ô is the density of the grain, here assumed to be 2.6.

284 10 9 8 7 6 5 4 spliti 3 split2 2 splits 1 0 44.2 31.3 22.1 15.6 11.1 7.81 5.52 3.91 2.76 1.95

particle diameter

Figure A2. Three splits of the same subsample were run three times to evaluate the precision of the Coulter Counter measurements.

15 14 13 □ split 1 12 ■ split 2 11 10 □ split 3 9 8 7 6 5 4 3 2 1 0 I -1-0 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 >9

Figure A3. Grain-size histogram after running three subsamples of sample PNW95-066 by Coulter Counter. The < 4 phi size fractions were determined by sieve analysis, the >9 phi fraction by pipet analysis (ony one subsample). Note poor reproducibility in Lhe 4-5 phi grain-size class by Coulter Counter in the test.

285 To investigate the precision of the measurements in the different size classes three splits

of one subsample are investigated (Figure A2). Variability is low except in the coarsest size

classes measured on the Coulter Multisizer, between 31.3 and 62.5 p.m. The measurements in the

smaller size classes are reproducible within 1.8 % per 0.5 phi interval, but the precision drops

rapidly for size-classes > 31.3 pm. To evaluate the effect on the grain-size histogram the results of 3 splits from the same subsample are included in Figure A3. Again, it shows that a problem occurs in the very coarse silt fraction, but that the finer fractions show quite good results.

After the sensitivity of the technique had been established, the silt fraction of 47 samples was characterized by Coulter Multisizer analysis from September-December 1998 by Mary Davis at BPRC. Techniques of subsampling were improved and one split per sample was run three times to increase the precision of the measurements.

28 6 COMPLETE RESULTS

500- 250 125 63 1-2 1000 -500 -250 -125 <63 Sample # mm (j.m ptm Hm fxm um sand% silt% clay %

PNW95-002 1.78 2.83 5.97 8.90 9.36 71.16 28.84 55.56 15.61 PNW95-003 3.83 6.10 8.93 11.51 11.75 57.88 42.12 50.14 7.74 PNW95-006 3.22 4.13 5.86 9.38 8.69 68.71 31.29 54.20 14.51 PNW95-053 5.77 6.17 7.79 11.14 11.16 57.97 42.03 50.64 7.33 PNW95-054 3.63 5.09 7.93 12.56 10.52 60.28 39.72 58.83 1.45 PNW95-055 6.15 5.72 7.32 9.70 10.67 60.43 39.57 51.19 9.25 PNW95-056 4.29 5.09 6.85 10.50 9.71 63.55 36.45 36.03 27.52 PNW95-057 3.86 4.87 6.48 9.90 8.36 66.52 33.48 49.35 17.16 PNW95-058 1.73 3.44 4.93 7.23 8.14 74.54 25.46 55.33 19.21 PNW95-059 3.64 4.46 9.03 16.52 13.13 53.21 46.79 38.15 15.06 PNW95-060 4.36 5.17 6.22 9.02 11.52 63.71 36.29 52.93 10.79 PNW95-061 3.53 4.69 5.89 8.07 8.77 69.05 30.95 52.74 16.30 PNW95-062 4.01 5.44 7.35 10.70 9.61 62.89 37.11 53.93 8.96 PNW95-063 4.80 5.61 7.44 10.95 9.37 61.83 38.17 48.76 13.06 PNW95-064 4.58 5.45 6.65 9.09 10.31 63.93 36.07 54.88 9.05 PNW95-065 4.20 4.91 6.65 10.20 11.13 62.90 37.10 49.83 13.07 PNW95-066 3.46 4.67 7.01 11.35 9.80 63.71 36.29 54.78 8.94 PNW95-067 4.46 5.06 7.43 10.80 11.32 60.93 39.07 50.76 10.17 PNW95-068 2.11 3.98 7.06 11.93 9.55 65.37 34.63 65.03 0.34 PNW95-070 4.29 5.26 7.00 9.65 12.00 61.81 38.19 54.56 7.25 PNW95-074 4.18 4.20 6.37 9.16 9.34 66.75 33.25 42.50 24.25 PNW95-074 4.28 4.23 6.47 9.61 9.18 66.24 33.76 41.89 24.35 DR13-2 4.06 4.15 6.52 11.98 10.67 62.62 37.38 43.57 19.05 DR13-4 3.88 4.63 7.63 13.73 11.12 59.01 40.99 47.65 11.36 DR13-6 5.99 5.58 7.13 10.43 11.22 59.65 40.35 44.70 14.95 DR13-8 2.26 3.18 5.86 9.90 9.32 69.49 30.51 45.89 23.60 DR5-10 9.34 10.28 10.13 10.95 10.25 49.05 50.95 39.16 9.90 DR5-L2 19.02 17.12 14.59 13.96 11.83 23.48 76.52 20.15 3.33 DR5-14 6.96 6.28 7.33 9.89 11.94 57.59 42.41 48.79 8.80 DR5-16 0.00 0.00 0.20 0.55 3.44 95.81 4.19 81.01 14.80 DR5-18 6.04 6.19 9.76 17.86 11.98 48.17 51.83 41.53 6.63 DR5-2 9.91 10.65 10.18 9.98 11.06 48.22 51.78 41.25 6.97 D R 5^ 10.03 11.65 10.89 11.85 11.48 54.75 55.91 46.62 8.13 DR5-6 5.88 6.64 10.19 14.12 10.54 52.63 47.37 41.30 11.32 DR5-8b 1.08 4.34 10.96 17.66 15.17 50.79 49.21 40.35 10.43 DR8-1 4.44 5.29 8.72 15.55 21.16 44.83 55.17 40.81 4.02

Table A2. Sieve results for the sand fraction and sand-silt-clay results. Clay % determined by pipet analysis. (continued)

287 Table A2. (continued)

500- 250 125 63 1-2 1000 -500 -250 -125 <63 Sample # mm pm pm pm pm pm sand% silt% clay %

MtSirius top 5.01 8.72 12.95 14.06 9.15 50.10 49.90 40.06 10.04 S-1 6.78 14.62 38.28 19.12 5.01 16.20 83.80 11.74 4.45 S-12 2.57 4.74 8.18 9.08 6.66 68.77 31.23 46.61 22.16 S-I3 2.64 4.06 8.92 11.45 6.91 66.02 33.98 48.12 17.90 S-14 0.84 1.53 9.75 16.88 9.71 61.28 38.72 47.52 13.76 S-2 0.83 1.34 6.02 12.72 9.00 70.09 29.91 48.79 21.30 S-3 4.29 5.48 11.62 15.15 9.39 54.05 45.95 37.88 16.17 SA 4.57 5.93 13.71 18.28 10.47 47.04 52.96 32.55 14.49 S-5 5.62 7.32 14.37 17.21 9.85 45.62 54.38 32.56 13.06 S-6 9.14 10.53 12.69 17.03 12.44 38.18 61.82 28.49 9.69 S-8 1.54 5.59 29.95 32.37 9.65 20.90 79.10 16.62 4.28 S-9 2.91 3.60 7.73 10.84 8.49 66.44 33.56 45.51 20.92

Sieve phi -1-0 0-1 1-2 2-3 3-4

PNW95-012 2.71 3.67 5.91 9.93 10.64 PNW95-013 3.34 3.84 6.28 10.63 10.35 PNW95-022 2.44 3.60 8.83 15.48 13.26 PNW95-023 2.87 4.29 8.66 13.31 10.27 PNW95-032 3.53 3.80 6.05 10.16 8.95

Pipet phi 4-5 5-6 6-7 7-8 8-9 >9 sand% silt% clay%

PNW95-012 13.43 12.16 8.36 7.54 6.71 18.94 32.85 48.21 18.94 PNW95-013 9.54 6.98 8.01 8.16 7.93 24.93 34.44 40.63 24.93 PNW95-022 9.68 8.52 7.55 7.40 7.22 16.02 43.61 40.37 16.02 PNW95-023 12.16 10.81 8.18 7.61 5.77 16.06 39.40 44.54 16.06 PNW95-032 10.81 8.29 8.99 8.88 9.66 20.88 32.49 46.63 20.88

Table A3. Sieve and pipet data (%) for Sirius Group samples.

288 1-1.1 1-1.2 1-1.3 1-2.1 1-2.2 1-2.3 lL-3.1 1-3.2 1-3.3 pm 1.95 12052 11750 11809 11214 11482 11531 IL4925 15053 15404 2.76 4191 4220 4163 4100 4120 4104 5449 5398 5505 3.91 1380 1419 1439 1411 1503 1423 L951 1910 1824 5.52 530 521 501 551 557 547 7^38 759 703 7.81 194 211 194 184 195 219 271 243 242 11.05 65 76 66 54 74 74 L25 103 98 15.63 18 33 34 15 31 27 ? 6 33 37 22.1 5 8 12 5 8 6 L6 11 9 31.25 1 4 2 0 1 0 3i 1 3 44.19 0 0 0 0 0 1 0# 2 1 62.5 0 0 0 0 0 0 0» 0 0

2-1.2 2-1.3 2-1.4 2-2.1 2-2.2 2-2.3 Z-3.1 2-3.2 2-3.3 pm 1.95 9672 9874 9825 11050 11277 11098 9:250 9367 9360 2.76 3407 3419 3478 4071 4080 3958 3382 3434 3393 3.91 1303 1228 1306 1402 1479 1523 1233 1252 1241 5.52 465 480 482 576 561 546 5 2 6 447 494 7.81 163 166 189 202 168 205 131 168 164 11.05 68 75 71 85 68 64 6 1 55 56 15.63 19 18 20 31 17 23 2i6 22 20 22.1 4 6 3 10 5 10 7 3 7 31.25 1 0 2 2 0 3 2 0 1 44.19 0 1 0 0 0 0 1 0 1 62.5 0 0 0 0 0 0 0 0 0

3-1.1 3-1.2 3-1.3 3-2.1 3-2.2 3-2.3 3-3.1 3-3.2 3-3.3 pm 1.95 11900 11848 11737 11022 10968 10653 11124 11098 11505 2.76 4537 4458 4599 4195 4076 3905 3989 4161 4128 3.91 1557 1662 1610 1554 1486 1400 1489 1523 1542 5.52 575 589 592 592 541 552 518 489 539 7.81 212 208 241 191 221 205 184 176 174 11.05 71 60 92 71 57 57 6 5 65 69 15.63 30 16 30 26 13 17 2 1 28 29 22.1 9 5 4 4 2 6 2 5 8 31.25 0 0 0 0 0 0 2 0 3 44.19 0 1 0 0 0 0 0 1 0 62.5 0 0 0 0 0 0 0 0 0

Table A4. Coulter Counter test data, sample PNW95-066 (continued)

289 Table A4, (continued)

5-1.1 5-1.2 5-1.3 5-2.1 5-2.2 5-2.3 5-3.1 5-3.2 5-3.3 pim 1.95 4695 4637 4681 6356 6683 6602 6402 6438 6485 2.76 1811 1821 1820 2592 2665 2517 2589 2516 2585 3.91 680 718 741 1036 1004 1017 1000 993 968 5.52 262 297 271 365 394 391 376 412 407 7.81 101 111 110 155 149 152 137 151 166 11.05 42 45 31 42 69 64 60 54 49 15.63 17 11 11 14 18 16 18 21 26 22.1 0 2 5 6 7 3 2 4 2 31.25 2 0 0 1 0 1 1 0 1 44.19 0 0 0 0 1 0 0 0 0 62.5 0 0 0 0 0 0 0 0 0

6-1.1 6-1.2 6-1.3 6-2.1 6-2.2 6-2.3 6-3.1 6-3.2 6-3.3 p.m 1.95 15884 16163 16471 6340 6210 6217 15337 15184 15145 2.76 6107 6198 6183 2082 2104 2050 5658 5570 5554 3.91 1903 1919 1962 626 652 606 1601 1701 1573 5.52 662 661 681 225 227 197 590 568 581 7.81 226 256 213 90 72 76 198 200 210 11.05 85 81 99 31 28 31 86 74 78 15.63 32 31 37 8 10 11 24 23 33 22.1 5 12 10 4 2 3 1 8 12 31.25 1 2 0 0 3 0 0 2 1 44.19 1 2 0 0 0 1 0 0 0 62.5 0 0 0 0 0 0 0 0 0

2 9 0 PNW95-002 PNW95-059 Diameter Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

1.95 9771 9670 9765 10849 11015 11262 2.76 4092 3962 4044 4093 4120 4113 3.91 1584 1594 1611 1461 1489 1458 5.52 632 639 645 514 571 574 7.81 240 240 288 191 195 196 11.05 98 94 104 75 97 68 15.63 26 30 35 32 34 28 22.1 14 15 13 11 9 16 31.25 4 1 5 8 11 8 44.19 0 0 0 1 1 1 62.5 0 0 0 0 0 0

PNW95-G58 PNW95-061 Diameter Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

1.95 13110 13422 12678 7965 7810 7900 2.76 4909 4859 4921 3138 3141 3144 3.91 1726 1723 1689 1299 1282 1269 5.52 632 629 618 548 515 509 7.81 251 228 233 244 201 211 11.05 87 81 97 100 79 91 15.63 44 30 44 32 41 25 22.1 10 7 13 7 10 10 31.25 2 1 4 5 4 3 44.19 1 1 0 2 2 0 62.5 0 0 0 0 0 0

PNW95-006 PNW95-031 Diameter Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

1.95 7722 7788 7528 11522 11468 11507 2.76 3393 3366 3333 5738 5704 5686 3.91 1429 1414 1476 2308 2276 2204 5.52 584 646 642 683 628 659 7.81 255 249 240 184 150 172 11.05 118 108 108 46 42 32 15.63 40 32 46 23 13 10 22.1 24 5 19 3 2 6 31.25 2 5 1 2 2 3 44.19 1 0 1 0 1 1 62.5 0 0 0 0 0 0

Table A5. Complete Coulter Counter data for the Sirius Group. (continued)

291 Table A5. (continued)

PNW95-062 PNW95-063 Diameter Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

1.95 6751 6712 6914 7992 7697 7368 2.76 2980 3091 2992 3149 3166 3092 3.91 1222 1276 1195 1272 1198 1270 5.52 533 566 519 471 495 517 7.81 223 239 226 200 198 180 11.05 105 103 90 84 65 79 15.63 38 49 36 33 27 32 22.1 19 20 20 29 7 10 31.25 6 3 5 2 7 2 44.19 0 0 0 1 1 0 62.5 0 0 0 0 0 0

PNW95-053 PNW95-054 Diameter Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

1.95 9364 9224 9292 8153 8202 8214 2.76 3910 3930 3911 3551 3618 3652 3.91 1529 1511 1549 1518 1583 1490 5.52 611 599 631 658 621 625 7.81 225 249 243 241 265 276 11.05 72 107 87 79 78 100 15.63 38 30 31 38 25 40 22.1 15 8 15 7 6 7 31.25 1 2 1 9 1 6 44.19 0 0 0 1 0 2 62.5 0 0 0 0 0 0

PNW95-064 PNW95-065 Diameter Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

1.95 7593 7693 7574 12270 12897 12920 2.76 3153 3183 3173 5153 5311 5347 3.91 1272 1363 1310 2035 2126 2130 5.52 545 589 589 898 911 900 7.81 281 284 260 394 353 386 11.05 107 108 110 162 153 159 15.63 43 42 40 65 61 55 22.1 13 19 20 17 24 23 31.25 8 7 6 11 10 11 44.19 2 3 3 1 2 2 62.5 0 0 0 0 0 0

292 Table A5. (continued)

PNW95-055 PNW95-056 Diameter Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

1.95 7854 7909 7802 11331 11355 11348 2.76 3279 3271 3335 4207 4279 4259 3.91 1345 1319 1291 1516 1508 1500 5.52 559 564 517 512 531 514 7.81 231 202 203 172 192 204 11.05 99 103 78 70 56 61 15.63 31 29 35 20 25 18 22.1 19 11 6 7 3 5 31.25 6 3 2 4 0 2 44.19 1 0 1 0 0 0 62.5 0 0 0 0 0 0

PNW95-068 PNW95-057 Diameter Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

1.95 8971 9215 9315 9559 9783 9615 2.76 3126 3318 3237 3869 3844 3783 3.91 1060 1041 1066 1372 1520 1435 5.52 417 397 379 556 565 543 7.81 163 172 163 227 207 203 11.05 60 52 70 100 82 78 15.63 22 16 27 36 28 31 22.1 10 8 7 13 14 14 31.25 8 0 4 5 2 5 44.19 0 1 1 1 0 1 62.5 0 0 0 0 0 0

PNW95-060 PNW95-067 Diameter Run 1 Run 2 Run 1 Run 2 Run 3 Run 4

1.95 6427 6293 5626 5646 5457 5446 2.76 2748 2612 2295 2229 2216 2268 3.91 1133 1149 929 920 919 913 5.52 475 470 368 404 348 379 7.81 195 196 150 137 147 144 11.05 84 104 62 47 68 71 15.63 28 37 19 17 21 18 22.1 9 17 6 8 12 16 31.25 1 6 3 1 3 6 44.19 1 1 0 0 1 0 62.5 0 0 0 0 0 0

293 Table A5. (continued)

PNW95-070 S-1 Diameter Run 1 Run 2 Run 3 1st run 2nd run 3rd run

1.95 4904 5048 4996 7881 7883 7879 2.76 2196 2239 2160 3338 3356 3265 3.91 911 930 958 1316 1274 1265 5.52 470 398 424 507 463 457 7.81 200 170 185 172 191 169 11.05 68 72 95 67 59 61 15.63 32 22 33 25 23 22 22.1 11 8 10 8 4 7 31.25 5 5 5 3 2 2 44.19 3 0 2 1 1 0 62.5 0 0 0 0 0 0

S-2 S-3 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 15138 15507 16086 12061 11672 12013 2.76 7229 7261 7263 4959 4810 5085 3.91 2879 2939 2824 1944 1933 1957 5.52 1059 1040 992 716 682 710 7.81 349 365 329 217 220 243 11.05 99 107 91 78 78 91 15.63 40 27 34 34 25 29 22.1 17 13 13 7 6 6 31.25 5 1 1 1 7 3 44.19 1 0 0 0 0 0 62.5 0 0 0 0 0 0

S-4 S-5 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 14376 12551 12576 7144 6831 7103 2.76 5924 5624 5681 2772 2522 2690 3.91 2103 2151 2120 887 884 969 5.52 727 768 753 300 316 294 7.81 270 231 265 94 85 102 11.05 81 75 98 35 29 34 15.63 37 24 31 10 16 8 22.1 12 10 14 2 3 4 31.25 1 3 2 1 0 0 44.19 2 0 0 2 0 0 62.5 0 0 0 0 0 0

294 Table A5. (continued)

S-6 S -8 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 8474 8646 8920 6049 5943 6162 2.76 3598 3625 3633 2575 2658 2604 3.91 1358 1407 1431 1075 1060 1078 5.52 517 559 558 491 445 389 7.81 210 209 193 184 178 165 11.05 67 69 60 78 60 53 15.63 27 30 19 38 31 28 22.1 13 6 5 9 16 16 31.25 0 1 2 3 6 5 44.19 0 0 0 0 0 0 62.5 0 0 0 0 0 0

S-9 S-12 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 15277 15460 15275 13640 13426 13949 2.76 6757 6559 6578 5569 5593 5666 3.91 2440 2515 2419 1971 1993 1864 5.52 781 864 825 696 647 629 7.81 306 268 257 220 230 234 11.05 95 96 87 55 65 83 15.63 33 33 26 24 15 25 22.1 14 10 8 6 12 4 31.25 3 1 0 4 5 5 44.19 1 0 0 0 1 1 62.5 0 0 0 0 0 0

S-13 S-14 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 13864 13846 13829 10906 10799 10965 2.76 6509 6585 6566 5065 4904 4888 3.91 2557 2580 2675 2080 2113 2030 5.52 1062 1001 1021 827 874 786 7.81 338 370 383 295 317 303 11.05 106 118 124 127 96 112 15.63 29 35 48 49 42 28 22.1 12 6 15 11 7 14 31.25 2 1 2 0 1 3 44.19 1 1 0 0 1 1 62.5 0 0 0 0 0 0

295 Table A5. (continued)

Top of Mt. Sirius Diameter 1st run 2nd run 3rd run

1.95 10775 10957 10432 2.76 4191 4157 4162 3.91 1582 1564 1481 5.52 579 511 511 7.81 194 167 188 11.05 60 58 74 15.63 17 18 21 22.1 8 5 11 31.25 4 1 0 44.19 0 1 1 62.5 0 0 0

DR5-2 DR5-4 Diameter 1st run 2nd run 3rd run 1st run2nd run 3rd run

1.95 4674 4654 4706 7901 7608 7562 2.76 2072 1996 1983 3305 3323 3320 3.91 814 833 831 1241 1236 1275 5.52 332 287 321 495 481 496 7.81 102 130 117 188 211 213 11.05 69 50 72 72 87 75 15.63 25 25 21 34 38 22 22.1 10 6 12 6 11 15 31.25 7 1 2 5 3 3 44.19 1 0 0 2 1 2 62.5 0 0 0 0 0 0

