UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Petrology and geochemistry of the 25 MA Mt. Marion dufresne basaltic section on the Kerguelen Archipelago… Annell, Heidi 2005

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

Item Metadata

Download

Media
831-ubc_2005-0153.pdf [ 21.36MB ]
Metadata
JSON: 831-1.0052331.json
JSON-LD: 831-1.0052331-ld.json
RDF/XML (Pretty): 831-1.0052331-rdf.xml
RDF/JSON: 831-1.0052331-rdf.json
Turtle: 831-1.0052331-turtle.txt
N-Triples: 831-1.0052331-rdf-ntriples.txt
Original Record: 831-1.0052331-source.json
Full Text
831-1.0052331-fulltext.txt
Citation
831-1.0052331.ris

Full Text

P E T R O L O G Y A N D G E O C H E M I S T R Y O F THE 25 M A MT . M A R I O N D U F R E S N E BASALTIC SECTION O N THE K E R G U E L E N A R C H I P E L A G O : C O N S T R A I N I N G THE TRANSIT ION F R O M THOLEIITIC T O MILDLY ALKAL IC V O L C A N I S M O N A M A J O R O C E A N I C ISLAND by HEIDI A N N E L L B.Sc. University of Calgary, 2002 A THESIS SUBMITTED FOR PARTIAL FULFILMENT O F THE REQUIREMENTS FOR THE DEGREE O F MASTER O F SCIENCE in THE F A C U L T Y O F G R A D U A T E STUDIES ( G E O L O G I C A L SCIENCES) THE UNIVERSITY O F BRITISH C O L U M B I A March 2005 © Heid i Annel l , 2005 ABSTRACT Flood basalts on the - 6 5 0 0 k m 2 Kerguelen Archipelago in the southern Indian Ocean formed in an intraplate location on the Antarctic plate between 29-24 M a as the Southeast Indian Ridge migrated from - 2 2 5 to - 4 0 0 km northeast relative to the Kerguelen hotspot. A - 7 0 0 m section of basalts exposed at Mt. Mar ion Dufresne, in the southern part of the Plateau Central, contains both transitional-tholeiit ic and mi ldly alkal ic lavas (A.I. -0.8 to 2.0) that are geochemical ly and isotopically variable, and captures an important period of time (-1 Myr) near the end of f lood basalt volcanism on the Kerguelen Archipelago. The basal 300 m of the section consists of l o w - M g O (<5.2 wt. %) predominantly aphyric lavas characterized by a limited range of "enr iched" isotopic composit ions ( 1 7 6 Hf / l 7 7 Hf = .0 .28282-0 .28287; 8 7 Sr / 8 6 Sr = 0.7048-0.7050; 1 4 3 N d / 1 4 4 N d = 0.5126-0.5127; 2 0 7 P b / 2 0 4 P b = 15.54-15.56) similar to composit ions observed in the enriched 24-25 M a mi ldly alkal ic lavas to the east and southeast (e.g. Mt. Crozier). This aphyric sequence of basalts appears to represent a "steady state" magmatic system where differentiation occurred primari ly in a sol id-dominated (mush or slurry) environment. Highly ca lc ic (>An 8 0) plagioclase composit ions observed in intercalated plagioclase-phyric and ultraphyric basalts from the upper part of this interval signal a change in crystall ization environments to relatively shal low (-5-6 km) magma storage and relatively hydrous condit ions (>3 wt.% H 2 0 ) . The l o w - M g O basaltic lavas are overlain by a thick ( -400 m) succession of ol iv ine-phyr ic h igh -MgO basalts (7.1-11.4 wt. %) that is interpreted to represent an interval of increased magma supply and eruptive flux. O l i v ine whole-rock Fe/Mg relations indicate that ol iv ine phenocrysts with ~ F o 8 0 . 8 6 are in equi l ibr ium with parental magma composit ions of 8-10 wt% M g O . Un l ike the l o w - M g O ii lavas, these h igh -MgO basalts span a range of "depleted" isotopic composit ions ( 1 7 6 Hf / 1 7 7 Hf = 0.28292-0.28300, 8 7 S r / 8 6 S r = 0.7040-0.7046, 1 4 3 N d / 1 4 4 N d = 0.5127-0.5128, and 2 0 7 P b / 2 0 4 P b = 15.49-15.52) and represent the first documented occurrence of a depleted mantle component in mi ldly alkal ic basalts from the Kerguelen Archipelago. The presence of this depleted mantle component, wh ich is composit ional ly similar to that observed in the older 28-29 M a basalts in the northern part of the archipelago, reflects a short term increase in the extent of melting beneath the oceanic lithosphere and renewed interaction between Southeast Indian Ridge MORB- t ype asthenosphere and the enriched component of Kerguelen mantle plume at 25 M a . iii TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES vii LIST OF FIGURES viii ACKNOWLEDGEMENTS x CHAPTER 1: EVOLUTION OF KERGUELEN HOTSPOT VOLCANISM AND CONTEXT OF THE MARION DUFRESNE SECTION 1 1.1 INTRODUCTION 2 1.2 OVERVIEW 16 1.3 REFERENCES 19 CHAPTER 2: PETROLOGY AND PHENOCRYST MINERAL CHEMISTRY IN FLOOD BASALTS FROM THE THOLEIITIC-ALKALIC TRANSITION, MT. MARION DUFRESNE, KERGUELEN ARCHIPELAGO, SOUTHERN INDIAN OCEAN 27 2.1 INTRODUCTION 28 2.2 GEOLOGIC SETTING OF THE KERGUELEN ARCHIPELAGO 28 2.3 G E O L O G Y OF THE MARION DUFRESNE BASALTIC SECTION 30 2.4 ANALYTICAL TECHNIQUES 36 2.5 RESULTS 37 2.5.1 OLIVINE 37 2.5.2 PYROXENE 46 2.5.2.1 CLINOPYROXENE PHENOCRYSTS 46 2.5.2.2 PYROXENE CORONAS 52 2.5.3 PLAGIOCLASE 57 2.6 DISCUSSION 70 2.6.1 OLIVINE A N D CLINOPYROXENE M G - N U M B E R RELATIONS: IMPLICATIONS FOR THE PARENTAL M A G M A S OF THE MARION DUFRESNE SECTION 70 2.6.2 PYROXENE C O R O N A S A N D SIEVE-TEXTURED PLAGIOCLASE IN QUARTZ-BEARING BASALTIC ANDESITES: EVIDENCE FOR M A G M A MIXING 75 2.6.3 SIGNIFICANCE OF HIGH-AN PLAGIOCLASE COMPOSITIONS A N D P LAG IOC LAS E-ULTRAPHYRIC BASALTS 79 2.7 CONCLUSIONS 85 2.8 ACKNOWLEDGEMENTS 87 2.9 REFERENCES 88 iv CHAPTER 3: EVIDENCE FOR A DEPLETED MANTLE COMPONENT IN 25 MA MILDLY ALKALIC BASALTS FROM MT. MARION DUFRESNE, KERGUELEN ARCHIPELAGO, SOUTHERN INDIAN OCEAN 93 3.1 INTRODUCTION 94 3.2 GEOLOGIC SETTING OF THE KERGUELEN ARCHIPELAGO 95 3.3 G E O L O G Y OF THE MT. MARION DUFRESNE BASALTIC SECTION 99 3.4 ANALYTICAL TECHNIQUES 102 3.4.1 4 0 A R / 3 9 A R G E O C H R O N O L O G Y 102 3.4.2 MAJOR A N D TRACE ELEMENT COMPOSITIONS 104 3.4.3 ISOTOPIC COMPOSITIONS 106 3.5 RESULTS 108 3.5.1 AGE OF THE MARION DUFRESNE SECTION 108 3.5.2 MAJOR ELEMENT VARIATIONS 112 3.5.3 TRACE ELEMENT VARIATIONS 125 3.5.4 RADIOGENIC ISOTOPIC VARIATIONS 133 3.6 DISCUSSION 142 3.6.1 IMPLICATIONS FOR THE VOLCANIC STRATIGRAPHY OF THE PLATEAU CENTRAL 142 3.6.2 THE TRANSITION FROM THOLEIITIC TO MILDLY ALKALIC VOLCANISM O N THE KERGUELEN ARCHIPELAGO 148 3.6.3 SIGNIFICANCE OF THE DEPLETED MANTLE C O M P O N E N T 156 3.7 CONCLUSIONS 162 3.8 ACKNOWLEDGEMENTS 164 3.9 REFERENCES 165 CHAPTER 4: IMPLICATIONS FOR THE MAGMATIC EVOLUTION OF THE KERGUELEN ARCHIPELAGO 172 4.1 SUMMARY 173 4.2 REFERENCES 178 APPENDICES 179 APPENDIX I: EXAMPLE OF A MASS-BALANCE CALCULATION FOR CORRECTION OF OLIVINE A C C U M U L A T I O N . . . . 180 APPENDIX II: MAJOR (WT. % OXIDES) A N D TRACE ELEMENT ABUNDANCES IN MARION DUFRESNE LAVAS 181 APPENDIX 111: TRACE ELEMENT A B U N D A N C E S (PPM) IN MARION DUFRESNE SAMPLES BY HR-ICP-MS 185 APPENDIX IV: DUPLICATE ANALYSES OF TRACE ELEMENT ABUNCANCES BY HR-ICP-MS 190 V APPENDIX V: COMPARISON OF TRACE ELEMENT ABUNDANCES (PPM) IN SAMPLE 1140-31R 193 APPENDIX VI: COMPARISON OF TRACE ELEMENT A B U N D A N C E S DETERMINED BY XRF A N D HR-ICP-MS 194 ELECTRONIC APPENDICES SEE ATTACHED C D ELECTRONIC APPENDIX I: ALL OLIVINE PHENOCRYST CORE A N D RIM COMPOSITIONS BY EPMA ELECTRONIC APPENDIX II: ALL CLINOPYROXENE PHENOCRYST CORE A N D RIM COMPOSITIONS BY EPMA ELECTRONIC APPENDIX III: ALL PYROXENE C O R O N A COMPOSITIONS BY EPMA ELECTRONIC APPENDIX IV: ALL PLAGIOCLASE PHENOCRYST COMPOSITIONS BY EPMA vi LIST OF TABLES TABLE 2.1: Representative olivine phenocryst core and rim compositions from the Marion Dufresne section 40 TABLE 2.2: Representative clinopyroxene phenocryst core and rim compositions from the Marion Dufresne section 49 TABLE 2.3: Representative pyroxene corona compositions from the Marion Dufresne section 58 TABLE 2.4: Representative plagioclase phenocryst core and rim compositions from the Marion Dufresne section 60 TABLE 2.5: Starting compositions for MELTS runs and experimental studies 80 TABLE 3.1: Detailed 4 0 A r / 3 9 A r step4ieating results from the Marion Dufresne section, Kerguelen Archipelago 111 TABLE 3.2: Major (wt. % oxide) and trace element abundances (ppm) in Marion Dufresne lavas 118 TABLE 3.3: Hf, Sr, and Nd isotopic ratios in Marion Dufresne samples 134 TABLE 3.4: Pb isotopic ratios in Marion Dufresne samples 135 vii LIST OF FIGURES FIGURE 1.1: Bathymetric map of the Indian Ocean .-...3 FIGURE 1.2: Three-dimensional P-wave velocity image of the Kerguelen hotspot .....4 FIGURE 1.3: Plate reconstruction of eastern Gondwana at -140 Ma 5 FIGURE 1.4: Tectonic evolution of the Indian Ocean from -131 Ma to -24 Ma 7 FIGURE 1.5: Simplified geologic map of the Kerguelen Archipelago 10 FIGURE 1.6: Plate reconstruction for the southern Indian Ocean at -34 Ma 14 FIGURE 2.1: Simplified geologic map of the Kerguelen Archipelago 29 FIGURE 2.2: Stratigraphy of the >700 m high Mt. Marion Dufresne section 32 FIGURE 2.3: Total alkalis vs. silica diagram 35 FIGURE 2.4: Stratigraphic variations in phenocryst compositions from the Marion Dufresne section 38 FIGURE 2.5: Representative photomicrographs of olivine phenocrysts in high-MgO lavas from the Marion Dufresne section 42 FIGURE 2.6: Histogram of the forsterite content of all olivine analyses 44 FIGURE 2.7: Representative photomicrographs of clinopyroxene phenocrysts and microphenocrysts from the Marion Dufresne section 47 FIGURE 2.8: Pyroxene quadrilateral showing pyroxene compositions from the Marion Dufresne samples 50 FIGURE 2.9: Co-variation of non-quadrilateral components (Al, Na and Ti) in pyroxene 53 FIGURE 2.10: Representative photomicrographs of quartz xenocrysts rimmed by fine-grained pyroxene coronas 55 FIGURE 2.11: Representative photomicrographs of plagioclase phenocrysts 62 FIGURE 2.12: Feldspar ternary diagram 65 FIGURE 2.13: Plagioclase zoning profiles 67 FIGURE 2.14: Histograms of all plagioclase compositions 69 FIGURE 2.15: Marion Dufresne olivine (Fo) and clinopyroxene (mg-number) phenocryst and microphenocryst compositions vs. whole-rock mg-number 72 FIGURE 2.16: Textural evidence for magma mixing in quartz-bearing basaltic andesites 77 FIGURE 2.17: Variations in plagioclase compositions in Kerguelen Archipelago basalts vs. H 2 0 content, temperature, and pressure based on recent experimental studies and MELTS calculations 82 FIGURE 3.1: Bathymetric map showing the location of the Kerguelen Archipelago on the Kerguelen Plateau in the southern Indian Ocean 96 FIGURE 3.2: Simplified geologic map of the Kerguelen Archipelago 97 FIGURE. 3.3. Stratigraphy of the >700 m high vertical section of lavas exposed at Mt. Marion Dufresne '. 100 FIGURE 3.4: Step-heating spectra and inverse isochron diagrams for samples of leached whole rock from the base and top of the Marion Dufresne section 109 vi i i FIGURE 3.5: Total alkalis vs. silica diagram 113 FIGURE 3.6: Stratigraphic variations of major element abundances in Marion Dufresne lavas 114 FIGURE 3.7: MgO variation diagrams for samples from the Marion Dufresne section 116 FIGURE 3.8: Nb variation diagrams for samples from the Marion Dufresne section 126 FIGURE 3.9: Primitive mantle-normalized incompatible element abundances in Marion Dufresne samples 129 FIGURE 3.10: Chondrite-normalized rare earth element abundances in the Marion Dufresne samples 131 FIGURE 3.11: Variations in Hf-Sr-Nd-Pb initial isotopic compositions (i.e. at 25 Ma) with stratigraphic height in the Marion Dufresne section 136 FIGURE 3.12: Isotopic co-variation diagrams for Marion Dufresne lavas 138 FIGURE 3.13: Simplified geologic map of the Plateau Central region of the Kerguelen Archipelago 143 FIGURE 3.14: Stratigraphic variations of Alkalinity Index (A.!.), A l 2 0 3 (wt. %), MgO (wt. %), and Nb/Zr in the Marion Dufresne, Tourmente, and Capitole sections 145 FIGURE 3.15: Isotopic variations in basaltic sections from the Kerguelen Archipelago 149 FIGURE 3.16: Zr/Nb variations in Marion Dufresne samples compared with other Kerguelen Archipelago lavas 153 FIGURE 3.17: (La/Yb)N vs Zr/Nb diagram for Kerguelen Archipelago samples 157 FIGURE 3.18: , 4 3 N d / H 4 N d vs. Zr/Nb diagram for Kerguelen Archipelago samples 160 FIGURE 3.19: Nb/Y vs Zr/Y variations in Kerguelen Archipelago samples 161 FIGURE A1: Forsterite content of olivine vs. whole rock mg-number in sample BOB93-527 ..........180 ix ACKNOWLEDGEMENTS Years of hard work by many dedicated researchers established the necessary context for this study, and I thank Dominique Weis, James Scoates, Andre Giret, Fred Frey, and their students, for their achievements on the intricacies of Kerguelen geology. This thesis was greatly improved by the leadership, staff, and students of the Pacific Centre for Isotopic and Geochemical Research (PCIGR) at UBC, including Wilma Pretorius, Bruno Kieffer, Tom Ullrich, Jane Barling, Bert Mueller, Rob Mackie, Caroline-Emmanuelle Morriset, Diane Hanano, and Gwen Williams, who have all made significant contributions to this work and are acknowledged for sharing their time and expertise. My warmest thanks go to Claude Maerschalk at ULB, who came to my rescue with his expertise in chemical separation procedures. I thank my advisors, Dominque Weis and James Scoates, for all they have taught me. I am grateful to them for arranging for me to have the Marion Dufresne section as my thesis area, because it has been a puzzling and intriguing topic since the first day we looked at the thin sections, the first time the XRF data was plotted, and the first isotopic compositions were determined. Also, I will always appreciate the opportunity they gave me to visit Kerguelen in 2004, which was one of the great highlights of my entire life. I know that I am tremendously fortunate to be among the few that have visited Kerguelen, but I do regret that they were not there at the same time so I could benefit from the unique perspectives they both offer. I thank James Scoates for his extraordinary commitment to this manuscript- for reading and re-reading it, and for always looking for ways that it might be improved. Andre Giret at Universite Jean Monnet is gratefully acknowledged for providing access to samples from the Kerguelen Archipelago, for his extensive field work on the Kerguelen Archipelago, and for enabling continued field work there. I thank the French Polar Institute (IPEV/ IFRTP) for supporting research in the southern Indian Ocean; I am grateful to the crew and scientific party of the KerguePlac mission aboard the Marion Dufresne II, and in particular I thank commandant Jean-Marc Lefevre, chef de missions Walter Roest, and chef d'operations Bernard Ollivier. Field work and detailed sampling by Olivier Brisse and his colleagues in 1993 provided the opportunity to do an in-depth study of the Marion Dufresne section. Field work on the Kerguelen Archipelago in 2004 benefited from the help and organization of Francois Grosvalet, Alain Lamalle, Roland Pagni, Henri Perau, Pierre Camps, Thierry Poidras and many X others. I also thank the hivernants in Port-aux-Francais for their hospitality to our crew and for many memorable events. The greater EOS community has been wonderful to me during my time at UBC. I am fortunate to work with so many people that I get along with and admire. Mati Raudsepp and Elisabetta Pani have my gratitude and respect for running an exemplary microbeam facility at UBC that was enjoyable to work in. Constant help and encouragement from Maggie Harder and Liane Boyer have made this process easier and made the problems seem smaller. My family has always enthusiastically supported my goals and I thank them for their continued encouragement. Pat Hayman deserves a lot of credit for being an example of good work ethic and for being reliably thoughtful and generous. Leslie Reid ensured that I started off on the best path, and has remained an important voice of wisdom and motivation through my years at university. I also thank Nikolai and Esther Feilberg, who valued and inspired education, hard work, and adventure. xi CHAPTER 1 EVOLUTION OF KERGUELEN HOTSPOT VOLCANISM AND CONTEXT OF THE MARION DUFRESNE BASALTIC SECTION 1 1.1 INTRODUCTION The focus of this thesis is a 700 m high section of lava flows exposed at Mt. Marion Dufresne, in the southern Plateau Central region of the Kerguelen Archipelago. The Kerguelen Archipelago, which is an oceanic island located at approximately 49°S, 070°E in the southern Indian Ocean (FIG. 1.1), is an overseas territory of France and part of the Terres Australes et Antarctiques Franchises (TAAF). With a total surface area of >6500 km 2, the Kerguelen Archipelago is the largest oceanic island after Iceland, >100 000 km2, and Hawai'i, >10 000 km2, both of which have been the subject of extensive studies related to the origin and evolution of hotspot volcanism. The Kerguelen Archipelago, and the Northern Kerguelen Plateau on which it is situated, are interpreted to represent part of the Cenozoic volcanic record of the long-lived Kerguelen mantle plume. This interpretation is consistent with recent tomographic inverse modeling of the Kerguelen hotspot to 2350 km depth, which shows a robust thermal anomaly situated beneath the Kerguelen Archipelago region with a minimum radius of 400 km in the lower mantle (Montelli et a/., 2003) (FIG.1.2). The Kerguelen mantle plume has been active for ~132 Myr (Frey et al., 1996), and its magmatic products have a wide geographic distribution across the Indian Ocean basin (FIG. 1.1), due in part to the rapid expansion of the Indian Ocean throughout the magmatic record of the Kerguelen plume. Tectonic and paleomagnetic reconstructions indicate that eastern Gondwana comprised India, Australia, and Antarctica (FIG. 1.3) prior to the onset of rifting at -133 Ma that opened the nascent eastern Indian Ocean (Royer & Sandwell, 1989), which preceded the earliest known activity of the Kerguelen plume. Volumetrically minor volcanism on the continental margins of Australia, India, and Antarctica that has been attributed to the early Kerguelen plume includes the -1000 km3 Bunbury basalts of southwestern Australia (-132 and 123 Ma; Frey eta/., 1996), the 3 x 104 km3 Rajmahal basalts in northeastern India (-118 Ma; Kent et al., 2002), and lamprophyres in India and Antarctica (-114 Ma; Storey et al., 1989). Although there are global correlations between continental flood basaltic volcanism and continental breakup throughout the geologic record (e.g. Morgan, 1971; Courtillot et al., 1999), current age constraints do not conclusively indicate the influence of the Kerguelen mantle plume on continental breakup between India and the conjoined Australia-Antarctica. 2 FIG. 1.1. Bathymetric map of the Indian Ocean (after Smith & Sandwell, 1997). The major physiographic features of the Indian Ocean basin are labelled. The location of the Kerguelen Archipelago on the Northern Kerguelen Plateau (NKP) is indicated, as are the major magmatic products of the long-lived Kerguelen plume, which include the Southern Kerguelen Plateau (SKP), the Central Kerguelen Plateau (CKP), Broken Ridge, Ninetyeast Ridge, the Rajmahal traps, and the Bunbury basalts. O D P drilling sites on the Kerguelen Plateau are also indicated (e.g. Site 1140 on the Northern Kerguelen Plateau). 6 O S H 300 km 650 km 1000 km 1450 km 1900 km 2350 km 2800 km Approximate location of the Kerguelen hotspot 1 5 13 1.1 0.9 i - ' C 07 O <" 0.5 _Q 0.3 3 0.1 Q_ -0.1 & u -0.3 .2 m -0.5 > -0.7 % -0.9 JF -1.1 - " 3 -15 FIG. 1.2. Three-dimensional P-wave velocity image of the Kerguelen hotspot (after Montelli et a I., 2003). The surface map is approximately 40° by 40° and is scaled with increasing depth (note vertical exaggeration). Depth spacing increases at 1000 km depth. Colours in subsurface represent the average P-wave velocity perturbation (in %). The Kerguelen hotspot is imaged as a columnar low-velocity anomaly that extends from the surface into the lower mantle; lateral variations in wave velocities are interpreted to represent lower seismic wave velocities corresponding to higher temperatures in the plume conduit. The anomaly attributed to the Kerguelen hotspot is visibly stronger in the upper mantle (above 650 km) and becomes attenuated at depth. The location of the Kerguelen hotspot as indicated on the map is in the region of the Northern Kerguelen Plateau. 4 FIG. 1.3. Plate reconstruction of eastern Gondwana at - 140 Ma (after Ingle et al., 2004, and Boger et al., 2001). Prior to rifting at -133 Ma, India was situated in the southern hemisphere adjacent to Antarctica and Australia. The dashed line indicates a possible northern boundary for Greater India prior to the Himalayan orogeny. The major Proterozoic terranes on each continental block (dark shaded areas) are labelled. The future sites of the Bunbury Basalts and Rajmahal Traps, which are interpreted to be the earliest magmatic products of the Kerguelen mantle plume, are indicated by diamond symbols. 5 The vast basaltic Kerguelen Plateau (area, -2 x10 6 km2; volume, -2.4 x10 7 km3) in the southern Indian Ocean, which is the largest oceanic plateau after the Ontong-Java Plateau in the Western Pacific Ocean (area, -2 x10 6 km2; volume, 5.8 x10 7 km3; Coffi n and Eldholm, 1994; Fitton eta/., 2004), was formed subsequent to the impingement of the Kerguelen plume at the base of relatively young and thin oceanic lithosphere (e.g. Morgan, 1981; Duncan & Storey, 1992; Weis eta/., 1992), which resulted in the ascent and eruption of large quantities of basaltic magmas. The formation of the Southern and Central Kerguelen Plateaus and Broken Ridge (FIG. 1.4) is attributed to voluminous plume 'head' volcanism between -120 and -95 Ma (e.g. Whitechurch, 1992; Duncan, 2002; Frey eta/., 2000a) that was characterized by relatively high magma flux rates (-0.8-0.9 kmVyr; Coffin et al., 2002). Seismic refraction imaging (Operto & Charvis, 1996) and geochemical studies of the Southern Kerguelen Plateau (Mahoney et al., 1995; Neal et al., 2002; Ingle et al., 2003) indicate that the oldest parts of the Kerguelen Plateau may contain fragments of continental lithosphere that were stranded by continental breakup. Elan Bank, which extends west from the Kerguelen Plateau, is interpreted to be a microcontinent capped with basalt (Weis et al., 2001) (FIG. 1.1). Gneiss clasts recovered from fluvial sandstones on Elan Bank at Site 1137 of the Ocean Drilling Program (ODP) (Coffin eta/., 2000) have Neoproterozoic and Archean U-Pb zircon ages and may have been derived from the Eastern Ghats belt in India or the Rayner Complex in Australia (Nicolaysen eta/., 2001; Ingle eta/., 2002) (FIG. 1.3). This indicates that the early Kerguelen Plateau probably evolved near continental margins subsequent to the breakup of Gondwana (Weis et al., 2001). Plume 'tail' volcanism is represented by the <90 Ma magmatic products of the Kerguelen hotspot system (FIG. 1.4), which were erupted at significantly reduced magma flux rates (-0.1 km3/yr; Coffin et al. 2002). The Ninetyeast Ridge, a 5000 km long volcanic chain and the longest linear feature on Earth, formed between -82-37 Ma as a result of the northward migration of the Indian Plate over the Kerguelen hotspot (Mahoney, 1983; Frey et al., 1991; Weis et al., 1992; Frey & Weis, 1995). Spreading along the Southeast Indian Ridge (SEIR) at approximately 43 Ma rifted the Ninetyeast Ridge and the -95 Ma Broken Ridge from the Central Kerguelen Plateau (e.g. Mutter & Cande, 1983; Royer & Coffin, 1992; Tikku & Cande, 2000), and Broken Ridge is now situated -1800 km northeast of the Kerguelen 6 FIG. 1.4. The tectonic evolution of the Indian Ocean from -131 Ma to -24 Ma, (after Coffin et al., 2002; see references cited therein for complete explanations of the events illustrated in this figure). Kerguelen hotspot volcanism is indicated in black. The location of the Kerguelen hotspot is indicated by a white star (if the Kerguelen Archipelago is used as the present location of the hotspot) and a grey star (if Heard Island is used). Lamprophyres on Antarctica and India are indicated by small diamond symbols and labelled "L." The dashed line in A through D indicates the possible northern boundary of Greater India. A. 130.9 Ma. Seafloor spreading between India (IND) and Antarctica (ANT) occurred at -133 Ma, and Australia (AUS) and Antarctica were separated at -125 Ma. Rifting may have caused stranding of microcontinents (blocks of continental lithosphere) in the nascent Indian Ocean, e.g. Elan Bank (EB), which may have been attached to India at this time, and the Naturaliste Plateau (NP), off the southwestern coast of Australia. B. 118.7 Ma. The Indian Ocean was enlarged by continued rifting between India, Antarctica, and Australia. Early Kerguelen hotspot volcanism produced the Bunbury Basalts (BB) between -132 Ma and -123 Ma. C. 110 Ma. More voluminous magmatism attributed to the Kerguelen plume head produced the Southern Kerguelen Plateau (-118 Ma), the Rajmahal Traps (-118 Ma) and lamprophyres on India and Australia (114 Ma). D. 95 Ma. By 95 Ma, the Central Kerguelen Plateau (-105 Ma to -110 Ma) and Broken Ridge (-100 Ma to -95 Ma) had formed, and Elan Bank had separated from India. India moved to the north relative to the hotspot. The Wallaby Plateau (WP), which has not been dated, is inferred to have formed. E. 83 Ma. India continued to move northward relative to the hotspot. The northernmost extent of the Ninetyeast Ridge may have formed prior to -83 Ma, but it is interpreted to be buried beneath the Bengal Fan and has never been sampled. F. 62.5 Ma. The northward migration of India over the plume tail produced the Ninetyeast Ridge (NER), a linear chain of volcanoes between India and the conjoined Central Kerguelen Plateau-Broken Ridge, between -83 Ma and -38 Ma. Skiff Bank (SB), the western salient and oldest portion of the Northern Kerguelen Plateau, formed at -68 Ma. G. 46.3 Ma. By -46 Ma, magmatism from the Kerguelen mantle plume had formed an archetypal large igneous province: a vast oceanic plateau with a nearly 5000 km long hotspot track. H. 23.4 Ma. Subsequent to the formation of the southern portion of the 5000 km long Ninetyeast Ridge, a new mid-ocean ridge system propagating from the southeast (the Southeast Indian Ridge) intersected the Kerguelen large igneous province and rifted Broken Ridge and the Ninetyeast Ridge from the Kerguelen Plateau. Since -40 Ma, the distance between the hotspot and the ridge has increased, and the Kerguelen hotspot has formed the Northern Kerguelen Plateau and the Kerguelen Archipelago, as well as Heard and McDonald Islands and 21-18 Ma seamounts on the Central Kerguelen Plateau. 