DR5-6 DR5-10 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 6961 6645 6883 5973 6004 5803 2.76 2978 2889 2874 2570 2559 2482 3.91 1215 1159 1209 1021 1028 1039 5.52 487 488 463 389 432 442 7.81 191 184 183 168 176 149 11.05 73 52 88 55 51 71 15.63 32 22 30 17 27 13 22.1 11 10 9 5 9 4 31.25 5 2 1 2 3 3 44.19 4 2 0 1 0 0 62.5 0 0 0 0 0 0

2 9 6 Table A5. (continued)

DR5-12 D R 5 -1 4 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 2286 2300 2273 8050 7884 8099 2.76 901 943 926 3658 3520 3476 3.91 371 357 331 1509 1532 1475 5.52 128 157 163 612 651 637 7.81 66 46 43 276 253 256 11.05 22 19 19 99 114 103 15.63 9 7 5 45 44 51 22.1 5 3 2 10 17 16 31.25 0 0 2 8 7 7 44.19 1 0 1 1 3 0 62.5 0 0 0 0 0 0

DR5-16 DR5-18 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 9525 9656 9680 11389 11212 11299 2.76 4896 4789 4831 4650 4755 4760 3.91 2293 2213 2270 1825 1825 1860 5.52 865 864 923 656 688 776 7.81 ,276 300 313 268 260 303 11.05 91 101 103 103 99 110 15.63 26 28 29 37 40 47 22.1 10 4 5 12 13 19 31.25 2 0 0 1 6 8 44.19 0 0 0 2 0 3 62.5 0 0 0 0 0 0

DR5-8b DMH051 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 8694 8461 8537 6029 5676 5912 2.76 3682 3615 3692 2532 2556 2553 3.91 1538 1507 1470 1114 1007 1065 5.52 613 624 594 413 421 415 7.81 232 214 240 175 185 176 11.05 92 80 86 64 79 75 15.63 25 29 30 23 31 34 22.1 10 11 10 15 5 15 31.25 3 6 7 1 7 2 44.19 3 0 1 1 0 0 62.5 0 0 0 0 0 0

297 Table A5. (continued)

DR13-2 D R 1 3 -4 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 12253 12582 12568 6362 6456 6423 2.76 5016 5208 5165 2620 2687 2640 3.91 1970 1949 2013 1021 1112 1060 5.52 694 714 702 415 446 436 7.81 274 278 218 159 180 162 11.05 103 88 102 88 56 91 15.63 29 23 30 26 16 26 22.1 13 8 9 11 9 16 31.25 3 5 4 2 5 5 44.19 0 2 1 0 4 1 62.5 0 0 0 0 0 0

DR13-6 DR13-8 Diameter 1st run 2nd run 3rd run 1st mn 2nd runL 3rd run

1.95 13290 13331 13229 13809 14079 14210 2.76 5644 5404 5619 5466 5472 5319 3.91 2244 2201 2142 1897 1911 1892 5.52 825 824 812 642 573 669 7.81 308 276 308 203 221 195 11.05 104 96 80 75 78 79 15.63 35 44 28 16 36 17 22.1 12 14 8 4 3 8 31.25 0 0 3 5 2 ■ 1 44.19 0 1 0 0 1 0 62.5 0 0 0 0 0 0

PNW95-074 PNW95-074a Diameter 1st run 2nd run 3rd run 1st run 2nd run

1.95 9825 10252 10245 12021 12289 2.76 3836 3937 3989 4672 4589 3.91 1463 1379 1412 1754 1709 5.52 540 545 517 658 601 7.81 201 198 208 234 206 11.05 101 63 73 76 60 15.63 22 38 32 22 21 22.1 6 10 8 5 10 31.25 2 3 3 7 2 44.19 0 0 0 1 1 62.5 0 0 0 0 0

298 T a b le A 5. (continued)

PNW95-074b PNW95-003 Diameter 1st run 2nd run 3rd run 1st run 2nd run 3rd run

1.95 11262 11422 11609 11392 11463 11390 2.76 4328 4243 4312 5283 5466 5336 3.91 1498 1515 1526 2133 2156 2105 5.52 571 504 536 847 865 846 7.81 219 183 190 286 287 304 11.05 67 77 63 113 105 101 15.63 30 21 25 33 41 39 22.1 8 10 11 9 7 13 31.25 4 1 1 3 7 3 44.19 0 1 0 2 3 0 62.5 0 0 0 0 0 0

299 APPENDIX B

BULK X-RAY DIFFRACTION

Powdered samples were pressed into aluminum sample holders at random orientation and glycolated (ethylene glycol) for 12 hours prior to the analysis. Samples were analyzed on a

Rigaku Miniflex XRD system at the Crary Laboratory in McMurdo Station, Antarctica. The scan range was from 3 to 65° 20 with a step size of 0.02° and a dwell time of 2 seconds per step. The digital data were interpreted using Jade 3.0 software, which comprises a search-match routine based on a Powder Diffraction File on CD-ROM provided by the International Center for

Diffraction Data. During the Jade 3 procedure, background was removed and the profiles were calibrated using known quartz peak positions. Reproducible results were obtained for the mineral/quartz ratios listed in Table Bl.

Mineral Peaks used Quartz peak Intensity/Area (A) (A) ratio Feldspar 4.04, 3.19, 3.24, 2.77 3.34 Intensity ratio Pyroxene 2.99, 2.95, 2.89 4.25 Intensity ratio Amphibole 8.40-8.50 4.25 Intensity ratio Mica/Illite 10, 4.99 4.25 Intensity ratio Total clays 4.47 4.25 Area ratio Chlorite 7.05-7.20 4.25 Area ratio Calcite 3.03 4.25 Area ratio Chabazite 9.30-9.35 4.25 Area ratio

Table Bl. Peaks used to calculate relative abundances of minerals from X-ray diffractograms.

300 Quartz d-spacing (Â) 4.25 3.34 4.25 3.34 Quartz Quartz ratio Quartz Quartz ratio

Sample I I AA

PNW95-070 282 1416 0.20 61 276 0.22 PNW95-070 454 2115 0.21 62 382 0.16

PNW95-068 528 2962 0.18 106 538 0.20 PNW95-068 451 2790 0.16 100 528 0.19

DR8-0 264 1172 0.23 60 240 0.25 DR8-0 217 1282 0.17 46 252 0.18

Feldspars d-spacing (Â) 3.34 4.04 3.76* 3.24 3.19 2.77 2.16* F/Q Quartz Plag. Fsp K-fsp Plag. K-fsp. K-fsp. w/o *

Sample I II IIII I

PNW95-070 1416 205 219 255 870 31 53 57 1.19 0.96 PNW95-070 2115 332 153 405 737 34 125 52 0.87 0.71

PNW95-068 2962 231 318 0 755 0 32 28 0.46 0.33 PNW95-068 2790 331 144 0 429 0 62 0 0.35 0.27

DR8-0 1172 204 159 0 857 0 117 60 1.19 0.90 DR8-0 1282 183 206 0 1072 0 47 75 1.23 0.98

Feldspars d-spacing (Â) 3.34 4.04 3.76 3.24 3.18 2.77 2.16 2.11 F/Q Quartz Plag. Fsp. K-fsp Plag. K-fsp. K-fsp.

Sample A AA A AAA

PNW95-070 276 41 39 55 198 3 7 15 1.30 PNW95-070 382 48 29 169 234 3 19 25 1.38

PNW95-068 538 38 57 0 296 0 18 6 0.77 PNW95-068 528 49 17 0 164 0 20 G 0.47

DR8-0 240 37 33 0 196 0 23 6 1.14 DR8-0 252 35 39 0 219 0 13 28 1.33

Table B2. Based on three duplicate analyses the peaks used for the XRD mineralogy studies were chosen. The duplicates were aliquots of the same ground sample prepared, glycolated, and scanned at a different times, so the variability between duplicates is a product of sample heterogenities, variability in glycolation and instrument drift. (continued)

301 Table B2. (continued)

Pyroxenes d-spacing (Â) 4J25 2.99 2.95 2.89 Pyx/Q d-spacing (Â) 8.42 Quartz Pyx. Pyx. Pyx. ratio Hb.

Sample I I I I I+A

PNW95-070 282 338 194 204 0.87 0 PNW95-070 454 645 177 232 0.77 0

PNW95-068 528 178 163 47 0.24 0 PNW95-068 451 153 106 84 0.25 0

DR8-0 264 206 148 298 0.82 0 DR8-0 217 180 124 275 0.89 0

Pyroxenes d-spacing (Â) 4.25 2.99 2.95 2.89 Pyx/Q Quartz ratio

Sample AAAA

PNW95-070 61 78 29 50 0.86 PNW95-070 62 113 48 46 1.11

PNW95-068 106 76 56 19 0.47 PNW95-068 100 48 53 31 0.44

DR8-0 60 82 29 60 0.95 DR8-0 46 98 24 48 1.23

Illite/Micas Chlorites d-spacing (Â) 4.25 10 4.99 4.47 IlliteM/Q d-spacing (Â) 7.05 Chlorite/Q Quartz ratio -7.20 ratio

Sample II11 I

PNW95-07Q 282 64 64 69 0.23 64 0.23 PNW95-070 454 0 73 53 0.09 48 0.11

PNW95-068 528 0 49 99 0.09 91 0.17 PNW95-068 451 0 39 109 0.11 80 0.18

DR8-0 264 70 58 0 0.16 45 0.17 DR8-0 217 64 35 0 0.15 57 0.26

302 Table B2. (continued)

niite/Micas Chlorites d-spacing (Â) 4.25 10 4.99 4.47 niiteM/Q d-spacing (Â) 7.05 Chlorite/Q Quartz ratio -7.20 ratio

Sample A AA A A

PNW95-070 61 3 8 7 0.10 9 0.15 PNW95-070 62 0 5 7 0.06 4 0.06

PNW95-068 106 0 5 22 0.08 10 0.09 PNW95-06S 100 0 7 32 0.13 7 0.07

DR8-0 60 4 7 0 0.06 5 0.08 DR8-0 46 15 4 0 0.14 5 0.11

Zeolites Calcite d-spacing (Â) 4.25 8.9-9.1 9.35 9.49 Zeolite/Q d-spacing (Â) 3.03 Calcite/Q Quartz Zeolite Chah. Laum. ratio ratio

Sample I II I I

PNW95-070 282 72 122 0 0.23 88 0.31 PNW95-070 454 148 70 0 0.16 111 0.24

PNW95-068 528 0 0 170 0.11 77 0.15 PNW95-068 451 64 0 0 0.05 59 0.13

DR8-0 264 162 0 0 0.20 173 0.66 DR8-0 217 0 160 0 0.25 91 0.42

Zeolites Calcite d-spacing (Â) 4.25 8.9-9.1 9.35 9.49 Zeolite/Q d-spacing (Â) 3.03 Calcite/Q Quartz Zeolite Chab. Laum. ratio ratio

Sample AAA AA

PNW95-070 61 4 20 0 0.13 22 0.36 PNW95-070 62 11 5 0 0.09 29 0.47

PNW95-068 106 0 0 16 0.05 20 0.19 PNW95-068 100 8 0 0 0.03 14 0.14

DR8-0 60 28 0 0 0.16 30 0.50 DR8-0 46 0 21 0 0.15 21 0.46

303 9.9-10.1 9.9-10.1 9.45 9.3-9.35 8.9-9.2 8.4-S.5 8 .. mite/ mite/ Laum. Chabaz. Zeolites Amph. Ai Micas Micas Sample I A A A A IA PNW95-070 64 3 0 20 4 0 0 PNW95-071 0 0 34 0 0 0 0 PNW95-072 69 4 22 0 0 0 0 PNW95-073 100 31 12 0 0 0 0 PNW95-074 109 4 32 0 0 0 0 PNW95-053 0 0 0 0 35 0 0 PNW95-054 94 8 23 0 0 0 0 PNW95-055 59 16 111 0 0 0 0 PNW95-056 0 0 31 0 0 0 0 PNW95-057 81 19 16 0 4 0 0 PNW95-058 100 56 0 46 0 52 5 PNW95-059 80 11 63 0 0 0 0 PNW95-060 63 7 46 0 5 0 0 PNW95-061 107 4 12 0 0 0 0 PNW95-062 87 5 0 13 0 0 0 PNW95-063 0 0 0 0 9 0 0 PNW95-064 63 4 43 0 9 0 0 PNW95-Ü65 63 15 10 0 0 0 0 PNW95-066 0 0 31 0 4 89 4 PNW95-067 58 23 39 0 13 50 5 PNW95-068 0 0 16 0 0 0 0 PNW95-049 66 32 44 0 0 0 0 PNW95-050 0 0 58 0 12 0 0 PNW95-016 0 0 130 0 69 0 0 PNW95-017 0 0 0 54 21 0 0 PNW95-018 91 5 34 0 22 0 0 13-9 70 4 8 0 4 0 0 13-8 50 4 51 0 4 0 0 13-7 98 5 9 0 3 0 0 13-6 63 4 6 0 0 0 0 13-5 54 4 18 0 0 0 0 13-4 98 5 13 0 0 0 0 13-3 54 4 18 0 0 0 0 13-2 153 19 7 27 0 0 0 13-1 51 8 0 11 0 0 0 5-19 96 15 15 0 0 0 0 5-18 43 3 0 12 0 0 0 5-17 45 4 14 0 3 0 0 5-16 0 0 23 16 0 48 4 5-15 0 0 15 0 0 0 0 5-14 0 0 14 0 0 51 5 5-13 0 0 0 41 0 0 0 5-12 0 0 0 35 0 0 0 5-11 0 0 21 0 0 64 3

Table B3. Results of Intensity (I) and Area (A) counts of bulk X-ray mineralogy analysis. (continued)

304 Table B3. (continued)

9.9-10.19.9-10.19.45 9.3-9.35 8.9-9.2 S.4-8.5 S.4-8.5 Elite/ Elite/ Laum. Chabaz. Zeolites Amph. An Micas Micas Sample I AA A AIA 5-10 80 5 27 0 0 0 0 5-9 110 5 18 0 3 0 0 5-8b 0 0 0 14 0 0 0 5-6 66 4 17 0 0 0 0 5-5 53 5 0 36 4 0 0 5-4 32 3 16 0 0 90 10 5-2 77 4 0 42 0 0 0 5-1 36 3 18 0 3 0 0 8-5z 0 0 31 0 5 0 0 8-5y 191 43 11 0 3 0 0 8-5x 91 14 19 0 0 0 0 8-5w 94 4 11 0 0 0 0 8-5u 55 5 24 0 11 34 4 8-5t 39 4 5 0 3 54 3 8-5s 40 3 0 0 0 0 0 8-5r 40 5 10 0 0 0 0 8-5q 63 4 7 0 3 0 0 8-5 0 0 0 0 0 0 0 8-4 34 5 22 0 0 0 0 8-3a/b 53 10 49 35 10 0 0 8-2 43 4 0 39 5 38 3 8-1 67 27 0 41 0 0 0 8-0 64 15 0 21 0 0 0 7-1 0 0 23 0 15 0 0 7-4 59 5 0 27 2 0 0 Mt.Sirius-top 91 10 0 0 0 0 0 S-7 101 15 11 0 0 0 0 S-6 174 35 6 0 0 0 0 S-5 131 31 0 0 0 0 0 S-4 141 18 20 0 7 0 0 S-3 86 7 0 0 4 0 0 S-2 155 24 0 0 7 0 0 S-9 181 38 5 0 0 0 0 S-10 182 36 20 0 0 0 0 S-12 58 6 0 0 37 0 0 S-13 109 17 0 0 0 0 0 S-14 103 18 0 0 8 0 0 PNW95-002 92 17 29 0 0 0 0 PNW95-003 0 0 13 0 0 0 0 PNW95-004 0 0 0 0 0 0 0 PNW95-005 0 0 0 0 0 0 0 PNW95-006 55 10 9 0 0 0 0 PNW95-012 40 6 0 33 0 0 0

305 Table B3. (continued)

9.9-lG.l 9.9-lG.l 9.45 9.3-9.35 8.9-9.2 8.4-8.5 8.4-8.S mite/ mite/ Laum. Chabaz. Zeolites Amph. Amph. Micas Micas Sample I A A A A IA PNW95-013 111 9 31 G 7 G G PNW95-014 G G 31 G G G G PNW95-019 G G 65 G G GG PNW95-020 89 26 29 G G GG PNW95-G21 53 8 28 G G GG PNW95-022 G G 24 G 7 GG PNW95-023 9G 14 13 G GGG PNW95-G31 97 36 G 35 GGG PNW95-G32 148 61 26 G 6 GG PNW95-G33 157 14 7G G GGG PNW95-G34 131 67 26 G G GG PNW95-G35 6G 4 23 G GG 0 Prosmesa 44 3 0 G G 127 IG MtFeath 3G8 85 49 G GGG TM-5 GG G 66 G GG TM-6 GG G 5G G 0 G TM-7a G G G 23 G GG TM-7b G G G 34 G GG MtFletn2 292 89 28 G G G G MtRem3 3G6 71 32 0 7 G G DVDP11-12G 371 77 0 G G 4G2 71 DVDPl 1-156 456 119 6 G G 98 21 DVDP11-17G 243 56 G G 3 233 35 DVDPl 1-2G4 253 83 G G 2 86 IG DVDPl 1-221 128 38 G G 2 143 17 DVDPl 1-242 232 49 G G 2 485 8 G DVDPl 1-253 215 38 G G G 174 21 DVDPl 1-269 151 19 G G 4 286 31 DVDPl 1-291 149 2G 3 4 2 111 15 DVDPl 1-311 163 44 G GG 2G5 37 DVDPl 1-323 179 15 G GG 439 60 94TS33 478 134 G GG GG 94TS31 4G3 1G5 0 GG GG 94TS29 6G1 178 G G G G G 94TS28 437 1G8 G GG GG 94TS27 457 14G G GG GG 94TS25 4G4 77 G G G GG 94TS24 212 5G G GG GG 94TS23 284 83 G G GG G 94TS22 764 217 G G GGG 94TS12 562 15G G G GG G 94TS11 1282 355 G G GG G 94TS14 258 49 G G GG G 94TS19 268 54 G GGG G

306 Table B3. (continued)

9.9-10.19-9-10.19.45 9.3-9.35 8.9-9.2 S.4-8.5 8.4-8.5 Illite/ Illite/ Laum. Chabaz. Zeolites Amph. Amph. Micas Micas Sample I AA AAI A 94TS18 1299 354 0 0 0 0 0 94TS16 1731 396 0 0 0 0 0 94TS08 838 230 0 0 0 0 0 94TS06 565 227 0 0 0 0 0 94TS05 1819 519 0 0 0 0 0 94TS02 1464 353 0 0 0 0 0 94TS01 328 137 0 0 0 0 0 94QH46 460 208 0 30 0 0 0 94QH45 646 283 5 0 3 50 3 94QH44 1233 460 20 0 36 87 6 94QH43 557 226 12 0 5 59 10 94QH40 801 234 0 19 10 56 8 94QH39 830 352 12 4 4 76 9 94QH38 764 171 0 16 0 192 19 94QH37 675 243 12 0 8 0 0 94QH36 1323 619 28 0 12 108 17 94QH34 879 192 7 0 8 264 34 94QH33 465 113 0 5 7 157 9 94QH31 1417 431 0 0 0 0 0 94QH30 643 213 0 6 0 107 20 94QH29 1273 464 12 0 5 248 23 94QH24 908 384 7 0 0 59 4 94QH13 910 411 6 0 0 71 12

307 Table B3. (continued)

7.G5-7.2 4.99 4.99 4.47 4.25 4.25 Chlorite llite/ Illite/ Illite Quartz Quartz Micas Micas -Mont. Sample AI A A I A PNW95-070 9 64 8 7 282 61 PNW95-071 40 45 6 31 749 161 PNW95-072 34 0 0 18 635 125 PNW95-073 19 37 8 40 716 151 PNW95-074 26 35 4 56 501 116 PNW95-053 0 71 5 19 583 122 PNW95-054 31 47 7 15 940 157 PNW95-055 19 40 7 13 671 130 PNW95-056 8 80 14 6 1130 214 PNW95-057 26 0 0 15 691 153 PNW95-058 13 53 14 44 649 111 PNW95-059 10 52 6 21 692 156 PNW95-060 21 97 15 9 553 104 PNW95-061 31 34 10 16 471 110 PNW95-062 29 32 7 33 659 116 PNW95-063 0 0 0 16 486 96 PNW95-064 17 37 10 14 691 136 PNW95-G65 22 30 6 11 430 111 PNW95-066 23 60 5 11 779 166 PNW95-067 17 0 0 0 784 178 PNW95-068 10 49 5 22 528 106 PNW95-G49 0 66 10 0 780 154 PNW95-G5G 14 43 7 11 1028 221 PNW95-G16 22 134 36 3 575 99 PNW95-G17 7 43 5 6 567 122 PNW95-G18 12 102 11 8 530 113 13-9 29 0 0 35 481 83 13-8 48 0 0 61 442 84 13-7 14 0 0 40 671 138 13-6 46 28 4 40 679 117 13-5 23 30 4 21 387 87 13-4 24 0 0 36 600 125 13-3 8 0 0 46 522 109 13-2 11 43 4 52 625 128 13-1 6 0 0 67 484 73 5-19 29 0 0 21 702 156 5-18 7 0 0 36 335 91 5-17 11 47 4 16 485 104 5-16 3 31 3 28 512 110 5-15 19 68 4 10 431 125 5-14 15 37 2 9 454 82 5-13 11 25 4 22 469 121 5-12 11 57 6 4 357 76 5-11 0 38 8 36 422 121

308 Table B3. (continued)

7.05-7.2 4.99 4.99 4.47 4.25 4.25 Chlorite Elite/ Elite/ Elite Quartz Quai Micas Micas -Mont. Sample A I AA I A 5-10 7 31 4 33 443 100 5-9 15 87 13 14 577 113 5-8b 9 66 4 17 423 135 5-6 14 0 0 45 542 103 5-5 3 0 0 4 261 35 5-4 10 0 0 13 310 68 5-2 0 49 9 8 368 65 5-1 14 33 0 10 500 98 8-5z 25 0 0 36 450 79 8-5y 70 66 15 55 641 100 8-5x 79 40 3 16 362 61 8-5w 32 0 0 47 483 133 8-5u 24 39 5 16 415 64 8-5t 0 32 5 43 330 63 8-5s 7 0 0 29 170 41 8-5r 16 0 0 8 385 93 8-5q 16 31 3 7 331 75 8-5 8 0 0 13 179 43 8-4 4 45 4 20 409 103 8-3a/b 5 62 11 6 616 125 8-2 5 76 5 9 433 83 8-1 9 44 8 7 581 129 8-0 5 35 4 0 217 46 7-1 8 28 4 19 638 122 7-4 6 56 9 9 240 64 Mt.Sirius top 22 73 12 40 1012 180 S-7 54 0 0 31 869 181 S-6 95 71 17 19 776 198 S-5 36 42 8 30 965 194 S-4 38 37 7 36 765 175 S-3 22 37 8 21 686 137 S-2 35 50 15 85 556 146 S-9 50 61 17 60 579 154 S-10 45 54 13 49 792 162 S-12 30 43 6 59 606 126 S-13 37 42 10 57 573 140 S-14 7 47 9 18 972 202 PNW95-002 14 57 5 79 705 192 PNW95-003 0 0 0 12 1441 293 PNW95-004 10 0 0 40 1520 300 PNW95-005 7 0 0 43 1504 299 PNW95-006 21 0 0 57 1122 246 PNW95-012 37 58 6 16 462 134

309 Table B3. (continued)

7.05-7.2 4.99 4.99 4.47 4.25 4.25 Chlorite niite/ Illite/ Illite Quartz Quai Micas Micas -Mont. Sample A I AA I A PNW95-0I3 53 107 19 32 678 151 PNW95-014 39 49 10 51 642 162 PNW95-0I9 22 0 0 27 1422 262 PNW95-02G 22 53 4 19 893 220 PNW95-02I 42 23 6 12 1335 237 PNW95-022 27 0 0 23 895 196 PNW95-023 32 0 0 25 903 178 PNW95-03I 28 75 17 24 464 88 PNW95-G32 52 34 12 28 914 178 PNW95-G33 85 52 12 7 935 220 PNW95-034 64 54 11 17 917 226 PNW95-G35 29 40 10 41 681 197 Prosmesa 12 0 0 0 581 86 MtFeath 97 107 23 40 1022 208 TM-5 0 147 21 0 1550 294 TM-6 0 115 19 0 2278 378 TM-7a 11 70 13 21 1709 306 TM-7b 18 82 18 37 1390 264 MtPlem2 109 74 25 58 1021 160 MtPlem3 114 101 38 41 1068 236 DVDPl 1-120 13 0 0 3 467 92 DVDPl 1-156 4 43 3 4 386 75 DVDPl 1-170 14 0 0 0 335 69 DVDPl 1-204 4 22 0 0 114 28 DVDPl 1-221 8 0 0 9 148 29 DVDPl 1-242 17 0 0 0 268 61 DVDPl 1-253 77 40 4 4 366 65 DVDPl 1-269 3 0 0 9 256 62 DVDPl 1-291 13 0 0 7 195 48 DVDPl 1-311 7 0 0 0 530 99 DVDPl 1-323 9 24 4 0 581 109 94TS33 18 106 17 24 1356 266 94TS31 15 99 18 22 1748 332 94TS29 24 146 23 28 1401 270 94TS28 25 102 19 27 955 200 94TS27 22 112 12 31 1758 326 94TS25 7 106 28 54 1074 220 94TS24 0 66 14 33 1263 229 94TS23 0 75 14 49 1092 220 94TS22 19 189 23 47 936 177 94TS12 0 136 16 56 740 155 94TS11 9 218 49 37 929 181 94TS14 8 73 14 25 899 196 94TS19 10 68 9 47 1618 264

310 Table B3. (continued)

7.05-7.2 4.99 4.99 4.47 4.25 4.25 Chlorite Illite/ mite/ mite Quartz Quai Micas Micas -Mont. Sample AIA AI A 94TS18 0 253 41 25 696 162 94TS16 51 352 59 14 741 169 94TS08 30 204 34 28 882 179 94TS06 62 109 22 87 875 192 94TS05 168 283 50 57 780 159 94TS02 27 277 47 0 1449 256 94TS01 0 85 16 50 981 204 94QH46 59 83 9 11 340 76 94QH45 6 92 14 5 427 93 94QH44 5 130 26 0 609 136 94QH43 24 82 7 0 437 91 94QH40 25 106 19 11 339 69 94QH39 28 95 9 5 606 105 94QH38 13 222 38 0 927 158 94QH37 0 99 12 0 507 115 94QH36 14 162 27 11 492 107 94QH34 11 82 17 8 434 90 94QH33 0 62 10 6 978 191 94QH31 41 142 16 8 626 108 94QH30 14 67 6 7 596 109 94QH29 24 133 24 8 413 87 94QH24 16 131 18 7 284 68 94QH13 25 85 13 7 435 86

311 Table B3. (continued)

3.18 4.04 4.04 3.76 3.34 3.24 -3.20 3.03 Plag. Plag. Fsp Quartz Microcl. Plag. Calcite Sample IAI II I A PNW95-070 205 41 219 1416 255 870 22 PNW95-07I 186 38 178 3303 109 704 25 PNW95-072 203 40 206 2973 105 519 24 PNW95-073 118 29 147 3387 0 480 32 PNW95-074 157 28 177 2816 0 526 31 PNW95-053 183 37 168 3397 0 743 16 PNW95-054 174 33 170 3670 162 872 0 PNW95-055 164 38 266 3471 0 822 28 PNW95-056 184 31 150 3640 0 608 12 PNW95-057 171 31 165 3136 0 640 11 PNW95-058 111 38 162 3638 0 534 10 PNW95-059 136 31 191 3317 814 1359 17 PNW95-060 165 34 143 2975 193 588 21 PNW95-061 131 31 209 3830 0 479 32 PNW95-062 115 21 123 2786 253 605 104 PNW95-063 157 34 179 2514 567 507 31 PNW95-064 177 34 123 2745 292 537 44 PNW95-065 203 37 188 2481 328 665 17 PNW95-066 199 42 167 3770 0 722 22 PNW95-067 166 37 143 3522 162 833 11 PNW95-068 231 38 318 2962 0 755 20 PNW95-049 207 41 199 3338 310 1225 18 PNW95-050 220 47 172 5123 0 657 0 PNW95-016 222 73 145 2358 169 481 46 PNW95-017 167 49 90 2555 281 516 27 PNW95-018 234 48 128 2944 472 634 90 13-9 128 28 148 2666 92 667 17 13-8 165 29 113 2854 0 964 18 13-7 123 23 153 2739 143 1370 0 13-6 176 37 102 2912 222 481 11 13-5 178 33 93 2932 338 840 33 13-4 122 22 186 2944 427 656 5 13-3 195 29 120 3258 123 608 4 13-2 149 29 118 2887 0 763 12 13-1 143 26 112 3135 0 933 40 5-19 139 24 155 3898 466 1138 19 5-18 108 21 111 1849 0 409 11 5-17 205 38 ISO 2778 230 624 22 5-16 226 37 145 2301 192 486 13 5-15 171 28 163 2377 0 485 16 5-14 185 32 164 1607 181 703 33 5-13 181 31 302 2141 0 727 18 5-12 168 32 216 1655 209 1017 50 5-11 157 31 161 2247 286 540 14

3 1 2 Table B3. (continued)

4.04 4.04 3.76 3.34 3.24 -3.20 3.03 Plag. Plag. Fsp Quartz Microcl. Plag. Calcite Sample I A III I A 5-10 224 41 208 2114 465 1009 22 5-9 113 13 158 2542 219 565 29 5-8b 237 43 138 2801 0 542 29 5-6 163 30 162 3405 0 631 15 5-5 264 40 230 2092 277 595 45 5-4 184 33 157 1536 350 934 45 5-2 160 34 176 1942 0 624 20 5-1 330 51 171 1731 432 905 20 8-5z 202 37 211 2534 0 1054 19 8-5y 143 27 98 2142 109 641 158 8-5x 180 40 170 1844 210 759 43 8-5w 126 28 179 2914 282 650 21 8-5u 347 60 125 1978 501 721 31 8-5t 144 33 114 1878 0 1859 38 8-5s 108 18 91 999 227 486 482 8-5r 196 36 234 2153 206 577 76 8-5q 226 46 194 1947 0 893 20 8-5 79 13 57 919 242 283 558 8-4 138 32 127 2301 233 694 30 8-3a/b 148 25 227 2708 308 683 26 8-2 169 33 181 2179 0 502 34 8-1 178 38 137 2112 193 1231 45 8-0 183 35 206 1282 0 1172 30 7-1 178 40 142 3250 136 677 11 7-4 278 45 187 1315 0 1410 56 Mt.Sirius top 101 21 101 3915 0 622 44 S-7 166 29 130 5840 271 3628 24 S-6 124 22 142 4751 498 1475 23 S-5 195 36 97 5952 356 1168 24 S-4 135 29 119 3535 240 574 8 S-3 253 32 110 4041 197 677 20 S-2 122 19 119 2876 2435 551 6 S-9 132 20 108 4630 2113 738 15 S-10 80 12 66 3606 344 407 171 S-12 96 18 91 3388 346 473 14 S-13 138 23 65 3516 246 569 36 S-14 135 23 130 5551 166 553 12 PNW95-002 107 19 131 2987 215 417 14 PNW95-003 160 27 191 7002 0 560 0 PNW95-004 158 25 122 7154 247 802 0 PNW95-005 130 19 148 6035 0 488 0 PNW95-006 150 26 138 4507 0 551 26 PNW95-012 304 53 154 2980 0 682 13

313 Table B3. (continued)

4.04 4.04 3.76 3.34 3.24 -3.20 3.03 Plag. Plag. Fsp Quartz Microcl. Plag. Calcite Sample I A I II I A PNW95-0I3 133 28 102 2416 193 570 21 PNW95-0I4 139 29 109 3839 217 625 14 PNW95-0I9 211 42 137 4570 0 863 18 PNW95-020 174 41 189 3489 137 736 15 PNW95-021 177 33 158 4111 289 919 13 PNW95-022 219 51 157 5700 182 946 23 PNW95-023 220 40 188 4798 366 916 23 PNW95-031 121 28 126 2084 192 402 0 PNW95-032 218 36 210 5047 264 906 69 PNW95-033 165 26 184 4591 0 667 19 PNW95-034 170 35 176 4439 212 4259 9 PNW95-035 168 30 136 3275 198 589 14 Prosmesa 257 55 366 1600 511 824 13 MtFeath 157 26 120 5014 332 754 0 TM-5 0 0 112 11526 0 396 70 TM-6 192 22 0 8395 178 358 0 TM-7a 0 0 0 8432 159 70 0 TM-7b 0 0 0 7951 187 149 9 MtFIem2 201 34 127 4326 268 629 10 MtFlem3 220 46 155 5149 195 1370 0 DVDPl 1-120 165 28 145 2179 631 1661 110 DVDPl 1-156 257 42 174 1740 532 1156 19 DVDPl 1-170 213 36 181 2847 653 1577 0 DVDPl 1-204 141 25 145 997 206 548 0 DVDPl 1-221 128 21 109 606 282 293 0 DVDPl 1-242 178 25 240 1388 746 919 0 DVDPl 1-253 181 31 184 1576 387 768 26 DVDPl 1-269 482 67 230 1408 410 616 100 DVDPl 1-291 189 33 132 1119 316 416 0 DVDPl 1-311 223 41 317 2346 370 1049 44 DVDPl 1-323 269 46 192 2277 820 1489 35 94TS33 193 31 135 6860 855 1181 0 94TS31 121 22 118 5976 247 929 0 94TS29 103 16 101 6066 422 696 0 94TS2S 117 16 88 5255 454 568 0 94TS27 115 17 110 7025 274 1431 0 94TS25 116 18 127 6491 467 700 0 94TS24 74 11 108 5181 195 409 0 94TS23 122 17 87 6786 626 387 38 94TS22 127 22 147 5255 183 1139 0 94TS12 83 9 85 5279 301 428 0 94TS11 128 17 129 6690 999 974 8 94TS14 146 20 78 6618 423 703 0 94TS19 65 11 75 3893 179 383 0

31 4 Table B3. (continued)

4.04 4.04 3.76 3.34 3.24 -3.20 3.03 Plag. Plag. Fsp Quartz Microcl. Plag. Calcite Sample IAI II I A 94TS18 156 21 85 5647 1199 728 0 94TS16 134 15 174 6366 2064 592 0 94TS08 76 11 99 5571 672 465 0 94TS06 84 12 78 3463 135 524 0 94TS05 0 0 87 5068 4889 902 0 94TS02 57 9 110 5183 1436 894 7 94TS01 66 9 94 6290 175 285 0 94QH46 141 29 120 2716 758 1039 5 94QH45 191 30 124 2217 453 1202 90 94QH44 643 87 222 4220 717 3247 17 94QH43 254 42 131 2171 275 453 4 94QH40 155 22 117 2243 278 601 0 94QH39 112 26 98 2512 293 661 3 94QH38 192 30 255 4075 860 913 10 94QH37 330 52 338 6639 565 890 0 94QH36 160 37 117 2800 1344 1236 26 94QH34 370 54 168 2916 558 872 31 94QH33 220 47 166 4359 450 2152 30 94QH31 527 63 128 3401 1731 1680 6 94QH30 440 74 127 2269 534 1085 9 94QH29 139 25 136 2385 390 553 5 94QH24 150 29 78 3087 339 625 13 94QH13 171 39 121 2764 690 1523 9

315 Table B3. (continued)

2.99 2-95 2.89 2.77 2.16 2.11 Cpx Cpx Cpx K-fsp. K-fsp. K-fsp Sample I I II II PNW95-070 338 194 204 31 53 57 PNW95-07I 130 70 107 0 0 0 PNW95-072 144 140 0 0 49 0 PNW95-073 110 0 0 0 35 36 PNW95-074 75 0 0 35 29 180 PNW95-053 357 160 133 0 0 37 PNW95-054 145 0 77 0 40 41 PNW95-055 156 111 117 0 41 48 PNW95-056 129 0 138 0 32 0 PNW95-057 139 0 48 0 54 32 PNW95-058 110 65 0 0 0 0 PNW95-059 151 155 86 0 32 0 PNW95-060 374 109 132 26 41 48 PNW95-06I 387 119 107 0 33 62 PNW95-062 133 50 71 0 36 37 PNW95-063 222 184 94 0 40 35 PNW95-064 145 97 98 31 38 37 PNW95-065 190 173 104 0 44 0 PNW95-066 162 0 108 11 38 0 PNW95-067 156 149 97 0 37 0 PNW95-068 178 163 47 0 32 28 PNW95-049 147 0 153 0 52 49 PNW95-050 147 0 85 0 45 0 PNW95-016 218 0 159 69 42 0 PNW95-017 186 142 155 37 47 57 PNW95-018 171 0 120 40 40 0 13-9 69 35 0 0 36 0 13-8 76 52 0 0 0 33 13-7 79 148 66 0 29 0 13-6 215 48 80 0 37 0 13-5 140 91 101 0 0 0 13-4 102 48 89 0 27 0 13-3 65 0 0 27 69 0 13-2 62 0 0 0 29 0 13-1 132 0 0 0 0 0 5-19 85 108 0 0 46 0 5-18 140 0 84 0 0 0 5-17 137 75 63 0 0 0 5-16 192 98 52 0 0 0 5-15 243 0 103 40 45 36 5-14 220 102 37 0 39 65 5-13 170 81 105 0 37 25 5-12 157 209 131 0 65 81 5-11 127 87 76 0 68 0

316 Table B3. (continued)

2.99 2.95 2.89 2.77 2.16 2.11 Cpx Cpx Cpx K-fsp. K-fsp. K-fsp, Sample I r r III 5-10 200 0 0 0 0 0 5-9 161 0 112 0 37 50 5-8b 261 96 142 0 39 0 5-6 0 0 0 32 29 0 5-5 165 0 138 20 22 48 5-4 159 136 103 41 71 38 5-2 260 141 104 0 67 0 5-1 226 137 118 0 37 50 8-5z 134 119 70 0 27 29 8-5y 109 47 46 0 0 40 8-5x 107 0 55 0 55 49 8-5 w 142 81 123 0 175 39 8-5u 368 131 75 0 63 45 8-5t 164 110 88 0 35 31 8-5s 130 142 46 0 22 0 8-5r 157 185 98 0 20 48 8-5q 202 136 139 23 66 58 8-5 0 66 64 0 0 0 8-4 185 93 111 0 44 39 8-3a/b 176 249 138 0 61 51 8-2 186 44 203 0 40 48 8-1 300 68 173 0 36 0 8-0 ISO 124 275 0 47 75 7-1 245 82 61 0 42 36 7-4 279 99 233 0 90 57 Mt.Sirius top 0 0 120 0 64 51 S-7 77 0 149 0 48 0 S-6 80 0 0 0 34 45 S-5 66 82 73 0 37 0 S-4 101 0 57 0 36 47 S-3 122 0 63 0 0 37 S-2 69 0 0 30 38 0 S-9 78 0 43 0 55 56 S-10 84 0 0 0 46 0 S-12 108 0 49 33 23 0 S-13 67 0 0 0 40 0 S-14 62 0 35 0 16 0 PNW95-002 103 66 78 0 31 0 PNW95-003 138 107 105 0 0 0 PNW95-004 0 0 43 0 0 0 PNW95-005 93 0 56 37 0 36 PNW95-006 98 0 44 46 0 0 PNW95-012 57 101 120 53 33 70

317 Table B3. (continued)

2.99 2.95 2.89 2.77 2.16 2.11 Cpx Cpx Cpx K-fsp. K-fsp. K-fsp. Sample I II III PNW95-013 97 0 0 36 30 0 PNW95-014 236 0 82 0 26 0 PNW95-019 111 0 50 0 57 43 PNW95-020 104 118 67 0 64 0 PNW95-021 90 88 67 0 37 0 PNW95-022 94 0 0 0 0 0 PNW95-023 137 0 74 0 0 0 PNW95-031 286 92 128 32 35 34 PNW95-032 107 50 50 0 43 51 PNW95-033 101 0 172 41 38 0 PNW95-034 0 0 93 0 48 30 PNW95-035 93 0 114 33 44 0 Prosmesa 156 104 168 44 70 128 MtFeath 101 79 0 0 36 0 TM-5 109 0 129 0 32 0 TM-6 91 284 154 0 38 0 TM-7a 74 0 89 0 0 0 TM-7b 84 0 129 0 106 0 MtFlem2 80 119 0 0 44 0 MtFIem3 84 108 0 0 45 0 DVDPLl-120 178 511 97 0 85 40 DVDP11-I56 252 90 230 0 93 0 DVDPIl-170 130 340 68 0 100 59 DVDPlI-204 124 89 55 33 50 42 DVDPll-221 97 109 68 27 36 38 DVDP11-242 126 110 57 30 61 57 DVDP11-253 202 114 82 44 55 58 DVDP 11-269 150 143 77 41 57 39 DVDPl 1-291 117 0 63 0 30 44 DVDP 11-311 152 187 90 33 36 59 DVDP 11-323 117 204 109 0 71 53 94TS33 108 0 87 0 68 0 94TS31 81 0 0 45 70 0 94TS29 61 0 0 36 0 0 94TS28 68 35 0 39 48 0 94TS27 0 48 0 0 36 0 94TS25 85 82 73 0 100 0 94TS24 60 0 0 0 37 0 94TS23 45 63 0 0 40 0 94TS22 74 0 0 45 52 0 94TS12 69 0 0 0 50 0 94TS11 97 0 73 35 125 0 94TS14 129 48 0 39 44 0 94TS19 47 0 0 0 67 0

318 Table B3. (continued)

2.99 2.95 2.89 2.77 2.16 2.11 Cpx Cpx Cpx K-fsp. K-fsp. K-fsp Sample I II I I I 94TS18 118 0 105 46 218 0 94TS16 71 0 0 0 81 52 94TS0S 75 0 0 38 44 0 94TS06 50 0 0 0 0 0 94TS05 58 0 0 0 128 0 94TS02 156 66 149 40 186 0 94TS0I 0 0 0 0 0 0 94QH46 51 64 0 0 187 0 94QH45 44 85 39 0 129 46 94QH44 0 72 0 0 109 94 94QH43 61 48 0 33 35 38 94QH40 43 43 0 28 55 0 94QH39 0 0 0 41 76 32 94QH38 0 30 0 0 89 69 94QH37 0 118 107 64 144 0 94QH36 104 88 79 0 82 44 94QH34 40 65 57 0 71 64 94QH33 63 108 112 0 116 0 94QH31 69 0 0 0 74 0 94QH30 92 59 78 0 54 0 94QH29 48 64 0 35 73 0 94QH24 69 0 0 59 206 0 94QH13 35 0 0 49 60 0

319 APPENDIX C

HEAVY MINERAL ANALYSIS

REPRODUCIBILITY

Heavy mineral analysis o f the fine to very fine sand fi-action is a particularly useful tool in provenance studies, because heavy minerals are resistant to glacial transport, and they are more diagnostic of a particular source rock than light minerals (Gwyn & Dreimanis, 1979). In a study o f heavy minerals in North-American tills Dreimanis and Vagners (1971) found that the terminal grades of heavy minerals in tills lie between 32 and 250 pm. Terminal grades are characteristic minimum grain-sizes for minerals at which they are resistant to further size reduction during long-distance transport. The fine sand and coarse silt fractions therefore comprise a sample of all rocks eroded by the glacier along its flow line. Studies on heavy minerals in tills fi’om the

Transantarctic Mountains, including the Sirius Group, have been carried out before: Faure et al.