8 Plateau. Since -40 Ma, the Kerguelen hotspot has been situated in an intraplate location on the nearly stationary Antarctic Plate and has formed the Northern Kerguelen Plateau (area, 3.6 x 105 km2; volume, 2.3 x 106 km3). The Kerguelen Archipelago (<30 Ma), which is the focus of this study, is the subaerial expression of the Kerguelen mantle plume on the Northern Kerguelen Plateau (FIG. 1.5). Historical volcanic activity at Heard Island, as well as the presence of a chain of seamounts (18-21 Ma) between the Kerguelen Archipelago and Heard Island on the Central Kerguelen Plateau, appears to suggest that the latter is the present location of the Kerguelen hotspot (Weis et al., 2002), although hot springs and fumaroles are currently present on the Kerguelen Archipelago, and volcanic activity has occurred as recently as -26 ka at Rallier-du-Baty, in the southwestern part of the archipelago (Gagnevin et al., 2003) (FIG. 1.5). Although the majority of the magmatic record of the Kerguelen plume is now submerged on the Indian Ocean floor, the Kerguelen Archipelago is a unique environment where nearly 30 Myr of magmatism on the Northern Kerguelen Plateau remain subaerially exposed. A simplified geologic map of the Kerguelen Archipelago adapted from Nougier (1970) (FIG. 1.5) shows that Cenozoic (29-24 Ma; Nicolaysen et al., 2000) flood basalts cover the majority (-85%) of the surface area of this large oceanic island, with the younger lavas exposed in the east and southeast. Numerous plutonic complexes (<24 Ma) intrude the flood basalts and are exposed by erosion; these intrusive rocks include gabbros, syenites, and quartz-saturated monzogranites (i.e. Giret, 1983; Giret & Lameyre, 1983; Weis & Giret, 1994; Gagnevin et al., 2003; Scoates et al., 2005a). In addition, numerous small bodies of 6-10 Ma trachyte, phonolite, and basanite are observed in the Southeast Province (the Upper Miocene Series of Weis et al., 1993). The highest point on the archipelago is the 1850 m Mt. Ross stratovolcano, which is also the youngest volcanic edifice on the archipelago (0.1 -2 Ma; Weis et al., 1998). As indicated in FIG.1.5, large areas of the archipelago are covered by glacial till and Quaternary sediments (e.g. the 'Great Moraine' on the eastern Courbet Peninsula) due to the numerous glaciations and abundant precipitation that characterize this subantarctic environment. Glaciers, the largest of which is the Cook Glacier, occur in the western part of the archipelago as well as on the summit region of Mt. Ross. 9 0 10 20 30 40 50 km i i FIG. 1.5. Simplified geologic map of the Kerguelen Archipelago after Nougier (1970). The Kerguelen Archipelago is dominantly covered by flood basalts that range in age from - 2 9 Ma in the northern part of the archipelago to -24 Ma in the east and southeast. Mt. Marion Dufresne (black filled circle) is located in the southern part of the Plateau Central region of the archipelago. The locations and ages of previously studied basaltic sections are also indicated. 70 Glaciation has carved deep valleys and fjords across the Kerguelen Archipelago, in which the layers of volcanic rock that form this large oceanic island are exposed in thick (< 1000 m) sections. Flood basalts on the archipelago are typically 1-5 m thick and have a regional dip of 2-5° to the southeast, which exposes the oldest horizons in the northwest and youngest in the southeast. In general, flood basalts on the archipelago are dominantly phenocryst-poor with fractionated compositions (-3-6 wt. % MgO). Despite the compositional range observed on the Kerguelen Archipelago (from picrites to trachyandesites), the most widespread rock type is basalt and as such, all Cenozoic lavas on the archipelago are typically referred to as 'flood basalts' in recognition of their nearly flat-lying character and extensive areal coverage (Scoates & Giret, 2000). The geochemical evolution of flood basaltic volcanism on the Kerguelen Archipelago over time is the focus of a major ongoing study, involving researchers from Canada, Belgium, France and the United States. The primary goal of this project, to which this thesis represents a contribution, is to understand the magmatic processes that formed this large oceanic island, including the degrees and depths of melting and fractionation, magma mixing relationships, and mantle source compositions. The petrology and geochemistry of thick basaltic sections (300-1000 m high) from across the archipelago have been examined on the scale of individual lava flows. To date, studies have focused on Mt. Bureau and Mt. Rabouillere (Yang etal., 1998), Ravin du Charbon and Ravin Jaune (Frey etal., 2000b), Mt. des Ruches and Mt. Fontaine (Doucet et al., 2002), Mt. Tourmente (Frey et al., 2002), Mt. Crozier (Damasceno, 1996; Damasceno etal., 2002), Baie Charrier (Hanano, 2004), and high-MgO basaltic and picritic cobbles (not observed in outcrop) collected from across the southeast (Doucet et al., 2005); one additional study is in progress (Le Capitole; Xu et al., in prep) (FIG. 1.5). Based on geographic location and age (from 4 0 Ar/ 3 9 Ar geochronology by Nicolaysen etal.; 2000), these sections can be grouped into three categories: (1) -29-28 Ma transitional-tholeiitic lavas in the northern part of the archipelago (Bureau, Rabouillere, Ruches and Fontaine sections); (2) transitional lavas on the Plateau Central, which include Tourmente (-26 Ma), Capitole, and this study; and (3) 25-24 Ma mildly alkalic lavas in the east and southeast (Crozier, Charrier, Charbon, Jaune, and the high-MgO basalts and picrites) (FIG. 1.5). Although large areas of the archipelago remain unstudied, clear geographic, physical, and geochemical trends are evident based on the results of the studies cited in the previous paragraph. With decreasing eruption age (from -29 to -24 Ma) from northwest to southeast, flood basalts on the Kerguelen Archipelago are generally observed (1) to be more alkalic, (2) to contain more evolved (trachybasaltic to trachyandesitic) compositions, (3) to be increasingly interbedded with sedimentary layers, indicating longer interruptions between volcanic events, (4) to be more plagioclase-phyric, reflecting a stage of differentiation at relatively low pressures, (5) to be controlled by extensive high-pressure crystallization of high-AI clinopyroxene, and (6) to be the result of more shield-like volcanism (e.g. at 25-24 Ma on the Courbet Peninsula and in the Southeast Province). These observations reflect variations in numerous magmatic processes, but overall indicate an increase in the depth of partial melting, a decrease in the degree of partial melting, and a decrease in magma supply from the plume over time (Weis et al., 1998; Frey et al., 2000b; Scoates et al., 2005b). Importantly, the isotopic compositions of these basalts also changed significantly over this period of time; with decreasing eruption age, isotopic compositions in Kerguelen Archipelago flood basalts are more "enriched" (higher 8 7Sr/8 6Sr, 2 0 6 Pb/ 2 0 4 Pb, 2 0 7 Pb/ 2 0 4 Pb, 2 0 8 Pb/ 2 0 4 Pb and lower 1 7 6 H f / , 7 7 H f and , 4 3 Nd/ 1 4 4 Nd) and show smaller variations. Detailed geochemical studies of the 29-28 Ma tholeiitic-transitional basalts from the northern part of the archipelago reveal the occurrence of basalts with "depleted" isotopic compositions that may reflect an asthenospheric or lithospheric mantle component (with higher , 7 6 Hf/ 1 7 7 Hf, , 4 3 N d / 1 4 4 N d and lower 8 7Sr/ 8 6Sr, 2 0 6 Pb/ 2 0 4 Pb, 2 0 7 Pb/ 2 0 4 Pb, and 2 0 8Pb/ 2 0 4Pb) in their source (Yang et al., 1998; Doucet et al., 2002). Younger (<26 Ma) transitional lavas typically have more restricted and slightly more enriched isotopic compositions (Frey et al., 2000b; 2002; Doucet et al., 2005). The 25-24 Ma mildly alkalic lavas in the eastern part of the archipelago have a narrow range of distinctly more radiogenic isotopic compositions that are interpreted to represent the enriched component of the Kerguelen mantle plume source (Weis & Frey, 2002; Weis etal., 1998; Mattielli et al., 2002; Doucet et al., 2005), which is best represented by the composition of the -25-24 Ma Crozier basalts on the Courbet Peninsula (FIG. 1.5). 12 Given that basalts with depleted isotopic signatures occur within a restricted time interval and geographic location, Doucet et al. (2002) proposed that the depleted component is not intrinsic to the plume source, as is suggested for the depleted source in ocean island basalts from Iceland (e.g. Fitton et al., 1997) and Galapagos (e.g. Harpp & White, 2001), but rather that it was derived from depleted asthenospheric mantle. The -34 Ma submarine basalts recovered from ODP Site 1140, located approximately 270 km north of the Kerguelen Archipelago on the Northern Kerguelen Plateau (FIG. 1.1), contain a significant depleted component (greater than 90% in Unit 1 for all isotopic systems; Weis & Frey, 2002), which is interpreted to represent mixing of Kerguelen plume-sourced magmas with depleted asthenospheric mantle-derived magmas from the Southeast Indian Ridge (SEIR). It is important to note that at -34 Ma, the SEIR was located only -50 km from the Kerguelen hotspot (FIG. 1.6). Spreading along the SEIR over the past -40 Myr has progressively increased the distance between the SEIR axis and the Kerguelen hotspot as the ridge moved to the northeast relative to the hotspot. At -40 Ma, the hotspot was in a ridge-centred tectonic environment (comparable to present-day Iceland). The Kerguelen hotspot is currently situated on the Antarctic plate, -1200 km southeast of the SEIR, in an intraplate location comparable to that of the present-day Hawaiian hotspot. As a result, the Kerguelen Archipelago is an excellent natural laboratory in which to study plume-ridge interactions and the chemical and isotopic expressions of the transition to intraplate volcanism. The SEIR was -225 km from the Kerguelen hotspot when the oldest flood basalts on the archipelago were erupted, but was up to -400 km away by 24 Ma (Doucet et al., 2005). The occurrence of a clear depleted mantle component in the older lavas has been interpreted to represent communication between the plume and the ridge along sublithospheric pathways (Doucet eta/., 2002). As previously noted, the 25-24 Ma lavas from the Crozier section are interpreted to represent the enriched component of the Kerguelen mantle plume source and do not show evidence for a depleted mantle component in their source; this may indicate that the distance between the plume and the SEIR was too great to allow for continued magmatic interactions along these pathways at <25 Ma. The youngest flood basalts on the Kerguelen Archipelago were erupted at -24 Ma. Some small plutonic complexes (e.g. Val, Montagnes Vertes) on the archipelago with ages of 24-25 Ma likely 13 FIG. 1.6. Plate reconstruction for the southern Indian Ocean at -34 Ma (after Weis & Frey, 2002). Courtesy of the PLATES Project, University of Texas Institute for Geophysics. At 34 Ma, the Southeast Indian ridge crest (thick black line) lies midway between isochrons on the Antarctic and Australian plates (blue and red, respectively). The location of O D P Site 1140, which was - 5 0 km southeast of the SEIR at 34 Ma, is indicated by the circled cross. The locations of the future Kerguelen Archipelago (white star) and Heard Island (black star) are also indicated. The extent of the <40 Ma Northern Kerguelen Plateau at 34 Ma is unknown and its present-day area was used in this reconstruction. Abbreviations used: BR, Broken Ridge; NKP, Northern Kerguelen Plateau; KA, Kerguelen Archipelago; SB, Skiff Bank (-68 Ma), CKP, Central Kerguelen Plateau; SKP, Southern Kerguelen Plateau; EB, Elan Bank. 14 represent the high-level remnants of magma storage systems for the flood basalts (Scoates et al., 2005a). Strongly alkalic magmatism on the archipelago is represented by 6-10 Ma basanitic to phonolitic intrusions in the Southeast Province (Weis et al., 1993) (FIG. 1.5). Larger, high-level volcano-plutonic complexes (Rallier-du-Baty, Societe de Geographie) ranging from gabbro to granite and syenite were emplaced from -15 Ma to -26 ka (e.g. Giret & Lameyre, 1983; Weis & Giret, 1994; Gagnevin et al., 2003). The large 1850 m high Mt. Ross stratovolcano (0.1-2 Ma) consists of trachybasalts and trachytes (Weis et al., 1998). The linear chain of seamounts (basalts and picritic basalts) that extends from the Kerguelen Archipelago to Heard Island is interpreted to represent the 21-18 Ma hotspot track of the Kerguelen plume, which would indicate that the focus of hotspot volcanism shifted to the south after -24 Ma (Weis et al., 2002). However, coeval volcanism on the Kerguelen Archipelago and Heard Island after -24 Ma also suggests that hotspot volcanism became more diffuse over time, and the evolved compositions of Miocene lavas and intrusions on the Kerguelen Archipelago probably resulted from increasing magmatic stagnation and fractionation within the lithosphere (Weis etal., 2002). To better constrain the effects of the increasing distance between the SEIR and the Kerguelen hotspot on basalt composition and petrogenesis, this study examines the Mt. Marion Dufresne section, which is the southernmost section from the Plateau Central region of the Kerguelen Archipelago to be studied to date. Mt. Marion Dufresne (hereafter referred to as simply Marion Dufresne) is situated south of Mt. Tourmente (-26 Ma) and west of the 25-24 Ma mildly alkalic lavas that form the Southeast Province and the Courbet Peninsula (FIG. 1.5). A total of 47 samples from Marion Dufresne were collected in 1993 during a French regional mapping program (CartoKer). This section was chosen as the basis for this study due to (1) its location in a previously unstudied area of the archipelago, (2) its position oblique to the NW-SE trends defined by previously studied sections, (3) the height of the section (-700 m), (4) the careful sampling of individual lava flows, dikes, and sills, (5) the diversity of phenocryst types noted during sampling (olivine, clinopyroxene, and plagioclase), and (6) the approximate time interval represented by this section (between 26-25 Ma), which is critical for understanding the temporal evolution of magmatism on the Kerguelen Archipelago. 15 This study is intended to complement previous work on the flood basalts of the Kerguelen Archipelago and of the Plateau Central region (e.g. Frey et al., 2002). Because the transition from transitional-tholeiitic to mildly alkalic volcanism on the Kerguelen Archipelago is interpreted to occur within the Plateau Central region (Frey etal., 2002), the results of this study will provide important constraints for the timing and nature of this transition and the magmatic processes that it represents. In addition, the nature and contribution of the depleted mantle component in Kerguelen Archipelago flood basalts from the Plateau Central region is poorly understood. Although there is some evidence for a depleted mantle component in the source of the Tourmente basalts (Frey et al., 2002), only two sections from the vast Plateau Central region have been studied to date, and both are located north of Marion Dufresne (Mt. Tourmente: Frey etal., 2002; Le Capitole: Xu etal., in prep). The Tourmente and Capitole sections consist of transitional basalts and volumetrically minor mildly alkalic basalts with very homogeneous isotopic compositions. Basalts from these sections also have distinctly lower 8 7Sr/ 8 6Sr and higher 1 4 3 N d / l 4 4 N d compared to the 25-24 Ma lavas, which has been interpreted to reflect efficient mixing of plume-derived melts with a constant proportion or small volume of depleted mantle material, or heterogeneities intrinsic to the plume (Frey et al., 2002). 1.2 OVERVIEW OF THE THESIS This thesis comprises two distinct, yet complementary studies, which are presented in Chapters 2 and 3. Chapter 2 describes the petrologic characteristics of samples from the Marion Dufresne section, and contains a detailed examination of phenocryst mineral chemistry (olivine, clinopyroxene, and plagioclase; see below). Chapter 3 is a geochemical evaluation of the major and trace element and radiogenic isotopic compositions of Marion Dufresne samples (see below). Chapter 3 expands upon the findings in Chapter 2, and some of the results presented in Chapter 3 (e.g. major element compositions) are used in a limited way in Chapter 2. Both chapters were prepared in a manuscript format appropriate for submission to international scientific journals. As such, there are some areas of overlap between these chapters, but repetition is limited to key introductory figures and more general descriptions of the Kerguelen Archipelago and the Marion Dufresne section, which are presented 16 within a context relevant to each chapter. Chapter 4 is a brief summary of the major findings of this thesis, with implications for the magmatic evolution of the Kerguelen Archipelago and of the Indian Ocean basin. The results of an extensive examination of phenocryst mineral chemistry in 26 samples from the Marion Dufresne section are presented in Chapter 2. This study consists of compositional data determined by electron microprobe, including 432 plagioclase analyses, 222 olivine analyses, 56 clinopyroxene analyses, and 120 pyroxene corona analyses. This work was initially the focus of a term project for a course entitled 'Microbeam and Diffraction Methods for the Characterization of Minerals and Materials' (EOSC 521), and preliminary results from olivine and clinopyroxene mineral chemistry were included in the report required for partial completion of this course. This is only the second major mineral chemistry study to focus on flood basalts from the Kerguelen Archipelago (the first being Damasceno et ah, 2002), which is due primarily to the relative scarcity of phenocrysts in most Kerguelen Archipelago lavas. In ocean island settings, phenocryst mineral chemistry has been successfully used to determine parental magma compositions, to constrain depths and temperatures of fractionation, and to determine magma contamination through mixing or interaction with wall rocks (e.g. Hawaii: Thornber, 2001; Garcia, 1996; Garcia et al., 2003; Yang et al., 1999; Galapagos: Geist et al., 2002; Cullen etal., 1989; Iceland: Hansen & Gronvold, 2000). The Marion Dufresne section contains a greater diversity and abundance of phenocrysts than most Kerguelen Archipelago lavas, and therefore represents an exceptional opportunity to improve our understanding of magmatic processes affecting basalt petrogenesis on the Kerguelen Archipelago. To constrain the maximum MgO content of the parental magma for the Marion Dufresne lavas, Fe-Mg mineral-melt equilibrium relationships are determined for samples from a thick (-400 m) sequence of olivine-phyric high-MgO (7-12 wt% MgO) basalts. To reveal magma mixing relationships in three samples of quartz-bearing basaltic andesite, the compositions of pyroxene crystals from reaction coronas surrounding quartz xenocrysts and plagioclase phenocrysts with disequilibrium characteristics (e.g. sieve textures and reverse zoning) are used to characterize the distinct endmembers involved. Finally, the significance of high-An plagioclase (>An80) in plagioclase-phyric and plagioclase-ultraphyric basalts from the Marion Dufresne section is addressed, with implications for the existence of shallow magma storage and increasing depths of fractionation at 25 Ma. In Chapter 3, the age, major and trace element, and radiogenic isotopic compositions (Hf-Sr-Nd-Pb) of lavas from the Marion Dufresne section are presented and compared with results from other basaltic sections on the archipelago (1) to identify the geochemical trends that distinguish the Plateau Central region, (2) to constrain and characterize the transition from tholeiitic to mildly alkalic volcanism on the Kerguelen Archipelago, (3) to characterize the mantle source components for this basaltic section, and (4) to constrain the chemical evolution of the Kerguelen mantle plume overtime. Lavas from the Marion Dufresne section have highly variable major and trace element abundances and isotopic compositions compared with previously studied sections from the Plateau Central (Tourmente and Capitole). The differences between these three sections with respect to source composition and temporal changes in magmatic processes are examined. The results from all three Plateau Central sections are subsequently combined to identify the nature and timing of the transition to mildly alkalic volcanism on the archipelago, and the factors involved in this transition, with implications for degrees and depths of melting associated with the Kerguelen mantle plume. Finally, the importance of the highly variable isotopic compositions of the Marion Dufresne lavas is examined, and the findings from this study are compared to previously studied basaltic sections from the Kerguelen Archipelago (29-24 Ma), along with SEIR mid-ocean ridge basalts and submarine basalts on the Northern Kerguelen Plateau, to better define the occurrence and composition of the depleted mantle source component in Kerguelen Archipelago flood basalts. Chapter 4 provides a summary of the principal findings of this thesis. This study has produced many unexpected and exciting results that, when considered within the context of volcanism across the Kerguelen Archipelago, provide important new insights into the evolution of the Kerguelen mantle plume during the Cenozoic. 18 1.3 REFERENCES Boger, S.D., Wilson, C.J.L., & Fanning, C M . (2001): Early Paleozoic tectonism within the East Antarctic craton: the final suture between east and west Gondwana? Geology 29, 463-466. Coffin, M.F. & Eldholm, O. (1994): Large igneous provinces: crustal structure, dimensions, and external consequences. Reviews of Geophysics 32, 1-36. Coffin, M.F., Frey, F.A., Wallace, P.J., et al. (eds) (2000): Proceedings of the Ocean Drilling Program, Scientific Results, 183. College Station, TX: Ocean Drilling Program, 101 p. Coffin, M.F., Pringle, M.S., Duncan, R.A., Gladczenko, T.P., Storey, M. , Muller, R.D. & Gahagan, L.A. (2002): Kerguelen hotspot output since 130 Ma. Journal of Petrology 43, 1121-1140. Courtillot, V., Jaupart, C , Manighetti, I., Tapponnier, P. & Besse, J. (1999): On causal links between flood basalts and continental breakup. Earth and Planetary Science Letters 166, 177-195. Cullen, A., Vicenzi, E. & McBirney, A.R. (1989): Plagioclase-phyric basalts of the Galapagos Archipelago. Journal of Volcanology and Geothermal Research 37, 325-337 Damasceno, D. (1996): Evolution magmatique de la serie basaltique du Mont Crozier (Archipel de Kerguelen, ocean Indien): etude petrologique et chimique. Unpublished B.Sc. thesis, Universite Libre de Bruxelles, Brussels, Belgium, 47 p. Damasceno, D. ( Scoates, J.S., Weis, D., Frey, F.A., & Giret, A. (2002): Mineral chemistry of mildly alkalic basalts from the 25 Ma Mont Crozier section, Kerguelen Archipelago: constraints on phenocryst crystallization environments. Journal of Petrology 43, 1389-1413. Doucet, S., Weis, D., Scoates, J.S., Nicolaysen, K., Frey, F.A. & Giret, A. (2002): The depleted mantle component in Kerguelen Archipelago basalts: petrogenesis of tholeiitic-transitional basalts from the Loranchet Peninsula. Journal of Petrology 43, 1341 -1366. Doucet, S., Scoates, J.S., Weis, D. & Giret, A. (2005): Constraining the components of the Kerguelen mantle plume: a Hf-Pb-Sr-Nd isotopic study of picrites and high-MgO basalts from the Kerguelen Archipelago. Geochemistry Geophysics Geosystems (in press). 19 Duncan, R.A. & Storey, M . (1992): The life cycle of Indian Ocean hotspots. In: Duncan, R.A., Rea, D.K., Kidd, R.B., von Road, U. & Weissel (eds) Synthesis of Results from Scientific Drilling in the Indian Ocean. Geophysical Monograph 70, American Geophysical Union, pp. 91-103. Duncan, R.A. (2002): A time frame for construction of the Kerguelen Plateau and Broken Ridge. Journal of Petrology 43, 1109- 1120. Fitton, J.G., Saunders, A.D., Norry, M.J., Hardarson, B.J. & Taylor, R.N. (1997): Thermal and chemical structure of the Iceland plume. Earth and Planetary Science Letters 153, 197-208. Fitton, J.G., Mahoney, J.J., Wallace, P.J. & Saunders, A.D. (2004): Origin and evolution of the Ontong Java Plateau: Introduction. In: Fitton, J.G., Mahoney, J.J., Wallace, P.J. & Saunders, A.D. (eds) Origin and Evolution of the Ontong Java Plateau. Geological Society, London, Special Publication 229, pp. 1-8. Frey, F.A., Jones, W.B., Davies, H. & Weis, D. (1991): Geochemical and petrologic data for basalts from Sites 756, 757, and 758: Implications for the origin and evolution of Ninetyeast Ridge. In: Weissel, J., Peirce, J., Taylor, E. & Alt, J. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 121. College Station, TX: Ocean Drilling Program, pp 611-659. Frey, F.A., & Weis, D. (1995): Temporal evolution of the Kerguelen plume: geochemical evidence from -39 to 82 Ma lavas forming the Ninetyeast Ridge. Contributions to Mineralogy and Petrology 121, 12-28. Frey, F.A., McNaughton, N.J., Nelson, D.R., deLaeter, J.R. & Duncan, R.A. (1996): Petrogenesis of the Bunbury basalt, western Australia: Interaction between the Kerguelen plume and Gondwana lithosphere? Earth and Planetary Science Letters 144, 163-183. Frey, F.A., Coffin, M.F., Wallace, P.J., Weis, D., Zhao, X., Wise, S.W., Wahnert, V., Teagle, D.A.H., Saccocia, P.J., Reusch, D.N., Pringle, M.S., Nicolaysen, K.E., Neal, C.R., Muller, R.D., Moore, C.L., Mahoney, J.J., Keszthelyi. L., Inokuchi, H., Duncan, R.A., Delius, H., Damuth, J.E., Damasceno, D., Coxall, H.K., Borre, M.K., Boehm, F., Barling, J., Arndt, N.T. & Antretter, M . (2000a): Origin and evolution of a submarine large igneous province: the Kerguelen Plateau and Broken Ridge, southern Indian Ocean. Earth and Planetary Science Letters 176, 73-89. Frey, F.A., Weis, D., Yang, H.-J., Nicolaysen, K., Leyrit, H. & Giret, A (2000b): Temporal geochemical trends in Kerguelen Archipelago basalts: evidence for decreasing magma supply from the Kerguelen plume. Chemical Geology 164, 61-80. Frey, F.A., Nicolaysen, K., Kubit, B.K., Weis, D. & Giret, A. (2002): Flood basalt from Mont Tourmente in the Central Kerguelen Archipelago: the change from transitional to alkalic basalt at -25 Ma. Journal of Petrology 43, 1367-1387. Gagnevin, D., Ethien, R., Bonin, B., Moine, B., Feraud, G., Gerbe, M.C., Cottin, J.-Y., Michon, G., Tourpin, S., Mamias, G., Perrache, C. & Giret, A. (2003): Open-system processes in the genesis of silica-oversaturated alkaline rocks of the Rallier-du-Baty Peninsula, Kerguelen Archipelago (Indian Ocean). Journal of Volcanology and Geothermal Research 123/ 267-300. Garcia, M.O. (1996): Petrography and olivine and glass chemistry of lavas from the Hawaiian Scientific Drilling Project. Journal of Geophysical Research 101, 11701-11713. Garcia, M.O., Pietruszka, A.J., & Rhodes, J.M. (2003): A petrologic perspective of Kilauea Volcano's summit magma reservoir. Journal of Petrology 44, 2313- 2339. Geist, D., White, W.M. , Albarede, F., Harpp, K., Reynolds, R., Blichert-Toft, J. & Kurz, M.D. (2002): Volcanic evolution in the Galapagos: the dissected shield of Volcan Ecuador. Geochemistry Geophysics Geosystems 3, 10.1029/2002GC000355. Giret, A. (1983): Le plutonisme oceanique intraplaque, exemple de I'archipel Kerguelen. These doctorat d'Etat, Universite Pierre et Marie Curie, Paris, France, 290 p. Giret, A. & Lameyre, J., (1983): A study of Kerguelen plutonism: petrology, geochronology and geological implications. In: Oliver, R.L., James, P.R. & Jago, J.B. (eds.) Antarctic Earth Science. Australia: Australian Academy of Sciences, pp. 646-651. Hanano, D. (2004): Geochemistry of the Baie Charrier basaltic section, Courbet Peninsula, Kerguelen Archipelago: Implications for the composition of the Kerguelen mantle plume. Unpublished B.Sc. thesis, University of British Columbia, Vancouver, Canada, 92 p. Hansen, H., & Gronvold, K. (2000): Plagioclase ultraphyric basalts in Iceland: the mush of the rift. Journal of Volcanology and Geothermal Research 98, 1 -32 Harpp, K.S. & White, W.M. (2001): Tracing a mantle plume: Isotopic and trace element variations of Galapagos seamounts. Geochemistry Geophysics Geosystems 2, 10.1029/2000GC000137. Ingle, S., Weis, D., Scoates, J.S., & Frey, F.A. (2002): Relationship between the early Kerguelen plume and continental flood basalts of the paleo-Eastern Gondwanan margins. Earth and Planetary Science Letters 197, 35-50. Ingle, S., Weis, D., Doucet, S. & Mattielli, N . (2003): Hf isotope constraints on mantle sources and shallow-level contaminants during Kerguelen hotspot activity since -120 Ma. Geochemistry Geophysics Geosystems 4, 10.1029/2002GC000482. Ingle, S., Scoates, J.S., Weis, D., Brijgmann, G., & Kent, R.W. (2004): Origin of cretaceous continental tholeiites in southwestern Australia and eastern India: Insights from Hf and Os isotopes. Chemical Geology 209, 83-106. Kent, R.W., Pringle, M.S., Muller, R.D., Saunders, A.D., & Ghose, N.C. (2002): 4 0 Ar/ 3 9 Ar geochronology of the Rajmahal basalts, India, and their relationship to the Kerguelen Plateau. Journal of Petrology 43, 1141-1153. Lameyre, J., Marot, A., Zimine, S., Cantagrel, J.-M., Sosso, L. & Vidal, P. (1976): Chronological evolution of the Kerguelen islands syenite-granite ring complex. Nature 263, 306-307. Mahoney, J.J., McDougall, J.D., Lugmair, G.W. & Gopalan, K. (1983): Kerguelen hot spot source for the Rajmahal traps and Ninetyeast Ridge? Nature 303, 385-389. Mahoney, J.J., Jones, W.B., Frey, F.A., Salters, V.J.M., Pyle, D.G. & Davies, H.L. (1995): Geochemical characteristics of lavas from Broken Ridge, the Naturaliste Plateau, and southernmost Kerguelen Plateau: Cretaceous volcanism in the southeast Indian Ocean. Chemical Geology 120,315-345. Mattielli, N. , Weis, D., Blichert-Toft, J. & Albarede, F. (2002): Hf isotope evidence for a Miocene change in the Kerguelen mantle plume composition. Journal of Petrology 43, 1327-1339. Montelli, R., Nolet, G., Dahlen, F.A., Masters, G., Engdahl, E.R. & Hung, S.-H. (2003): Finite-frequency tomography reveals a variety of plumes in the mantle. Science 303, 338-343. Morgan, W.J. (1971): Convection plumes in the lower mantle. Nature 230, 42-43. Morgan, W.J. (1981): Hotspot tracks and the opening of the Atlantic and Indian Oceans. In: Emiliani, C. (ed), The Sea, 7. New York: Wiley, pp. 443-487. Mutter, J.C. & Cande, S.C. (1983): The early opening between Broken Ridge and Kerguelen Plateau. Earth and Planetary Science Letters 65, 369-376. Neal, C.R., Mahoney, J.J. & Chazey, W.J. (2002): Mantle sources and the highly variable role of continental lithosphere in basalt petrogenesis of the Kerguelen Plateau and Broken Ridge LIP: Results from ODP Leg 183. Journal of Petrology 43, 1177-1205. Nicolaysen, K., Frey. F.A., Hodges, K.V., Weis, D & Giret, A. (2000): 4 0 Ar/ 3 9 Ar geochronology of flood basalts from the Kerguelen Archipelago, southern Indian Ocean: implications for Cenozoic eruption rates of the Kerguelen plume. Earth and Planetary Science Letters 174, 313-328. Nicolaysen, K., Bowring, S., Frey, F., Weis, D., Ingle, S., Pringle, M.S., & Coffin, M.F. (2001): Provenance of Proterozoic garnet-biotite gneiss recovered from Elan Bank, Kerguelen Plateau, southern Indian Ocean. Geology 29, 235-238. Nougier, J. (1970): Carte geologique au 1/200 000 de I'archipel des Kerguelen. Paris: Institut Geographique National. Operto, S., & Charvis, P. (1996): Deep structure of the southern Kerguelen Plateau (southern Indian Ocean) from ocean bottom seismometer wide-angle seismic data. Journal of Geophysical Research 101, 25077-25103. Royer, J.-Y. & Sandwell, D.T. (1989): Evolution of the eastern Indian Ocean since the late Cretaceous: constraints from Geosat altimetry. Journal of Geophysical Research 94, 13755-13782. Royer, J.-Y. & Coffin M.F. (1992): Jurassic to Eocene plate tectonic reconstructions in the Kerguelen Plateau region. In: Wise, W.S., Schlich, R. eta/., (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 120. College Station, TX: Ocean Drilling Program, pp 917-930. Scoates, J.S. & Giret, A. (2000): Physical volcanology of flood basalts on the Kerguelen Archipelago. EOS 81, S429. 23 Scoates, J.S., Weis, D., Franssens, M. , Mattielli, N. , Annell, H., Frey, F.A., Nicolaysen, K. & Giret, A. (2005a): The Val gabbro plutonic suite, Kerguelen Archipelago: evolution and significance of a volcanic feeder system in an ocean island. To be submitted to Journal of Petrology. Scoates, J.S., Lo Cascio, M. , Weis, D., & Lindsley, D.H. (2005b): Experimental constraints on the origin and evolution of mildly alkalic basalts from the Kerguelen Archipelago, southeast Indian Ocean. Submitted to Contributions to Mineralogy and Petrology. Smith, W.H.F. & Sandwell, D.T. (1997): Global seafloor topography from satellite altimetry and ship depth soundings. Science 277, 1956-1962. Storey, M. , Saunders, A.D., Tarney, J., Gibson, I.L., Norry, M.J., Thirwall, M.F., Leat, P., Thompson, R.N. & Menzies, M.A. (1989): Contamination of Indian Ocean asthenosphere by the Kerguelen-Heard mantle plume. Nature 338, 574-576. Thornber, C.R. (2001): Olivine-liquid relations of lava erupted by Kilauea Volcano from 1994 to 1998: implications for shallow magmatic processes associated with the ongoing East Rift Zone eruption. Canadian Mineralogist 39, 239-266. Tikku, A.A. & Cande, S.C. (2000): On the fit of Broken Ridge and Kerguelen plateau. Earth and Planetary Science Letters 180, 117-132. Weis, D. & Frey, F.A. (1991). Isotope geochemistry of Ninetyeast Ridge basalts: Sr, Nd, and Pb evidence for the involvement of the Kerguelen hot spot. In: Weissel, J., Peirce, J., Taylor, E., Alt, J., et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 121. College Station, TX: Ocean Drilling Program, pp. 591-610. Weis, D., White, W.M. , Frey, F.A., Duncan, R.A., Dehn, J., Fisk, M. , Ludden, J., Saunders, A & Storey, M . (1992): The influence of mantle plumes in generation of Indian oceanic crust. In: Duncan, R.A., Rea, D.K., Kidd, R.B., von Road, U. & Weissel (eds) Synthesis of Results from Scientific Drilling in the Indian Ocean. Geophysical Monograph 70, American Geophysical Union, pp. 57-89. Weis, D., Frey, F.A., Leyrit, H. & Gautier, I. (1993): Kerguelen Archipelago revisited: geochemical and isotopic study of the SE Province lavas. Earth and Planetary Science Letters 118, 101-119. Weis, D. & Giret, A. (1994): Kerguelen plutonic complexes: Sr, Nd, Pb isotopic study and inferences about their sources, ages and geodynamic setting. Bulletin de la Societe Ceologique, France, 166,47-59. Weis, D., Frey, F.A., Giret, A. & Cantagrel, J.-M. (1998): Geochemical characteristics of the youngest volcano (Mount Ross) in the Kerguelen Archipelago: inferences for magma flux, lithosphere assimilation, and composition of the Kerguelen Plume. Journal of Petrology 39, 973-994. Weis, D., Ingle, S., Nicolaysen, K., Frey, F.A., Damasceno, D., Barling, J., and Leg 183 Shipboard Scientific Party (2001): Origin of continental components in Indian Ocean Basalts: Evidence from Elan Bank (Kerguelen Plateau, ODP Leg 183, Site 1137). Geology 29, 147-150. Weis, D. & Frey, F.A. (2002): Submarine basalts of the Northern Kerguelen Plateau: interaction between the Kerguelen plume and the Southeast Indian ridge revealed at ODP Site 1140. Journal of Petrology 43, 1287-1309. Weis, D., Frey, F.A., Schlich, R., Schaming, M. , Montigny, R., Damasceno, D., Mattielli, N. , Nicolaysen, K.E. & Scoates, J.S. (2002): Trace of the Kerguelen mantle plume: evidence from seamounts between the Kerguelen Archipelago and Heard Island, Indian Ocean. Geochemistry Geophysics Geosystems 3, 10.1029/2001GC000251. Whitechurch, H., Montigny, R., Sevigny, J., Storey, M . & Salters, V. (1992): K-Ar and 4 0 Ar/ 3 9 Ar ages of central Kerguelen Plateau basalts. In: Wise, W.S., Schlich, R. etal., (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 120. College Station, TX: Ocean Drilling Program, pp 71-77. Xu, G., Frey, F., Weis, D., Scoates, J. S. & Giret, A. (2005): Geochemical characteristics of flood basalts from Mont Capitole in the central Kerguelen Archipelago: further constraints on the Kerguelen plume. In preparation. Yang, H.-J., Frey, F.A., Weis, D., Giret, A., Pyle, D. & Michon, G. (1998): Pedogenesis of the flood basalts forming the Northern Kerguelen Archipelago: implications for the Kerguelen plume. Journal of Petrology 39, 711 -748. 25 Yang, H.-J., Frey, F.A., Clague, D.A., & Garcia, M.O. (1999): Mineral chemistry of submarine lavas from Hilo Ridge, Hawaii: implications for magmatic processes within Hawaiian rift zones. Contributions to Mineralogy and Petrology 135, 355-372. 26 CHAPTER 2 PETROLOGY AND PHENOCRYST MINERAL CHEMISTRY IN FLOOD BASALTS FROM THE THOLEIITIC-ALKALIC TRANSITION, MT. MARION-DUFRESNE, KERGUELEN ARCHIPELAGO, SOUTHERN INDIAN OCEAN 27 2.1. INTRODUCTION The assemblage and composition of phenocrysts contained within a basaltic rock are an expression of the many processes acting on the magma during ascent and crystallization. Phenocryst compositions, compositional gradients, and mineral-melt equilibria can be used to constrain the conditions under which different minerals were crystallized, and the extent to which parental magmas were fractionated, mixed with other magmas, or interacted with wall rocks. Few studies of phenocryst chemistry have been carried out in the -6500 km2 Kerguelen Archipelago (i.e. Damasceno et al., 2002) due primarily to the relative scarcity of phenocrysts in these lavas. In this study, we present a detailed evaluation of the petrology and phenocryst compositions from the -25 Ma Mt. Marion Dufresne section in the Plateau Central region of the archipelago. These lavas are geochemically variable with atypically high phenocryst volumes (up to -60 vol. %). The compositions of olivine and clinopyroxene phenocrysts are combined with Fe-Mg mineral-melt equilibria to constrain the maximum MgO content of the parental magma for basalts from this section, and to enable a characterization of the effects of phenocryst and xenocryst accumulation on observed whole rock compositions. We present textural and compositional evidence of magma mixing between high-MgO basaltic and quartz-bearing evolved (trachytic) magmas in a three quartz-bearing basaltic andesite flows, which suggests the occurrence of local variations in magma flux. We also address the petrologic significance of high-An plagioclase phenocrysts (>An80) in the relatively evolved basaltic magmas of the Marion Dufresne section. Temporal changes in magma flux and storage processes evident from mineral chemistry are used to explain the origin of the geochemical diversity observed at Marion Dufresne in comparison to other Kerguelen Archipelago basaltic sections and allow for a better understanding of magmatism associated with the Kerguelen mantle plume at -25 Ma. 2.2 GEOLOGIC SETTING OF THE KERGUELEN ARCHIPELAGO The Kerguelen Archipelago is a large oceanic island located at approximately 49°S, 070°E in the southern Indian Ocean. It comprises a central island of -100 km width surrounded by over 300 smaller islands (FIG. 2.1). With a total surface area of >6500 km2, it is the third largest oceanic island 28 0 10 20 30 40 50 km 1 i i i i i FIG. 2.1. Simplified geologic map of the Kerguelen Archipelago after Nougier (1970) showing the location of the Marion Dufresne section in the southern Plateau Central region (black filled circle). The locations and ages of previously studied basaltic sections are indicated. Mt. Tourmente and Mt. Capitole, the most proximal sections to Mt. Marion Dufresne, are the only other sections from the Plateau Central to be studied to date. 29 after Iceland and Hawaii. The Kerguelen Archipelago is the emergent part of the Northern Kerguelen Plateau, which is an oceanic plateau up to 20-25 km thick (Recq et al., 1994; Charvis et al., 1995; Charvis & Operto, 1999). The Kerguelen Plateau, the second largest oceanic plateau after the Ontong-Java Plateau in the western Pacific Ocean, is attributed to volcanism from the Kerguelen mantle plume, which has produced -2.5 x 107 km3 of mafic oceanic crust in the past 130 million years (Coffin etal., 2002). Cenozoic (29-24 Ma) flood basalts cover approximately 85% of the surface area of the Kerguelen Archipelago (FIG. 2.1). Numerous younger alkalic volcanic edifices, largely concentrated in the southeast, include smaller phonolite and basanite high-level intrusions and the recent (2-0.1 Ma) Mt. Ross stratovolcano that forms the highest peak on the archipelago at 1850 m (Weis et al., 1993; 1998). Large plutonic complexes exposed by erosion represent shallow sub-volcanic feeder systems and the remnants of magma reservoirs (e.g. Weis & Giret, 1994; Scoates et al., 2005b). The many deep, narrow valleys and fjords marking the extensive coastline of the Kerguelen Archipelago reflect the numerous glaciations and abundant precipitation that characterize this sub-Antarctic environment. The resulting erosion has exposed thick sections through the layers of volcanic rock, permitting direct observation and sampling of individual basalt flows across a major oceanic island. Basalt flows on the archipelago have an overall regional dip of 2-5° to the southeast, which exposes the oldest successions in the northwest and the youngest in the southeast. Based on 4 0 Ar/ 3 9 Ar dating of basaltic rocks from across the archipelago, there is a clear trend from older 29-28 Ma transitional-tholeiitic basalts in the northern part of the archipelago to the overlying younger (25-24 Ma) mildly alkalic basalts in the southeast (e.g. Nicolaysen et al., 2000; Frey et al., 2000; Doucet et al., 2002). 2.3 GEOLOGY OF THE MARION DUFRESNE BASALTIC SECTION In this study, we examine lavas from Mt. Marion Dufresne, a prominent east-west trending ridge located on the east side of the Cook Glacier (FIG. 2.1). Marion Dufresne is the southernmost section from the Plateau Central to be studied to date, and was chosen based on its location in a previously unstudied area of the archipelago, its position oblique to the NW-SE trends defined by 30 previously studied sections, its relative age, its height (>700 m), and the abundance of olivine ± clinopyroxene ± plagioclase phenocrysts noted during sampling. Glaciation has exposed a nearly 700 m stratigraphic section of subaerially erupted lava flows with thicknesses of 1-15 m (FIG. 2.2). These flows are separated by intervals of debris, talus, or recessive-weathering units (i.e. scoriaceous tuffs and other pyroclastic deposits, highly vesiculated or rubbly lava flows, and oxidized flow tops). The massive cores of all prominent lava flows in the Marion Dufresne section were sampled (FIG. 2.2). A subset of 47 samples was chosen for petrographic and bulk chemical analysis (on the basis of sufficient sample material), and this data is presented and discussed in detail in Chapter 3. Seven petrographically and geochemically distinct rock types occur in the section, which are listed below according to relative order of eruption or emplacement (i.e. bottom to top): 1) Aphyric trachybasalts (3.7-4.4 wt. % MgO) 2) Aphyric basalts (4.3-5.2 wt. % MgO) 3) Plagioclase-phyric basalts (4.4-4.8 wt. % MgO) with 5-25 vol. % plagioclase phenocrysts 4) Plagioclase-ultraphyric basalts (3.0-3.8 wt. % MgO) with >50 vol. % plagioclase phenocrysts and >20 wt. % A l 2 0 3 5) High-MgO basalts (7.1 -11.4 wt. % MgO) with 5-20 vol. % olivine and 5-10 vol. % clinopyroxene phenocrysts 6) Quartz-bearing basaltic andesites (5.1 -6.7 wt. % MgO) with <5 vol. % plagioclase and 1-2 vol. % resorbed quartz crystals rimmed by pyroxene-dominant coronas 7) Dikes and sills (8.1-9.3 wt. % MgO) emplaced after the main pulse of volcanism (diabase) Previous geochemical and geochronological studies suggest that the Plateau Central region has an intermediate age and composition (Nicolaysen et al., 2000; Frey et al., 2002; Xu et al., in prep). A transition from tholeiitic to alkalic volcanism occurs in the northern part of the Plateau Central at Mt. Tourmente (-26 Ma), 20 km north of Mt. Marion Dufresne, where -500 m of transitional-tholeiitic 31 800 526-525 529 • 7f530l S ^ T l ,25 ± 0.7 Ma Sample selected for EPMA 500 No EPMA data FIGURE 2.2 Olivine-phyric basalts 5-20 vol% olivine phenocrysts <10 vol. % clinopyroxene phenocrysts high MgO (6.7-11.4 wt. %) low Si02 (46-48.5 wt. %) Quartz-bearing basaltic andesites <5 vol. % plagioclase crystals 1 -2 vol. % quartz crystals relatively high MgO (5.1-6.7 wt. %) high S i0 2 (54.4-55.9 wt. %) Plagioclase-phyric basalts 5-25 vol% plagioclase phenocrysts low MgO (4.4-4.8 wt. %) 48.4-49.7 wt. % Si02 & Plagioclase-ultraphyric basalts 50-60 vol. % plagioclase phenocrysts low MgO (3-3.8 wt. %) 47.8-48.6 wt. % S i 0 2 ^ s z o —24 + 1.6 Ma ~2500 m Aphyric basalts and trachybasalts <5 vol. % phenocrysts low MgO (3.7-5.2 wt. %) wide range of Si0 2 (46.9-50.2 wt. %) FIG. 2.2. Stratigraphy of the >700 m high Mt. Marion Dufresne section. Samples with olivine ± clinopyroxene ± plagioclase phenocrysts that were analyzed in this study are indicated by grey boxes. All prominent lava flows in the section (indicated by horizontal solid lines) are between 1-15 m thick. Grey lines indicate trachybasalts. Flows are separated by intervals of debris, talus, or recessive-weathering units, including scoriaceous tuffs and other pyroclastic deposits, highly vesiculated or rubbly lava flows, and oxidized flow tops, which were not sampled. The dashed lines indicate mafic dikes (cross-cutting lines) and sills (i.e. sample 534; horizontal dashed line). Samples from the base and top of the section dated by 4 0 Ar/ 3 9 Ar geochronology are indicated. Ages cited for samples 532 and 569 are 4 0 Ar/ 3 9 Ar plateau ages of leached whole rocks (see Chapter 3 for detailed explanation). Three major stratigraphic units occur in the section: aphyric basalts and trachybasalts in the lower 200 m, plagioclase-phyric and plagioclase-ultraphyric basalts in the interval of 200-400 m, and olivine-phyric basalts in the upper 400 m of the section. The three distinct quartz-bearing basaltic andesite flows occur within the olivine-phyric basalts in the upper part of the section. 