(1995) found that the highest abundance of heavy minerals occurs in the 500 to 250 pm fiaction of tills fi"om the Transantarctic Mountains. However, identification of minerals using optical properties is virtually impossible for grains > 250 pm in diameter and the abundance of unidentifiable rock fragments is also greater. In addition grains > 250 :m may distort the process of heavy mineral separation due to interaction with the heavy liquid. The fraction 63-125 pm of some tills from the Transantarctic Mountains is virtually fi"ee o f heavy minerals (Faure et a i.

320 1995), but almost 20 % of all heavy minerals in the sand fraction occur in the 125 - 250 nm

fraction. Heavy mineral assemblages from Ontario in the size range 37 to 250 |im show no

significant difference in composition between three size classes; 37 - 63 pm, 63 — 125 pm, and

125 — 250 pm (Gwyn & Dreimanis, 1979). Therefore, the heavy minerals in the size range 125 —

250 pm of the Sirius Group were analyzed and the heavy mineral composition of the fine sand

fiaction was regarded as representative to determine the sand provenance of the deposits. A test

was carried out to determine how many grains should be counted to produce good results and

300-350 counts was regarded as sufBcient (Figure Cl).

Sample PNW95-057

♦ D iop sid e

X Augite ■ Epidote A Tremolite X Hypersthene • Silllmanite

0 100 200 300 400 500 number of grains counted

Figure Cl. From a test count on sample PNW95-057 it was concluded that 300-350 counts would be sufGcient to obtain reproducible data.

321 COMPLETE RESULTS

Sample PNW95 PNW95 PNW95 PNW95 PNW95 PNW95 PNW95 PNW95 PNW95 -002 -012 -013 -022 -023 -032 -053 -055 -057

Brown Augite 5 39 16 10 8 20 25 45 45 Green Augite 39 19 10 9 10 25 2 6 0 Diopside 7 114 158 124 153 116 138 108 111 Diallage 63 3 1 14 0 0 7 5 10 Hypersthene 3 27 17 7 15 12 30 16 19 Enstatite 5 1 2 0 3 0 10 0 1 Tremo 1 i te/actinc 1 ite 30 2 0 3 1 0 5 4 22 Anthophyllite 38 9 0 6 6 7 12 5 28 Anthophyllite (fibrous) 0 0 0 0 0 0 3 0 0 Green Hornblende I 0 1 0 0 0 0 0 0 Brown Hornblende 16 0 0 0 0 0 0 2 0 Amphiboles (other) 0 14 2 2 3 0 9 0 24 Olivine 0 0 0 0 0 0 0 0 0 Biotite 0 0 0 0 0 0 0 0 0 Chlorite I 0 0 0 0 0 0 0 0 Kyanite 0 0 0 0 0 0 0 0 2 Sillimanite 12 4 2 0 2 0 2 2 2 Staurolite 0 0 3 0 0 0 0 0 0 Epidote Group 70 29 21 53 22 57 31 29 29 Apatite 0 0 0 2 0 0 0 0 1 Monazite 1 1 0 0 0 0 0 0 1 Xenotime 0 0 0 0 0 0 0 0 0 Cassiterite 0 2 1 0 0 1 0 0 0 Tourmaline 0 0 1 0 0 0 0 0 0 Rutile 0 0 0 0 0 0 0 0 0 Zircon I 0 0 0 0 0 0 0 1 Clear garnet 0 1 1 2 1 0 0 0 1 Pink garnet 0 0 0 0 0 0 0 0 0 Dirty garnet 0 0 0 0 3 0 1 0 1 Sphene 0 0 0 1 0 1 0 0 0 Serpentine 0 0 0 0 0 0 0 0 0 Topaz 0 1 0 0 0 0 0 0 0 Beryl 0 0 0 0 0 0 0 0 0 Vesuvianite 0 0 0 0 0 0 0 0 0 unknown mineral 1 0 0 5 0 0 2 2 2 10 Opaque (non-magnetic) 5 4 8 12 7 6 4 1 5 Lithic fragments 21 52 27 14 30 23 58 69 49 Oxidized 34 31 48 44 49 37 41 51 63 Unidentified 4 4 4 4 3 5 5 5 6 grains counted 356 357 328 307 316 312 385 350 431 % unidentified 1 1 1 1 1 2 1 1.43 1 % heavies 13.8 17.1 12.7 5.4 7.4 6.9 22.2 25.1 13.6

Table C l. Complete results of heavy mineral counts. (continued)

322 Table C l. (continued)

Sample PNW95 PNW95 PNW95 PNW95 PNW95 PNW95 -059 -061 -063 -065 -066 -068 S-12 S-9 S-4

Brown Augite 50 25 46 21 23 70 41 2 3 36 Green Augite 6 3 4 9 0 10 0 1 3 Diopside 93 129 114 199 71 56 21 4 8 85 Diallage 8 18 0 8 48 0 0 0 0 Hypersthene 14 19 7 21 8 29 63 11 20 Enstatite 2 6 1 2 5 2 0 13 10 Tremolite/actinolite 2 3 3 3 13 1 0 13 0 Anthophyllite 4 4 7 7 20 0 8 11 1 Anthophyllite (fibrous) 1 I 0 0 9 10 0 0 0 Green Hornblende 0 0 0 0 0 1 1 0 0 Brown Hornblende 1 0 0 0 0 0 0 9 0 Amphiboles (other) 4 10 4 8 19 8 0 13 23 Olivine 0 0 0 0 0 8 0 0 5 Biotite 1 0 0 0 0 0 0 5 2 Chlorite 0 0 0 1 0 0 0 1 1 Kyanite 0 1 0 0 0 2 0 7 4 Sillimanite 3 4 1 0 7 1 0 20 10 Staurolite 0 0 0 0 0 0 0 0 0 Epidote Group 21 25 8 8 30 45 13 19 11 Apatite 0 0 0 0 0 0 4 3 1 Monazite 0 0 0 0 0 0 5 2 0 Xenotime 0 0 0 0 0 0 0 0 0 Cassiterite 0 0 0 0 0 0 2 2 1 Tourmaline 0 0 0 0 0 0 0 1 1 Rutile 0 0 0 0 0 0 2 0 2 Zircon 0 1 0 1 0 1 0 0 1 Clear garnet 0 0 0 1 1 0 39 5 0 64 Pink garnet 0 0 0 1 0 1 7 33 20 Dirty garnet 0 0 1 0 4 0 0 0 0 Sphene 0 0 0 0 0 0 0 0 0 Serpentine 0 0 0 0 0 0 2 2 0 Topaz 0 0 0 0 0 0 0 0 0 Beryl 0 0 0 0 0 0 0 3 1 Vesuvianite 0 0 0 0 0 0 0 0 0 unknown mineral 1 1 2 2 11 5 2 0 0 4 Opaque (non-magnetic) I 6 4 7 6 1 24 7 8 Lithic fragments 48 63 45 43 28 11 54 35 38 Oxidized 63 28 63 83 55 49 6 28 37 Unidentified 5 5 5 5 6 2 5 2 5 grains counted 328 353 315 439 358 310 297 36 2 394 % unidentified 1.52 I 1.59 1 2 1 2 1 1 % heavies 13 21 21.4 18.7 11.8 17.9 1.5 3.3 1.3

323 Table C l. (continued)

Sample S-6 13-2 13-4 13-8 5-2 5-4 5-6 5-8b 5-10

Brown Augite 27 31 83 11 23 84 24 19 64 Green Augite 0 0 0 1 2 0 1 3 0 Diopside 29 95 80 128 75 49 110 130 32 Diallage 39 0 0 0 1 0 10 11 0 Hypersthene 33 11 12 12 9 19 11 9 15 Enstatite 0 22 16 7 0 5 0 1 11 Tremolite/actinolite 13 2 0 10 0 0 2 3 0 Anthophyllite 21 0 0 8 8 0 21 16 0 Anthophyllite (fibrous) 0 0 0 0 0 0 0 0 0 Green Hornblende 0 0 0 0 0 0 0 0 0 Brown Hornblende 6 0 0 0 0 0 1 0 0 Amphiboles (other) 5 7 1 21 1 7 9 4 13 Olivine 0 0 5 0 0 0 0 0 0 Biotite 1 0 0 0 0 0 0 0 0 Chlorite 1 7 0 0 0 0 0 0 0 Kyanite 0 9 6 0 0 3 0 0 2 Sillimanite 10 4 2 2 2 3 5 4 7 Staurolite 0 0 0 0 0 2 0 0 2 Epidote Group 2 40 29 16 57 14 15 46 19 Apatite 5 0 1 0 0 0 3 0 0 Monazite 0 0 2 0 0 4 0 0 0 Xenotime 1 0 0 0 0 1 0 0 0 Cassiterite 0 1 0 0 0 1 0 0 0 Tourmaline 0 0 3 0 0 1 0 0 0 Rutile 1 0 0 0 0 0 0 0 0 Zircon 3 0 0 1 0 1 0 1 1 Clear garnet 72 3 2 5 0 1 5 0 1 Pink garnet 25 1 0 1 0 0 1 0 0 Dirty garnet 0 0 0 0 0 0 1 1 0 Sphene 0 0 0 0 0 1 0 0 0 Serpentine 0 0 0 0 0 0 0 0 0 Topaz 0 0 0 0 0 0 0 0 0 Beryl 0 0 0 0 0 0 6 0 0 Vesuvianite 0 0 0 0 0 0 0 0 0 unknown mineral 1 0 4 0 4 4 0 0 4 0 Opaque (non-magnetic) 15 7 2 4 5 0 6 4 11 Lithic fragments 19 29 24 84 89 79 52 79 66 Oxidized 39 46 58 58 81 31 74 70 37 Unidentified 4 5 5 4 4 7 9 4 6 grains counted 371 324 331 377 361 313 366 409 287 % unidentified 1 2 2 1 1.11 2 2.46 0.98 2 % heavies 2.1 8.2 10.1 4.3 17.3 25.7 7.1 7.1 8.7

32 4 Table C l. (continued)

Sample 5-12 5-14 5-18 95DMHDVDP DVDP DVDP DVDP DVDP -051 21 m 120 m 156 m 170 m 222 1

Brown Augite 14 30 31 47 31 23 32 15 25 Green Augite 1 4 0 0 2 2 0 0 11 Diopside 112 100 131 75 57 67 23 48 39 Diallage 0 1 0 0 0 2 0 0 0 Hypersthene 7 26 13 12 18 32 20 30 25 Enstatite 1 0 13 9 8 8 3 0 3 Tremolite/actinolite 2 3 3 2 0 0 0 4 5 Anthophyllite 14 17 23 11 1 3 0 6 9 Anthophyllite (fibrous) 0 0 0 0 0 1 0 3 0 Green Hornblende I 0 0 0 3 10 0 16 16 Brown Hornblende 0 0 1 5 2 22 0 21 36 Amphiboles (other) 6 6 0 0 2 20 81 37 29 Olivine 1 0 2 3 61 0 0 0 0 Biotite 0 0 0 0 1 6 3 3 13 Chlorite 0 0 0 0 0 3 0 3 0 Kyanite 0 0 5 2 0 0 0 0 0 Sillimanite 2 2 8 9 0 0 30 7 1 Staurolite 0 0 0 0 0 0 0 0 0 Epidote Group 33 36 23 26 3 11 1 1 7 Apatite 0 0 5 0 3 2 1 1 0 Monazite 0 0 1 1 0 0 0 0 0 Xenotime 0 0 0 0 0 0 0 0 0 Cassiterite 0 0 0 1 0 0 0 0 0 Tourmaline 0 0 0 0 12 19 10 14 18 Rutile 0 0 0 0 1 1 1 1 2 Zircon 1 1 0 1 1 2 1 3 4 Clear garnet 2 2 7 2 0 1 0 0 0 Pink garnet 0 0 0 0 0 0 0 0 0 Dirty garnet 0 0 0 0 0 0 0 2 0 Sphene 0 0 0 0 1 0 0 0 2 Serpentine 0 0 0 0 0 0 0 1 1 Topaz 0 0 0 0 1 0 1 0 0 Beryl 0 0 0 0 1 0 0 0 0 Vesuvianite 0 0 0 0 0 0 0 0 0 unknown mineral 1 0 2 0 0 0 0 0 0 1 Opaque (non-magnetic) 5 5 9 13 32 9 14 8 4 Lithic fragments 75 76 35 54 44 13 24 21 53 Oxidized 64 44 35 45 55 50 54 57 27 Unidentified 4 4 7 0 7 7 7 7 6 grains counted 345 359 352 318 347 314 306 309 337 % unidentified 1.16 1.11 2 0 2.02 2.23 2 2.27 1.78 % heavies 17.7 27.2 12.7 11.4 19 8 14 14 14

325 Table C l. (continued)

Sample DVDPDVDPDVDP Mount Mount Prospi 280 m 311 m 325 m Feather Flem#2 Mesa

Brown Augite 5 10 2 0 0 7 Green Augite 8 16 8 0 1 4 Diopside 75 84 93 1 2 132 Diallage 0 0 0 0 0 0 Hypersthene 19 55 38 2 1 37 Enstatite 0 28 38 0 0 12 Tremolite/actinolite 3 7 2 0 0 0 Anthophyllite 10 6 19 0 0 46 Anthophyllite (fibrous) 0 0 0 0 0 0 Green Hornblende 12 18 39 0 0 2 Brown Hornblende 22 26 10 0 0 0 Amphiboles (other) 8 15 7 0 0 5 Olivine 0 2 0 0 0 0 Biotite 16 8 3 2 1 0 Chlorite 1 4 0 1 0 0 Kyanite 0 0 0 0 0 0 Sillimanite 0 0 0 0 0 4 Staurolite 0 0 0 0 0 0 Epidote Group 5 8 7 17 4 10 Apatite 0 2 2 9 22 0 Monazite 0 0 1 0 0 0 Xenotime 0 1 0 0 0 0 Cassiterite 0 0 0 0 0 0 Tourmaline 7 2 0 0 2 1 Rutile 1 7 0 1 13 0 Zircon 3 4 10 6 6 0 Clear garnet 0 0 0 148 45 0 Pink garnet 0 0 0 40 10 0 Dirty garnet 0 0 1 3 0 0 Sphene 4 4 0 1 0 4 Serpentine 0 0 0 0 0 0 Topaz 0 0 0 0 0 1 Beryl 0 0 0 0 0 0 Vesuvianite 0 0 1 0 0 0 unknown mineral 1 0 2 0 0 0 0 Opaque (non-magnetic) 9 2 1 6 1 2 Lithic fragments 39 7 19 23 18 3 Oxidized 59 10 12 95 145 23 Unidentified 2 6 4 2 4 4 grains counted 308 334 317 357 275 297 % unidentified 0.65 1.80 1 1 1 1 % heavies 10 12 10 0.2 0.2 23.3

326 APPENDIX D

CHEMICAL ANALYSIS

ICP-OES ANALYSIS AND COMPLETE RESULTS

The method used here is Inductively Coupled Plasma-Optical Emission Spectroscopy

(ICP-OES). The analyses were carried out by XRAL Laboratories in Toronto, a certified commercial laboratory. For ICP-OES samples are prepared by fusion of 0.1 g of ground sample in lithium metaborate in a graphite crucible (ICP95 method of XRAL). The melt is dissolved in nitric acid and made up to volume. The working principle of the ICP method requires a liquid to be fed into a nebulizer and be sprayed into Argon gas, which carries it to an Argon plasma torch.

The high temperature of the torch (8000° C) causes sample ionisation. The atoms and ions contained in the plasma vapor are excited into a state of radiated light (photon) emission. The radiation emitted can be passed to the spectrometer optics via an optical fiber, where it is dispersed into its spectral components. From the specific wavelengths emitted by each element, the most suitable line for the application is measured by means of a detection device. The ICP-

OES at XRAL-Toronto is equipped with a CCD (charge coupled device). The wavelength is characteristic of the element, whereas the radiation intensity is proportional to the concentration of the element in the sample. XRAL runs standards, duplicates and blanks with each bacch to monitor accuracy, precision and contamination.

327 Sample Na Mg A1 Si P K Ca Ti Mn Fe % % % % % % % % % %

Detection Limit 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

PNW95-002 1.27 1.31 7.96 30.3 0.03 1.99 1.60 0.38 0.05 3.26 PNW95-012 1-41 1.43 7.81 30.8 0.03 1.81 2.51 0.41 0.07 3.92 PNW95-053 1.67 1.78 7.10 30.6 0.04 1.46 3.48 0.42 0.08 4.18 PNW95-055 1.67 1.63 6.93 30.4 0.04 1.43 3.25 0.47 0.08 4.43 PNW95-057 1.39 1.27 7.03 30.6 0.03 1.63 2.23 0.40 0.07 3.42 PNW95-059 1.40 1.35 6.81 30.9 0.03 1.53 2.48 0.40 0.07 3.46 PNW95-06I 1.36 1.61 7.31 28.8 0.04 1.56 3.03 0.44 0.08 4.41 PNW95-063 1.41 2.28 7.04 28.7 0.04 1.27 4.39 0.43 0.09 5.08 PNW95-066 1.62 1.17 6.14 31.7 0.04 1.41 2.33 0.43 0.06 3.05 PNW95-068 1.40 1.42 6.77 29.4 0.04 1.53 2.55 0.38 0.06 3.45 95DMH051 1.45 2.00 6.44 27.7 0.04 1.22 4.30 0.37 0.07 3.72 DR8-Unit2 1.42 1.69 6.60 27.6 0.05 1.28 3.79 0.37 0.07 3.44 DR5-4 1.26 2.71 7.05 26.4 0.04 1.23 4.94 0.33 0.08 4.50 DR5-9 1.29 1.47 6.90 27.7 0.03 1.51 2.85 0.35 0.06 3.61 DR5-11 1.35 1.39 7.18 27.5 0.04 1.63 2.86 0.35 0.06 3.25 DR5-I6 1.48 1.32 6.97 27.8 0.03 1.39 3.07 0.36 0.07 3.05 DR13-4 1.29 1.47 6.82 28.8 0.04 1.49 2.66 0.35 0.06 3.51 DR13-8 1.45 0.88 8.15 29.4 0.03 1.95 1.78 0.33 0.06 3.18 DR13-9 1.75 1.32 8.49 31.9 0.04 1.85 2.49 0.39 0.07 3.90 S-4 1.25 0.85 6.33 31.8 0.03 1.77 1.35 0.32 0.04 2.63 S-6 1.23 1.08 6.14 31.5 0.05 1.79 1.75 0.33 0.05 2.93 S-9 1.07 0.93 7.11 29.9 0.05 1.93 1.16 0.37 0.04 2.84 S-12 l.Ol 0.96 6.91 28.4 0.03 1.75 1.87 0.38 0.04 3.45 TM-5 0.45 1.25 2.52 36.6 0.01 0.39 2.19 0.15 0.03 1.94 TM-6 0.44 1.46 2.45 35.6 0.02 0.36 2.19 0.17 0.04 2.46 Mfeath-84 1.38 0.41 6.44 34.4 0.03 1.78 0.62 0.35 0.02 1.72 Mflem-83 1.59 0.43 7.24 32.3 0.03 2.13 0.70 0.31 0.02 1.84 94TS02 0.76 0.54 6.28 33.3 0.04 3.40 0.33 0.50 0.05 2.66 94TS18 0.81 0.48 7.37 32.7 0.05 3.70 0.28 0.33 0.03 2.34 94TS23 0.82 0.48 6.23 33.6 0.04 2.23 0.17 0.35 0.02 2.17 94QH13 1.54 1.43 8.00 28.0 0.12 3.48 1.90 0.53 0.08 4.17 94QH39 1.40 1.54 7.80 27.5 0.14 3.40 1.92 0.55 0.08 4.18

Duplicates PNW95-002 1.13 1.19 7.16 29.2 0.03 1.83 1.65 0.35 0.05 3.21 DR5-4 1.20 2.60 6.76 26.0 0.04 1.19 4.99 0.32 0.08 4.47 TM-6 0.44 1.44 2.40 36.6 0.01 0.36 2.17 0.17 0.04 2.44

Table Dl. Results of ICP-OES analysis for the Sirius Group, Transantarctic Mountains, major elements.