33 basalts with Alkalinity Index values (A.I., = (Na 20+K 20) - 0.37*SiO2 + 14.43; Macdonald & Katsura, 1964) ranging from -0.7 to 0.8 are capped by -100 m of mildly alkalic basalts (A.I. = 0.3 to 1.0; the "Upper Alkalic Croup" of Frey et al., 2002). In the younger Marion Dufresne section, all lava flows have Alkalinity Index (A.I.) values between +2 and -2, clustering about the alkalic-tholeiitic dividing line of Macdonald & Katsura (1964) (FIG. 2.3), which indicates that both silica-saturated and silica-undersaturated compositions are observed in this section (see Chapter 3 for a complete discussion of the major and trace element geochemistry of the Marion Dufresne section). Aphyric and plagioclase-phyric lavas in the lower 200 m of the Marion Dufresne section are the most alkalic (A.l.= -0.7 to +1.8) and trend toward lower alkalinity with increasing stratigraphic height (and thus decreasing age), which contrasts with the regional trends observed on the archipelago; this indicates that the tholeiitic-alkalic transition is not as abrupt as previously considered and may have occurred episodically over a relatively extended period of time (hundreds of thousands of years to several million years). The wide variety of phenocryst types, distributions, and groundmass textures observed in samples from Marion Dufresne has not yet been observed in a single section on the Kerguelen Archipelago. Although plagioclase-phyric basalts from Marion Dufresne resemble some of the younger mildly alkalic flood basalts to the east, in particular the plagioclase-phyric lavas observed in the 25-24 Ma Crozier section (Damasceno etal., 2002), olivine and clinopyroxene phenocrysts are only a minor component of the Crozier lavas (<6 and <2 vol. %, respectively). Olivine phenocrysts are observed in some older transitional-tholeiitic basalts to the north, especially those on the Loranchet Peninsula that contain up to 15 vol. % olivine and plagioclase phenocrysts (Doucet et al., 2002). Plateau Central flood basalts studied to date are typically aphyric, although several plagioclase-rich flows (up to 40 vol. %) are observed in the upper 180 m of the Capitole section (-14 km north of Marion Dufresne; Xu et al., in prep). Only two plagioclase- and clinopyroxene-phyric samples are observed in -500 m of lavas exposed at Mt. Tourmente (-16 km north of Marion Dufresne; Frey etal., 2002). Samples examined for this study were taken from the massive flow cores, which cool more slowly than flow margins and are typically non-vesicular. Groundmass textures observed in the Marion Dufresne lavas include local to pervasive trachytic textures, which are commonly observed in the 34 FIG. 2.3. Total alkalis vs. silica diagram modified from Le Bas et al. (1986) showing that all of the Marion Dufresne samples cluster about the Macdonald & Katsura (1964) tholeiitic-alkalic boundary. Deviations are limited to three aphyric trachybasalts, which are the most alkalic samples recovered from the section, and three quartz xenocryst-bearing basaltic andesites. 35 plagioclase-rich groundmass of fine-grained aphyric basalts and trachybasalts. These alignment domains are interpreted to represent shear zones that overprint homogenous aligned textures during flow (e.g. Smith, 1998). Groundmass olivine is rare to absent in the aphyric basalts and trachybasalts and plagioclase-phyric and -ultraphyric basalts. In contrast, olivine-phyric basalts typically contain abundant groundmass olivine, clinopyroxene, and plagioclase, and commonly display subophitic groundmass textures. The olivine-phyric basalts also contain small amounts (<1 vol. %) of altered intersertal volcanic glass; primary glass is not observed. Visible low-temperature alteration features observed in the Marion Dufresne lavas include local groundmass alteration to fine-grained clays and oxyhydroxides, alteration of olivine and/or plagioclase phenocrysts to iddingsite and sericite respectively, and zeolite recrystallization in samples with more vesicles. 2.4 ANALYTICAL TECHNIQUES Olivine, clinopyroxene and plagioclase phenocryst and pyroxene corona compositions were analyzed from 26 samples representative of the Marion Dufresne lavas (FIG. 2.2). All analyses were done on a Cameca SX50 Electron Microprobe at the University of British Columbia using an accelerating voltage of 15 KeV, a beam current of 20 nA, and a beam size of 5 pm. Fine-grained pyroxene coronas in samples BOB93-539, 540 and 544 were analyzed using a 1 pm beam. Counting times for peak and background were 20 seconds and 10 seconds, respectively. Natural and synthetic standards were used for calibration and procedural set-up. The "PAP" <|>(pZ) data reduction procedure of Pouchou & Pichoir (1991) was applied to all analyses. Olivine and clinopyroxene phenocrysts were not complexly zoned, so compositions were determined using one analysis closest to the crystal rim, one analysis in the core of the crystal, and at least one additional analysis at an intermediate distance between the core and rim. The small average size of pyroxene crystals in reaction coronas (generally <20 microns wide) did not allow for multiple analyses on individual crystals, so each corona analysis represents the composition of a single pyroxene crystal. Plagioclase compositions were collected on transects across individual crystals that typically consisted of 20 to 45 analyses, depending on the size, number of fractures, and alteration state of the 36 crystal. Crystals cut perpendicular to the c-axis, or that showed strong zoning, were preferentially selected for analysis. When variations in crystal size, morphology, or zoning characteristics were noted within one sample, at least one crystal from each population was analyzed. Representative compositional data for all types of phenocrysts analyzed are summarized in Table 2.1-2.4; complete tables of all analyses are available as an electronic supplement (Electronic Appendices l-IV). All analyses reported are consistent with mineral stoichiometry. 2.5 RESULTS 2.5. / Olivine A total of 222 analyses were collected on 69 olivine crystals in 15 high-MgO basalts from the upper part of the Marion Dufresne section as well as from two mafic dikes (FIG. 2.4 and Table 2.1). Photomicrographs of representative olivine phenocrysts analyzed are shown in FIG. 2.5. Olivine is relatively abundant (up to 20 vol. %) in the upper part of the section, with greater than 10 vol. % olivine phenocrysts observed in most flows. Olivine crystals (1 -5 mm long) are typically euhedral hexagonal to tabular prisms with weak zoning and numerous embayments (e.g. FIG. 2.5b,e). Evidence of deformation (i.e. deformation banding and subgrain boundaries) was rarely observed. Forsterite (Fo) contents in olivine cores range from Fo 6 9to Fo 8 8and nearly 50% of all olivine analyses are between Fo 8 0 and Fo 8 5 (FIG. 2.6). Olivine rims (Fo4 8 to Fo83) are significantly less forsteritic than crystal cores. No systematic variation in olivine composition is noted with stratigraphic height. Olivine phenocrysts in sample BOB93-543 (e.g. FIG. 2.5d; labelled as '543' on FIG. 2.2) are extremely forsterite-rich and have an average core composition of Fo 8 7; one olivine crystal from this sample has the highest forsterite content observed in olivine from the Kerguelen Archipelago to date (Fo88; see Table 2.1). Core and intermediate position compositions are typically within 1-2 mol % forsterite of each other. NiO content decreases systematically with decreasing forsterite (FIG. 2.6) and ranges from 0.06-0.42 wt. % in cores and 0.03-0.32 wt. % in rims. Olivine xenocrysts are not morphologically distinct from phenocrysts in the same sample, but may be recognized by their higher Fo core compositions, different rim compositions, and stronger zoning patterns. 37 45 50 55 60 65 70 75 80 85 90 800 i 1 1— 0.70 0.75 0.80 0.85 0.90 40 50 60 70 *3 Q. 2 750 700 650 600 550 500 450 | 400 4-1 1/1 350 300 250 200 150 + + *"i 1 1 r + ++QH + + n r a OP oca + + <rjcro + + -H-C D O O C?CD c C C C D + + + -+ +++ + • • D m DD T Lava Flows: oCore Rim Mafic Intrusions: •Core • Rim Olivine-phyric high-MgO basalts GOOO + + + O O C O 00 o o o f o ° ^ J ^ J o o o + + n n r m Plagioclase-phyric basalts Aphyric basalts 80 90 Quartz-bearing basaltic andesites O O O C O O D O O O C X Z X D o nrnrwrn C1XDO0C O nrrs-t ooflEfCa cani iuuo h4H<- trr.rm-mm ff;n:xirr> 800 - 750 - 700 650 600 550 £ S 500 3 •o 450 ? » (5* 400 % f 350 300 250 200 45 50 55 60 65 70 75 80 85 90 0.70 0.75 0.80 0.85 0.90 40 50 60 70 80 Olivine (%Fo) Clinopyroxene (mg-number) Plagioclase (%An) 90 150 FIGURE 2.4 FIG. 2.4. Stratigraphic variations in phenocryst compositions from the Marion Dufresne section. Olivine cores have a relatively restricted compositional range (Fo74.88) that does not change systematically with increasing elevation. Clinopyroxene from a high-MgO basalt (BOB93-543) has core mg-numbers of 0.79-0.88, whereas clinopyroxene occurring in aphyric and plagioclase-phyric basalts lower in the section has slightly lower core mg-numbers (0.74-0.87). Plagioclase phenocrysts the quartz-bearing basaltic andesites are significantly more sodic (An3 5.7 3) than plagioclase in the plagioclase-phyric and -ultraphyric basalts lower in the stratigraphy (An 6 4. 8 6). 39 Table 2.1: Representative olivine phenocryst core and rim compositions from the Marion Dufresne section R o c k type: H i g h - M g O ol iv ine-phyr ic basa l t S a m p l e : B O B 9 3 - 5 2 7 B O B 9 3 - 5 2 8 B O B 9 3 - 5 3 0 B O B 9 3 - 5 3 1 B O B 9 3 - 5 3 2 B O B 9 3 - 5 3 5 C r y s t a l : 5 2 7 - 2 527-2 528 -2 528 -2 530-4 530 -4 531-1 531-1 532-1 532-1 5 3 5 - 2 5 3 5 - 2 He igh t (m): 760 7 3 0 7 1 5 6 3 5 S i z e : P h e n o . P h e n o . P h e n o . P h e n o . M ic ro . M ic ro . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . T y p e : z z Z Z - - D B D B Z Z Z Z Z o n e : R i m C o r e R i m C o r e R i m C o r e R i m C o r e R i m C o r e R i m C o r e Oxides (wt%) S i 0 2 34 .15 39.30 3 5 . 5 3 39 .62 37 .32 3 8 . 5 9 3 5 . 5 5 39 .19 36.81 38 .76 3 6 . 6 0 4 0 . 5 4 C r 2 0 3 0.00 0.07 0 .02 0.11 0.02 0 .02 0 .03 0 .06 0.04 0 .05 0 .07 0 .02 F e O 4 2 . 4 8 16.13 34 .49 13 .82 21 .95 16.59 3 3 . 6 0 14.31 2 7 . 2 5 15 .00 30 .44 12 .12 M n O 0.68 0 .23 0.51 0.20 0.36 0 .20 0 .60 0 .18 0 .43 0 .25 0 .46 0 .15 M g O 2 2 . 0 0 4 3 . 8 6 2 8 . 6 7 4 5 . 7 6 38 .76 4 3 . 6 2 2 8 . 8 8 4 4 . 6 0 34 .64 4 4 . 2 4 31 .84 4 7 . 0 8 C a O 0 .40 0.26 0 .36 0 .25 0.31 0.30 0 .39 0.24 0.34 0.31 0 .27 0 .28 N i O 0 .07 0 .33 0 .07 0.39 0.13 0 .17 0 .07 0.32 0 .17 0 .28 0.11 0 .37 To ta l 99 .84 100.19 99 .64 100 .16 98 .85 99 .49 99.11 98 .90 9 9 . 6 8 9 8 . 8 8 9 9 . 7 9 100 .57 Cations (p.f.u.) S i 0 .993 0 .995 0 .992 0 .993 0 .987 0 .988 0 .994 0 .997 0.991 0.991 0 .998 1.003 C r 0 .000 0.001 0 .000 0 .002 0.001 0 .000 0.001 0.001 0.001 0.001 0 .002 0 .000 M g 0 .947 1.644 1.184 1.698 1.518 1.652 1.195 1.680 1.379 1.674 1.285 1.724 F e 2 * 1.033 0 .342 0 .805 0 .290 0.486 0 .355 0 .786 0 .305 0 .613 0.321 0 .694 0.251 M n 0 .017 0 .005 0 .012 0 .004 0.008 0 .004 0 .014 0 .004 0 .010 0 .005 0.011 0 . 0 0 3 C a 0 .013 0 .007 0.011 0 .007 0 .009 0 .008 0 .012 0 .006 0 .010 0 .008 0 .008 0 .007 Ni 0 .002 0 .007 0 .002 0 .008 0 .003 0 .003 0 .002 0 .007 0 .004 0 .006 0 .002 0 .007 S u m 3 .006 3 .002 3 .007 3 .002 3.011 3.011 3 .004 3 .000 3 .008 3 .007 2 .999 2 . 9 9 6 End members (%) F o 4 7 . 8 82.8 59 .5 85 .4 75.8 82 .3 6 0 . 3 84 .7 69 .2 83 .9 64 .9 8 7 . 3 F a 52 .2 17.2 4 0 . 5 14.6 24 .2 17.7 39 .7 15 .3 30 .8 16.1 35.1 12.7 Crystal size is indicated by Pheno. (phenocryst, >0.5mm) or Micro, (microphenocryst, <0.5mm). DB, deformation banding; Z, zoned 40 Table 2.1 (continued): Representative olivine phenocryst core and rim compositions from the Marion Dufresne section R o c k type: H i q h - M g O ol iv ine-phyric basal t Maf ic d ike S a m p l e : B O B 9 3 - 5 4 3 B O B 9 3 - 5 4 5 B O B 9 3 - 5 4 9 B O B 9 3 - 5 5 5 B O B 9 3 - 5 4 2 B O B 9 3 - 5 6 1 Crys ta l : 543-2 543-2 545-3 545-3 549-1 549-1 555-6 555-6 542-1 542-1 561-1 561-1 Height (m): 530 490 450 540 300 S i z e : P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o l T y p e : - - Z Z - - Z Z - - -• -Z o n e : R i m C o r e R i m C o r e R i m C o r e R i m C o r e R i m C o r e R i m C o r e Oxides (wt%) S i 0 2 38.51 39.88 37.79 39.39 37 .07 39.36 35.11 38.66 39 .12 40 .06 36 .34 38.98 C r 2 0 3 0.04 0.06 0.05 0.04 0.08 0.06 0.02 0 .07 0.00 0.07 0.00 0.02 F e O 21.27 11.59 21.42 15.03 27 .07 15.22 34.96 15.99 19.24 14.06 31 .53 17.90 M n O 0.33 0.15 0.28 0.20 0.36 0.18 0.52 0.24 0.41 0.19 0.52 0.19 M g O 39.42 46.86 39.22 44.28 34 .52 44.50 28.29 43 .86 40 .67 4 5 . 3 3 30 .63 42 .06 C a O 0.29 0.23 0.24 • 0.23 0 .33 0.24 0.30 0.27 0.27 0.23 0.34 0.19 N i O 0.23 0.28 0.18 0.29 0.14 0.32 0.13 0.22 0.09 0.31 0.08 0.18 Tota l 100.11 99.06 99.18 99.50 99 .59 99.88 99 .33 99.31 99.80 100.26 99 .47 99 .53 Cations (p.f.u.) S i 1.000 1.000 0.993 0.999 0.996 0.996 0.987 0.988 1.008 1.003 1.001 1.001 C r 0.001 0.001 0.001 0.001 0 .002 0.001 0.000 0.001 0.000 0.001 0.000 0 .000 M g 1.515 1.739 1.525 1.663 1.373 1.666 1.177 1.660 1.551 1.679 1.248 1.599 F e 2 * 0 .462 0.243 0.470 0.319 0.608 0.322 0.822 0 .342 0.415 0.294 0.726 0.384 M n 0.007 0.003 0.006 0.004 0.008 0.004 0.012 0 .005 0 .009 0 .004 0.012 0.004 C a 0.008 0.006 0.007 0.006 0.010 0.007 0.009 0 .007 0.008 0.006 0.010 0 .005 Ni 0 .005 0.006 0.004 0.006 0 .003 0.007 0.003 0 .005 0.002 0 .006 0.002 0.004 S u m 2.998 2.998 3.005 2.999 3.000 3.002 3.011 3 .009 2.992 2 .994 2.999 2 .998 End members (%) F o 76.6 87.7 76.4 83.9 69 .3 83.8 58.9 82 .9 78.9 85.1 63.2 80.6 F a 23.4 12.3 23.6 16.1 30.7 16.2 41.1 17.1 21.1 14.9 36.8 19.4 Crystal size is indicated by Pheno. (phenocryst, >0.5 mm) or Micro, (microphenocryst, <0.5 mm). D.B., deformation banding; Z., zoned 41 42 FIGURE 2.5 FIG. 2.5. Representative photomicrographs of olivine phenocrysts in high-MgO lavas from the Marion Dufresne section. A . Large euhedral olivine phenocryst from sample BOB93-528. The appearance and composition of this crystal (Foe 6 core, Fo 7 5 rim) are typical of other phenocrysts in this sample. B. Olivine phenocryst with a fractured and partly resorbed rim from sample BOB93-531 and strong normal zoning from Fo 8 5 (core) to Fo 6 3 (rim). C . Euhedral olivine phenocryst from sample BOB93-536 with exaggerated terminations and normal zoning from Fo 8 4 (core) to Fo 7 7 (rim). D. Olivine xenocryst from sample BOB93-543 illustrating the irregular shape and strong fracturing typical of most olivine crystals observed in this sample. This xenocryst has one of the most forsteritic core compositions observed in this study (Fo88). E. Large skeletal olivine phenocryst from sample BOB93-535. This phenocryst has a slightly lower core composition (Fo84) than smaller, rectangular olivine crystals from the same sample. F. Relatively small rectangular olivine crystal from sample BOB93-535 with a relatively high core composition (Fo86.87), distinct tabular morphology, and faint deformation banding. 43 125 100 (0 (0 > » c < 0) n E 75 50 olivine Core (n=151) 25 Rim (n=71) 40 45 50 55 60 65 70 F0 0 | jvine 1 1 r 80 85 90 95 FIGURE 2.6 FIG. 2.6. Histogram of the forsterite content of all olivine analyses reported in this study (n=222). The majority of the core compositions are in the range of Fo 8 0 . 8 5 and rim compositions are distributed between Fo 4 5 . 8 5 . NiO contents of the olivines decrease strongly with decreasing Fo 0 | i v i n e (see inset diagram). 45 2.5.2 Pyroxene 2.5.2.1 Clinopyroxene phenocrysts Clinopyroxene typically occurs as a minor (<2 vol. %) phenocryst phase in Marion Dufresne lavas, except in one sample of high-MgO basalt (sample BOB93-543), where it is more abundant (-10 vol. %). Clinopyroxene microphenocrysts (<0.5 mm) are much more abundant than phenocrysts (0.5-2 mm) and typically occur as glomerocrysts of subhedral, weakly zoned crystals. Clinopyroxene is associated with plagioclase phenocrysts or microphenocrysts in plagioclase-phyric basalts (FIG. 2.7c,d). A total of 56 analyses were collected for clinopyroxene phenocrysts and microphenocrysts in 6 samples with a wide range of compositions (4.7 to 9.8 wt. % MgO; A.l.= -1.5 to 0.8). Representative clinopyroxene phenocryst and microphenocryst compositions are reported in Table 2.2, and compositions for all 17 crystals analyzed are available in Electronic Appendix II. Cationic proportions and Fe3 +, AI I V , and AIV 1 were calculated stoichiometrically following the method of Lindsley (1983). Clinopyroxene phenocrysts from the Marion Dufresne section have a relatively restricted range of compositions with respect to quadrilateral components ( W O 4 0 j , 7 , En 4 1 . 5 0 , Fs8.,6) and overlap the ranges observed in mildly alkalic basalts on the Kerguelen Archipelago (FIG. 2.8). Microphenocrysts have slightly lower diopside (Di) and higher ferrosilite (Fs) contents than phenocrysts. The lowest core wollastonite (Wo) contents were observed in microphenocrysts from a basaltic andesite (BOB93-540; Wo 4 0 . 4 3), whereas the highest (Wo47) occur in mafic dikes (FIG. 2.7; Table 2.2). Most (70%) of the clinopyroxene phenocrysts analyzed have core mg-numbers (Mg27(Mg 2 ++ Fe2+)) between 0.75 and 0.85. The highest core mg-number (0.88) was observed in a phenocryst from sample BOB93-543 (e.g. FIG. 2.7a), which also has the most abundant clinopyroxene (10 vol. %) and the highest forsterite content observed in olivine. With respect to mg-number, most clinopyroxene crystals were normally zoned from core to rim, with the exception of two microphenocrysts that show weak reverse zoning (<2 mol %). Clinopyroxene phenocrysts in high-MgO basalts and mafic dikes (samples BOB93-543, 561 and 555) have distinct non-quadrilateral components, including higher cationic Al and Na compared to clinopyroxene from plagioclase-phyric basalts (BOB93-553, 554) and all rim analyses (FIG. 2.9). The 46 47 FIGURE 2.7 FIG. 2.7. Representative photomicrographs of clinopyroxene phenocrysts and microphenocrysts from the Marion Dufresne section. A. Strongly zoned euhedral clinopyroxene phenocryst from sample BOB93-543 showing extensive alteration of the core area (now occupied by secondary zeolites). B. Subhedral, twinned clinopyroxene phenocryst from sample BOB93-553 with strong normal zoning; mg-number decreases from 0.76 to 0.72 between core and rim. C. Nearly euhedral clinopyroxene crystal in large plagioclase glomerocryst (visible at top of image) from sample BOB93-554. D. Glomerocryst of small clinopyroxene crystals in association with a much larger plagioclase crystal (>6 mm long, visible at right) from sample BOB93-554. Only the largest microphenocryst shown here was analyzed in this study. E. Complexly zoned subhedral clinopyroxene phenocryst with resorbed rim, in contact with a similarly-sized euhedral olivine phenocryst (sample BOB95-555). F. Slightly elongate euhedral clinopyroxene phenocryst in diabase with doleritic groundmass. This crystal is normally zoned (mg-number decreases from 0.84-0.79 from core to rim) with a strongly resorbed core (sample BOB93-561). 48 Table 2.2: Representative clinopyroxene phenocryst core and rim compositions from the Marion Dufresne section R o c k type: P lag ioc lase-phyr i c basal t H i g h - M g O ol iv ine-phyr ic basal t Maf i c d ike S a m p l e : B O B 9 3 - 5 5 3 B O B 9 3 - 5 5 4 B O B 9 3 - 5 4 3 B O B 9 3 - 5 5 5 B O B 9 3 - 5 6 1 Height (m): 390 380 530 370 300 S i z e : P h e n o . Pheno . Mic ro . Micro. P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . T y p e : - - S Z S Z . - - R Z R Z R S R S - -Z o n e : R i m Co re R i m C o r e R i m C o r e R i m C o r e R i m C o r e R i m C o r e Oxides (wt%) S i 0 2 50 .80 51.92 50 .73 50.34 50 .85 52.21 49 .59 49 .44 50.12 48 .13 47 .74 49 .05 T i 0 2 1.52 0.97 1.58 1.42 1.27 1.00 1.81 0.99 1.09 0.64 1.78 1.76 A l 2 0 3 2.99 2.66 2.52 3.86 2.48 2.97 3.63 7.49 2.64 7.03 6.04 3.96 F e O 9.55 8.48 10.95 7.84 10.11 7.09 8.83 5.88 9.55 6.44 7.52 7.47 M n O 0.20 0.23 0.25 0.17 0.27 0.19 0.21 0.13 0.25 0.14 0.16 0.16 M g O 13.81 14.95 13.07 14.76 14.39 15.48 13.51 15.24 14.81 15.91 13.85 14.18 C a O 20 .54 20.74 20 .63 20.93 20 .13 21 .53 21.07 19.49 20 .32 19.01 2 1 . 1 3 21 .56 N a 2 0 0.28 0.31 0.38 0.32 0.32 0.30 0.41 0.49 0.28 0.44 0 .53 0.50 Tota l 99 .69 100.25 100.12 99.63 99 .83 100.78 99.06 99.16 99.05 97.74 98 .75 98.64 Cations (p.f.u.) S i 1.903 1.922 1.906 1.875 1.906 1.913 1.871 1.823 1.893 1.808 1.801 1.852 A l 1 7 0.097 0.078 0.094 0.125 0.094 0.087 0 .129 0.177 0.107 0.192 0 .199 0.148 A T 0 .034 0.038 0.018 0.045 0.016 0.041 0.033 0.148 0.010 0.119 0 .070 0.028 C r 0 . 0 0 0 " 0.000 0.002 0.000 0.002 0.002 0.000 0.001 0.000 0.003 0 .000 0 .002 F e 2 * 0 .000 0.008 0 .013 0.023 0.028 0.011 0.023 0 .009 0.055 0.067 0.067 0 .055 T i 0 .043 0.027 0 .045 0.040 0.036 0.028 0.051 0.027 0.031 0.018 0 .050 0 .050 M g 0.771 0.825 0.733 0.820 0.804 0.846 0.760 0.838 0.834 0.891 0 .779 0.798 F e 3 * 0 .299 0.254 0.331 0.221 0.289 0.207 0.256 0.173 0.246 0.135 0 .170 0.181 M n 0 .006 0.007 0.008 0.005 0.008 0.006 0.007 0.004 0.008 0.005 0 .005 0 .005 C a 0 .824 0.823 0.831 0.835 0.809 0.845 0.852 0.770 0.822 0.765 0.854 0 .872 N a 0 .020 0.022 0.028 0.023 0.024 0.022 0.030 0.035 0.020 0.032 0 .039 0 .037 S u m 3.999 4.004 4 .007 4 .012 4 .015 4 .006 4.011 4 .005 4 .028 4 .035 4 .034 4 .028 End members (%) W o 43 .5 43 .3 43 .9 44 .5 42 .5 44.6 45 .6 43.2 43.2 42 .7 47.4 47.1 E n 40 .7 43.4 38.7 43 .7 42 .3 44.6 40 .7 47.1 43.8 49 .7 43 .2 43.1 F s 15.8 13.4 17.5 11.8 15.2 10.9 13.7 9.7 13.0 7.6 9.4 9.8 mg- number 1 0.721 0.764 0.689 0.788 0.736 0.804 0.748 0 .829 0.772 0.868 0.821 0 .815 Crystal size is indicated by Pheno. (phenocryst, >0.5 mm) or Micro, (microphenocryst, <0.5 mm). RS, resorbed; SZ, sector zoned; RZ, reversely zoned 1 mg -number= M g 2 7 ( M g 2 * + F e 2 * ) 49 FIG. 2.8 FIG. 2.8. Pyroxene quadrilateral showing pyroxene compositions from the Marion Dufresne samples. Clinopyroxene phenocrysts and microphenocrysts from the Marion Dufresne section have compositions similar to clinopyroxene phenocrysts from Mt. Crozier (Damasceno et al., 2002), southeast high-MgO basalts and picrites (Doucet et al., 2005), and the Val gabbro plutonic suite (Scoates et al., 2005b). Clinopyroxene that occurs in coronas around quartz crystalss in the basaltic andesites is notably less calcic than clinopyroxene phenocrysts in the section, with pyroxene compositions that extend from augite to pigeonite. 57 lower Ti/AI and Na/Al concentrations observed in phenocrysts from the high-MgO basalts and mafic dikes are consistent with crystallization from a plagioclase-undersaturated magma where higher abundances of Al and Na were incorporated into clinopyroxene phenocrysts. In contrast, the plagioclase-phyric basalts contain clinopyroxene phenocrysts and microphenocrysts with comparatively low total Al and Na abundances, elements that are highly compatible in plagioclase, which suggests crystallization from plagioclase-saturated magmas. Most phenocryst rims analyzed have similar Ti/AI and Ti/Na concentrations to phenocryst cores in the plagioclase-phyric basalts (FIG. 2.9). Clinopyroxene phenocrysts in most samples contain domains with higher A l 2 0 3 (5-8 wt. %). The abundance of A l v l in clinopyroxene phenocrysts from this study varies from ~0 to 0.2 per formula unit (p.f.u.) and generally increases with increasing stratigraphic height in the high-MgO basalts. Clinopyroxene crystals in plagioclase-phyric basalts, which are mostly microphenocrysts, have lower, more limited values for A l v l (<0.065 p.f.u.) and do not fit this trend (Table 2.2). 2.5.2.2 Pyroxene coronas A chemically and morphologically distinct variety of pyroxene is observed in reaction coronas in three samples of quartz-bearing basaltic andesite (BOB93-539, 540, and 544). Quartz in BOB93-540 and 544 are more extensively resorbed, and coronas in these samples in samples are generally broader, than those in sample BOB93-539 (FIG. 2.10). These samples were recovered from the upper 400 m of the Marion Dufresne section, which is otherwise dominated by high-MgO basalts, and have significantly lower MgO contents (5.1-6.7 wt. %) than the adjacent flows. Reaction coronas consisting mainly of fine-grained pyroxene and altered volcanic glass are observed at the margins of extensively resorbed quartz and are interpreted to have developed in response to mixing between a high-MgO magma and a more evolved quartz-bearing magma (see Discussion). Pyroxene in reaction coronas typically occurs as elongate to acicular crystals (up to 0.25 mm long) that radiate from the quartz margins (FIG. 2.10c-e), and more rarely as larger (-0.2 mm) equant euhedral crystals (e.g. FIG. 2.10b). Irregularly-shaped vesicles are commonly observed in the outermost zone of the coronas (e.g. FIG. 2.10d-f). Quartz (1-2 vol. %) occurs as deeply embayed and resorbed single crystals or polycrystalline 52 0.00 1 1 1 1 1 1 1 1 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 Ti (p.f.u.) FIGURE 2.9 S3 FIG. 2.9. Co-variation of non-quadrilateral components (Al, Na and Ti) in pyroxene from Marion Dufresne. Units are cations per formula unit (p.f.u.). Na and Al are highly compatible in plagioclase, thus clinopyroxene that crystallized from a plagioclase-saturated magma will have lower Na and Al compared to clinopyroxene that crystallized from plagioclase-undersaturated magmas. Clinopyroxene phenocrysts in the high-MgO basalts (samples BOB93-543, 561 and 555) have cores with distinctly higher Na and total Al for given Ti values. In contrast, clinopyroxene from samples BOB93-553 and 554 has lower total Al and Na and overlaps with most rim compositions. Pyroxene that occurs in coronas in the quartz-bearing basaltic andesites is lower in Ti, Na and total Al . Pigeonite in these samples has lower Na for a given Ti, as indicated. Clinopyroxene microphenocrysts analyzed in one sample of quartz-bearing basaltic andesite (BOB93-540) have significantly higher Na and total Al than pyroxene in coronas from the same sample. The compositions of high-Al, high-pressure clinopyroxene phenocrysts from the 25-24 Ma Crazier section are indicated (~1:8; Damasceno et al., 2002). 54 5 5 FIGURE 2.10 FIG. 2.10. Representative photomicrographs of quartz crystals rimmed by fine-grained pyroxene coronas. A. Resorbed polycrystalline quartz with a narrow corona (<0.1 mm) of altered glass and fine-grained pyroxene (sample BOB93-539). B. Resorbed quartz surrounded by a large zone of altered glass and subhedral pyroxene crystals (sample BOB93-539). C. Nearly euhedral quartz with a thin, concentric corona of acicular pyroxene crystals (sample BOB93-540). D. Partially resorbed quartz with a corona of elongate pyroxene crystals and altered glass (sample BOB93-540). E. Extensively resorbed quartz with a broad corona of acicular pyroxene, altered glass, and numerous void spaces (sample BOB93-544). A number of pyroxene crystals from the outer part of this corona are Ca-poor with <14% Wo (i.e. pigeonite). F. Partially resorbed quartz with corona of elongate pyroxene and altered glass. Numerous void spaces are present in the outer rim of this corona (sample BOB93-544-2). One crystal from near the.outer edge of this corona is Ca-poor (<10% Wo). 56 clusters that locally exhibit relict hexagonal morphology (FIG. 2.10c); remnant rounded patches that contain only fine-grained pyroxene and altered glass likely represent complete dissolution of quartz crystals. A total of 114 pyroxene crystals were analyzed from 12 different coronas in these three samples. Representative compositions are shown in Table 2.3. Pyroxene in reaction coronas is distinctly less calcic than the clinopyroxene phenocrysts described above, with very little compositional overlap (FIG. 2.8). Some pyroxene corona compositions extend into the pigeonite field (Wo8.1 4). Pyroxene crystals at the outer margins of the coronas generally have higher mg-numbers that resemble groundmass clinopyroxene compositions. Pyroxene crystals in the broad coronas observed in samples BOB93-540 and 544 generally increase in mg-number (from -0.65 to -0.82) with increasing distance from the quartz crystals (FIG. 2.10c-f), whereas mg-numbers are relatively constant (variations of 0.74-0.85 or less) across the narrow coronas in sample BOB93-539 (e.g. FIG. 2.10a). Pyroxene in reaction coronas is further distinguished from phenocrystic clinopyroxene by its significantly lower non-quadrilateral components, including extremely limited cationic Ti, Na and Al values (FIG. 2.9). Fewer than 10 of the 114 corona compositions extend into typical "phenocrystic" values. All coronas have similar (overlapping) Na and Ti contents, although each sample forms a distinct trend with respect to Ti/AI values. Pigeonite crystals have significantly lower Na contents and are plotted separately on FIG. 2.9b. Two microphenocrysts analyzed in sample BO93-540 have slightly higher cationic Na and Al and comparable Ti contents to pyroxene in coronas from the same sample, and plot between the high-MgO and plagioclase-phyric basalts. 2.5.3 Plagioclase Plagioclase phenocrysts are rare in the upper part of the Marion Dufresne section, but are abundant in the lower part of the section where thick lava flows with up to 60 vol. % plagioclase are interlayered with aphyric basalt flows. Plagioclase was analyzed in 10 samples from a wide range of elevations and whole-rock compositions, including three quartz-bearing basaltic andesites, five plagioclase-phyric basalts (5-25 vol. %), and two plagioclase-ultraphyric basalts (50-60 vol. %). 5 7 Table 2.3: Representative pyroxene corona compositions from the Marion Dufresne section R o c k type: Quar tz -bear ing basalt ic andes i te S a m p l e : C rys ta l : Height (m): B O B 9 3 - 5 3 9 539-3 585 539-28 585 539-36 585 539-37 585 B O B 9 3 - 5 4 0 540-19 5 6 5 540-20 565 540-35 565 540-44 565 B O B 9 3 - 5 4 4 544-7 505 544-11 5 0 5 544 -16 5 0 5 544-26 5 0 5 Oxides (wt%) S i 0 2 51.68 52.15 52.58 52.45 52.50 51.13 52.06 51.90 53.14 52 .03 52.91 52 .23 T i 0 2 0.50 0.44 0.38 0.42 0.37 0.84 0.60 0.69 0.39 0 .43 0 .33 0.43 A l 2 0 3 2.21 1.67 0.52 1.14 0.58 2.10 0.53 1.40 0 .45 1.33 0.60 0.56 F e O 7.27 6.98 10.33 6.99 20 .98 11.04 11.94 8.77 16.74 9.01 9.12 20 .15 M n O 0.19 0.17 0.25 0.18 0.52 0.20 0.30 0.24 0.29 0.24 0.18 0.34 M g O 16.92 17.85 17.58 17.73 21.56 15.87 15.80 17.53 22 .35 17.47 17.05 21 .45 C a O 20 .07 19.13 17.43 19.59 3.72 17.79 17.73 18.23 6.30 18.49 19.33 4 .13 N a 2 0 0.24 0.27 0.23 0.24 0.05 0.23 0.20 0.22 0.10 0.19 0.23 0 .05 Tota l 99 .08 98.66 99.30 98.74 100.29 99.19 99.17 98.97 99 .75 99 .19 99 .74 99 .33 Cations (p.f.u.) S i 1.922 1.939 1.964 1.952 1.959 1.922 1.965 1.937 1.967 1.940 1.965 1.962 A I I V 0.078 0.061 0.036 0.048 0.041 0.078 0.035 0.063 0 .033 0 .060 0 .035 0.038 A I V I 0.019 0.013 0.000 0.002 0.000 0.015 0.000 0.000 0 .000 0 .000 0 .000 0 .000 C r 0 .003 0.002 0.002 0.001 0.001 0.001 0.001 0.000 0 .000 0 .002 0.001 0.001 F e ' * 0 .182 0.177 0.293 0.179 0.632 0.316 0.362 0.234 0.500 0 .234 0 .252 0.617 Ti 0 .014 0.012 0.011 0.012 0.010 0.024 0.017 0.019 0.011 0 .012 0 .009 0.012 M g 0.938 0.990 0.979 0.984 1.200 0.890 0.889 0.976 1.234 0.971 0.944 1.201 F e 3 * 0 .044 0.040 0.030 0.039 0.023 0.031 0.015 0.040 0.018 0 .047 0.031 0.016 M n 0.006 0.005 0.008 0.006 0.016 0.006 0.010 0.008 0.009 0.008 0.006 0.011 C a 0 .800 0.762 0.697 0.781 0.149 0.717 0.717 0.729 0.250 0 .739 0 .769 0.166 N a 0 .017 0.019 0.017 0.017 0.004 0.017 0.015 0.016 0.007 0.014 0.016 0.004 S u m 4 .024 4.021 4 .023 4 .020 4 .020 4 .016 4 .014 4 .020 4 .016 4 .025 4.021 4 .015 End members (%) W o 41.7 39 .5 35.4 40.2 7.5 37 .3 36.4 37.6 12.6 38.0 39.1 8.4 E n 48 .9 51.3 49.7 50.6 60.6 46 .3 45.2 50.3 62.2 50.0 48 .0 60 .5 F s 9.5 9.2 14.9 9.2 31.9 16.4 18.4 12.1 25.2 12.0 12.8 31.1 mg- number 1 0.838 0.849 0.770 0.846 0.655 0.738 0.711 0.806 0.712 0.806 0 .789 0.661 All crystals are <0.5 mm long. ' mg -number= M g ' 7 ( M g z * + F e " ) 58 Representative plagioclase compositions from the Marion Dufresne section are reported in Table 2.4. The length of the transect and number of analyses collected for each crystal was dependent on crystal size and alteration; transects were 0.7-1.4 mm long in basaltic andesites, 1-2.5 mm long in plagioclase-phyric basalts, and 2.5-5.2 mm long in plagioclase-ultraphyric basalts (as indicated in FIG. 2.11). A wide range of plagioclase compositions is observed in lavas from the Marion Dufresne section (FIGs. 2.4 and 2.12). Plagioclase-phyric and plagioclase-ultraphyric basalts are interlayered in the Marion Dufresne section and contain plagioclase phenocrysts and megacrysts with similar core compositions (An 7 5. 8 0) and fe-numbers (between 0.62-0.84 for all crystals, where /e-number = Fe27(Mg 2 ++ Fe2+), and thus appear to be genetically related. The relationship between crystal size and composition in the Marion Dufresne plagioclase phenocrysts is illustrated in FIG. 2.13. A significant proportion of plagioclase phenocrysts in these samples has compositions of A n 8 0 or higher (FIG. 2.14). Plagioclase-phyric basalts typically contain tabular crystals and glomerocrysts (0.5-5 mm) with strong normal zoning to more sodic rim compositions (AnS 0. 7 9) (FIG. 2.4). Plagioclase-ultraphyric basalts contain >50 vol. % large plagioclase phenocrysts (1 to 9 mm) that generally lack sodic rims and commonly exhibit oscillatory zoning, glomerophyric habits, and rounded forms, indicating resorption (FIG. 2.11 e,f). In contrast, plagioclase in quartz-bearing basaltic andesites (<2 mm long) is characterized by prominent reverse zoning and sieve textures, features that are not observed elsewhere in the section, as well as resorbed crystal forms (FIG. 2.11 a-d). Plagioclase in these samples is the most sodic observed in this study (An 3 5 . 7 3 in cores, An 5 2 _ 7 1 in rims) and also has the highest observed fe-numbers (0.56-0.98 in cores, 0.42-0.88 in rims) (FIG. 2.14). Feldspar compositions the quartz-bearing basaltic andesites have slightly elevated Or contents (up to Or5) (FIG. 2.12). As indicated in FIG. 2.12, feldspars with up to Or 5 5 in the 25-24 Ma Crozier basaltic section on the Courbet Peninsula (Damasceno et al., 2002) have low An content and plot along a clear fractionation trend. 59 Table 2.4: Representative plagioclase phenocryst core and rim compositions from the Marion Dufresne section R o c k type: P laq ioc lase-phyr i c basal t S a m p l e : B 0 B 9 3 - 5 5 3 B 0 B 9 3 - 5 5 4 B O B 9 3 - 5 5 7 B O B 9 3 - 5 6 3 B O B 9 3 - 5 6 6 Crys ta l 553-1 553-1 554-1 554-1 557-2 557-2 563-1 563-1 563 -3 563-3 566-2 566-2 Height (m): 390 390 380 380 330 330 290 290 2 9 0 290 200 200 S i z e : P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . T y p e : N Z R S , O Z C Z R S , N Z N Z G , N Z Z o n e : R i m Co re R i m C o r e R i m C o r e R i m C o r e R i m C o r e R i m C o r e Oxides (wt%) S i 0 2 53.86 48.93 53.30 47 .59 51 .92 45 .46 52.89 47.61 53 .56 4 9 . 2 3 47 .63 46 .82 A l 2 0 3 28.22 31.46 29 .13 32.50 28.89 33.24 28.71 32.51 28 .30 31 .18 31.82 32 .79 F e O 0.72 0.71 0.79 0.73 1.00 0.59 0.75 0.66 0.91 0.69 0.73 0.71 M g O 0.12 0.12 0.14 0.12 0.15 0.10 0.17 0 .13 0.11 0 .13 0.16 0.09 C a O 11.51 15.49 12.50 16.50 12.48 17.57 12.64 16.86 11.90 15.09 16.51 17.03 N a O 4 .73 2.67 4 .32 2.23 4.31 1.59 4 .23 1.99 4 .63 2 .83 2.26 1.85 K 2 0 0.42 0.17 0.35 0.10 0.30 0.07 0.23 0.06 0.38 0 .13 0.08 0.06 Tota l 99 .58 99.55 100.54 99.77 99.04 98 .62 99.61 99 .83 99 .79 99 .28 99.20 99 .35 Cations (p.f.u.) S i 2.451 2.253 2 .409 2.194 2.387 2 .128 2.412 2 .193 2 .437 2 .269 2.209 2 .170 A l 1.514 1.707 1.551 1.766 1.565 1.834 1.543 1.765 1.518 1.694 1.739 1.791 F e " 0 .025 0.025 0 .027 0.025 0.035 0.021 0.026 0 .023 0.031 0 .024 0.025 0 .025 M g 0.008 0.008 0 .010 0.008 0.010 0 .007 0.011 0 .009 0 .007 0 .009 0.011 0 .006 C a 0.561 0.764 0.606 0.815 0.615 0.881 0.618 0 .832 0 .580 0 .745 0.821 0 .846 N a 0 .417 0.238 0 .379 0.199 0.384 0 .145 0.374 0 .178 0 .409 0 .253 0.204 0.166 K 0 .025 0.010 0 .020 0.006 0.017 0.004 0.014 0 .004 0 .022 0.008 0.005 0.004 S u m 5.000 5.006 5.001 5.013 5.014 5.019 4 .997 5.004 5.004 5.002 5.013 5 .007 End members (%) A n 56.0 75.5 60 .3 79.9 60.5 85.6 61.5 82.1 57.4 74.1 79.7 83 .3 A b 41.6 23.5 37.7 19.5 37.8 14.0 37.2 17.5 40 .4 25 .2 19.8 16.3 O r 2.4 1.0 2.0 0.6 1.7 0.4 1.4 0.4 2.2 0.8 0.5 0.3 Fe-number 1 0.8 0.8 0.7 0.8 0.8 0.8 0.7 0.7 0.8 0.7 0.7 0.8 Crys ta l s i ze is indicated by P h e n o . (phenocryst , >0.5 mm) or Micro, (microphenocryst , <0.5 mm). S T , s i eve texture; G , g lomerocryst ic ; R S , resorbed ; N Z , normal zon ing ; R Z , reverse zon ing ; O Z , osci l latory zon ing ; C Z , comp lex zon ing ' fe -number= F e ' 7 ( F e " + M g z * ) 60 Table 2.4 (continued): Representative plagioclase core and rim compositions from the Marion Dufresne section Unit: P laq ioc lase-u l t raphyr ic basal t Quar tz -bear ing basal t ic andes i te S a m p l e : B O B 9 3 - 5 5 6 B O B 9 3 - 5 6 4 B O B 9 3 - 5 3 9 B O B 9 3 - 5 4 0 Crys ta l 556-1 556-1 556-3 556-3 564-1 564-1 564 -3 564 -3 539^t 539-4 540-1 540-1 Height (m): 360 360 360 360 240 240 240 2 4 0 585 585 5 6 5 565 S i z e : P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . P h e n o . T y p e : N Z G , N Z R S , G , N Z R S S T , R S , R Z N Z Z o n e : R i m C o r e R i m Co re R i m C o r e R i m C o r e R i m C o r e R i m C o r e Oxides (wt%) S i 0 2 49 .13 47.19 48 .64 47.80 49 .43 47 .22 49 .76 4 7 . 3 0 50.96 56 .67 53.90 51.18 A l 2 0 3 31.80 33.52 31.46 32.24 32.57 33 .15 31 .48 32.71 30.06 26 .79 28.66 30.24 F e O 0.67 0.69 0.76 0.74 0.66 0.57 0.70 0 .63 1.10 0.41 0.83 0.71 M g O 0.15 0.08 0.15 0.12 0.16 0.12 0.16 0 .13 0.07 0.03 0.12 0.09 C a O 15.72 17.59 15.59 16.51 15.87 17.28 15.31 17 .15 14.04 9.33 12.30 14.30 N a O 2.66 1.66 2.62 2.13 2.44 1.82 2 .90 1.83 3.41 5.62 4.42 3.51 K 2 0 0.13 0.07 0.11 0.09 0.09 0.08 0 .13 0 .07 0.25 0.80 0.39 0.26 Tota l 100.24 100.80 99.32 99.63 101.23 100.24 100.45 99 .82 99.91 99 .66 100.61 100.27 Cations (p.f.u.) S i 2 .247 2.157 2 .245 2.205 2.236 2.168 2 .268 2.181 2.331 2 .559 2 .433 2.331 A l 1.714 1.806 1.712 1.753 1.737 1.794 1.691 1.777 1.620 1.426 1.524 1.623 F e z t 0.023 0.024 0.026 0.026 0.022 0.020 0.024 0 .022 0.038 0.014 0.028 0.024 M g 0 .010 0.006 0.011 0.008 0.011 0.008 0.011 0 .009 0 .005 0.002 0.008 0.006 C a 0 .770 0.861 0.771 0.816 0.769 0.850 0.748 0 .847 0.688 0.451 0.595 0.698 N a 0 .236 0.147 0 .235 0.191 0.214 0.162 0 .257 0 .163 0.302 0.492 0.386 0.310 K 0.008 0.004 0.006 0.005 0.006 0.005 0.008 0.004 0 .015 0.046 0.023 0.015 S u m 5.007 5.004 5.006 5.004 4.994 5.008 5.006 5.004 4 .999 4.991 4 .996 5.007 End members (%) A n 76.0 85.1 76.2 80.7 77.8 83 .6 73 .9 8 3 . 5 68 .5 45 .6 59.2 68.2 A b 23 .3 14.5 23.2 18.8 21.6 15.9 25.4 16.1 30.1 49 .7 38.5 30.3 O r 0.7 0.4 0.6 0.5 0.6 0.5 0.8 0.4 1.5 4 .7 2.3 1.4 Fe-number ' 0.7 0.8 0.7 0.8 0.7 0.7 0.7 0.7 0.9 0.9 0.8 0.8 Crys ta l s i ze is indicated by P h e n o . (phenocryst , >0.5 mm) or Micro, (microphenocryst , <0.5 mm). S T , s ieve texture; G , g lomerocryst ic ; R S , resorbed ; N Z , normal zon ing ; R Z , reverse zon ing , O Z , osci l latory zon ing . ' fe -number= F e 2 7 ( F e " + M g " ) 61 62 FIGURE 2.11 63 FIGURE 2.11 (continued) FIG. 2.11. Representative photomicrographs of plagioclase phenocrysts from the Marion Dufresne section. Solid white lines indicate the locations of EPMA transects. A. Rounded plagioclase phenocryst with strongly sieve-textured outer rim. Core is oscillatorilly zoned (sample BOB93-539). B. Elongate plagioclase lath from sample BOB93-539 with a strongly sieve-textured core and thin clear outer rim. C. Euhedral plagioclase phenocryst with slight normal zoning and extensive development of sieve textures in most of the core (sample BOB93-544). D. Elongate, twinned plagioclase phenocryst with rounded core (sample BOB93-544). E. Plagioclase phenocryst from sample BOB93-554 exhibiting rounding due to resorption and weak oscillatory zoning. F. Large plagioclase glomerocryst in sample BOB93-554 with irregular subgrain boundaries. G. Elongate plagioclase glomerocryst in plagioclase ultraphyric basalt (sample BOB93-556-4), exhibiting weak normal zoning. H. Elongate plagioclase phenocryst from sample (BOB93-557) with complexly zoned, resorbed core (An76_86) and oscillatory zoned sodic rim (~An60). I. Euhedral plagioclase phenocryst with complexly zoned core (An 6 8. 7 6) and slightly resorbed, sodic margin (~An60) (sample BOB93-563). J. Complexly zoned twinned plagioclase phenocryst from sample BOB93-563 with a more calcic core (An 7 4. 8 1) and sodic rim (~An59). K. Large phenocryst in plagioclase-ultraphyric basalt (sample BOB93-563.2); crystal appears resorbed and generally lacks zoning (core A n 7 3 . 8 0 , rim ~An 7 5). L. Large plagioclase glomerocryst, composed of elongate twinned crystals, in aphyric basalt (sample BOB93-566) 64 FIG. 2.12. Feldspar ternary diagram illustrating the wide variation of plagioclase compositions observed in Marion Dufresne lavas. The trend defined by feldspar compositions from the 25-24 Ma Mt. Crozier basaltic section (Damasceno et al., 2002) is shown for comparison. The compositions of plagioclase phenocrysts in the plagioclase-phyric and -ultraphyric basalts from Marion Dufresne are indicated at left, whereas the compositions of feldspars in the quartz-bearing basaltic andesites are indicated at right. Plagioclase core compositions in the quartz-bearing basaltic andesites have lower An contents and are more evolved than those observed in the plagioclase-phyric and -ultraphyric basalts. 66 90 80 70 ^ 60 50 40 30 Unzoned plagioclase-ultraphyric basalt I i i i 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 Distance from rim (mm) 2.7 3.0 3.3 3.6 0.3 0.6 0.9 1.2 1.5 1.8 Distance from rim (mm) 2.1 2.4 40 30 Reversely zoned quartz-bearing basaltic andesite 0 0.3 0.6 0.9 1.2 1.5 Distance from rim (mm) FIG. 2.13 67 FIG. 2.13. Plagioclase zoning profiles for representative samples of plagioclase-ultraphyric basalt, plagioclase-phyric basalt, and quartz-bearing basaltic andesite from the Marion Dufresne section. A. The plagioclase-ultraphyric basalts contain large, relatively unzoned, high-An (>An80) plagioclase phenocrysts. B. The plagioclase-phyric basalts typically contain relatively high-An plagioclase with strong normal zoning. C. The rare quartz-bearing basaltic andesites, which occur in the upper part of the Marion Dufresne section, contain smaller, lower-An plagioclase crystals with more complex zoning profiles and a high proportion of reversely zoned crystals. 68 FIG. 2.14. Histograms of all plagioclase compositions reported in this study (n=436). Fe-number is calculated as Fe 27 (Fe2++ Mg 2 +). Similar population distributions are observed in the plagioclase-phyric and plagioclase-ultraphyric basalts with the latter showing a distinctive lack of lower-An rims. Quartz-bearing basaltic andesites, in comparison, contain plagioclase with a wider range of lower-An compositions and higher fe-numbers ranging from 0.40-0.98. Lines at A n 7 5 and fe-number 0.70 are for reference only. 70 2.6 DISCUSSION 2.6.7 Olivine and clinopyroxene mg-number relations: implications for the parental magmas of the Marion Dufresne section Although physical characteristics can be useful in distinguishing phenocrystic from xenocrystic olivine (e.g. Garcia, 1996), the relationship between the forsterite content of olivine and the whole rock mg-number can be reliably used to determine chemical equilibrium relationships between olivine and basaltic liquid. This technique has been successfully used to reveal magma mixing and crystal accumulation processes and to assess melt MgO contents in olivine-phyric basalts from a variety of oceanic islands (e.g. Hawaii: Garcia, 1996; Yang etal., 1999; Thornber, 2001; Garcia etal., 2003; Galapagos: Geist et al., 2002; Kerguelen: Damasceno et al., 2002; Doucet et al., 2002; 2005). Roeder & Emslie (1970) experimentally determined an equilibrium constant for Fe-Mg exchange between olivine and basaltic liquid of 0.30 ± 0.03 that is independent of changes in temperature, and that is valid at pressures below -10 kbars (Ulmer, 1989) and for intermediate to high mg-numbers (Toplis & Carroll, 1995). Whole-rock mg-numbers (Mg27(Mg 2 ++ Fe2+)) used in this study were determined from XRF data (see Chapter 3) assuming 10% Fe3 +, which is consistent with low-pressure crystallization near the FMQ (fayalite-magnetite-quartz) buffer (Damasceno et al., 2002). In FIG. 2.15, the grey band indicates the theoretical range of olivine compositions that would be in equilibrium with a given melt composition. Olivine accumulation in a particular sample will increase the whole rock mg-number at constant Fo 0 | i v i n e , causing the sample to plot below the equilibrium field. Horizontal arrows in FIG. 2.15 indicate the effect of crystal accumulation. Olivine cores for eight samples (BOB93-530, 531, 532, 536, 537, 542, 545, and 547) fall within or near the equilibrium field, indicating that olivine phenocrysts in these samples crystallized in equilibrium with the melt and that the whole rock composition of these samples closely represents that of the melt (8.8 to 9.7 wt. % MgO). Samples BOB93-527, 528, and 549 plot below the lower limit of the equilibrium field, indicating that these samples contain accumulated olivine. This is supported by the high whole rock MgO contents (10.7-11.4 wt. %) and high modal volumes of olivine (15-20 vol. %) observed in these samples. Based on mass-balance calculations using representative core compositions that closely 71 c > o O LL 70 O Phenocryst core • Xenocryst core Rim -1 I 1 I I I L 0.56 0.58 0.60 0.62 0.64 Whole rock mg-number 0.66 0.68 0.90 e o.7o 0.60 Phenocrysts: Microphenocrysts: 0.65 O Core © Core — Rim Rim -1 I I I I I I I I • 1 • -1 I I 1-0.40 0.45 0.50 0.55 0.60 Whole rock mg-number 0.65 0.68 FIGURE 2.15 72 FIG. 2.15. Marion Dufresne olivine (Fo) and clinopyroxene (mg-number) phenocryst and microphenocryst compositions (by EPMA) vs. whole-rock mg-number (calculated from whole rock XRF data). There is no whole-rock geochemical analysis for sample BOB93-555, which is omitted. The equilibrium curves for Fe-Mg exchange between basaltic liquid and olivine and clinopyroxene are indicated in A and B, respectively, with error bars defining the equilibrium field (KD= 0.3 ± 0.03 for olivine; Roeder & Emslie, 1970 and 0.25 ± 0.05 for clinopyroxene; Grove et al., 1982). Whole rock mg-numbers [Mg27(Mg 2 ++ Fe2+)] were calculated assuming 10% Fe 3 +. Horizontal and vertical arrows indicate the effects of accumulation and xenocryst incorporation, respectively. After calculations to correct for accumulated olivine (horizontal arrows), it appears that the majority of equilibrium olivine crystallized from magmas with 8-10 wt. % MgO. As a reference, olivine compositions from sample BOB93-561 plot within the equilibrium field after removal of 3 wt. % olivine from the whole rock composition. 73 resemble the average core composition from each sample, the removal of small amounts of accumulated olivine (1.5-6.4 wt. %) from these samples results in estimated melt compositions with 7-11 wt. % MgO (Appendix I). The relatively high-MgO melt compositions determined in this study (8-10 wt. % MgO) represent minimum values for the parental magmas of the Marion Dufresne suite. These values overlap with previous estimates for the composition of parental magmas for Kerguelen Archipelago flood basalts (8-10 wt. % MgO; Doucet et al., 2002; 2005). Olivine xenocrysts entrained during magma ascent will have forsterite contents higher than the equilibrium field in FIG. 2.15. Physical characteristics, including deformation banding and resorption, have been used to discriminate xenocrystic from phenocrystic olivine in Hawaii (Garcia, 1996), however these characteristics are rarely observed in olivine crystals in this study. Several samples in this study that likely contain xenocrystic olivine (samples BOB93-535, 543 and 550) are indicated by the filled symbols above the equilibrium field in FIG. 2.15. Sample BOB93-535 appears to have a more complex, multi-stage history involving xenocryst incorporation as well as olivine accumulation. Olivine populations are compositionally and morphologically distinct in this sample; phenocrysts occur as larger (>3 mm) embayed crystals, suggesting relatively rapid crystallization (see FIG. 2.5e) and xenocrysts are smaller (1.5- 2 mm) tabular crystals with weak deformation banding (FIG. 2.5f). Small olivine xenocrysts most likely crystallized earlier from hotter, higher-MgO melts and were entrained in the ascending magma. Clinopyroxene phenocrysts are relatively scarce in the Marion Dufresne lavas, but occur in both olivine-phyric and plagioclase-phyric lavas and provide information on equilibrium relations in more evolved magmas. The equilibrium constant for Fe-Mg partitioning between clinopyroxene and basaltic melt is only slightly dependant on pressure, and likely falls between 0.25 ± 0.05 (Grove et al., 1982) and 0.23 ± 0.05 (Toplis & Carroll, 1995). In addition to xenocrystic olivine, sample BOB93-543 contains high mg-number clinopyroxene, which plots within and below the equilibrium field in FIG. 2.15b. Although there is no evidence for xenocrystic clinopyroxene in this sample, clinopyroxene phenocrysts commonly exhibit reverse zoning in core regions. The <1 mm clinopyroxene crystals in two plagioclase-phyric samples (BOB93-554 and 553) generally plot above the equilibrium field and 74 are therefore likely xenocrystic. Many clinopyroxene crystals in these samples are in contact with large plagioclase laths (FIG.