328 Sample Cr Sr Y Zr Nb Ba ppm ppm ppm ppm ppm ppm

Detection Limit 10 10 10 10 10 10

PNW95-002 41 160 20 186 11 483 PNW95-0L2 76 208 21 159 11 479 PNW95-053 65 190 21 144 0 420 PNW95-055 47 158 21 156 11 391 PNW95-057 52 168 21 166 0 443 PNW95-059 53 179 20 182 10 420 PNW95-061 43 169 21 162 11 413 PNW95-063 55 172 21 138 13 367 PNW95-066 44 214 19 158 12 443 PNW95-068 55 163 19 141 15 411 95DMH051 78 237 17 147 0 389 DR8-Unit2 66 273 18 136 10 431 DR5-4 110 219 16 109 0 355 DR5-9 49 226 18 144 0 449 DR5-11 46 168 20 149 12 458 DR5-16 53 230 18 141 0 458 DR13-4 53 188 18 171 0 408 DR13-8 32 199 23 174 0 500 DR13-9 70 212 20 180 0 469 S-4 39 162 18 207 0 454 S-6 52 167 17 184 0 505 S-9 46 130 20 173 0 500 S-12 32 191 21 170 0 418 TM-5 42 47 0 124 0 114 TM-6 48 41 0 114 0 99 Mfeath-84 40 158 19 218 0 464 Mflem-83 33 179 19 168 11 602 94TS02 25 71 41 432 13 453 94TSI8 18 68 38 197 0 480 94TS23 43 55 21 229 12 353 94QH13 35 216 38 243 24 365 94QH39 43 155 54 283 24 373

Duplicates PNW95-002 38 148 19 152 10 446 DR5-4 120 214 17 115 0 343 TM-6 47 40 0 98 11 97

Table D2. Results of ICP-OES analysis for the Sirius Group, Transantarctic Mountains, trace elements.

329 Na Mg A1 Si P K Ca Ti Fe Mn % % % %% % % % %%

Detection Limit 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

1-025 0.61 0.7 8.01 26 0.06 2.66 0.35 0.46 4.58 0.33 6-014 0.76 0.72 7.25 28.1 0.05 2.61 0.3 0.44 3.84 0.04 8-001 0.75 0.57 6.45 31.5 0.05 2.87 0.46 0.4 2.73 0.06 1-006 1.13 0.91 9.33 24.1 0.06 3.16 1.67 0.5 3.98 0.04 9-001 0.5 0.66 7.97 28 0.06 3.24 0.43 0.53 3.45 0.05 3AA-027 0.34 0.4 7.75 30.8 0.03 3.14 0.22 0.46 1.48 -0.01 3AA-026 0.3 0.32 7.78 29.5 0.03 3.12 0.31 0.45 1.09 -0.01 3 AD-4 0.49 0.78 8.56 29.2 0.07 3.2 1.11 0.52 3.6 0.05 6-033 0.9 0.9 9.15 27.5 0.06 3.14 0.33 0.52 3.73 0.04 lOB-001 0.42 0.67 7.71 26.9 0.05 2.81 0.35 0.47 2.58 0.03 6-004 0.93 0.9 8.73 27.4 0.06 3.01 0.41 0.52 3.63 0.03 1-020 0.73 0.88 9.34 24.7 0.06 3.16 3.54 0.52 4.59 0.06 1-009 0.76 0.92 10.4 24.9 0.07 3.21 0.23 0.56 3.89 0.03 1-011 0.77 0.63 7.23 27.1 0.04 2.96 1.69 0.35 2.81 0.04 3AB-12 0.48 0.92 10.7 28.6 0.06 3.5 0.42 0.56 4.19 0.06 3AB-025 0.41 0.48 6.81 33.6 0.02 2.99 0.18 0.37 1.44 0.01 1-016 0.73 0.79 8.73 23.9 0.05 3.03 1.76 0.48 3.98 0.04 7-001 3.4 0.92 8.2 31.1 0.05 3.14 0.46 0.49 3.53 0.04 3AB-18 0.4 0.92 9.92 24.3 0.05 3.01 0.31 0.52 3.57 0.04 lOB-002 0.5 0.88 9.24 28.9 0.05 3.21 0.36 0.56 3.14 0.04

Duplicates 1-025 0.66 0.74 8.65 29.4 0.07 3.06 0.38 0.49 4.95 0.36 1-009 0.74 0.91 10.3 24.7 0.07 3.19 0.23 0.54 3.89 0.02

Table D3. Results of ICP-OES analysis for the Pagodroma Group, Prince Charles Mountains, major elements.

330 Cr Sr Y Zr Nb Ba ppm ppm ppm ppm ppm ppm

Detection Limit 10 10 10 10 10 10

1-025 77 154 29 207 -10 752 6-014 79 84 32 217 21 534 8-001 57 118 36 311 17 688 1-006 97 275 32 197 17 695 9-001 87 120 37 317 13 775 3AA-027 71 124 31 247 32 778 3AA-026 90 125 26 261 15 795 3AD-4 76 209 39 260 19 816 6-033 89 104 32 188 -10 682 lOB-001 85 139 28 279 29 735 6-004 89 116 41 235 19 829 1-020 104 412 33 243 15 798 1-009 94 159 32 197 16 663 1-011 72 159 23 141 -10 624 3AB-12 96 125 45 186 20 1130 3AB-025 50 132 33 212 -10 704 1-016 79 209 32 232 34 652 7-001 59 104 35 219 -10 659 3AB-18 101 105 33 149 31 717 lOB-002 84 162 36 319 -10 852

Duplicates 1-025 73 166 34 229 -10 811 1-009 101 150 35 196 17 661

Table D4. Results of ICP-OES analysis for the Pagodroma Group, Prince Charles Mountains, trace elements.

331 Major elements

Na Mg AI Si P K Ca Ti Fe Mn % % %% % % % % % %

Detection Limit 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

D-1 0.47 0.92 2.93 22.9 0.06 0.94 14.7 0.17 2.2 0.3

Trace elements

Cr Sr Y Zr Nb Ba ppm ppm ppm ppm ppm ppm

Detection Limit 10 10 10 10 10 10

D-1 19 216 14 168 -10 323

Table D5. Results of ICP-OES analysis for Eocene erratic D-I from the Mount Discovery coastal moraine.

CARBONATE ANALYSIS METHODS AND COMPLETE RESULTS

Carbonate analysis was aimed at estimating the amount of Ca-carbonate, in order to correct the values of the Chemical Index of Alteration based on whole-rock analysis, by eliminating Ca present in carbonate from the equation. The method used here is based on the emission of COt during reaction of carbonates with acid. For the procedure, samples are ground in a ball mill and 2-5 grams of sample are added to 250-ml Erlenmeyer flasks. Glass tubes with

10 % HCl are placed inside the Erlenmeyer flasks and the entire combination of flask, tube with acid and ground sample is weighed again. After determining the initial weight, the acid is allowed to flow out of the tube onto the sample and a reaction begins if carbonate is present. After the reaction with HCl the flasks are weighed again and the weight loss is regarded as the amount of

332 COt emitted. A blank consisting of an Erlenmeyer flask with only acid and no sample was run with each batch and used to correct for emission of other gases than CO?

To determine the appropriate reaction time ca. 600 mg of pure CaCOs was added to an

Erlenmeyer flask and analyzed using the method described above. The results of the calibration are presented in figure D l. Based on the calibration it was determined that three hours reaction time were sufficient to allow for the dissolution of nearly all calcium carbonate present in the samples. Longer reaction times will results in the dissolution of other carbonate phases, which do not contain much Ca, such as dolomite and siderite. Since the major objective here is to determine the amount of Ca present in carbonate, longer reaction times are not appropriate since they would allow low Ca-carbonate phases to contribute to the CO% emission.

120 100 O Ü Ü

20

0 5 10 15 20 Time (hours)

Figure Dl. Calibration curve for the reaction time of CaCOs in contact with 10% HCl.

Assuming that the CO 2 emission can be entirely attributed to the dissolution of calcium carbonate, wt. % CaCOs can be calculated using the following formula:

wt.% C aC O s = % weight lossCO2 * 100/44

333 Sample %COz % CaCO;

PNW95-002 0.1 0.2 PNW95-012 0.0 0.1 PNW95-053 0.2 0.4 PNW95-053 0.2 0.4 PNW95-055 0.1 0.1 PNW95-057 0.6 1.3 PNW95-059 0.3 0.8 PNW95-061 0.1 0.2 PNW95-063 0.6 1.4 PNW95-066 0.3 0.8 PNW95-068 0.3 0.7 95DMH051 0.8 1.8 DR8-Unit2 1.1 2.6 DR5-4 0.6 1.4 DR5-9 0.1 0.2 DR5-11 0.1 0.3 DR5-16 0.3 0.6 DR13-4 0.5 1.0 DR13-8 0.5 1.2 DR13-9 0.5 1.0 S-4 0.0 0.1 S-6 0.5 1.2 S-9 0.3 0.7 S-12 0.3 0.8 TM-5 0.4 0.9 TM-6 0.1 0.1 MFeath-84 0.3 0.6 MFIem-83 0.0 0.0 94TS02 0.0 0.0 94TSIS 0.0 0.0 94TS23 0.0 0.0 94QHI3 0.1 0.2 94QH39 0.1 0.2

Duplicate:

DR5-9 0.2 0.5

Table D6. Results of carbonate analysis of the Sirius Group.

334 Sample %COz % CaC03

1-025 0.4 1.0 6-014 0.1 0.2 8-001 0.1 0.2 1-006 1.2 2.7 9-001 0.2 0.5 3AA-027 0.0 0.0 3AA-026 0.1 0.2 3AD-4 0.8 1.9 6-033 0.2 0.4 lOB-001 0.3 0.7 6-004 0.1 0.2 1-020 3.0 6.7 1-009 0.1 0.2 1-011 0.5 1.2 3AB-12 0.3 0.7 3AB-025 0.1 0.2 1-016 0.9 2.1 7-001 0.5 1.0 3AB-18 0.2 0.5 lOB-002 0.3 0.7 D-1 15.5 35.3

Duplicate:

1-020 2.9 6.5

Table D7. Results carbonate analysis of the Bardin Bluffs Formation, and D-1 (Eocene erratic).

335 BIBLIOGRAPHY

Abreu, V.S., Anderson, J.B., 1998. Glacial eustasy during the Cenozoic: sequence stratigraphie implications. AAPG Bulletin, 82(7); 1385-1400.

Alley, R.B., Blankenship, D.D., Bentley, C.R., Rooney, S.T., 1986. Deformation of till beneath B, . Nature, 322: 57-59.

Alley, R.B., Guffey, K.M., Evenson, E.B., Straaser, J.C., Lawson, D.E., Larson, G.J., 1997. How glaciers entrain and transport basal sediment: physical constraints. Quaternary Science Reviews, 16: 1017-1038.

Anderson, S.P., Drever, J.I., Humphrey, N.F., 1997. Chemical weathering in glacial environments. Geology, 25(5): 399-402.

Anderson, S.P., Drever, J.I., , C D., Holden, P., 2000. Chemical weathering in the foreland of a retreating glacier. Geochimica et Cosmochimica Acta, 64(7): 1173-1189.

Askin, R.A., 1998. Palynological investigations of Mount Feather Sirius Group samples: recycled Triassic assemblages. In: G.S. Wilson and J.A. Barron (Editors), Mount Feather Sirius Group Core Workshop and Collaborative Sample Analysis. Report No. 14, Byrd Polar Research Center, Columbus, Ohio, pp. 59-65.

Askin, R.A., Markgraf, V., 1986. Palynomorphs from the Sirius Formation, Dominion Range, Antarctica. Antarctic Journal of the United States, 21(5): 34-35.

Baker, J.C., Fielding, C.R., 1998. Diagenesis of glaciomarine Miocene strata in CRP-1, Antarctica. Terra Antartica, 5(3): 647-653.

Bardin, V.I., 1982. Composition of East Antarctic moraines and some problems of Cenozoic history. In: C. Craddock (Editor), Third Symposium on Antarctic Geology and Geophysics. Madison, Wisconsin, U.S.A., The University of Wisconsin Press, pp. 1069- 1076.

Barker, P.F., Barrett, P.J., Camerlenghi, A., Cooper, A.K., Davey, F.J., Domack, E.W., Escutia, C., Kristoffersen, Y., O'Brien, P.E., 1998. Ice sheet history from Antarctic Continental margin sediments: the ANTOSTRAT approach. Terra Antartica, 5(4): 737-760.

Barrett, P.J., 1966. Petrology of some Beacon rocks between the Axel Heiberg and Shackleton Glaciers, Queen Maud Range, Antarctica. Journal of Sedimentary Petrology, 36(3): 794- 805.

336 Barrett, P.J., 1969. Photomicrographs of some sedimentary and volcaniclastic Permian and Triassic Beacon rocks from the Beardmore Glacier area, Antarctica. Report No. 31, Institute of Polar Studies, Columbus, Ohio.

Barrett, P.J., 1972. Stratigraphy and petrology of the mainly fluviatile Permian and Triassic part of the Beacon Supergroup, Beardmore Glacier area. In: R.J. Adie (Editor), Antarctic Geology and Geophysics. Oslo, Norway, Universitetsforlaget. pp. 365-372.

Barrett, P.J. (Editor), 1986. Antarctic Cenozoic history from the MSSTS-1 drillhole, McMurdo Sound, DSIR Bulletin, Volume 237, Wellington.

Barrett, P.J. (Editor), 1989. Antarctic Cenozic history from the CIROS-1 drilhole, McMurdo Sound, DSIR Bulletin, Volume 245, Wellington.

Barrett, P.J., 1991. The Devonian to Triassic Beacon Supergroup of the Transantarctic Mountains and correlatives in other parts of Antarctica. In: R.J. Tingey (Editor), The geology of Antarctica. Clarendon Press Oxford Monographs on Geology and Geophysics, Volume 17, Oxford, pp. 120-152.

Barrett, P.J., 1997. Cape Roberts Project Science Plan: Coring for Antarctic Tectonic and Climate History. Report No. 20, Antarctic Data Series, Antarctic Research Center, School of Earth Sciences, Victoria University of Wellington, Wellington, New Zealand.

Barrett, P.J., 1998. Grain size analysis of diamicton samples from Mount Feather. In: G. Wilson and J. Barron (Editors), Mount Feather Sirius Group Core Workshop and Collaborative Sample Analysis. Report No. 14, Byrd Polar Research Center, Columbus, Ohio, pp. 66- 68 .

Barrett, P.J., 1999. Antarctic climate history over the last 100 Million years. In: P.J. Barrett and G. Orombelli (Editors), Geological Records of Global and Planetary Changes. Siena, Terra Antarctica Reports, 3, Terra Antartica Publication, pp. 53-72.

Barrett, P.J., Elliot, D.H., 1972. The Early Mesozoic volcaniclastic Prebble Formation, Beardmore Glacier area. In: R.J. Adie (Editor), Antarctic Geology and Geophysics. Oslo, Norway, Universitetsforlaget. pp. 403-409.

Barrett, P.J., Powell, R.D., 1982. Middle Cenozoic glacial beds at Table Mountain, Southern Victoria Land. In: C. Craddock (Editor), Third Symposium on Antarctic Geology and Geophysics. Madison, Wisconsin, U.S.A., The University of Wisconsin Press, pp. 1059- 1067.

Barrett, P.J., McKelvey, B.C., 1989. Stratigraphy. In: P.J. Barrett (Editor), Antarctic Cenozoic history from the CIROS-1 drilhole, McMurdo Sound. DSIR Bulletin, Volume 245, Wellington, pp. 9-35.

Barrett, P.J., Adams, C.J., McIntosh, W.C., Swisher, C.C., Wilson, G.S., 1992. Geochronological evidence supporting Antarctic deglaciation three million years ago. Nature, 359: 816-818.

Barrett, P.J., Bleakley, N.L., Dickinson, W.W., Hannah, M.J., Harper, M.A., 1997. Distribution of siliceous microfossils on Mount Feather, Antarctica, and the age of the Sirius Group.

337 In: C.A. Ricci (Editor), VU International Symposium on Antarctic Earth Sciences, Siena, Italy. The Antarctic Region: Geological Evolution and Processes, Terra Antartica Publication, pp. 763-770.

Bart, P.J., deBatist, M., Jokat, W., 1999. Interglacial collapse of Crary Trough Mouth fan, Weddell Sea, Antarctica; implications for Antarctic glacial history. Journal of Sedimentary Research, 69(6): I276-I289.

Bart, P.J., Anderson, J.B., Trincardi, P., Shipp, S.S., 2000. Seismic data from the Northern basin, Ross Sea, record extreme expansions of the East Antarctic Ice Sheet during the late Neogene. Marine Geology, 166: 31-50.

Bartek, L.R., Henrys, S. A., Anderson, J.B., Barrett, P.J., 1996. Seismic stratigraphy of McMurdo Sound, Antartica: implications for glacially influenced early Cenozoic eustatic change? Marine Geology, 130: 79-98.

Behrendt, J.C., Cooper, A., 1991. Evidence of rapid uplift of the shoulder escarpment of the Cenozoic West Antarctic rift system and a speculation on possible climate forcing. Geology, 19: 315-319.

Bellanca, A., Neri, R., Palumbo, B., 1997. Geochemical features of Cenozoic sediments from CIROS-1 drillhole, McMurdo Sound, Antarctica: implications for provenance and diagenesis. Terra Antartica, 4(2): 127-132.

Berggren, W.A., Kent, D.V., Flynn, J.J., Couvering, J.A.V., 1985. Cenozoic geochronology. Geological Society of America Bulletin, 96: I407-14I8.

Bergman, L., 1982. Clastic dykes in the Aland Islands, SW Finland and their origin. Report No. 317, Geological Survey of Finland.

Billups, K., Ravelo, A.C., Zachos, J.C., 1998. Early Pliocene deep water circulation in the western equatorial Atlantic: Implications for high-latitude climate change. Paleoceanography, 13(1): 84-95.

Blatt, H., 1992. Sedimentary Petrology. W.H. Freeman and Company, New York.

Bohaty, S.M., Harwood, D.M., 1998. Southern Ocean paleotemperature variation from high- resolution silicoflagellate biostratigraphy. Marine Micropaleontology 33 (3-4): 241-272.

Bohrmann, G., Ehrmann, W., 1991. Analysis of sedimentary facies using bulk mineralogical characteristics of Cretaceous to Quaternary sediments from the Kerguelen Plateau: Sites 737, 738, and 744. In: J. Barron and B. Larsen (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 119, College Station TX, pp. 211-223.

Bohrmann, G., Greinert, J., Suess, E., Torres, M., 1998. Authigenic carbonates from the Cascadia subduction zone and their relation to gas hydrate stability. Geology, 26(7): 647-650.

Boulton, G.S., 1972. Modem Arctic glaciers as depositional models for former ice sheets. Journal of the Geological Society of London, 128: 361-393.

338 Boulton, G.S., 1978. Boulder shapes and grain-size distributions of debris as indicators of transport paths through a glacier and till genesis. Sedimentology, 25: 773-799.

Boulton, G.S., 1996. Theory of glacial erosion, transport and deposition as a consequence of subglacial sediment deformation. Journal of Glaciology, 42: 43-62.

Boulton, G.S., Hindmarsh, R.C.A., 1987. Sediment deformation beneath glaciers: rheology and geological consequences. Journal of Geophysical Research, 92(B9): 9059-9082.

Boulton, G.S., Caban, P., 1995. Groundwater flow beneath ice-sheets; Part H-Its impact on glacier tectonic structures and moraine formation. Quaternary Science Reviews, 14: 563- 587.

Brady, H.T., 1982. Late Cenozoic history of Taylor and Wright Valleys and McMurdo Sound inferred from diatoms in Dry Valleys Drilling Project cores. In: C. Craddock (Editor), Third Symposium on Antarctic Geology and Geophysics. Madison, Wisconsin, U.S.A., The University of Wisconsin Press, pp. 1123-1131.

Brady, H., McKelvey, B., 1979. The interpretation of a Tertiary tillite at Mount Feather, Southern Victoria Land, Antarctica. Journal of Glaciology, 22(86): 189-193.

Brady, H., McKelvey, B., 1983. Some aspects of the Cenozoic glaciation of Southern Victoria Land, Antarctica. Journal of Glaciology, 29(102): 343-349.

Brennand, T.A., Sharpe, D.R., 1993. Ice-sheet dynamics and subglacial meltwater regime inferred from form and sedimentology of glaciofluvial systems: Victoria Island, District of Franklin, Northwest Territories. Canadian Journal of Earth Sciences, 30: 928-944.

Breza, J.R., 1992. High-resolution study of Neogene ice-rafted debris. Site 751, southern Kerguelen Plateau. In: S.W. Wise and R. Schlich (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 120, College Station TX, pp. 207-221.

Bridle, I.M., Robinson, P.H., 1989. Diagenesis. In: P.J. Barrett (Editor), Antarctic Cenozioc history from the CIROS-1 drilhole, McMurdo Sound. DSIR Bulletin, Volume 245, Wellington, pp. 201-207.