*2.7c,d), indicating that clinopyroxene microphenocrysts were entrained with these larger crystals during ascent, however, the compositions of these microphenocrysts (e.g. high Ti/Al and Ti/Na) are not consistent with high-pressure crystallization. In the Crozier section, -30% of fractionated, plagioclase-phyric samples contained clinopyroxene with anomalously high mg-numbers that plotted above the equilibrium field (Damasceno et al., 2002), indicating crystallization at depth from less fractionated magmas; geothermobarometry calculations indicate that some clinopyroxene in the Crozier lavas formed at very high pressures (11-12 kbar). Although the Marion Dufresne section does not contain sufficient clinopyroxene for a detailed characterization of clinopyroxene crystallization environments, clinopyroxene phenocrysts from the olivine-phyric high-MgO basalts and mafic dikes at Marion Dufresne have Ti/Al abundances within the range (1:8) of the Crozier high-P phenocrysts (FIG. 2.9) and may therefore have a high-pressure origin (>5 kbar). Clinopyroxene is an important fractionating phase from Kerguelen Archipelago parental magmas that exerts a strong control on magma composition, but is rarely erupted, remaining instead at the base of the crust where it contributes to the high-velocity gradient crust-to-mantle transition zone (Damasceno et al., 2002; Scoates et al., 2005a). 2.6.2 Pyroxene coronas and sieve-textured plagioclase in quartz-bearing basaltic andesites: evidence for magma mixing Phenocryst populations in the quartz-bearing basaltic andesites from the Marion Dufresne section exhibit disequilibrium textures that are consistent with mixing between chemically diverse magmas. In particular, the occurrence of olivine pseudomorphs and numerous resorbed and embayed quartz crystals together in individual samples indicates that an evolved trachytic to rhyolitic, silica-saturated magma was likely mixed and assimilated into a high-MgO, olivine-phyric basaltic magma to produce the hybrid basaltic andesites observed. Pyroxene coronas surrounding quartz crystals are common in many different tectonic environments (such as island arcs) where felsic magmas are readily assimilated into more mafic magmas, and they are interpreted to represent diffusion-limited reactions in 75 dissolution boundary layers (e.g. Sato, 1975; Stimac & Pearce, 1992). The occurrence of such features in an ocean island environment such as the Kerguelen Archipelago, which is dominated by mafic magmatism and where there is no evidence of silica-saturated quartz-bearing evolved magmatism coeval with 29-24 Ma basaltic volcanism, is unique. Younger silica-saturated rocks are observed on the archipelago, but considerably post-date basaltic volcanism (e.g. 16.6-14.9 Ma, e.g. Rallier-du-Baty, Societe de Ceographie, and lie de I'Ouest; Lameyre et al., 1976; Dosso et al., 1979; Giret & Lameyre, 1983). Direct evidence for magma mixing exists in sample BOB93-539 (FIG. 2.16), where two separate domains are visible in thin section due to incomplete mixing between compositionally distinct magmas. One domain (at left) represents the mafic endmember, and contains abundant large (0.5-2.5 mm) serpentinized olivine pseudomorphs and numerous mm-scale vesicles in a groundmass dominated by elongate plagioclase laths (0.5-1 mm). Small (<1.5 mm) enclaves of basalt are observed in the domain at right, which represents the hybrid magma. Resorbed crystals that represent evolved compositions, including the evolved plagioclase and quartz described earlier, occur exclusively in the hybrid domain (FIG. 2.12b). Olivine pseudomorphs retain their euhedral crystal forms in the hybrid magma due to the relative stability of forsteritic olivine in basaltic andesite (e.g. Ussier & Glazner, 1989). Plagioclase in sample BOB93-539 exhibits features consistent with the incorporation of evolved plagioclase phenocrysts into magma in equilibrium with more calcic plagioclase compositions. Some rounded plagioclase crystals from this sample (e.g. FIG. 2.11a) exhibit rounding of strongly zoned evolved cores (An 3 5. 4 2) and secondary sieve-textured rims (An 5 0. 5 9), whereas most other plagioclase crystals are tabular with extensively sieve-textured cores with highly variable compositions (An 4 3. 7 2) and calcic rims of A n 6 0 . 7 0 . Sieve textures are interpreted to represent the partial reaction of evolved plagioclase to more calcic compositions (progressing inward from rim to core) in response to the change in magma composition. Resorbed crystal forms primarily reflect the effects of the higher temperature of the more mafic magma rather than its composition. The continuum of calcic to potassic feldspar compositions observed in the Crozier section (FIG. 2.12) indicates that magma compositions in 76 *5> , o cy: O f C3 6 mm > 5 \ \ J FIG. 2.10a;. 7 ^ FIG.2.11a?> B 6 mm | Plagioclase 1 1 0 l i v i n e 1 1 Quartz (with pyroxene ' 1 corona) | | High-MgO basalt 1 1 Hybrid basaltic ' ' andesite J ! Vesicle/ amygdule 77 FIGURE 2.16 FIG. 2.16. The contact between two distinct types of lava in sample BOB93-539 provides t evidence for magma mixing in the quartz-bearing basaltic andesites. A. Full thin section photomicrograph with crystals and vesicles outlined. B. Drawing of photomicrograph in A. highlighting the distribution, size, and shape of crystals and vesicles. Two distinct compositional zones are evident in thin section. The high-MgO basalt (at left) contains large serpentinized olivine pseudomorphs, whereas the hybrid basaltic andesite (at right) contains olivine, quartz, and feldspar crystals that display strong disequilibrium textures. All quartz crystals in the hybrid basaltic andesite are resorbed and surrounded by pyroxene coronas, and feldspar crystals display sieve textures, resorption, and reverse zoning. Photomicrographs of selected crystals shown in FIGs. 2.10 and 2.11 are noted. 78 equilibrium with <An 3 S may be generated by progressive fractionation of basaltic magmas in the Kerguelen Archipelago. The source of the quartz-bearing evolved magma is most likely an earlier formed highly differentiated magma body that resulted from extensive fractionation of a mafic parental magma. Such extreme magmatic differentiation has been reported in other intraplate oceanic islands (e.g. Hawaii: Fodor, 2001), and oceanic islands situated at or near ridges (e.g. Iceland: Marsh, 1996; Gunnarsson et al., 1998; Galapagos: Geistetal., 1995). The differentiated magma may have formed during an interval of lower magma flux, or when the center of the Kerguelen hotspot was farther from the Marion Dufresne area. With increasing magma flux, which is represented by the high-MgO basalts with which the basaltic andesites are interlayered, the crystal-rich remnants of this evolved chamber or sill were flushed to the surface during ascent of a mafic magma, and largely mixed with the mafic magma. As previously mentioned, discrete bodies of evolved magma products (e.g. trachytes, syenites) do begin to appear on the Kerguelen Archipelago at about 24 Ma as flood basalt volcanism waned (e.g. Southeast Province trachytes, Weis et al., 1993, Frey et al., 2000; Montagnes Vertes syenites, Giret & Lameyre, 1983, Weis & Giret, 1994; Val gabbros and microsyenites, Scoates etal., 2005b) 2.6.3 Significance ofhigh-An plagioclase compositions and plagioclase-ultraphyric basalts Many Marion Dufresne lavas contain large amounts of high-An plagioclase (>An80) (FIGs. 2.4, 2.13). Highly calcic plagioclase (An 8 0. 1 0 0) is observed in basalts and andesites from mid-ocean ridge and island arc environments, where plagioclase compositions are dependent on crystallization pressure, magmatic H 2 0 content, melt composition, and the oxidation state of the magma (e.g. Housh & Luhr, 1991; Sisson & Grove, 1993; Panjasawatwong etal., 1995), but are unusual in ocean island settings. To evaluate the effects of pressure and H 2 0 content on plagioclase compositions in Marion Dufresne magmas, we used the MELTS program (a free energy minimization algorithm; Ghiorso & Sack, 1995) to simulate fractional crystallization across a range of pressures (0.5, 1, 3 and 5 kilobars) and H 2 0 contents (0.75, 1.5 and 2.5 wt. %) using the major element abundances of a mildly alkalic aphyric basalt from the Marion Dufresne section (BOB93-553) as a starting composition (Table 2.5). This 79 Table 2.5: Starting compositions for MELTS calculations (this study) and experimental studies BOB-93-553 OB-93-190 OB-93-147 ( this s t u d y ) ( F r e i s e e f al., 2 0 0 5 ) ( S c o a t e s e f al., 2 0 0 5 b ) Wt % oxides 1 Si0 2 48.95 47.94 49.11 Ti0 2 3.06 2.69 2.72 Al 2 0 3 14.70 16.14 14.90 Fe 2 0 3 * 14.07 12.93 13.43 MnO 0.19 0.17 0.19 MgO 4.73 5.78 5.08 CaO 9.12 9.64 10.48 Na 20 2.89 3.19 2.88 K 2 0 1.41 1.16 0.96 P 2 0 5 0.46 0.37 0.34' TOTAL 99.58 100.01 100.09 1AII major element oxide data determined by XRF at University of Massachusetts Amherst *AII Fe as Fe 2 0 3 80 sample has a whole rock composition representative of the aphyric basalts from Marion Dufresne and a relatively high MgO content (4.7 wt. %). In general, Marion Dufresne lavas do not have whole-rock compositions consistent with high-An plagioclase crystallization as outlined by Panjasawatwong et al. (1995) (i.e. A l 2 0 3 <18 wt. % and CaO/Na 2 0 « 1 0 - 1 5 ) (see Chapter 3). MELTS runs were conducted at FMQ with removal of fractionated solids and were limited by a maximum H 2 0 content of 3 wt. %, above which MELTS failed to converge over the pressure-temperature range of interest. The results from the MELTS calculations are compared with plagioclase compositions from recent experimental studies on aphyric basalts from the Kerguelen Archipelago (FIG. 2.17, Table 2.5), which provide additional constraints across a wide variety of temperatures, water contents and oxygen fugacities (Freise eta/., 2005), and pressures (Scoates eta/., 2005a). Our MELTS calculations at 5 kbar and FMQ overlap many of the experimental results of Freise et al. (2005) at 5 kbar and FMQ, despite the different starting compositions used. The FMQ+4 experiments of Freise et al. (2005) are shown only for comparison. They are considered to represent extremely unlikely oxidizing conditions that are not supported by the compositions of Fe-Ti oxide microphenocrysts (ilmenite and titanomagnetite) observed in the Crozier lavas, which indicate crystallization conditions near the FMQ oxygen buffer (Damasceno et al., 2002). The high-pressure (0-14.3 kbars) experiments of Scoates et al. (2005a) were conducted on "dry" samples that had been dried at 800°C for 20 minutes as well as "no volatiles added" or "NVA" samples that contained approximately 1.2-1.3 wt. % structurally-bound H 2 0 . The Freise et al. (2005) experiments at -5 kbar, FMQ, and 900-1100°C contained 0.9-5.0 wt. % H 2 0 . The range of plagioclase phenocryst core compositions observed in Marion Dufresne lavas (An 6 4. 8 6) is best simulated by the 1 kbar MELTS calculations, particularly if these results are extrapolated to higher water contents (1 -5%) (FIG. 2.17). Plagioclase compositions predicted by the 5 kbar MELTS calculations (An 6 0. 6 5) overlap the lower limit of the Marion Dufresne data and fail to reproduce the range of compositions observed (as indicated by the grey band in FIG. 2.17). The 3 kbar MELTS runs result in more calcic compositions (An 6 6. 7 2) and describe a larger portion of the data than the 5 kbar runs, however each basaltic sample examined in this study contains plagioclase with >An 7 5 (FIGs. 2.4 and 2.13), which is only predicted by the largely overlapping lower pressure (0.5 and 1 kbar) runs 81 0 s O IN 10 8 6 4 2 0 u 1200 m 3 n o o (0 v. Q . 1000 E ,0) I— 900 A3 01 V) (A 0) 800 12 1 1 1 1 1 1 1 1 O Freise et al. (QFM+4) • Freise efo/.(QFM) " A Scoates et al. (d ry) A Scoates etal. (NVA) • MELTS (5 kbar) " ffl MELTS (3 kbar) 0 + MELTS (1 kbar) . • MELTS (0.5 kbar) O i 1 1 1 r 1 o o i • 1 ffl <3h' ffl -B | ^ - I • 1 1 A A A A ^ A * A £ • • • • 1 O O • o ffl +• | ffl t] O 1 o 1 1 1 1 1 1 1 1 • /A >^A/ < x s > | o s Ess | 20 40 60 % Anorthite 80 100 FIGURE 2.17 82 FIG. 2.17. Variations in plagioclase compositions in Kerguelen Archipelago basalts vs. H 2 0 content, temperature, and pressure based on recent experimental studies (Scoates et al., 2005a; Freise et al., 2005) and MELTS calculations (this study). The grey band indicates the range of plagioclase core compositions observed in Marion Dufresne plagioclase-phyric and -ultraphyric basalts (An 6 4. 8 6). The dashed line indicates a plagioclase composition of A n e o for reference only. The high An content observed in plagioclase appears to reflect the combined effects of relatively high magmatic H 2 0 contents (>3 wt % H 2 0), low temperatures (<1000°C), and low pressures (<2 kbars) on plagioclase fractionation. 83 (An 7 0. 7 8). Plagioclase compositions from Scoates et al. (2005a) and the relevant FMQ experiments of Freise et al. (2005) are below A n 6 7 . FIG. 2.17 shows that the range of plagioclase compositions observed in the Marion Dufresne lavas likely resulted from relatively low equilibrium pressures (i.e. <2 kbars) and high water contents (>3 wt. %), combined with low crystallization temperatures (-900-1125°C) (FIG. 2.17). These conditions could be satisfied by a body of volatile-rich magma that stalled and underwent extensive crystallization and slow cooling in a shallow subvolcanic reservoir to produce a crystal-rich mush layer composed of large plagioclase phenocrysts and glomerocrysts. The high water content and viscosity of the resulting crystal mush could significantly impede eruption (e.g. Sisson'& Grove, 1993; Hansen & Gronvold, 2000) until new magma was forced through the chamber by a sufficient increase in magma flux to entrain and erupt the unusual volumes of high-An plagioclase. The occurrence of plagioclase-ultraphyric basalts in this section is direct evidence for the existence of fractionating magma chambers and crystal mush bodies beneath Marion Dufresne prior to -25 Ma. The complex oscillatory zoning combined with limited compositional variance across many of these phenocrysts indicates that these large crystals (1 -9 mm) formed in a magmatic system where temperature variations were not large. The resorbed nature of most plagioclase crystals observed in the plagioclase-ultraphyric basalts indicates that they were transported to the surface by higher-temperature magmas, and the lack of new growth rims indicates a short residence time in the host magma. Plagioclase-ultraphyric basalts are observed on other oceanic islands (i.e. Galapagos: Cullen et al., 1989; Iceland: Hansen & Gronvold, 2000), where they are commonly associated with high-An contents (>An85) in plagioclase. In Iceland, the plagioclase phenocrysts (1-3 mm) observed in plagioclase-ultraphyric basalts are interpreted to originate in a "mush zone" situated at -10-20 km depth and are erupted only during rifting events triggered by inflow of new magma from deeper magma chambers. Our results suggest that similar processes were involved at Marion Dufresne, but that the "mush zone" was much shallower (5-6 km, or <2 kbar as indicated in FIG. 2.17). The plagioclase-ultraphyric basalts at Marion Dufresne provide evidence for a temporal increase in magma flux at -25 Ma, which then resulted in the eruption of the overlying succession of olivine-phyric high-MgO basalts. 84 2.7 CONCLUSIONS This study of phenocryst mineral chemistry in the 700 m high, 25 Ma Marion Dufresne section, in the southern part of the Plateau Central region of the Kerguelen Archipelago, uses olivine, clinopyroxene, and plagioclase phenocryst compositions to constrain the processes affecting magmatism and basalt petrogenesis on this large oceanic island. The Marion Dufresne section consists of a basal -300 m succession of aphyric, fractionated basalts (<5.2 wt. % MgO) that, with increasing height, are increasingly interlayered with plagioclase-phyric basalts (5-25 vol. %) and plagioclase-ultraphyric basalts (50-60 vol. %). The upper 400 m of the section consists of olivine-phyric (5-20 vol. %) high-MgO basalts (7.1-11.4 wt.%). Plagioclase phenocrysts and megacrysts in the plagioclase-phyric and -ultraphyric basaltic flows from the lower part of the section are characterized by high anorthite contents (An 7 5. 8 5). Comparisons between experimental studies of Kerguelen basalts and thermodynamically-constrained fractionation paths calculated using MELTS indicate that crystallization of this high-An plagioclase occurred at <5-6 km depth and that relatively high H 2 0 contents (>3 wt%) likely played an important role. The distribution of clinopyroxene compositions, particularly with respect to non-quadrilateral components (Ti, Al , and Na), indicates that the depth of fractionation increased during the eruption of the Marion Dufresne section. Clinopyroxene phenocrysts from the lower part of the section have relatively high Ti/AI contents and likely crystallized from plagioclase-saturated magmas at lower pressures, whereas clinopyroxene phenocrysts with lower Ti/AI contents are observed in the olivine-phyric high-MgO basalts, which indicates crystallization at higher pressures from plagioclase-undersaturated magmas. This observation is consistent with a clear change at roughly 400 m elevation, above which olivine is the dominant phenocryst phase. Fe-Mg mineral-melt equilibria of olivine-phyric (Fo80_86) high-MgO basalts indicates that the parental magmas for this section had a maximum MgO content of 8-10 wt. % and that samples with >10 wt. % MgO contain small amounts (<6%) of accumulated olivine. Three quartz-bearing basaltic andesite flows (5.1-6.7 wt. % MgO), which are interlayered with the olivine-phyric high-MgO basalts in the upper part of the section, contain plagioclase and quartz 85 phenocrysts that exhibit disequilibrium textures and evolved compositions (e.g. reversely zoned plagioclase with A n 3 5 . 7 2 cores and An 5 2 . 7 , rims; resorbed quartz with pyroxene reaction coronas) and are interpreted to be hybrid magmas that record mixing between an olivine-phyric high-MgO basaltic magma and an evolved, quartz-phyric magma (e.g. trachyte). The abundance and diversity of phenocrysts observed in the Marion Dufresne section reflects changing magma flux conditions at 25 Ma. The plagioclase-phyric and -ultraphyric lava flows represent relatively viscous magmas that erupted and entrained a low-pressure phenocryst assemblage that had accumulated during a period of reduced magma flux. This was followed by an abrupt increase in magma flux and the eruption of a 400 m sequence of olivine-laden high-MgO basalts. These magmas had undergone significantly less fractionation at deeper levels, and ascended rapidly enough to entrain relatively dense olivine from the magma conduit system. 86 2.8 ACKNOWLEDGEMENTS We are grateful to the Institut Polaire Francais Paul Emile Victor (IPEV) for supporting field work on the Kerguelen Archipelago and to Olivier Brisse for his careful sampling of the Marion Dufresne section. Mati Raudsepp is gratefully acknowledged for his assistance during sample preparation and training on the electron microprobe at UBC. The authors thank Elisabetta Pani for her assistance in sample preparation in the SEM/ Electron Microprobe facility at UBC and J. Michael Rhodes and Michael J. Vollinger at the University of Massachusetts for bulk chemical analyses by XRF. H. Annell was supported by an NSERC Discovery Grant to D. Weis and an NSERC PGS-M. Sample preparation and analytical costs for this study were provided by grants from the Actions de Recherche Concertee, Belgium (ARC 98/03-233 to D. Weis and J. Scoates) and from NSERC Discovery Grants to D. Weis and J. Scoates. 87 2.9 REFERENCES Charvis, P., Recq, M. , Operto, S. & Brefort, D. (1995): Deep structure of the northern Kerguelen Plateau and hotspot-related activity. Geophysical journal International M2, 899-924. Charvis, P., & Operto, S. (1999): Structure of the Cretaceous Kerguelen Volcanic Province (southern Indian Ocean) from wide-angle seismic data, journal of Geodynamics 28, 51-71. Coffin, M.F., Pringle, M.S., Duncan, R.A., Gladczenko, T.P., Storey, M. , Muller, R.D. & Gahagan, L.A. (2002): Kerguelen hotspot output since 130 Ma. Journal of Petrology 43, 1121-1140. Cullen, A., Vicenzi, E. & McBirney, A.R. (1989): Plagioclase-phyric basalts of the Galapagos Archipelago. Journal of Volcanology and Geothermal Research 37, 325-337. Damasceno, D., Scoates, J.S., Weis, D., Frey, F.A., & Giret, A. (2002): Mineral chemistry of mildly alkalic basalts from the 25 Ma Mont Crozier section, Kerguelen Archipelago: constraints on phenocryst crystallization environments. Journal of Petrology 43, 1389-1413. Doucet, S., Weis, D., Scoates, J.S., Nicolaysen, K., Frey, F.A. & Giret, A. (2002): The depleted mantle component in Kerguelen Archipelago basalts: petrogenesis of tholeiitic-transitional basalts from the Loranchet Peninsula. Journal of Petrology 43, 1341 -1366. Doucet, S., Scoates, J.S., Weis, D. & Giret, A. (2005): Constraining the components of the Kerguelen mantle plume: a Hf-Pb-Sr-Nd isotopic study of picrites and high-MgO basalts from the Kerguelen Archipelago. Geochemistry Geophysics Geosystems (in press). Fodor, R.V. (2001): The role of tonalite and diorite in Mauna Kea Volcano, Hawaii, magmatism: petrology of summit-region leucocratic xenoliths. Journal of Petrology 42, 1685-1704. Frey, F.A., Weis, D., Yang, H.-J., Nicolaysen, K., Leyrit, H. & Giret, A. (2000): Temporal geochemical trends in Kerguelen Archipelago basalts: evidence for decreasing magma supply from the Kerguelen plume. Chemical Geology 164, 61-80. Frey, F.A., Nicolaysen, K., Kubit, B.K., Weis, D. & Giret, A. (2002): Flood basalt from Mont Tourmente in the Central Kerguelen Archipelago: the change from transitional to alkalic basalt at -25 Ma. Journal of Petrology 43, 1367-1387. 88 Freise, M. , Holz, F., & Scoates, J.S. (2005): Differentiation of LIP basalts from the Kerguelen Plateau: an experimental study. To be submitted to Journal of Petrology. Garcia, M.O. (1996): Petrography and olivine and glass chemistry of lavas from the Hawaiian Scientific Drilling Project. Journal of Geophysical Research 101, 11701-11713. Garcia, M.O., Pietruszka, A.J., & Rhodes, J.M. (2003): A petrologic perspective of Kilauea Volcano's summit magma reservoir. Journal of Petrology 44, 2313- 2339. Geist, D., Howard, K.A. & Larson, P. (1995): The generation of oceanic rhyolites by crystal fractionation: the basalt-rhyolite association at Volcan-Alcedo, Galapagos Archipelago. Journal of Petrology 36, 965-982. Geist, D., White, W.M. , Albarede, F., Harpp, K., Reynolds, R., Blichert-Toft, J. & Kurz, M.D. (2002): Volcanic evolution in the Galapagos: the dissected shield of Volcan Ecuador. Geochemistry Geophysics Geosystems 3, 10.1029/2002GC000355. Ghiorso, M.S., & Sack, R.O. (1995): Chemical mass transfer in magmatic processes. IV. A revised and internally consistent thermodynamic model for the interpolation and extrapolation of liquid-solid equilibria in magmatic systems at elevated temperatures and pressures. Contributions to Mineralogy and Petrology 119, 197-212. Giret, A. & Lameyre, J., (1983): A study of Kerguelen plutonism: petrology, geochronology and geological implications. In: Oliver, R.L., James, P.R. & Jago, J.B. (eds.) Antarctic Earth Science. Australia: Australian Academy of Sciences, pp. 646-651. Grove, T.L., Gerlach, D.C., & Sando, T.W. (1982): Origin of calc-alkaline series lavas at Medicine Lake Volcano by fractionation, assimilation and mixing. Contributions to Mineralogy and Petrology 80, 160-182. Gunnarsson, B., Marsh, B.D. & Taylor, H.P. (1998): Generation of Icelandic rhyolites: silicic lavas from the Torfajokull central volcano. Journal of Volcanology and Geothermal Research 83, 1-45. Hansen, H., & Gronvold, K. (2000): Plagioclase ultraphyric basalts in Iceland: the mush of the rift. Journal of Volcanology and Geothermal Research 98, 1 -32. Housh, T.B., & Luhr, J.F. (1991): Plagioclase-melt equilibria in hydrous systems. American Mineralogist 76; 477-492. Lameyre, J., Marot, A., Zimine, S., Cantagrel, J.-M., Sosso, L. & Vidal, P. (1976): Chronological evolution of the Kerguelen islands syenite-granite ring complex. Nature 263, 306-307. Le Bas, M.J., Le Maitre, R.W., Streckeisen, A., & Zanettin, B. (1986): A chemical classification of volcanic rocks based on the total alkali-silica diagram. Journal of Petrology 27, 745-750. Lindsley, D.H. (1983): Pyroxene thermometry. American Mineralogist 86, 477-493. Macdonald, G.A. & Katsura, T. (1964): Chemical composition of Hawaiian lavas. Journal of Petrology 5, 82-133. Marsh, B.D. (1996): Solidification fronts and magmatic evolution. Mineralogical Magazine 60, 5-40. Nicolaysen, K., Frey. F.A., Hodges, K.V., Weis, D & Giret, A. (2000): 4 0 Ar/ 3 9 Ar geochronology of flood basalts from the Kerguelen Archipelago, southern Indian Ocean: implications for Cenozoic eruption rates of the Kerguelen plume. Earth and Planetary Science Letters 174, 313-328. Nougier, J. (1970): Carte geologique au 1/200 000 de l'archipel des Kerguelen. Paris: Institut Geographique National. Panjasawatwong, Y., Danyushevky, L.V., Crawford, A.J. & Harris, K.L. (1995): An experimental study of the effects of melt composition on plagioclase-melt equilibria at 5 kbar and 10 kbar: implications or the origin of magmatic high-An plagioclase. Contributions to Mineralogy and Petrology 118, 420-432. Pouchou, J.L. & Pichoir, F. (1991): Quantitative analysis of homogeneous or stratified microvolumes applying the model "PAP". In: Heinrich, K.F.J. & Newbury, D.E. (eds.). Electron Microprobe Quantitation. Plenum Press, New York, pp 31-75. Recq, M. , Le Roy, I., Charvis, P., Goslin, J. & Brefort, D. (1994): Structure profode du mont Ross d'apres la refraction sismique (Ties Kerguelen, ocean Indien austral). Canadian Journal of Earth Sciences 31, 1806-1821. Roeder, P.L. & Emslie, R.F. (1970): Olivine-liquid equilibrium. Contributions to Mineralogy and Petrology 29, 275-289. Sato, H. (1975): Diffusion coronas around quartz xenocrysts in andesite and basalt from Tertiary volcanic region in northeastern Shikoku, Japan. Contributions to Mineralogy and Petrology 50, 49-64. Sisson, T.W. & Grove, T.L. (1993): Temperatures and H 2 0 contents of low-MgO high-alumina basalts. Contributions to Mineralogy and Petrology 113:167-184. Scoates, J.S., Lo Cascio, M. , Weis, D., & Lindsley, D.H. (2005a): Experimental constraints on the origin and evolution of mildly alkalic basalts from the Kerguelen Archipelago, southeast Indian Ocean. Submitted to Contributions to Mineralogy and Petrology. Scoates, J.S., Weis, D., Franssens, M. , Mattielli, N. , Annell, H., Frey, F.A., Nicolaysen, K. & Giret, A. (2005b): The Val gabbro plutonic suite, Kerguelen Archipelago: evolution and significance of a volcanic feeder system in an ocean island. To be submitted to Journal of Petrology. Smith, J.V. (1998): Interpretation of domainal groundmass textures in basalt lavas of the southern Lamington Volcanics, eastern Australia. Journal of Geophysical Research 103, 27383-27391. Stimac, J.A. & Pearce, T.H. (1992): Textural evidence of mafic-felsic magma interaction in dacite lavas, Clear Lake, California. American Mineralogist 77, 795-809. Thornber, C.R. (2001): Olivine-Iiquid relations of lava erupted by Kilauea Volcano from 1994 to 1998: implications for shallow magmatic processes associated with the ongoing East Rift Zone eruption. Canadian Mineralogist 39, 239-266. Toplis, M.J. & Carroll, M.R. (1995): An experimental study of the influence of oxygen fugacity on Fe-Ti oxide stability, phase relations, and mineral-melt equilibria in ferro-basaltic systems. Journal of Petrology 36, 1137-1170. Ulmer, P. (1989): The dependence of the Fe 2 +-Mg cation-partitioning between olivine and basaltic liquid on pressure, temperature and composition: an experimental study to 30 kbars. Contributions to Mineralogy and Petrology 101, 261-273. Ussier, W. & Glazner, A.F. (1989): Phase equilibria along a basalt-rhyolite mixing line: implications for the origin of calc-alkaline intermediate magmas. Contributions to Mineralogy and Petrology 101, 232-244. Weis, D., Frey, F.A., Leyrit, H. & Gautier, I. (1993): Kerguelen Archipelago revisited: geochemical and isotopic study of the SE Province lavas. Earth and Planetary Science Letters 118, 101-119. Weis, D. & Giret, A. (1994): Kerguelen plutonic complexes: Sr, Nd, Pb isotopic study and inferences about their sources, ages and geodynamic setting. Bulletin de la Societe Geologique, France, 166, 47-59. Weis, D., Frey, F.A., Giret, A. & Cantagrel, J.-M. (1998): Geochemical characteristics of the youngest volcano (Mount Ross) in the Kerguelen Archipelago: inferences for magma flux, lithosphere assimilation, and composition of the Kerguelen plume, journal of Petrology 39, 973-994. Weis, D., Frey, F.A., Schlich, R., Schaming, M. , Montigny, R., Damasceno, D., Mattielli, N. , Nicolaysen, K.E. & Scoates, J.S. (2002): Trace of the Kerguelen mantle plume: evidence from seamounts between the Kerguelen Archipelago and Heard Island, Indian Ocean. Geochemistry Geophysics Geosystems 3, 10.1029/2001GC000251. Xu, O , Frey, F., Weis, D., Scoates, J. S. & Giret, A. (2005): Geochemical characteristics of flood basalts from Mont Capitole in the central Kerguelen Archipelago: further constraints on the Kerguelen plume. In preparation. Yang, H.-J., Frey, F.A., Weis, D., Giret, A., Pyle, D. & Michon, G. (1998): Petrogenesis of the flood basalts forming the Northern Kerguelen Archipelago: implications for the Kerguelen plume. Journal of Petrology 39, 711 -748. Yang, H.-J., Frey, F.A., Clague, D.A., & Garcia, M.O. (1999): Mineral chemistry of submarine lavas from Hilo Ridge, Hawaii: implications for magmatic processes within Hawaiian rift zones. Contributions to Mineralogy and Petrology 135, 355-372. 92 CHAPTER 3 EVIDENCE FOR A DEPLETED MANTLE COMPONENT IN 25 MA MILDLY ALKALIC BASALTS FROM MT. MARION-DUFRESNE, KERGUELEN ARCHIPELAGO, SOUTHERN INDIAN OCEAN 93 3.1 INTRODUCTION The origin of the depleted isotopic compositions observed in many ocean island settings is currently the subject of much interest (e.g. Iceland: Fitton et al., 1997; 2003; Hanan et al., 2000; Kempton etal., 2000; Hawaii: Frey etal., 2005; Galapagos: Harpp & White, 2001; Kerguelen: Doucet et al., 2002; Frey et al., 2002). On the Kerguelen Archipelago, in the southern Indian Ocean, basaltic lavas with a clear depleted component (i.e. nonradiogenic Sr, Pb and radiogenic Hf, Nd isotopic compositions) were observed in early reconnaissance studies of the 29-24 Ma flood basalts (Storey et al., 1988; Gautier, 1990). The depleted component was attributed to interactions between Kerguelen mantle plume-derived melts and Southeast Indian Ridge (SEIR) asthenospheric mantle that occurred when the hotspot was ridge-centred (-43 Ma; Tikku & Cande, 2000) and as the ridge moved to the northeast relative to the hotspot. Recent studies of basaltic sections from across the Kerguelen Archipelago have been systematically carried out to develop a detailed time-integrated geochemical characterization of the Kerguelen hotspot over the -5 Myr of flood basaltic volcanism exposed on the archipelago. To date, it is evident that there are strong correlations between age, geographic location, alkalinity, and the isotopic composition of flood basalts on the archipelago (e.g. Weis etal., 1993; Yang et al., 1998; Nicolaysen et al., 2000; Frey et al., 2000; Frey et al., 2002; Doucet et al., 2002). In this study, we evaluate the major, trace element, and radiogenic isotopic compositions (Hf-Sr-Nd-Pb) of samples from a 700 m high section of -25 Ma flood basalts from Mt. Marion Dufresne, which is located in a previously unstudied area of the Plateau Central region of the Kerguelen Archipelago. Lavas from the Marion Dufresne section have major and trace element abundances and radiogenic isotopic compositions that are highly variable compared with other sections of similar age. In particular, a thick sequence of olivine-phyric high-MgO basalts from the top of the Marion Dufresne section is characterized by "depleted" (i.e. low Sr, Pb and high Hf, Nd) isotopic compositions that to date have only been observed in older (29-28 Ma) basalts from the northern part of the archipelago. We compare the geochemical characteristics of the Marion Dufresne lavas to previously studied sections from the Plateau Central (Mt. Tourmente and Mt. Capitole) to explain the differences between these three sections in terms of temporal changes in magmatic processes and source components. The 94 transition from tholeiitic-transitional to mildly alkalic volcanism, which is observed in all three Plateau Central sections, is addressed in detail with implications for degrees and depths of melting associated with the Kerguelen mantle plume. Finally, we examine the significance of "depleted" isotopic signatures in the -25 Ma Marion Dufresne lavas to better constrain the nature and composition of the depleted mantle source component in Kerguelen Archipelago flood basalts. The occurrence of 25 Ma lavas with depleted isotopic compositions remarkably similar to those observed in older (29-28 Ma) tholeiitic-transitional lavas has prompted a re-evaluation of the plume-ridge interaction model proposed by Doucet et al. (2002) to explain the controls on isotopic composition. This study provides key constraints for an important interval of magmatism on the Kerguelen Archipelago and thus contributes to a better understanding of the evolution of this large oceanic island and of Indian Ocean volcanism. 3.2 GEOLOGIC SETTING OF THE KERGUELEN ARCHIPELAGO The Kerguelen Archipelago is located at approximately 49°S, 070°E in the southern Indian Ocean (FIG. 3.1). It comprises a large central island of -100 km width surrounded by over 300 smaller islands (FIG. 3.2). With a total surface area of over 6500 km2, it is the third largest oceanic island after Iceland and Hawaii. The Kerguelen Archipelago is the emergent part of the <40 Ma Northern Kerguelen Plateau (FIG. 3.1), which is up to 20-25 km thick based on the results of wide-angle seismic refraction studies (Recq, 1994; Charvis, 1995; Charvis & Operto, 1999); the Northern Kerguelen Plateau represents an estimated volume of 2.3 x 106 km 3 of intraplate mafic magmatism associated with the Kerguelen hotspot (Coffin et al., 2002). Cenozoic flood basalts cover -85% of the surface of the Kerguelen Archipelago (FIG. 3.2); the remaining area is occupied by glacio-fluvial sedimentary deposits (-10%) and intrusive rocks (-5%). A notable feature of the archipelago is the presence of plutonic complexes (e.g. Rallier-du-Baty and Societe de Geographie) and smaller intrusive bodies (e.g. Val and Montagnes Vertes) exposed by erosion (e.g. Giret, 1983; Giret & Lameyre, 1983; Weis & Giret, 1994; Scoates etal. 2005a). A comprehensive 4 0 Ar/ 3 9 Ar dating study by Nicolaysen etal. (2000) on the flood basalts on the archipelago demonstrated that older (29-28 Ma) tholeiitic-transitional lavas are dominant in the 95 FIG. 3.1. Bathymetric map of the Indian Ocean (after Smith & Sandwell, 1997). The major physiographic features of the Indian Ocean basin are labelled. The location of the Kerguelen Archipelago on the Northern Kerguelen Plateau (NKP) is indicated, as are the major magmatic products of the long-lived Kerguelen plume, which include the Southern Kerguelen Plateau (SKP), the Central Kerguelen Plateau (CKP), Broken Ridge, Ninetyeast Ridge, the Rajmahal traps, and the Bunbury basalts. O D P drilling sites on the Kerguelen Plateau are also indicated (e.g. Site 1140 on the Northern Kerguelen Plateau). 96 UMBO'S ILES NUAGEUSES Quaternary deposits ^ Francals ^Ravin du Charbon \ & Ravin Jaune 24-25 Ma SOUTHEAST PROVINCE ValGabbro 170° \70°30'E 0 10 20 30 40 50 km FIG. 3.2. Simplified geologic map of the Kerguelen Archipelago after Nougier (1970). Mt. Marion Dufresne (black filled circle) is located in the southern part of the Plateau Central region of the archipelago. The locations and ages of previously studied basaltic sections are indicated. The age of Kerguelen Archipelago basalts ranges from -29.5 Ma in the northern part of the archipelago to -24 Ma in the east and southeast. 97 northern part of the archipelago, whereas increasingly younger and more alkalic lavas are exposed in the Plateau Central region (-26 Ma) and eastern parts of the archipelago (25-24 Ma). Detailed geochemical studies of the 29-28 Ma tholeiitic-transitional basalts from the northern part of the archipelago reveal the occurrence of basalts that are characterized by the presence of a depleted asthenospheric or lithospheric mantle component (with higher 1 7 6 Hf/ 1 7 7 Hf and 1 4 3 N d / 1 4 4 N d and lower 8 7Sr/ 8 6Sr, Sr and 2 0 8 Pb/ 2 0 4 Pb, 207Pb/2 0 4Pb, and 2 0 6Pb/ 2 0 4Pb). in their source (Yang et al., 1998; Doucet et al., 2002). Younger (<26 Ma) lavas typically have more restricted and slightly more enriched isotopic compositions (Frey et al., 2000; 2002; Doucet et al., 2005). The youngest flood basalts are the 25-24 Ma mildly alkalic lavas in the eastern part of the archipelago, which have a narrow range of more radiogenic isotopic compositions that are interpreted to represent the enriched component of the Kerguelen mantle plume source (Weis etal., 1998; Mattielli et al., 2002; Doucet etal., 2002). The correlation between the occurrence of basalts with depleted isotopic signatures with eruption age and geographic location strongly suggests that a depleted component is not intrinsic to the plume source, but rather that results from mixing with melts derived from depleted asthenospheric mantle (Doucet et al., 2002; Doucet et al., 2005). Spreading along the Southeast Indian Ridge (SEIR) separated Broken Ridge from the Kerguelen Plateau at -40 Ma (FIG. 3.2). SEIR-derived melts are interpreted to represent a major component (63-99%) of -34 Ma submarine basalts recovered at Site 1140 during Leg 183 of the international Ocean Drilling Program (ODP). Site 1140 was located on the Northern Kerguelen Plateau, approximately 270 km north of the Kerguelen Archipelago, and submarine basalts recovered from this site were erupted when the SEIR was -50 km from the Kerguelen hotspot (Weis & Frey, 2002). The SEIR was -350 km from the Kerguelen hotspot during the eruption of the oldest flood basalts on the archipelago, and grew increasingly distant (>500 km) over the -5 Myr of flood basaltic volcanism (Doucet et al., 2002; 2005). The nature and contribution of the depleted mantle component in Kerguelen Archipelago flood basalts from the Plateau Central region is relatively poorly constrained, yet this time period (-26-25 Ma) is critical for understanding the transition to mildly alkalic volcanism on the archipelago. Only two sections from the northern part of the vast Plateau Central region have been studied to date (Mt. Tourmente: Frey et al., 2002; Mt. Capitole: Xu et al., in 98 prep). These sections consist of transitional basalts and volumetrically minor mildly alkalic basalts with very homogeneous isotopic compositions. They also have distinctly lower 8 7Sr/ 8 6Sr and higher 1 4 3 N d / 1 4 4 N d compared to the 25-24 Ma lavas, which may reflect efficient mixing of plume-derived melts with a constant proportion or small volume of depleted mantle material (Frey et al., 2002). 3.3 GEOLOGY OF THE MT. MARION DUFRESNE BASALTIC SECTION Mt. Marion Dufresne, a prominent E-W trending ridge located on the east side of the Cook Glacier, is the southernmost section from the Plateau Central region of the archipelago to be studied (FIG. 3.2). Glaciation has exposed a -700 m section of subaerially erupted lava flows with thicknesses of 1-15 m. These flows are separated by debris, talus, or recessive-weathering units (i.e. scoriaceous tuffs and other pyroclastic deposits, highly vesiculated or rubbly lava flows, and oxidized flow tops) as indicated in FIG. 3.3. One mafic sill and two near-vertical mafic dikes that intrude the section were also sampled. The massive cores of all prominent lava flows (n=44) in the Marion Dufresne section were examined, of which 19 are aphyric (<5 vol. % phenocrysts), 16 are olivine-phyric (with 5-20% olivine ± <5% plagioclase ± <10% clinopyroxene phenocrysts), 6 are plagioclase-phyric (5-25 vol. % plagioclase), and 3 are plagioclase-ultraphyric (>50 vol. % plagioclase). In general, aphyric and plagioclase-phyric flows occur in the lower part of the section and olivine-phyric flows are dominant in the upper part of the section (FIG. 3.3). Although plagioclase-ultraphyric flows are thicker than other flows (up to 12-15 m), they are more readily eroded and altered. Three quartz-bearing basaltic andesites that occur in the upper part of the Marion Dufresne section contain olivine phenocrysts in addition to sieve-textured plagioclase and rounded, resorbed quartz crystals surrounded by pyroxene reaction coronas resulting from mixing between a basaltic magma and an evolved quartz-bearing magma (see Chapter 2). The abundance and diversity of phenocrysts in basalts from the Marion Dufresne section is unusual compared with the majority of the other studied basaltic sections from the Kerguelen Archipelago (n = 7), many of which contain significant amounts of aphyric basalts. Olivine 99 800 -L -526-525 ,527 ,25.0 ±0 .7 Ma 500 No geochemical data 500 Major/trace elements by XRF [5001 Additional traces by HR-ICP-MS E23 Full geochemical dataset including Hf, Sr, Nd, Pb isotopes FIGURE 3.3 Olivine-phyric basalts 5-25 vol% olivine phenocrysts minor clinopyroxene phenocrysts high MgO (6.7-11.4 wt. %) low Si02 (46-48.5 wt. %) Quartz-bearing basaltic andesites <5 vol. % plagioclase crystals 1-2 vol. % quartz crystals relatively high MgO (5.1-6.7 wt. %) high S i 0 2 (54.4-55.9 wt. %) Plagioclase-phyric basalts 5-25 vol% plagioclase phenocrysts low MgO (4.4-4.8 wt. %) 48.4-49.7 wt. % Si02 & Plagioclase-ultraphyric basalts 50-60 vol. % plagioclase phenocrysts low MgO (3-3.8 wt. %) 47.8-48.6 wt. % S i 0 2 !Sa— 23.8 ± 1.6 Ma -575 -2500 m Aphyric basalts and trachybasalts <5 vol. % phenocrysts low MgO (3.7-5.2 wt. %) wide range of Si02 (46.9-50.2 wt. %) FIG. 3.3. Stratigraphy of the >700 m high section of lavas exposed at Mt. Marion Dufresne. All prominent lava flows in the section (indicated by horizontal solid lines) are between 1-15 m thick. Flows are separated by intervals of debris, talus, or recessive-weathering units including scoriaceous tuffs and other pyroclastic deposits, highly vesiculated or rubbly lava flows, and oxidized flow tops, which were not sampled. The dashed lines indicate mafic dikes (cross-cutting lines) and sills (i.e. sample 534; horizontal dashed line). Samples from the base and top of the section dated by 4 0 Ar/ 3 9 Ar geochronology are indicated; ages cited for samples 532 and 569 are 4 0 Ar/ 3 9 Ar plateau ages Of leached whole rock. Three major stratigraphic units occur in the section: aphyric basalts and trachybasalts in the lower 200 m, plagioclase-phyric and plagioclase-ultraphyric basalts in the interval of 200-400 m, and olivine-phyric basalts in the upper 400 m of the section. The grey solid lines indicate trachybasaltic flows. The quartz-bearing basaltic andesites are three distinct flows in the upper part of the Marion Dufresne section that record mixing between an olivine-phyric high-MgO basaltic magma and an evolved quartz-bearing trachytic (?) magma (see Chapter 2). 707 phenocrysts occur in some older tholeiitic basalts to the north (e.g. <15 vol. % olivine in lavas from the Loranchet Peninsula; Doucet et al., 2002), whereas plagioclase phenocrysts are increasingly common in younger, mildly alkalic basalts (e.g. <17 vol. % plagioclase in lavas from Mt. Crozier; Damasceno et al., 2002). Cobbles of high-MgO basalt and picrite recovered from moraines across the southeast part of the archipelago contain significant phenocryst abundances (up to 20% olivine, 20% clinopyroxene, and 15% plagioclase by volume), including some samples of high-MgO basalt and picrite collected around the Armor site that may have been derived from erosion of Plateau Central flows (Doucet et al., 2005). Low-temperature alteration of Marion Dufresne lavas is demonstrated by local alteration of volcanic glass and groundmass minerals to fine-grained clays and oxyhydroxides, particularly adjacent to vesicles, as well as minor alteration of olivine and plagioclase phenocrysts to iddingsite and sericite, respectively. In addition, zeolite recrystallization is noted in many vesicular samples. These observations of post-magmatic alteration are similar to other studies of Kerguelen Archipelago flood basalts (e.g. Nativel, 1994; Nicolaysen etal., 2000). 3.4 ANALYTICAL TECHNIQUES 3.4.1 40Ar/39Ar geochronology Crystallization ages of Marion Dufresne basalts were determined by the 4 0 Ar/ 3 9 Ar incremental step-heating method, which readily identifies sources of inaccuracy such as 4 0 A r loss due to alteration, excess 4 0 Ar from undegassed mantle sources, and recoil problems in fine-grained samples. For extensive descriptions of the different spectra associated with these effects in basaltic rocks, see Duncan et al. (1997), McDougall & Harrison (1999), and Duncan (2002). The 4 0 Ar/ 3 9 Ar method is appropriate for Kerguelen basalts given that all samples contain groundmass plagioclase, an Ar-retentive mineral, and that the decay constant for 4 0 K is appropriate for the time frame of volcanism on the Kerguelen Archipelago (Nicolaysen etal., 2000). Apparent ages can be considered as accurate estimates of the sample crystallization age if there is a well-defined plateau (see results for criteria used) and inverse isochron, which are concordant, and if the calculated 4 0 Ar/ 3 6 Ar intercept reflects atmospheric values 102 ( 4 0Ar/ 3 6Ar= 295.5±1; McDougall & Harrison, 1999) (e.g. Duncan, 2002; Duncan etal., 1997; and Sharp etal., 1996). Samples were selected on the basis of microscopic appearance, chemical composition, low loss-on-ignition (LOI) values, and stratigraphic position. Samples with minimal post-magmatic alteration were chosen in an attempt to limit the loss of 4 0 Ar due to low-temperature zeolite facies burial metamorphism, surface weathering, and seawater alteration; secondary minerals such as zeolites have higher K/Ca than primary igneous phases, which can result in measured ages significantly younger than crystallization ages. The samples were crushed in a hydraulic piston crusher between WC-plates (no grinding) using the percussion method to obtain a coarse fraction (-1-5 mm), then hand-crushed using a stainless steel mortar and pestle and sieved to 250-450 pm. Mafic minerals and oxides were removed using a Frantz magnetic separator due to the tendency of olivine and clinopyroxene to trap excess magmatic Ar (e.g. McDougall & Harrison, 1999). The least magnetic fraction was subsequently hand-picked to remove stained and altered grains to produce a -100-200 mg whole rock separate. All samples were leached with cold 3N HCI in an ultrasonic bath for 15 minutes then rinsed with deionized water following the procedure of Nicolaysen et al. (2000). This process was repeated, typically 1-2 times, until a colourless supernatant was achieved. The samples were then rinsed with 1 N nitric acid and finally with deionized water, dried at room temperature, then wrapped in aluminium foil and stacked in an irradiation capsule with similar-aged samples and neutron flux monitors (Fish Canyon Tuff sanidine, 28.02 Ma; Renne et al., 1998). Neutron irradiation to produce 3 9 Ar from 4 0 K was done at the McMaster Nuclear Reactor in Hamilton, Ontario for 28 megawatt hours (MWh), with a neutron flux of approximately 3x10 1 6 neutrons/cm2. Analyses (n=48) of 16 neutron flux monitor positions produced 2o errors of 0.4% in the J value, which is a measure of neutron flux in the flux monitor. Samples were step-heated at incrementally higher intensities in the defocused beam of a 10-W C 0 2 laser (New Wave Research MIR10) until fused. The gas evolved from each step was analyzed by a VG5400 mass spectrometer equipped with an ion-counting electron multiplier at the Pacific Centre for Isotopic and Geochemical Research (PCIGR), University of British Columbia. All measurements were corrected for total system blank, mass spectrometer sensitivity, mass discrimination, radioactive decay during and 103 subsequent to irradiation, as well as for interfering Ar from atmospheric contamination and the irradiation of Ca, CI and K using the following measured isotope production ratios: ( 4 0Ar/ 3 9Ar)K = 0.03 02, ( 3 7Ar/ 3 9Ar)Ca = 1416.4306, ( 3 6Ar/ 3 9Ar)Ca = 0.3952, and Ca/K = 1.83(37ArCa/39ArK)). 