Brodzikowski, K., Van Loon, A.J., 1985. Penecontemporaneous non-tectonic brecciation of unconsolidated silts and muds. Sedimentary Geology, 41: 269-282.

Burckle, L.H., Potter, N., 1996. Plio-PIeistocene diatoms in Paleozoic and Mesozoic sedimentary and igneous rocks from Antarctica: A Sirius problem solved. Geology, 24: 235-238.

Burckle, L.H., Mortlock, R., Rudolph, S., 1996. No evidence for extreme, long term warming in early Pliocene sediments of the Southern Ocean. Marine Micropaleontology, 27: 215- 226.

Burckle, L.H., Kellogg, D.E., Kellogg, T.B., Fastook, J.L., 1997. A mechanism for emplacement and concentration of diatoms in glacimarine deposits. Boreas, 26: 55-60.

339 Campbell, I.B., Claridge, G.G.C., 1987. Antarctica: soils, weathering processes and environment. Developments in Soil Science, 16. Elsevier Science Publishers B.V., Amsterdam.

Cape Roberts Science Team, 1998. Initial Report on CRP-1, Cape Roberts Project, Antarctica. Terra Antartica, 5(1): 1-187.

Cape Roberts Science Team, 1999. Studies from the Cape Roberts Project, Ross Sea, Antarctica - Initial Report on CRP-2/2A. Terra Antartica, 6(1/2): 1-173.

Carol, H., 1943. Beobachtungen zur Entstehung der Rundhoker (Observations on the development of whaleback forms). Die Alpen, 19: 173-180.

Christiansen, E.A., 1971. Tills in southern Saskatchewan, Canada. In: R.P. Goldthwait (Editor), Till / a symposium. Ohio State University Press, Columbus, pp. 167-183.

Church, M., Gilbert, R., 1975. Proglacial fluvial and lacustrine environments. SEPM Special Publication, 23: 22-100.

Claps, M., Coren, P., DeSantis, L., Masetti, D., Sarti, M., 1997. Deep-water deposits from CIROS-1 drillhole, McMurdo Sound, and implications for Ross Sea late Eocene-early Oligocene paleogeography (western Ross Sea, Antarctica). Terra Antartica, 4(2): 133- 147.

Claridge, G.G.C., Campbell, I B., 1968. Soils of the Shackleton Glacier region. Queen Maud Range, Antarctica. New Zealand Journal of Science, 11(2): 171-218.

Cooper, A.K., Webb, P.-N., 1994. The ANTOSTRAT project: an international effort to investigate Cenozoic glacial history, climates and sea-level. Terra Antartica, 1(2): 239- 242.

Cooper, A., Eittreim, S., tenBrink, U., Zayatz, I., 1993. Cenozoic glacial sequences of the Antarctic continental margin as recorders of Antarctic ice-sheet fluctuations. In: J.P. Kennett and D.A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change Part Two. American Geophysical Union, Antarctic Research Series, Volume 60, Washington D C., pp. 75-89.

Crowley, T.J., 1991. Modeling Pliocene warmth. Quaternary Science Reviews, 10: 272-282.

Crowley, T.J., 1996. Pliocene climates: the nature of the problem. Marine Micropaleontology, 27: 3-12.

Denton, G.H., Prentice, M.L., Kellogg, D.E., Kellogg, T.B., 1984. Late Tertiary history of the Antarctic ice sheet: evidence from the Dry Valleys. Geology, 12: 263-267.

Denton, G.H., Prentice, M.L., Burckle, L.H., 1991. Cainozoic history of the Antarctic ice sheet. In: R.J. Tingey (Editor), The geology of Antarctica. Clarendon Press Oxford Monographs on Geology and Geophysics, Volume 17, Oxford, pp. 365-433.

Denton, G.H., Sugden, D.E., Marchant, D R., Hall, B.L., Wilch, T.I., 1993. East Antarctic ice sheet sensitivity to Pliocene climate change from a Dry Valleys perspective. Geografiska Annaler, 75 A(4): 155-204.

340 DeSantis, L., Barrett, P.J., 1998. Grain size analysis of samples from CRP-1. Terra Antartica, 5(3): 375-382.

DeSantis, L., Brancolini, G., Busetti, M., Marchetti, A., 1997. Seismic sequences and Late Cenozoic glacial history in the Ross Sea. In: C.A. Ricci (Editor), VII International Symposium on Antarctic Earth Sciences. Siena, The Antarctic Region: Geological Evolution and Processes, Terra Antartica Publication, pp. 781-790.

Dickinson, W., 1998. Modal analysis of COMRAC samples - Mount Feather. In: G.S. Wilson and J.A. Barron (Editors), Mount Feather Sirius Group Core Workshop and Collaborative Sample Analysis. Report No. 14, Byrd Polar Research Center, Columbus, Ohio, pp.74- 76.

Dickinson, W.W., Grapes, R.H., 1997. Authigenic chabazite and implications for weathering in Sirius Group diamictite. Table Mountain, dry valleys, Antarctica. Journal of Sedimentary Research, 67(5): 815-820.

Dietrich, H.G., Klosa, D., 1998. Carbonate contents in CRP-1 samples-initial results. Terra Antartica, 5(3): 645-646.

Dionne, J.-C., Shilts, W.W., 1974. A Pleistocene clastic dike. Upper Chaudiere Valley, Quebec. Canadian Journal of Earth Sciences, 11: 1594-1605.

Dowsett, H., Barron, J., Poore, R., 1996. Middle Pliocene sea surface temperatures: a global reconstruction. Marine Micropaleontology, 27: 13-25.

Dowsett, H.J., Cronin, T.M., Poore, R.Z., Thompson, R.S., Whatley, R.C., Wood, A.M., 1992. Micropaleontological evidence for increased meridional heat transport in the North Atlantic Ocean during the Pliocene. Science, 258: 1133-1135.

Dreimanis, A., 1988. Tills: Their genetic terminology and classification. In: R.P. Goldthwait and C.L. Matsch (Editors), Genetic classification of glacigenic deposits. A.A. Balkema, Rotterdam, pp. 17-83.

Dreimanis. A., 1993. Small to medium-sized glacitectonic structures in till and in its substratum and their comparison with mass movement structures. Quaternary International, 18: 69- 79.

Dreimanis, A., Vagners, U.J., 1971. Bimodal distribution of rock and mineral fragments in basal tills. In: R.P. Goldthwait (Editor), Till / a symposium. Ohio State University Press, Columbus, pp. 237-250.

Dreimanis, A., Vagners, U.J., 1972. The effect of lithology upon the texture of till. In: E. Yatsu and A. Falconer (Editors), Research methods in Pleistocene geomorphology. Geo Abstracts Ltd., pp. 66-82.

Dreimanis, A., Rappol, M., 1997. Late Wisconsinan sub-glacial clastic intrusive sheets along Lake Erie bluffs, at Bradtville, Ontario, Canada. Sedimentary Geology, 111: 225-248.

Echelmeyer, K., Wang, Z., 1987. Direct observations of and deformation of basal drift at sub-freezing temperatures. Journal of Glaciology, 33(113): 83-98.

341 Eggers, A.J., 1979. Scallop Hill Formation, Brown Peninsula, McMurdo Sound, Antarctica. New Zealand Journal of Geology and Geophysics, 22(3): 353-361.

Ehlers, J., Grube, P., 1983. Meltwater deposits in north-west Germany. In: J. Ehlers (Editor), Glacial deposits in North-West Europe. A. A. Balkema, Rotterdam, pp. 249-256.

Ehlers, J., Stephan, H.-J., 1983. Till fabric and ice-movement. In: J. Ehlers (Editor), Glacial deposits in North-West Europe. A.A. Balkema, Rotterdam, pp. 267-274.

Ehrmann, W.U., 1991. Implications of sediment composition on the southern Kerguelen Plateau for paleoclimate and depositional environment. In: J. Barron and B. Larsen (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 119, College Station TX, pp. 185-210.

Ehrmann, W.U., 1997. Smectite concentrations and crystallinities: indications for Eocene age of glaciomarine sediments in the CIROS-1 drill hole, McMurdo Sound, Antarctica. In: C.A. Ricci (Editor), VU International Symposium on Antarctic Earth Sciences, Siena. The Antarctic Region: Geological Evolution and Processes, Terra Antartica Publication, pp. 771-780.

Ehrmann, W., Polozek, K., 1999. The heavy mineral record in the Pliocene to Quaternary sediments of the CIROS-2 drill core, McMurdo Sound, Antarctica. Sedimentary Geology, 128: 223-244.

Elliot, D.H., 1970. Mesozoic Volcanic Rocks of the Beardmore Glacier area, Antarctica, Second Gondwana Symposium. Pretoria, South Africa, Council for Scientific and Industrial Research, Scientia. pp. 421-431.

Elliot, D.H., 1972. Major oxide chemistry of the Kirkpatrick Basalt, Central Transantarctic Mountains. In: R.J. Adie (Editor), Antarctic Geology and Geophysics. Oslo, Norway, Universitetsforlaget. pp. 413-418.

Elliot, D.H., Fleming, T.H., Larsen, D., Buchanan, D.M., 1986. Ferrar and uppermost Beacon rocks, Beardmore Glacier region. Antarctic Journal of the United States, 21(5): 22-23.

Elliot, D.H., Fleming, T.H., Gero, C.M., 1996. Geologic studies on rocks from the Jurassic Ferrar Group. Antarctic Journal of the United States, 31(2): 37-38.

Evans, D.J.A., 1989. Apron entrainment at the margins of sub-polar glaciers, north-west Ellesmere Island, Canadian high arctic. Journal of Glaciology, 35(121): 317-323.

Eyles, N., Eyles, C.H., Miall, A.D., 1983. Lithofacies types and vertical profile models: an alternative approach to the description and environmental interpretation of glacial diamict and diamictite sequences. Sedimentology, 30: 393-410.

Faure, G., Taylor, K.S., 1981. Provenance of some glacial deposits in the Transantarctic Mountains based on Rb-Sr dating of feldspars. Chemical Geology, 32: 271-290.

Faure, G., Bowman, J.R., Elliot, D.H., Jones, L.M., 1974. Strontium isotope composition and petrogenesis of the Kirkpatrick basalt. Queen Alexandra Range, Antarctica. Contributions to Mineralogy and Petrology, 48: 153-169.

342 Faure, G., Taylor, K.S., Mercer, J.H., 1983. Rb-Sr provenance dates of feldspar in glacial deposits of the , Transantarctic Mountains. Geological Society of America Bulletin, 94: 1275-1280.

Faure, G., Hagen, E.H., Johnson, K.S., Strobel, M.L., Whiting, K.S., 1995. Geological exploration of East-Antarctica: iron, manganese, and titanium in the heavy mineral fractions of till in the Transantarctic Mountains. In: D.H. Elliot (Editor), Contributions to Antarctic Research IV. American Geophysical Union, Antarctic Research Series, Volume 67, Washington D C., pp. 19-31.

Feeser, V., 1988. On the mechanics of glaciotectonic contortion of clays. In: D.G. Croot (Editor), Glaciotectonics: Forms and Processes. A.A. Balkema, Rotterdam, pp. 63-76.

Fielding, C.R., Woolfe, K.J., Purdon, R.G., Lavelle, M., Howe, J.A., 1997. Sedimentological and stratigraphical re-evaluation of the CIROS-1 core, McMurdo Sound Antarctica. Terra Antartica, 4: 149-160.

Fitzgerald, P.G., 1992. The Transantarctic Mountains of southern Victoria Land: the application of apatite fission track analysis to a rift shoulder uplift. Tectonics, 11: 634-662.

Fitzgerald, P.G., 1994. Thermochronology constraints on post-Paleozoic tectonic evolution of the central Transantarctic Mountains, Antarctica. Tectonics, 13: 818-836.

Fitzsimons, S.J., 1990. Ice-marginal depositional processes in a polar maritime environment, Vestvold Hills, Antarctica. Journal of Glaciology, 36(124): 279-286.

Fitzsimons, S.J., McManus, K.J., Lorrain, R.D., 1999. Structure and strength of basal ice and substrate of a dry-based glacier: evidence for substrate deformation at sub-freezing temperatures. Annals of Glaciology, 28: 236-240.

Fleming, R.F., Barron, J.A., 1996. Evidence of Pliocene Nothofagus in Antarctica from Pliocene marine sedimentary deposits (DSDP Site 274). Marine Micropaleontology, 27: 227-336.

Flower, B.P., 1999. Cenozoic deep-sea temperatures and polar glaciation: the oxygen isotope record. In: P. Barrett and G. Orombelli (Editors), Geological Records of Global and Planetary Changes. Siena, Terra Antartica Reports, Terra Antartica. pp. 27-42.

Flugel, E., 1978. Mikrofazielle Untersuchingsmethoden von Kalken. Springer-Verlag, Berlin.

Forsberg, C.F., Solheim, A., Elverhoi, A., Jansen, E., Cannell, J.E.T., Andersen, E.S., 1999. The depositional environment of the western Svalbard margin during the Pliocene and the Pleistocene: sedimentary facies changes at site 986. In: M.E. Raymo, E. Jansen, P. Blum and T.D. Herbert (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 162, College Station, TX, pp. 233-246.

Fralick, P.W., Kronberg, B.I., 1997. Geochemical discrimination of clastic sedimentary rock sources. Sedimentary Geology, 113: 111-124.

Francis, J.E., 2000. Fossil wood from Eocene high latitude forests, McMurdo Sound, Antarctica. In: J.D. Stilwell and R.M. Feldmann (Editors), Paleobiology and paleoenvironment of

343 Eocene rocks, McMurdo Sound, East Antarctica. American Geophysical Union, Antarctic Research Series, Volume 76, Washington D.C., pp. 253-260.

Francis, J.E., Hill, R.S., 1996. Fossil plants from the Pliocene Sirius Group, Transantarctic Mountains: Evidence for climate from growth rings and fossils leaves. Palaios, 11(4): 389-396.

Fyffe, G.J., 1990. The effect of water depth on ice-proximal glaciolacustrine sedimentation: Salpausselka I, southern Finland. Boreas, 19: 147-164.

Gersonde, R., Kyte, F.T., Bleil, U., Diekmann, B., Flores, J.A., Gohl, K., Grahl, G., Hagen, R., Kuhn, G., Sierro, F.J., Voelker, D., Abelmann, A., Bostwick, J.A., 1997. Geological record and reconstruction of the late Pliocene impact of the Eltanin asteroid in the Southern Ocean. Nature, 390: 357-363.

Gibson, E.K., Wentworth, S.J., McKay, D.S., 1983. Chemical weathering and diagenesis of a cold desert soil from Wright Valley, Antarctica: an anolog of martian weathering processes. Journal of Geophysical Research, 88(supplement): A912-A928.

Goldthwait, R.P., 1971. Introduction to till, today. In: R.P. Goldthwait (Editor), Till / a symposium. Ohio State University Press, Columbus, pp. 3-26.

Goldthwait, R.P., 1979. Giant grooves made by concentrated basal ice streams. Journal of Glacialogy, 23(89): 297-307.

Gravenor, C.P., 1979. The nature of the Late Paleozoic glaciation of Gondwana as determined from an analysis of garnets and other heavy minerals. Canadian Journal of Earth Sciences, 16(6): 1137-1153.

Grindley, G.W., 1963. The geology of the Queen Alexandra Range, Beardmore Glacier, , Antarctica; with notes on the correlation of Gondwana sequences. New Zealand Journal of Geology and Geophysics, 6(3): 307-360.

Gross, D.L., Moran, S.R., 1971. Grain-size and mineralogical gradations within tills of the Allegheny Plateau. In: R.P. Goldthwait (Editor), Till / a symposium. Ohio State University Press, Columbus, pp. 251-274.

Gunn, B.M., 1962. Differentiation in Ferrar Dolerites. New Zealand Journal of Geology and Geophysics, 5(5): 820-863.

Gunner, J.D., 1969. Petrography of metamorphic rocks from the Miller Range, Antarctica. Report No. 32, Institute of Polar Studies, Columbus, Ohio.

Gwyn, Q.H.J., Dreimanis, A., 1979. Heavy mineral assemblages in tills and their use in distinguishing glacial lobes in the Great Lakes region. Canadian Journal of Earth Sciences, 16: 2219-2235.

Haldorsen, S., 1981. Grain-size distribution of subglacial till and its relation to glacial crushing and abrasion. Boreas, 10: 91-105.

344 Haldorsen, S., J0rgensen, P., Rappel, M., Riezebos, P.A., 1989. Composition and source of the clay-sized fraction of Saalian till in The Netherlands. Boreas, 18: 89-97.

Hambrey, M.J., 1994. Glacial Environments. UBC Press, Vancouver.

Hambrey, M.J., 1997. Cape Roberts Core Logging Manual. Report No. 21, Antarctic Data Series, Victoria University of Wellington, Wellington.

Hambrey, M.J., Barrett, P.J., Robinson, P.H., 1989a. Stratigraphy. In: P.J. Barrett (Editor), Antarctic Cenozoic history from the CIROS-1 drillhole, McMurdo Sound. DSIR Bulletin, Volume 245, Wellington, pp. 23-48.

Hambrey, M.J., Larson, B., Ehrmann, W.U., 1989b. Forty million years of Antarctic glacial history yielded by Leg 119 of the Ocean Drilling Program. Polar Record, 25(153): 99- 106.

Hambrey, M.J., Ehrmann, W.U., Larsen, B., 1991. Cenozoic glacial record of the Prydz Bay continental shelf. East Antarctica. In: J. Barron and B. Larsen (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 119, College Station, TX, pp. 77-132.

Hambrey, M.J., McKelvey, B., 2000. Neogene fjordal sedimentation on the western margin of the Lambert Crraben, East Antarctica. Sedimentology, 47(3): 577-608.

Hamilton, R.J., Sorlien, C.C., Luyendyk, B.P., Bartek, L.R., Henrys, S.A., 1998. Tectonic regimes and structural trends off Cape Roberts, Antarctica. Terra Antartica, 5(3): 261- 272.

Hampton, M.A., Lee, H.J., Locat, J., 1996. Submarine landslides. Reviews of Geophysics, 34(1): 33-59.

Hancock, P.L., 1985. Brittle microtectonics: principles and practice. Journal of Structural Geology, 7(3/4): 437-457.

Haq, B.U., 1998. Natural gas hydrates: searching for long-term climatic and slope-stability records. In: J.-P. Henriet and J. Mienert (Editors), Gas Hydrates: Relevance to World Margin Stability and Climate Change. Geological Society London Special Publication, Volume 137, London, pp. 303-318.

Haq, B.U., Hardenbol, J., Vail, P R., 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235: 1156-1167.

Hart, J.K., Roberts, D.H., 1994. Criteria to distinguish between subglacial glaciotectonic and glaciomarine sedimentation, 1. Deformation styles and sedimentology. Sedimentary Geology, 91: 191-213.

Harwood, D.M., 1986. Diatom biostratigraphy and paleoecology with a Cenozoic history of Antarctic ice sheets. Ph.D. Thesis, The Ohio State University, Columbus, Ohio.

Harwood, D.M., Maruyama, T., 1992. Middle Eocene to Pleistocene diatom biostratigraphy of Southern Ocean sediments from the Kerguelen Plateau. In: S.W. Wise and R. Schlich

345 (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 120, College Station, TX.

Harwood, D.M., Rose, S_A., 1998. Report: on diatom analysis of Mount Feather COMRAC cores. In: G.S. Wilson auid J.A. Barron (Editors), M ount Feather Sirius Group Core Workshop and Collaborative Sample Analysis. Report No. 14, Byrd Polar Research Center, Columbus, OH, 79-89.

Harwood, D.M., Webb, P.-N., 1998. Glacial transport of diatoms in the Antarctic Sirius Group: Pliocene refrigerator. GSA Today, 8(4): 1-8.

Harwood, D.M., Levy, R_H., 2000. The McMurdo Sound erratics: introduction and overview. In: J.D. Stilwell and R.M. Feldmann (Editors), Paleobiology and paleoenvironment of Eocene rocks, McMurdo Sound, East Antarctica. American Geophysical Union, Antarctic Research Series, Volume 76, Washington D.C., pp. 1-18.

Harwood, D.M., Barrett, P.J., Edwards, A.R., Rieck, H.J., Webb, P.-N., 1989. Biostratigraphy and chronology. Ea: P.J. Barrett (Editor), Antarctic Cenozoic history from the ClROS-1 drillhole, McMurdo Sound. DSIR Bulletin, Volume 245, Wellington.

Hayes, D.E., Frakes, L.A., 1975. Initial reports of the Deep Sea Drilling Project. US Government Printing Office, Volume 28, Washington D C., pp. 1017.

Haywood, A.M., Valdes, iP.J., Sellwood, B.W., 2000. Global scale paleoclimate reconstruction of the middle Pliocene climate using the UKMO GCM: initial results. Global and Planetary Change, 25: 239-256.

Hicock, S.R., Goff, J.R., Dickinson, W.W., 1999. Mesosopic analysis of Sirius Group in core from Table Mountain, Dry Valleys Area, Antarctica. 8th IS AES Programme & Abstracts: 141.

Hiemstra, J.F., 1999. Microscopic evidence of grounded ice in the sediments of the ClROS-1 core, McMurdo Sound, Antarctica. Terra Antartica, 6 (4): 365-376.

Hiemstra, J.F., Van der Meer, J.J.M., 1997. Pore-water controlled grain fracturing as an indicator for subglacial shearing in tills. Journal of Glaciology, 43: 446-454.

Hill, R.S., Truswell, E.M., 1993. Nothofagus fossils in the Sirius Group, Transantarctic Mountains: leaves; and pollen and their climatic implications. In: J.P. Kennett and D.A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change Part Two. American Geophysical Union, Antarctic Research Series, Volume 60, Washington D C., pp. 67-73.

Hill, R.S., Harwood, D.M ., Webb, P.-N., 1996. Nothofagus beardmorensis {Nothofagaceae), a new species based on leaves from the Pliocene Sirius Group, Transantarctic Mountains, Antarctica. Review of Palaeobotany and Palynology, 94: 11-24.

Hodell, D.A., Venz, K., 1992. Toward a high-resolution stable isotope record of the Southern Ocean during the Pliocene-Pleistocene (4.8 to 0.8 Ma). In: J.P. Kennett and D.A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change Part One.

346 American Geophysical Union, Antarctic Research Series, Volume 56, Washington D.C., pp. 265-310.

Hofer, J.W., Szabo, J.P., 1993. Port Bruce ice-flow directions based on heavy-mineral assemblages in tills from the south shore of Lake Erie in Ohio. Canadian Journal of Earth Sciences, 30(6): 1236-1241.

Hooke, R.L., Iverson, N.R., 1995. Grain-size distribution in deforming subglacial tills: role of grain fracture. Geology, 23(1): 57-60.

Homer, T.C., 1992. Sedimentologic, minéralogie, and geochemical evaluation of the provenance and paleoclimatic record of Permian mudrocks from the Beardmore Glacier area, Antarctica. Ph.D. Thesis, Ohio State University, Columbus, Ohio.

Huybrechts, P., 1993. Glaciological modeling of the Late Cenozoic East Antarctic ice sheet: stability or dynamism. Geografiska Annaler, 75A: 221-238.

Ishman, S.E., Rieck, H.J., 1992. A late Neogene Antarctic glacio-eustatic record, Victoria Land Basin margin, Antarctica. In: J.P. Kennett and D A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change Part One. American Geophysical Union, Antarctic Research Series, Volume 56, Washington D.C., pp. 327-347.

Iverson, N.R., Hooyer, T.S., Hooke, R.L., 1996. A laboratory study of sediment deformation: stress heterogeneity and grain-size evolution. Annals of Glaciology, 22: 167-175.

Johnson, S.J., Clausen, H.B., Dansgaard, W., Führer, K., Gundestrup, N., Hammer, C.U., Iversen, P., Jouzel, J., Stauffer, B., Steffenson, J.P., 1992. Irregular glacial insterstadials in a new Greenland . Nature, 359: 311-313.

Johnsson, M.J., 1993. The system controlling the composition of clastic sediments. Geological Society of America Special Paper, 284: 1-19.

Johnsson, W.H., Gross, D.L., Moran, S.R., 1971. Till stratigraphy of the Danville region, east- central Illinois. In: R.P. Goldthwait (Editor), Till / a symposium. Ohio State University Press, Columbus, pp. 184-216.

J0rgenson, P., 1977. Some properties of Norwegian tills. Boreas, 6: 149-157.

Kellogg, D.E., Kellogg, T.B., 1996. Diatoms in ice: Implications for eolian contamination of Sirius Group deposits. Geology, 24: 1I5-II8.

Kempton, J.P., DuMontelle, P.B., Glass, H.D., 1971. Subsurface stratigraphy of the Woodfordian tills in the McLean County region, Illinois. In: R.P. Goldthwait (Editor), Till / a symposium. Ohio State University Press, Columbus, pp. 217-233.

Kennett, J.P., 1977. Cenozoic evolution of Antarctic glaciation, the Circum-Antarctic ocean, and their impact on global paleoceanography. Journal of Geophysical Research, 82(27): 3843-3860.

Kennett, J.P., Hodell, D.A., 1993. Evidence for relative climate stability of Antarctica during the early Pliocene: a marine perspective. Geografiska Annaler, 75A(4): 205-220.

347 Kerr, A., Huybrechts, P., 1999. The response of the East Antarctic ice-sheet to the evolving tectonic configuration of the Transantarctic Mountains. Global and Planetary Change, 23: 213-229.

ICinny, P.D., Black, L.P., Sheraton, J.W., 1997. Zircon U-Pb ages and geochemistry of igenous and metamorphic rocks in the northern Prince Charles Mountains, Antarctica. AGSO Journal of Australian Geology and Geophysics, 16(5): 637-654.

Kjær, K.H., 1999. Mode of subglacial transport deduced from till properties, MyrdalsjokuU, Iceland. Sedimentary Geology, 128: 271-292.

Koomen, A., Passchier, S., Van Tatenhove, F., 1993. Progress report granulometry GIMEX samples. Report, University of Amsterdam, Department of Physical Geography and Soil Science FGBL, Amsterdam.

Korsch, R.J., 1974. Pétrographie comparison of the Taylor and Victoria Groups (Devonian to Triassic) in South Victoria Land, Antarctica. New Zealand Journal of Geology and Geophysics, 17(3): 523-541.

Krantz, D.E., 1991. A chronology of Pliocene sea-level fluctuations: the U.S. middle Atlantic coastal plain record. Quaternary Science Reviews, 10: 163-174.

Krissek, L.A., 1989. Bulk mineralogy of nonbiogenic sediments from ODP Sites 642 and 643, Norwegian Sea: implications for sediment provenance and recycling. In: O. Eldholm, J. Tiiiede and E. Taylor (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 104, College Station TX, pp. 29-39.

Krissek, L.A., Kyle, P.R., 1998. Geochemical indicators of weathering and Cenozoic paleoclimates in sediments from CRP-1 and CIROS-1, McMurdo Sound, Antarctica. Terra Antartica, 5(3): 673-680.

Krook, L., 1994. Heavy mineral analysis of glacial sediments from northern Victoria Land, Antarctica. In: F.M. Van der Wateren, A.L.L.M. Verbers and F. Tessensohn (Editors), LIRA Workshop on Landscape Evolution: a multidisciplinary approach to the relationship between Cenozoic climate change and tectonics in the Ross Sea area, Antarctica. Haarlem, The Netherlands, Geological Survey of The Netherlands, pp. 141.

Kurz, M.D., Ackert, R.P., 1997. Stability of the East Antarctic Ice Sheet?: New chronological evidence from Bennett Platform, Antarctica. EOS, Transactions of the American Geophysical Union, 78(17): S185.

Kvenvolden, K.A., 1993. Gas hydrates - geological perspective and global change. Reviews of Geophysics, 31(2): 173-187.

Kwiek, P.B., Ravelo, A.C., 1999. Pacific Ocean intermediate and deep water circulation during the Pliocene. Paeogeography, Palaeoclimatology, Palaeoecology, 154: 191-217.

Kyle, P R., 1980. Development of heterogeneities in the subcontinental mantle: evidence from the Ferrar Group, Antarctica. Contributions to Mineralogy and Petrology, 73: 89-104.

348 LaPrade, K.E., 1982. Petrology and petrography of the Beacon Supergroup, Shackleton Glacier area. Queen Maud Range, Transantarctic Mountains, Antarctica, hi: C. Craddock (Editor), Third Symposium on Antarctic Geology and Geophysics. Madison, Wisconsin, U.S.A., The University of Wisconsin Press, pp. 581-590.

Larsen, E., Mangerud, J., 1992. Subglacially formed clastic dykes. Sveriges Geologiska Undersoking, 81: 163-170.

Lawyer, L.A., Gahagan, L.M., Coffin, M.F., 1992. The development of paleoseaways around Antarctica. In: J.P. Kennett and D.A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change Part One. American Geophysical Union, Antarctic Research Series, Volume 56, Washington D.C.

Laznicka, P., 1988. Breccias and Coarse Fragmentites-petrology, environments, associations, ores. Developments in Economic Geology, 25, 832p.

Leckie, R.M., Webb, P.N., 1979. Scallop Hill Formation and associated Pliocene marine deposits of southern McMurdo Sound. Antarctic Journal of the United States, 14(5): 54-56.

Levy, R.H., Harwood, D.M., 2000. Sedimentary lithofacies of the McMurdo Sound erratics. In: J.D. Stilwell and R.M. Feldmann (Editors), Paleobiology and paleoenvironment of Eocene rocks, McMurdo Sound, East Antarctica. American Geophysical Union, Antarctic Research Series, Volume 76, Washington D C., pp. 39-61.

Lewis, K.J., Fountain, A.G., Dana, G.L., 1999. How important is terminus cliff melt? A study of the terminus, Taylor Valley, Antarctica. Global and Planetary Change, 22(1-4): 105-115.

Linkletter, G.O., 1974. Authigenic phillipsite in lacustrine sediments, Taylor Valley Antarctica. New Zealand Journal of Geology and Geophysics, 17(3): 603-609.

Lopatin, B.G., 1972. Basement complex of the McMurdo 'Oasis', South Victoria Land. In: R.J. Adie (Editor), Antarctic Geology and Geophysics. Oslo, Norway, Universitetsforlaget. pp. 287-292.

Mabin, M.C.G., 1986. Sirius Formation basal contacts in the Beardmore Glacier region. Antarctic Journal of the United States, 21(5): 32-33.

Marchant, D.R., Denton, G.H., Swisher, C.S., 1993a. Miocene-Pliocene-Pleistocene glacial history of Arena Valley, Quatermain Mountains, Antarctica. Geografiska Annaler, 75A(4): 269-302.

Marchant, D.R., Denton, G.H., Sugden, D.E., HI, C.C.S., 1993b. Miocene glacial stratigraphy and landscape evolution of the western Asgard Range, Antarctica. Geografiska Armaler, 75A(4): 303-330.

Margolis, S.V., Kennett, J.P., 1970. Antarctic glaciation during the Tertiary recorded in sub- Antarctic deep-sea cores. Science, 170: 1085-1087.

Martinson, O., 1994. Mass Movements. In: A. Maltman (Editor), The Geological Deformation of Sediments. Chapman & Hall, London, pp. 127-165.

349 Maslin, M., Mikkelsen, N., Vilela, C., Haq, B.U., 1998. Sea-level and gas-hydrate controlled catastrophic sediment failures of the Amazon Fan. Geology, 26(12): 1107-1110.

Matthews, R.K., 1984. Oxygen isotope records of ice-volume history: 100 million years of glacio-eustatic sea-level fluctuation. AAPG Memoir, 36: 97-107.

Mayewski, P.A., 1975. Glacial geology and Late Cenozoic history of the Transantarctic Mountains, Antarctica. Report No. 56, Institute of Polar Studies, The Ohio State University, Columbus, Ohio.

Mayewski, P.A., Goldthwait, R.P., 1985. Glacial events in the Transantarctic Mountains: a record of the East Antarctic ice sheet. American Geophysical Union, Antarctic Research Series, Volume 36, pp. 275-324.

Mazzarini, P., DellaVedova, B., Salvini, P., 1997. Crustal segmentation of the Transantarctic Mountains rift shoulder along the David Glacier lineament, Victoria Land (Antarctica). In: C.A. Ricci (Editor), VH International Symposium on Antarctic Earth Sciences, Siena, Italy. The Antarctic Region: Geological Evolution and Processes, Terra Antartica Publication, pp. 565-569.

McGregor, V.R., 1965. Geology of the area between the Axel Heiberg and Shackleton Glaciers, Queen Maud Range, Antarctica. New Zealand Journal of Geology and Geophysics, 8(2): 314-343.

McKelvey, B.C., 1981. The lithologie logs of DVDP cores 10 and 11, eastern Taylor Valley. In: L.D. McGinnis (Editor), Dry Valley Drilling Project. American Geophysical Union, Antarctic Research Series, Volume 33, Washington D.C., pp. 63-94.

McKelvey, B.C., 1982. Late Cenozoic marine and terrestrial glacial sedimentation in eastern Taylor Valley, Southern Victoria Land. In: C. Craddock (Editor), Third Symposium on Antarctic Geology and Geophysics. Madison, Wisconsin, U.S.A., The University of Wisconsin Press, pp. 1109-1116.

McKelvey, B., 1994. The Pagodroma Tillite and the Sirius Group - a comparison. In: P.M. Van der Wateren, A.L.L.M. Verbers and P.Tessensohn (Editors), LIRA Workshop on Landscape Evolution: a multidisciplinary approach to the relationship between Cenozoic climate change and tectonics in the Ross Sea area, Antarctica. Haarlem, The Netherlands, Geological Survey of The Netherlands, pp. 113-116.

McKelvey, B.C., Mercer, J.H., Harwood, D.M., Stott, L.D., 1984. The Sirius Formation: further considerations. Antarctic Journal of the United States, 19(5): 42-43.

McKelvey, B.C., Webb, P.-N., Harwood, D.M., Mabin, M.G.C., 1991. The Dominion Range Sirius Group: a record of the late Pliocene - early Pleistocene Beardmore Glacier. In: M.R.A. Thomson, J.A. Crame and J.W. Thomson (Editors), Fifth International Symposium on Antarctic Earth Sciences. Cambridge, England, Cambridge University Press, pp. 675-682.

35 0 McLennan, S.M., Hemming, S., McDaniel, D.K., Hanson, G.N., 1993. Geochemical approaches to sedimentation, provenance and tectonics. Geological Society of America Special Paper, 284: 21-40.

Menzies, J., 1990. Brecciated diamictons from Mohawk Bay, S. Ontario. Sedimentology, 37: 481-493.

Mercer, J.H., 1968. Glacial geology of the Reedy Glacier area, Antarctica. Geological Society of America Bulletin, 79: 471-486.

Mercer, J.H., 1972. Some observations on the glacial geology of the Beardmore Glacier area. In: R.J. Adie (Editor), Antarctic Geology and Geophysics. Oslo, Norway, Universitetsforlaget. pp. 427-433.

Minshew, V.H., 1966. Stratigraphy of the Wisconsin Range, Horlick Mountains, Antarctica. Science, 152(3722): 637-638.

Morton, A.C., Hallsworth, C.R., 1999. Processes controlling the composition of heavy mineral assemblages in sandstones. Sedimentary Geology, 124: 3-29.

Mudrey, M.G., 1975. Basement rocks in DVDP-12, Lake Leon, Taylor Valley. Report No. 5, Bulletin, Department of Geology, Northern Illinois University, DeKalb, Illinois.

Mulholland, J.W., 1976. Texture of tills, central Massachusetts. Journal of Sedimentary Petrology, 46(4): 778-787.

Munksgaard, N.C., Hensen, B.J., Thost, D.E., 1992. Geochemistry of Proterozoic granulites from the northern Prince Charles Mountains, East Antarctica. Antarctic Science, 4(1): 59-69.

Murray, T., 1997. Assessing the paradigm shift: deformable glacier beds. Quaternary Science Reviews(l6): 995-1016.

Nechaev, V.P., Isphording, W.C., 1993. Heavy-mineral assemblages of continental margins as indicators of plate-tectonic environments. Journal of Sedimentary Petrology, 63(6): 1110- 1117.

Nesbitt, H.W., Young, G.M., 1982. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature, 299: 715-717.

Nesbitt, H.W., Young, G.M., 1984. Prediction of some weathering trends of plutonic and volcanic rocks based on thermodynamic and kinetic considerations. Geochimica et Cosmochimica Acta, 48: 1523-1534.

Nesbitt, H.W., Young, G.M., 1989. Formation and diagenesis of weathering profiles. The Journal of Geology, 97(2): 129-147.

Nesbitt, H.W., Young, G.M., 1996. Petrogenesis of sediments in the absence of chemical weathering: effects of abrasion and sorting on bulk composition and mineralogy. Sedimentology, 43: 341-358.

351 Odonne, F., Menard, L, Massonat, G.J., Rolando, J.-P., 1999. Abnormal reverse faulting above a depleting reservoir. Geology, 27(2): 111-114.

Oglesby, R.J., 1999. Use of climate models to extend paleoclimatic data. In: P. Barrett and G. Orombelli (Editors), Geological Records of Global and Planetary Changes. Siena, Terra Antartica Reports, Terra Antartica. pp. 131-149.

Oppenheimer, M., 1998. Global warming and the stability of the West Antarctic Ice Sheet. Nature, 393: 325-232.

Orheim, O., Lucchitta, B., 1990. Investigating climate change by digital analysis of blue ice extent on satellite images of Antarctica. Annals of Glaciology, 14: 211-214.

Passchier, S., Wilson, T.J., Paulsen, T.S., 1998. Origin of breccias in the CRP-1 core. Terra Antartica, 5(3); 401-409.

Paterson, W.S.B., 1994. The physics of glaciers. 3"* edition, Elsevier Science, Oxford, 480 p.

Pauli, C.K., Buelow, W.J., HI, W.U., Borowski, W.S., 1996. Increased continental margin slumping frequency during sea level lowstands above hydrate-bearing sediments. Geology, 24(2): 143-146.

Peryt, T.M., Jasionowski, M., 1994. In situ formed redeposited gypsum breccias in the Middle Miocene Badenian of southern Poland. Sedimentary Geology, 94: 153-163.

Pettijohn, F.J., Potter, P.E., Siever, R., 1987. Sand and Sandstone. Springer-Verlag, New York.

Pickard, J., Adamson, D.A., Harwood, D.M., Miller, G.H., Quilty, P.G., Dell, R.K., 1988. Early Pliocene marine sediments, coastline, and climate of East Antarctica. Geology, 16: 158- 161.

Polozek, K., Ehrmann, W., 1998. Distribution of heavy minerals in CRP-1. Terra Antartica, 5(3): 633-638.

Porter, S.C., Beget, J.E., 1981. Provenance and depositional environments of Late Cenozoic sediments in permafrost cores from lower Tayor Valley. In: L.D. McGinnis (Editor), Dry Valley Drilling Project. American Geophysical Union, Antarctic Research Series, Volume 33, Washington D.C., pp. 351-363.

Powell, R.D., 1981. Sedimentation conditions in Taylor Valley, Antarctica, inferred from textural analysis of DVDP cores. In: L.D. McGinnis (Editor), Dry Valley Drilling Project. American Geophysical Union, Antarctic Research Series, Volume 33, Washington D.C., pp. 331-349.

Powell, R.D., Molnia, B.F., 1989. Glacimarine sedimentary processes, facies and morphology of the south-east Alaska shelf and Qords. Marine Geology, 85: 359-390.

Prentice, M.L., Denton, G.H., Lowell, T.V., Conway, H.C., Heusser, L.E., 1986. Pre-late Quaternary glaciation of the Beardmore Glacier region, Antarctica. Antarctic Journal of the United States, 21(5): 95-98.

352 Prentice, M.L., Bockheim, J.G., Wilson, S.C., Burckle, L.H., Hodell, D.A., Schluechter, C., Kellogg, D.E., 1993. Late Neogene Antarctic glacial history: evidence from Central Wright Valley. In: J.P. Kennett and D.A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change Part Two. American Geophysical Union, Antarctic Research Series, Volume 60, Washington D C., pp. 207-250.

Prentice, M., Ishman, S., Clemens, S., McIntosh, B., Clarke, K., 1999. Late Neogene stable isotope records from Qords within the McMurdo Dry Valleys, Antarctica. 8th IS AES Programme & Abstracts: 249.

Quilty, P.G., 1993. Coastal East Antarctic Neogene sections and their contribution to the ice sheet evolution debate. In: J.P. Kennett and D.A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change. American Geophysical Union, Volume 60, Washington D C., pp. 251-264.

Raymo, M.E., 1992. Global climate change: a three million year perspective. In: G.J. Kukla and E. Went (Editors), Start of a Glacial. Springer, Volume 13, Berlin-Heidelberg, pp. 207- 223.

Retallack, G.J., Krull, E.S., 1997. Neogene paleosols of the Sirius Group, Dominion Range, Antarctica. Antarctic Journal of the United States, 32(5): 10-13.

Riezebos, P.A., 1983. A pétrographie study of a till profile near Lunteren (the Netherlands). In: J. Ehlers (Editor), Glacial deposits in North-West Europe. A.A. Balkema, Rotterdam, pp. 399-404.

Rijsdijk, K.F., Owen, G., Warren, W.P., McCarroll, D., Van der Meer, J.J.M., 1999. Clastic dykes in over-consolidated tills: evidence of subglacial hydrofracturing at Killiney bay, eastem Ireland. Sedimentary Geology, 129: 111-126.

Robert, C., Chamley, H., 1992. Late Eocene - early Oligocene evolution of climate and marine circulation: deep-sea clay mineral evidence. In: J.P. Kennett and D.A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change Part One. American Geophysical Union, Antarctic Research Series, Volume 56, Washington D C., pp. 97- 117.

Robert, C., Kennett, J.P., 1997. Antarctic continental weathering changes during Eocene- Oligocene cryosphere expansion: Clay mineral and oxygen isotope evidence. Geology, 25(7): 587-590.

Roser, B.P., Pyne, A.R., 1989. Wholerock geochemistry. In: P.J. Barrett (Editor), Antarctic Cenozoic history from the CIROS-1 drillhole, McMurdo Sound. DSIR Bulletin, Volume 245, Wellington, pp. 175-184.

Ruegg, G.H.J., 1983. Glaciofluvial and glaciolacustrine deposits in the Netherlands. In: J. Ehlers (Editor), Glacial deposits in North-West Europe. A.A. Balkema, Rotterdam, pp. 379-392.

Sasttem, J., 1994. Glaciotectonic structures along the southern Barents shelf margin. In: W.P. Warren and D.G. Croot (Editors), Formation and Deformation of Glacial Deposits. A.A. Balkema, Rotterdam, pp. 95-113.

353 Sagnotti, L-, Florindo, F., Verosub, K.L., Wilson, G.S., 1998. Environmental magnetic record of Antarctic palaeoclimate from Eocene/OIigocene glaciomarine sediments, Victoria Land basin. Geophysical Journal International, 134: 653-662.

Saito, S., 1998. Major and trace element geochemistry of sediments from east Greenland continental rise: an implication for sediment provenance and source area weathering. In: A.D. Saunders, H.C. Larsen and S.W. Wise (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 152, College Station TX, p p .19-28.

Salvini, P., Storti, P., 1999. Cenozoic tectonic lineaments of the Terra Nova Bay region, Ross Embayment, Antarctica. Global and Planetary Change, 23(1-4): 129-144.

Schack-Pedersen, S.A., 1988. Glacitectonite: Brecciated sediments and cataclastic sedimentary rocks formed subglacially. In: R.P. Goldthwait and C.L. Matsch (Editors), Genetic classification of glacigenic deposits. A.A. Balkema, Rotterdam, pp. 89-91.

Schafer, J., Ivy-Ochs, S., Wieler, R., Leya, I., Baur, H., Denton, G., Schliichter, C., 1999. Cosmogenic noble gas studies in the oldest landscape on earth: surface exposure ages of the Dry Valleys, Antarctica. Earth and Planetary Science Letters, 167(3): 215-226.

Shackleton, N.J., Hall, M.A., Pate, D., 1995. Pliocene stable isotope stratigraphy of Site 846. In: N.G. Pisas, L.A. Mayer, T.R. Janecek, A. Palmer-Julson and T.H. Van Andel (Editors), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 138, College Station TX, pp. 337-355.

Shanmugam, G., Bloch, R.B., Mitchell, S.M., Beamish, G.W.J., Hodgkinson, R.J., Damuth, J.E., Straume, T., Syvertsen, S.E., Shields, K.E., 1995. Basin floor fans in the North Sea: sequence stratigraphie models vs. sedimentary facies. AAPG Bulletin, 79(4): 477-512.

Shaw, J., 1977. Tills deposited in arid polar environments. Canadian Journal of Earth Sciences, 14(6): 1239-1245.

Shaw, S.E., 1962. Petrography of Beacon sandstone samples from Beacon Height West, upper Taylor Glacier, Antarctica. New Zealand Journal of Geology and Geophysics, 5: 733- 739.

Sheraton, J.W., Tindle, A.G., Tingey, R.J., 1995. Geochemistry, origin, and tectonic stetting of granitic rocks of the Prince Charles Mountains, Antarctica. AGSO Journal of Australian Geology and Geophysics, 16(3): 345-370.

Siders, M.A., Elliot, D.H., 1985. Major and trace element geochemistry of the Kirkpatrick Basalt, Mesa Range, Antarctica. Earth and Planetary Science Letters, 72: 54-64.

Skinner, D.N.B., Ricker, J., 1968. The geology of the region between Mawson and Priestley Glaciers, North Victoria Land, Antarctica. New Zealand Journal of Geology and Geophysics, 11(4): 1009-1076.

Sloan, L.C., Crowley, T.J., Pollard, D., 1996. Modeling of middle Pliocene climate with the NCAR GENESIS general circulation model. Marine Micropaleontology, 27: 51-61.

354 Steiger, J.R., Holowaychuk, N., 1971. Particle-size and carbonate analysis of glacial till and lacustrine deposits in western Ohio. In: R.P. Goldthwait (Editor), Till / a symposium. Ohio State University Press, Columbus, pp. 275-289.

Strobel, M., Faure, G., 1987. Transport of indicator clasts by ice-sheets and the transport half­ distance: a contribution to prospecting for ore deposits. Journal of Geology, 95: 687-697.

Stroeven, A.P., 1997. The Sirius Group of Antarctica: age and environments. In: C.A. Ricci (Editor), Vn International Symposium on Antarctic Earth Sciences, Siena, Italy. The Antarctic Region: Geological Evolution and Processes, Terra Antartica Publication, pp. 747-761.

Stroeven, A.P., Prentice, M.L., 1997. A case for Sirius Group alpine glaciation at Mount Fleming, South Victoria land, Antarctica: A case against Pliocene East Antarctic Ice Sheet reduction. Geological Society of America Bulletin, 109(7): 825-840.

Stroeven, A.P., Kleman, J., 1999. Age of Sirius Group on Mount Feather, McMurdo Dry Valleys, Antarctica, based on glaciological inferences from the overridden mountain range of Scandinavia. Global and Planetary Change, 23: 231-247.

Stroeven, A.P., Prentice, M.L., Kleman, J., 1996. On marine microfossil transport and pathways in Antarctica during the late Neogene: Evidence from the Sirius Group at Mount Fleming. Geology, 24(8): 727-730.

Stroeven, A., Burckle, L.H., Kleman, J., Prentice, M.L., 1998. Atmospheric transport of diatoms in the Antarctic Sirius Group: Pliocene Deep Freeze. GSA Today, 8(4): 1-5.

Stump, E., 1995. The Ross Orogen of the Transantarctic Mountains. Cambridge University Press, Cambridge, U.K.

Stump, E., Sheridan, M.F., Borg, S.G., Sutter, J.F., 1980. Early Miocene subglacial basalts, the East Antarctic ice sheet, and uplift of the Transantarctic Mountains. Science, 207: 757- 759.

Sugden, D.E., John, B.S., 1976. Glaciers and Landscape. 2"** edition. Edward Arnold, London, England.

Sugden, D.E., Marchant, D.R., Denton, G.H., 1993. The case for a stable East Antarctic ice sheet: the background. Georgrafiska Annaler, 75 A(4): 151-154.

Sugden, D.E., Summerfield, M.A., Denton, G.H., Wilch, T.I., McIntosh, W.C., Marchant, D R., Rutford, R.H., 1999. Landscape development in the Royal Society Range, southern Victoria Land, Antarctica: stability since the mid-Miocene. Geomorphology, 28.

Swithinbank, C., 1988. Antarctica. U. S. Geological Survey Professional Paper 1044-9612: B l- B278.

Syvitski, J.P.M., Burrell, D.C., Skei, J.M., 1987. , Processes and Products. Springer Verlag, New York.

355 Tessensohn, F., 1994. Structurai evolution of the northern end of the Transantarctic Mountains. In: F.M. Van der Wateren, A.L.L.M. Verbers and F. Tessensohn (Editors), LIRA Workshop on Landscape Evolution: a multidisciplinary approach to the relationship between Cenozoic climate change and tectonics in the Ross Sea area, Antarctica. Haarlem, The Netherlands, Geological Survey of The Netherlands, pp. 57-61.

Tingey, R.J., 1991. Mesozoic tholeiite igneous rocks in Antarctica: the Ferrar (Super) Group and related rocks. In: R.J. Tingey (Editor), The Geology of Antarctica. Clarendon Press, Oxford, U.K., pp. 153-174.

Tulaczyk, S., Kamb, B., Scherer, R.P., Engelhardt, H.F., 1998. Sedimentary processes at the base of a West Antarctic Ice Stream: constraints from textural and compositional properties of subglacial debris. Journal of Sedimentary Research, 68(3): 487-496.

Ugolini, F.G., Jackson, M.L., 1982. Weathering and Mineral Synthesis in Antarctic Soils. In: C. Craddock (Editor), Third Symposium on Antarctic Geology and Geophysics. Madison, Wisconsin, U.S.A., The University of Wisconsin Press, pp. 1101-1108.

Ugolini, F.C., Deutsch, W., Harris, H.J.H., 1981. Chemistry and clay mineralogy of selected cores from the Antarctic Dry Valley Drilling Project. In: L.D. McGiimis (Editor), Dry Valley Drilling Project. American Geophysical Union, Volume 33, Washington D.C., pp. 315-329.

Van der Meer, J.J.M. (Editor), 1987. Tills and Glaciotectonics, A.A. Balkema. Rotterdam.

Van der Meer, J.J.M., 1993. Microscopic evidence of subglacial deformation. Quaternary Science Reviews, 12: 553-587.

Van der Meer, J.J.M., 1996. Chapter 12. Micromorphology. In: J. Menzies (Editor), Glacial Environments: Past Glacial Environments - Sediments, forms and techniques. Volume 2, Butterworth-Heinemann, pp. 335-355.

Van der Meer, J.J.M., Hiemstra, J.F., 1998. Micromorphology of Miocene diamicts, indications of grounded ice. Terra Antartica, 5(3): 363-366.

Van der Meer, J.J.M., Evenson, E.B., Rabassa, J.O., 1994. Observations on the sedimentology of some clastic dykes from glacial environments, IGCP Symposium: The termination of the Pleistocene in South America. Tierra del Fuego, Argentina.

Van der Meer, J.J.M., Miicher, H.J., Hofle, H.C., 1994. Micromorphological observations on some till samples from Antarctica. In: F.M. Van der Wateren, A.L.L.M. Verbers and F. Tessensohn (Editors), LIRA Workshop on Landscape Evolution: a multidisciplinary approach to the relationship between Cenozoic climate change and tectonics in the Ross Sea area, Antarctica. Haarlem, The Netherlands, Geological Survey of the Netherlands, pp. 143-145.

Van der Meer, J.J.M., Hiemstra, J.F., Stroeven, A.P., 1996. Micromorphology of two Sirius Group core diamict samples from Mt. Feather, Dry Valleys, Antarctica. In: G.S. Wilson and J.A. Barron (Editors), Mount Feather Sirius Group Core Workshop and Collaborative

356 Sample Analysis. Report No. 14, Byrd Polar Research Center, Ohio State University, Columbus, Ohio, pp. 96-108.

Van der Wateren, P.M., 1994. Proglacial subaquatic outwash fan and delta sediments in push moraines; indicators of subglacial meltwater activity. Sedimentary Geology, 91(1-4); 145-172.

Van der Wateren, F.M., 1995. Structural geology and sedimentology of push moraines, Mededelingen van de Rijks Geologische Dienst. Report No. 54, Rijks Geologische Dienst, Haarlem, The Netherlands.

Van der Wateren, P.M., Verbers, A.L.L.M., 1994. Differential tectonic uplift of fault blocks in the West Antarctic rift system and their landscape evolution. In: P.M. Van der Wateren, A.L.L.M. Verbers and P. Tessensohn (Editors), LIRA Workshop on Landscape Evolution: a multidisciplinary approach to the relationship between Cenozoic climate change and tectonics in the Ross Sea area, Antarctica. Haarlem, The Netherlands, Geological Survey of The Netherlands, pp. 125-128.

Van der Wateren, P.M., and 11 others, 1996. Glaciation and deglaciation of the uplifted margins of the Cenozoic West Antarctic rift system, Ross Sea, Antarctica. Geologisches Jahrbuch, 89: 123-155.

Van der Wateren, P.M., Dunai, T., Klas, W., 1997. Evidence of differential neotectonic and glacial histories of regions within the Transantarctic Mountains, based on measurements of in situ cosmogenic nuclides ^He, ^°Be, “‘Ne and “®A1, EUG, 9, Terra Nova. pp. 219.

Van der Wateren, P.M., Dunai, T.J., VanBalen, R.T., Klas, W., Verbers, A.L.L.M., Passchier, S., Herpers, U., 1999. Contrasting Neogene denudation histories of different structural regions in the Transantarctic Mountains rift flank constrained by cosmogenic isotope measurements. Global and Planetary Change, 23: 145-172.

Verbers, A.L.L.M., Van der Wateren, P.M., 1992. A glacio-geological reconnaissance of the southern Prince Albert Mountains, Victoria Land, Antarctica. In: Y. Yoshida, K. Kaminuma and K. Shiraishi (Editors), Recent Progress in Antarctic Earth Science. TERRAPUB, Tokyo, Japan, pp. 715-719.

Verbers, A.L.L.M., Damm, V., 1994. Morphology and late Cenozoic (<5Ma) glacial history of the area between David and Mawson Glaciers, Victoria Land, Antarctica. Annals of Glaciology, 20: 55-60.

Vermeulen, P.J.M., 1994. Till provenance in North Victoria Land: a pilot study based on geochemistry. In: P.M. Van der Wateren, A.L.L.M. Verbers and P. Tessensohn (Editors), LIRA Workshop on Landscape Evolution: a multidisciplinary approach to the relationship between Cenozoic climate change and tectonics in the Ross Sea area, Antarctica. Haarlem, The Netherlands, Geological Survey of the Netherlands, pp. 125- 128.

Visser, J.N.J., Colliston, W.P., 1984. The origin of soft-sediraent deformation structures in Permo-Carboniferous glacial and proglacial beds. South Africa. Journal of Sedimentary Petrology, 54(4): 1183-1196.

357 Von Brunn, V., Talbot, C.J., 1986. Formation and deformation of subglacial intrusive clastic sheets in the Dwyka formation of northern Natal, South Africa. Journal of Sedimentary Petrology, 56(1): 35-44.

Wamke, D.A., Allen, C.P., Mueller, D.W., Hodell, D.A., Brunner, C., 1992. Miocene-Pliocene Antarctic glacial evolution: a synthesis of ice-rafted debris, stable isotope, and planktonic foraminiferal indicators, ODP Leg 114. In: J.P. Kennett and D.A. Wamke (Editors), The Antarctic Paleoenvironment: A Perspective on Global Change Part One. American Geophysical Union, Antarctic Research Series, Volume 56, Washington D C., pp. 311- 325.

Webb, P.-N., 1990. The Cenozoic and its global impact. Antarctic Science, 2(1): 3-21.

Webb, P.-N., 1994. Paleo-drainage systems of East Antarctica and sediment supply to west Antarctic rift system basins. Terra Antartica, 1(2): 457-461.

Webb, P.-N., Wrenn, J.H., 1982. Upper Cenozoic micropaleontology and biostratigraphy of eastem Taylor Valley, Antarctica. In: C. Craddock (Editor), Third Symposium on Antarctic Geology and Geophysics. Madison, Wisconsin, U.S.A., The University of Wisconsin Press, pp. 1117-1122.

Webb, P.N., Andreasen, J.E., 1986. Potassium/argon dating of volcanic material associated with the Pliocene Pecten Conglomerate (Cockbum Island) and Scallop Hill Formation (McMurdo Sound). Antarctic Joumal of the United States, 21(5): 59.

Webb, P.N., Harwood, D.M., 1987. Terrestrial flora of the Sirius Formation: Its significance for Late Cenozoic glacial history. Antarctic Joumal of the United States, 22(5): 7-9.

Webb, P.-N., Harwood, D.M., 1991. Late Cenozoic glacial history of the Ross embayment, Antarctica. Quatemary Science Reviews, 10: 215-223.

Webb, P.-N., Harwood, D.M., 1993. Pliocene fossil Nothofagus (southem beech) from Antarctica: phytogeography, dispersal strategies, and survival in high latitude glacial- deglacial environments. In: J. Alden (Editor), Forest Development in Cold Climates. Plenum Press, New York, pp. 135-165.

Webb, P.N., Harwood, D.M., McKelvey, B.C., Mercer, J.H., Stott, L.D., 1984. Cenozoic marine sedimentation and ice-volume variation on the East Antarctic craton. Geology, 12: 287- 291.

Webb, P.-N., Harwood, D.M., Mabin, M.G.C., McKelvey, B.C., 1986a. Late Cenozoic tectonic and glacial history of the Transantarctic Mountains. Antarctic Joumal of the United States, 21(5): 99-100.

Webb, P.-N., McKelvey, B.C., Harwood, D.M., Mabin, M.C.G., Mercer, J.H., 1986b. Sirius Formation of the Beardmore Glacier region. Antarctic Joumal of the United States, 22(1/2): 8-13.

358 Webb, P.-N., Harwood, D.M., Mabin, M.G.C., McKelvey, B.C., 1994. Late Neogene uplift of the Transantarctic Mountains in the Beardmore Glacier region. Terra Antartica, 1(2): 463- 467.

Webb, P.-N., Harwood, D.M., Mabin, M.G.C., McKelvey, B.C., 1996a. A marine and terrestrial Sirius Group succession, middle Beardmore Glacier-Queen Alexandra Range, Transantarctic Mountains, Antarctica. Marine Micropaleontology, 27: 273-297.

Webb, P.-N., Harwood, D.M., Hambrey, M.J., Krissek, L.A., Ashworth, A C., Mabin, M.C.G., Fabel, F.G., 1996b. The Late Cenozoic Sirius Group of the upper Shackleton Glacier region, Transantarctic Mountains. Antarctic Joumal of the United States, 31(2): 12-17.

Webb, P.-N., Harwood, D.M., Hambrey, M.J., Krissek, L.A., Ashworth, A.C., Mabin, M.C.G., 1996c. The sub-Sirius Group erosion surface at Roberts Massif, upper Shackleton Glacier region, Transantarctic Mountains. Antarctic Joumal of the United States, 21(2): 14-15.

Webb, P.-N., Harwood, D.M., Hambrey, M.J., Krissek, L.A., Ashworth, A.C., Mabin, M.C.G., 1996d. Stratigraphy of the Sirius Group, upper Shackleton Glacier region, Transantarctic Mountains. Antarctic Joumal of the United States, 21(2): 16-17.

White, W.A., 1972. Deep erosion by continental ice sheets. Geological Society of America Bulletin, 83: 1037-1056.

Wilson, G., Barron, J., 1996. Section 1: The Mount Feather Project. In: G. Wilson and J. Barron (Editors), Mount Feather Sirius Group Core Workshop and Collaborative Sample Analysis. Report No. 14, Byrd Polar Research Center, Columbus, Ohio, pp. 1-7.

Wilson, G.S., Harwood, D.M., Askin, R.A., Levy, R.H., 1998a. Late Neogene Sirius Group strata in Reedy Valley, Antarctica: a multiple-resolution record of climate, ice-sheet and sea- level events. Joumal of Glaciology, 44(148): 437-447.

Wilson, G.S., Roberts, A.P., Verosub, K.L., Florindo, F., Sagnotti, L., 1998b. Magnetobiostratigraphic chronology of the Eocene-Oligocene transition in the ClROS-1 core, Victoria Land margin, Antarctica. GSA Bulletin, 110: 35-47.

Wilson, T.J., 1995. Cenozoic transtension along the Transantarctic Mountain-West Antarctic rift boundary, southem Victoria Land, Antarctica. Tectonics, 14(2): 531-545.

Wilson, T.J., 1999. Cenozoic structural segmentation of the Transantarctic Mountains rift flank in southem Victoria Land. Global and Planetary Change, 23: 105-127.

Wilson, T.J., Paulsen, T.S., 1998. CRP-1 fracture arrays: constraints on the Neogene-Quatemary stress regime along the Transantarctic Mountain Front, Antarctica. Terra Antartica, 5(3): 327-335.

Winslow, M.A., 1983. Clastic dike swarms and the stmctural evolution of the foreland fold and thrust belt of the southem Andes. GSA Bulletin, 94: 1073-1080.

Wise, S.W., Breza, J.R., Harwood, D.M., Wei, W., Zachos, J.C., 1992. Paleogene glacial history of Antarctica in light of Leg 120 drilling results. In: S.W. Wise and R. Schlich (Editors),

359 Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, Volume 120, College Station, TX, pp. 1001-1030.

Womer, G., 1992. Kirkpatrick lavas. Exposure Hill Formation and Ferrar Sills in the Prince Albert Mountains, Victoria Land, Antarctica. Polarforschung, 60(2): 87-90.

Young, D.J., Rybum, R.J., 1968. The geology of Buckley and Darwin nunataks, Beardmore Glacier, Ross Dependency, Antarctica. New Zealand Joumal of Geology and Geophysics, 11: 922-939.

Young, G.M., Nesbitt, H.W., 1998. Processes controlling the distribution of Ti and A1 in weathering profiles, siliciclastic sediments and sedimentary rocks. Joumal of Sedimentary Research, 68(3): 448-455.

Zachos, J.C., Flower, B.P., Paul, H., 1997. Orbitally paced climate oscillations across the Oligocene-Miocene boundary. Nature, 388: 567-570.

Zandstra, J.G., 1983. Fine gravel, heavy mineral and grain-size analysis of Pleistocene, mainly glacigenic deposits in the Netherlands. In: J. Ehlers (Editor), Glacial deposits in North- West Europe. A.A. Balkema, Rotterdam, pp. 361-377.

Zhao, J., Ellis, D.J., Kilpatrick, J.A., McCulloch, M.T., 1997. Geochemical and Sr-Nd study of chamockites and related rocks in the northem Prince Charles Mountains, East Antarctica; implications for chamockite petrogenesis and Proterozoic cmstal evolution. Precambrian Research, 81(1-2): 37-66.

36 0