3.4.2 Major and Trace Element Compositions Initial sample preparation of 47 samples from the Marion Dufresne section was carried out at the Universite Libre de Bruxelles in Belgium. The samples were cut using a diamond-embedded saw to remove surface alteration and the cut surfaces were then abraded with sandpaper to eliminate saw traces and residual alteration. The samples were coarse-crushed in a hydraulic piston crusher between WC-plates using the percussion method and subsequently reduced to powder in an agate planetary mill. All 47 samples were analyzed for major element compositions and some trace elements (Rb, Sr, Ba, V, Cr, Ni, Zn, Ga, Y, Zr, Nb, Ce and La) by X-ray fluorescence (XRF) at the University of Massachusetts, Amherst (see Rhodes, 1996 for analytical procedure). Major and trace element compositions by XRF are the mean of duplicate analyses and are reported in Appendix II. A subset of 26 samples with low LOI values (typically <2.2 wt. %) representative of the sample suite were prepared for trace element analysis by high resolution inductively coupled plasma mass spectrometer (HR-ICP-MS) at the PCIGR. To dissolve the powders, 2 mL of concentrated sub-boiled HF and 500 pL of concentrated Seastar® H N 0 3 were added. The samples were sealed and placed on a hotplate for 48 hours at a temperature of approximately 120°C. Twice during this time interval, the samples were placed in an ultra-sonic bath for 15 minutes to ensure complete dissolution of the powder. The samples were then uncapped and placed on a hotplate until dry. Once dry, 2 mL of Seastar® concentrated H N 0 3 were added. After 24 hours on the hotplate, the samples were uncapped and dried once again. Approximately 4 g of a stock solution of 1 % H N 0 3 spiked with 1 ppb Indium (In) were added and the samples were placed on a hotplate for 24 hours. Twice during this time interval, the samples were placed in an ultra-sonic bath for 15 minutes to ensure complete redissolution. The samples were then transferred to clean 125 mL high density polyethylene (HDPE) sample bottles. Two separate dilutions were carried out. Samples were diluted to 10OOx for the rare 104 earth elements (REEs), and to 2000x for all other elements, after an initial dilution to 1000 times their initial weight with the stock solution of 1% H N 0 3 with 1 ppb In. A portion of the 1000x diluted mixture was transferred to clean 60 mL HDPE sample bottles and subsequently diluted to 2000x. A proportional amount of concentrated HCI was added to each of the 2000x dilutions so that the resulting solutions would have a concentration of 1% HCI. The samples were analyzed on a Thermo Finnigan Element2 HR-ICP-MS at the PCICR. The REEs were measured in high resolution, while Pb and U were measured in low resolution. Most remaining elements were measured in medium resolution mode, with the exception of Cd, Sn, Sb, Cs, Ta, W, and Bi, which were also measured in low resolution mode. A series of 6 standards, obtained by series dilution from 1000 ppm High Purity® stock standard solutions using 1% H N 0 3 with 1 ppb In, was used for external calibration and concentration calculation. All trace element abundances by HR-ICP-MS are shown in Appendix III. Data reproducibility is demonstrated by the analysis of three complete procedural duplicates of samples BOB93-528, 552 and 561; see Appendix IV. Repeated measurements of a USGS basaltic reference material (BHVO-1) and an internal reference material (1140A-31 R-1) were also made to ensure the accuracy of the concentration results; see Appendix V. Procedural blank values were in the parts-per-trillion (ppt) range, and are considered negligible in comparison to the sample concentrations, which are in the parts-per-million (ppm) range. A solution of 4% Aqua regia + 0.05% HF was used to rinse the instrument between each sample to minimize memory effects from the previously analyzed sample. Background values were checked periodically during the course of analysis, especially after high abundance samples or standards, to detect and correct for memory effects. Indium, which is spiked at 1 ppb in all blank, standard, and sample solutions, was used as an internal standard to monitor and correct for sensitivity drift. Low-level (5 ppb) REE standard solutions were run after every fifth or sixth sample to detect and correct for instrumental mass drift. Trace element abundances by XRF and HR-ICP-MS show excellent agreement (R2 values are >90 for all elements; Appendix VI). 105 3.4.3 Isotopic Compositions A total of 22 samples with LOI <2.2 wt. % were selected for Hf-Pb-Nd-Sr isotopic analyses. For Pb-Nd-Sr analyses, all samples were acid-leached in cold 6N HCI following the procedure of Weis & Frey (1991) to remove any secondary alteration products such as zeolites and carbonates, which are common in basalts from the Kerguelen Archipelago. For the Marion Dufresne samples, between 8-12 leaching steps were carried out to achieve a clear solution, and -30-70% of the initial sample weight was lost. The leached residues were digested in a solution of -8:1 concentrated HF:HN0 3 on a hotplate for 48 hours, and then samples were uncapped and dried on a hotplate. Subsequently, 6N HCI was added to all samples for an additional 24 hours on the hotplate, and then samples were again dried. To ensure complete digestion, the samples were placed in an ultrasonic bath for 15 minutes at regular intervals throughout this procedure. Samples were redissolved in 1 mL 0.5N HBr on the hotplate and centrifuged prior to Pb column chemistry. Pb separations were made on anion exchange columns in an HBr-HCI medium using 0.5N HBr and sub-boiled 6N HCI. Each Pb separate was then analyzed on the Elements HR-ICP-MS to determine the abundance of Pb. Depending on the abundance of Pb, aliquots of between 25% and 100% were taken from the remaining 0.9 mL solutions and diluted to 2 mL with 0.05M H N 0 3 to ensure that a minimum 2V 2 0 8 Pb signal would be exceeded in all the samples during analysis. The samples were spiked with a solution of 5 ppm Thallium (TI). For every ng of Pb present in the samples, 0.25 ng of TI was added to achieve a Pb/TI ratio of 4, which is critical for proper correction for mass fractionation. The samples were analyzed on a Nu Instruments Nu Plasma MC-ICP-MS (Nu 021). A desolvation nebulizer (DSN) was used with a nebulizer pressure of 37 psi, hot gas flow of 0.23 L/min, and membrane gas flow of 3.01 L/min. Membrane and spray temperatures were 110°C. The uptake rate of the Class Expansion Micromist nebulizer was between 150-160 pL/min. Standard-sample bracketing was used to monitor and correct for instrument drift during the course of analyses. Pb isotopic compositions were measured in the static mode with an interference correction for 2 0 4 H g . Pb compositions in each sample are derived from the mean of 60 analyses (3 blocks of 20 cycles). Pb isotopic data was corrected for mass fractionation using 2 0 5TI/ 2 0 3TI = 2.3885. Average triple spike values 106 for NBS 981 were 2 0 6 Pb/ 2 0 4 Pb = 16.9414 ± 0.001 7 (2a), 2 0 7 Pb/ 2 0 4 Pb = 15.4967 ± 0.0016 (2a), and 208p b /204p b _ 36 7 1 5 1 ± 0.0041 (2a). These values fall within 2a error of the accepted triple spike values ( 2 0 6Pb/ 2 0 4Pb = 16.9405, 2 0 7Pb/ 2 0 4Pb = 15.4963, and 2 0 8 Pb/ 2 0 4 Pb = 36.7219) (Caler & Abouchami, 1998). Sr and Nd were separated using two separate, sequential columns in an HCI medium (after Weis & Frey, 1991). Prior to Sr column chemistry, 5 mL of 1.5N HCI were added to the dried samples, which were then placed on a hotplate for 5 minutes, then in an ultrasonic bath for 5 minutes, and finally centrifuged before being transferred to the columns. For Nd separations, the dried samples were redissolved in 0.5 mL of 0.16N HCI and placed in an ultrasonic bath for 5 minutes, then added to the columns using a pipette. The Sr and Nd isotopic measurements were carried out on a Thermo Finnigan Triton-TI thermal ionization mass spectrometer (TIMS). Sr and Nd isotopic compositions were measured in the static mode with relay matrix rotation. Sr and Nd compositions in each sample are the mean of 135 analyses (9 blocks of 15 cycles). During the course of analyses, 15 analyses of the NBS987 Sr standard and 16 analyses of the La Jolla Nd standard were made, with mean values of 8 7Sr/ 8 6Sr = 0.710250 ± 0.000011 (2a) and , 4 3 N d / 1 4 4 N d = 0.511853 ± 0.000009 (2a), respectively, which are within the range of accepted values. The Sr isotopic data were corrected for mass fractionation using 8 6Sr/ 8 8Sr = 0.1194, while the Nd isotopic data were corrected using l 4 6 N d / 1 4 4 N d = 0.7219. Sr and Nd abundances in procedural blanks are negligible in comparison to sample abundances. For the Hf isotopic analyses, 250-300 mg of whole rock powder for each of the 24 samples was dissolved following the procedure described in Blichert-Toft et al. (1997). The powders were dissolved using sub-boiled HF and H N 0 3 in Savilex Teflon vials. After drying down the dissolved sample, concentrated HF was added to precipitate REE fluoride salts and separate the REEs from the remaining sample. High field strength elements (HFSE) were then separated from the matrix using an anion exchange column, and Hf and Zr were isolated from the HFSE concentrate using a cation exchange column. The Hf isotopic compositions were analyzed in the static mode on a Nu Plasma MC-ICP-MS (Nu 021) at the PCICR in the "wet" plasma mode. Both Lu and Yb beams were monitored during the course of analysis for interference corrections on mass 176; the Yb beam was negligible. The 107 measured Hf isotopic ratios were corrected for isobaric interference with Lu at mass 176 by monitoring the isotope 1 7 5 Lu; the , 7 6 L u interference was subtracted using a value of 37.69969 for , 7 6 L u / , 7 5 L u (Rosman & Taylor, 1998). During the analyses, replicate measurements of the Hf JMC475 in-house standard gave 0.282159 ± 0.000015 (2om on 23 measurements), which is within the range of previously published values for this standard (e.g., Blichert-Toft et al., 1997; Chauvel & Blichert-Toft, 2001; Coolaerts et al., 2004). Replicate analyses of two samples agree within 0.08 epsilon units. 3.5 RESULTS 3.5.1 Age of the Marion Dufresne Section Details of the analyses, including plateau spectra and inverse correlation plots, are presented in FIG. 3.4 and Table 3.1. The plateau and correlation ages were calculated using Isoplot 3.09 (Ludwig, 2003). Errors are quoted at the 95% confidence level (2o) and were propagated from all sources, except mass spectrometer sensitivity and age of the flux monitor. The best statistically-justified plateau and plateau age were picked based on the following criteria (after Ludwig, 2003): 1. a well-defined, mid- to high-temperature plateau was formed by at least three concordant, contiguous steps, representing a50% of the 3 9 Ar released; 2. the probability of fit of the weighted mean age was greater than 5%; 3. the slope of the error-weighted line through the plateau ages equalled zero at 95% confidence; 4. when six or more steps were used, the ages of the two outermost steps on a plateau were not significantly different (1.80) from the weighted-mean plateau age. A sample of aphyric basalt from near the base of the section (BOB93-569; 170 m) yields an 4 0 Ar/ 3 9 Ar plateau age of 23.8 ± 1.6 Ma and an inverse isochron age of 23.3 ± 4.5 Ma (FIG. 3.4). Sample BOB93-532, from near the top of the section (715 m), is an olivine-phyric high-MgO (9.34 wt. %) basalt and has a plateau age of 25.0 ± 0.7 Ma and an inverse isochron age of 25.4 ± 1.3 Ma. Both samples are transitional basalts with similar Alkalinity Index (A.I.) values of 0.01 and 0.02 respectively, where A.I. = (Na 20+K 20) - 0.37*SiO2+ 14.43 (Macdonald & Katsura, 1964). The overlapping ages of samples from the base and the top of Marion Dufresne indicate that this section was erupted in less than 1 Myr. These ages fall within the range of ages previously reported for Kerguelen Archipelago 108 50 40 I 3 0 oi < 20 II 10 en < 50 4 0 ro 3 0 20 10 Plateau age= 25.03±0.69 Ma MSWD= 0.59, probabi l i ty= 0.67 Steps 6-10 ( includes 73.8% of the 3»Ar) BOB93-532 715m Leached whole rock 20 4 0 60 Cumulative % Ar 80 100 Plateau aga= 23.8+1.6 Ma MSWD= 0.57, p r o b a b i l i t y 0.76 Steps 1 -7 (includes 100% of the " A r ) BOB93-569 170 m Leached whole rock 2 0 4 0 6 0 Cumulative % Ar 80 100 3.4 1.0 Inverse Isochron age= 25.4+1.3 Ma Initial " A r / 3 6 / ^ 292.2±9.1 M S W D = 0 . 5 3 BOB93-532 715m Leached whole rock 0.02 0 .04 0 .06 0 .08 0 .10 0.12 0.14 0 .16 3 9 Ar/ 4 0 Ar Inverse isochron age= 23.3±4.5 Ma Initial " A f / 3 * A r - 2 9 8 * 1 8 MSWD= 0.16 1.8 1.4 BOB93-569 170 m Leached whole rock 0.00 0 .02 0.04 0 .06 0 .08 0.10 0 .12 3 9 Ar/ 4 0 Ar FIGURE 3.4 FIG. 3.4. Step-heating spectra and inverse isochron diagrams for samples of leached whole rock from the base and top of the Marion Dufresne section (BOB93-569 and 532, respectively) dated by the 4 0 Ar/ 3 9 Ar method. Box heights and data point error ellipses are 2a and include a calculated J-error of 0.4%. Eruption ages cited in the text are calculated from age spectra plateaus, which are defined by at least three consecutive steps that represent >50% of the total Ar released that have ages that agree within 2o. Plateau steps used in the age calculations are filled, and rejected steps are open. All heating steps were used in the inverse isochron age calculations. The 3 6 Ar/ 4 0 Ar intercept indicates the initial ratio of 4 0 Ar/ 3 6 Ar in each sample, which is within error of the atmospheric value (295.5 ± 1; McDougall & Harrison, 1999). Mean square weighted deviations (MSWD) for the ages of these two samples are <0.6. 110 Table 3.1: Detailed "Ar/^Ar step-heating results from the Marion Dufresne section, Kerguelen Archipelago. L a s e r intensity (%)' 4 0 A r / 3 9 A r ±2o " A r / ^ A r ±2o 3 7 A r / 3 9 A r ±2o 4 ° A r ' / 3 9 A r ±2o % 4 °Ar* B O B 9 3 - 5 3 2 , o l iv ine-phyr ic high M g O basalt (715 m) 2J = 0 .003489 ± 0 .000008 (2a) 1.8 1843.108 0 .075 1.356 0.091 1.577 0.083 35 .195 55.503 98.09 208 .95 2.0 830 .589 0.032 0.607 0.059 1.429 0.041 14 .048 15.933 98.31 86 .32 2.2 190.126 0.012 0.152 0.039 1.219 0.027 7.580 3.714 96.01 47 .09 2.4 94 .868 0 .010 0.081 0.048 1.028 0.023 5.461 1.718 94 .23 34 .05 2.6 135.714 0.022 0.133 0.098 0.797 0.040 7.246 3.544 94 .65 45 .04 2.8 24 .755 0.012 0.033 0 .035 0.996 0.019 3 .923 0.502 84.06 24 .53 3.0 11 .353 0 .019 0.021 0.081 1.073 0.021 3 .812 0.324 65 .83 23 .84 3.3 7 .839 0.018 0.019 0 .060 1.278 0.023 4 .077 0.184 47 .02 25 .48 3.6 8 .845 0.016 0.019 0.068 1.463 0.022 3.979 0.187 54.02 24 .87 3.9 7.946 0.017 0.022 0 .075 1.885 0.023 4 .067 0.307 47 .75 25 .42 4 .3 8.261 0 .019 0.024 0.068 2.618 0.025 3 .585 0.150 55 .46 2 2 . 4 3 B O B 9 3 - 5 6 9 , aphyr ic basa l t (170 m) 2J = 0.003474 ± 0 .000008 (2o) 1.8 1858.360 0 .049 1.249 0.091 8.166 0.054 10.851 39 .35 99.42 66 .76 2.0 65 .839 0.017 0.059 0.100 9.333 0.020 4.401 2.513 93 .40 27 .37 2.4 23 .699 0.011 0.031 0.074 9.586 0.018 4 .003 0.698 83.21 24 .92 2.7 10.461 0.026 0.026 0.097 9.315 0.027 3.510 0.507 65.91 21 .87 3.2 11.517 0 .022 0.030 0.073 8.737 0.022 3 .765 0.544 66 .79 23 .44 3.7 11.595 0.017 0.036 0.063 8.931 0.022 4.141 0.575 63 .53 25 .77 4.2 13.556 0.021 0.027 0.113 10.441 0.023 3.846 0.661 71 .09 23 .94 M e a s u r e d va lues of 4 0 A r / 3 9 A r , 3 8 A r / ^ A r , 3 7 A r / 3 9 A r , and 4 0 A r * / 3 9 A r , as wel l a s ca lcu lated a g e s , are s h o w n for e a c h increment of step heat ing. A l l measu remen ts were cor rec ted for interfering A r frorr irradiation of C a , CI and K us ing measu red isotope production ratios: f*°Ar/ 3 9 Ar)K = 0 .0302, ( 3 7 A r / 3 9 A r ) C a = 1416.4306, ( 3 6 A r / 3 9 A r ) C a = 0 .3952 , a n d C a / K = 1 . 8 3 ( 3 7 A r C a / 3 9 A r K ) . ' Intensity in the d e f o c u s e d b e a m of a 10-W C 0 2 laser . 2 T h e J va lue ( -0 .4%) ind icates the irradiation flux for e a c h samp le , wh ich w a s ca lcu la ted relative to a neutron flux moni tor (F ish C a n y o n Tuff san id ine ; 28 .02 M a ; R e n n e et a l . ,1998) . Er rors a re 2 •As te r i sks indicate unrad iogen ic (i.e. a tmospher ic) a rgon; % 4 C A r * indicates the percentage of m e a s u r e d 4 0 A r that is unrad iogen ic . 111 flood basalts (29.3 ± 0.9 Ma to 24.5 ± 0.3 Ma) and are comparable with those from the Mont Tourmente section in the northern Plateau Central (25.3 ± 0.6 Ma to 26.0 ± 1.0 Ma) (Nicolaysen et al., 2000). The isochron ages determined for both samples are concordant with plateau ages, but have slightly larger uncertainties due to the small dispersion of very radiogenic step compositions. Thus, the calculated plateau ages are interpreted to represent the age of crystallization of the two dated basalts. 3.5.2 Major element variations Based on the total alkalis vs. silica classification diagram of Le Bas et al. (1986), most of the Marion Dufresne lavas are basalts (n=38) with minor trachybasalts (n=3) and basaltic andesites (n=3) (FIG. 3.5). All Marion Dufresne samples cluster about the Macdonald & Katsura (1964) tholeiitic-alkalic dividing line and span a range of A. l . values, from transitional-tholeiitic to mildly alkalic (-2 to +2). The highest A. l . values (1.4-2.0) are observed in two trachybasaltic flows marking the base of the section and one thick (-10 m) trachybasaltic flow from near the top of the section. A. l . decreases sharply with increasing stratigraphic height over the lower 200 m of the section, from 1.8 to 0.1 (FIG. 3.6). This temporal shift to decreasing alkalinity with decreasing age contrasts with regional trends observed on the archipelago. These mildly alkalic lavas are overlain by 200 m of transitional basalts; A. l . values increase with stratigraphic height (from -0.8 to +0.8) across this interval. No systematic difference in alkalinity is noted with increasing MgO content, although olivine-phyric high-MgO basalts (7.1 -11.4 wt. %) from this section have a more limited range of A. l . values (-0.8 to +0.5). All quartz-bearing basaltic andesites are tholeiitic with anomalously low A. l . values (-1.5 to -2). All dikes and sills sampled are mildly alkalic (A.l. = 0.5 to 1). Olivine phenocryst chemistry indicates that MgO is a valid index of magma differentiation for the Marion Dufresne suite in samples with <10 wt. % MgO (i.e. samples with >10 wt. % MgO contain accumulated olivine; see Chapter 2). The Marion Dufresne lavas span a wide range of MgO contents (3.0-11.4 wt. %) (Table 3.2, and FIGs. 3.6 and 3.7) and can be divided into two broad geochemical groups based on MgO content and stratigraphic height. The "low-MgO" group comprises the lower 300 m of the section and consists of aphyric basalts and trachybasalts, plagioclase-phyric basalts, and 112 FIG. 3.5. Total alkalis vs. silica classification diagram modified from Le Bas et al. (1986) showing that all of the Marion Dufresne samples cluster about the Macdonald & Katsura (1964) tholeiitic-alkalic boundary. Deviations are limited to three aphyric trachybasalts, which are the most alkalic samples recovered from the Marion Dufresne section, and three quartz xenocryst-bearing basaltic andesites, which are "tholeiitic". Compositions of the Marion Dufresne lavas overlap fields for previously studied 29-24 Ma Kerguelen Archipelago lavas. (Data sources: Weis et al., 1993; Damasceno, 1996; Yang et al., 1998; Chauvel & Blichert-Toft, 2001, Frey et al., 2000; Frey et al., 2002; Doucet et al., 2002, Weis & Frey, 2002; Xu et al., in prep.) 113 Stratigraphic height (m) o o ro w A cn o o o o o o o o 8 -J CO o o o o i N m ui as ' " 0 L X l IN) ^ m o o o • o o o <o8 s? o - V > • • > • -» is) S 3 co 8 o o Cn o o o> -J. co o o o o o o Stratigraphic height (m) to CO cn oi co g g g § Q O O Q o o 1 ^ O u O I on 3D ft (fc <b > O ^3 _ «L a H—I—H -1—I—h- H—I—I-M Pun o fl) o IS) o o o o co A cn O) -vi co o o o o o o o o o o o o Stratigraphic height (m) IS) CO cn o -J. co 8 8 8 8 8 8 8 6 ro o 4 > f > H 1 1 1 1-h o I 1 . • • • i u > t> O 0% °o o - * N) co A cn O) co o o o o o o o o o o o o o o o o FIG. 3.6. Stratigraphic variations of major element abundances in Marion Dufresne lavas. The Marion Dufresne section is primarily divided on the basis of MgO content, reflecting relative degrees of fractionation and accumulation of olivine. MgO contents of >6 wt. %, which are observed in the olivine-phyric basalts from the upper part of the section, are relatively rare in Kerguelen Archipelago basalts. Alkalinity index (A.l.) is the deviation from the Macdonald & Katsura (1964) line ((Na 20+K 20)-S i0 2 x 0.37 +14.43) where negative A. l . values indicate transitional-tholeiitic samples and positive A. l . values indicate mildly alkalic samples. All major element compositions were determined by XRF. Legend as in FIG. 3.5. 115 FIG. 3.7. MgO variation diagrams for samples from the Marion Dufresne section. The high-MgO and low-MgO basalts form distinct trends. Note that the plagioclase-ultraphyric basalts have characteristically high A l 2 0 3 and low MgO contents, indicating plagioclase accumulation. The quartz-bearing basaltic andesites have distinctly higher S i0 2 and lower T i 0 2 and Fe 2 0 3 , and the aphyric trachybasalts have higher Na 2 0 and P 2 0 5 contents relative to the aphyric basalts. The linear trends in Ni and Cr vs. MgO indicate the effects of olivine fractionation and olivine accumulation (in the highest MgO samples). All major and trace element abundances were determined by XRF. Legend as in FIG. 3.5. 117 Table 3.2: Major (wt. % oxides) and trace element abundances (ppm) in Marion Dufresne lavas. Rock type: Olivine-phyric high-MgO basalt Sample: BOB-93-527 BOB-93-528 BOB-93-530 BOB-93-531 BOB-93-532 BOB-93-535 BOB-93-536 BOB-93-537 BOB-93-538 Height (m): 760 755 740 730 715 635 620 600 595 Major elements (wt%) Si0 2 46.81 45.89 46.51 46.14 46.31 47.21 47.47 48.29 47.90 T i 0 2 1.83 1.82 1.83 1.83 1.92 1.74 1.94 1.99 1.93 A l 2 0 3 14.36 14.01 15.23 15.08 15.13 13.83 14.98 15.60 15.81 Fe 2 0 3 " 12.74 12.99 12.74 12.58 12.60 12.81 12.39 12.37 12.20 MnO 0.19 0.19 0.20 0.20. 0.20 0.19 0.18 0.18 0.18 MgO 10.74 11.44 9.77 9.77 9.34 11.43 9.66 7.43 7.61 CaO 9.92 9.96 10.72 11.05 10.95 10.10 9.79 10.85 10.80 Na 2 0 2.38 2.63 1.92 2.29 2.30 1.86 2.20 2.44 2.42 K 2 0 0.62 0.39 0.40 0.40 0.42 0.49 0.84 0.48 0.48 P 2 O s 0.23 0.23 0.24 0.25 0.27 0.27 0.35 0.31 0.31 TOTAL 99.82 99.55 99.56 99.59 99.44 99.93 99.80 99.94 99.64 LOI 6.79 2.20 3.90 1.65 0.88 4.09 5.00 1.13 3.15 mg- number1 0.65 0.66 0.63 0.63 0.62 0.66 0.63 0.57 0.58 A.I. 2 0.11 0.47 -0.46 0.05 0.02 -0.69 -0.09 -0.52 -0.39 Trace elements (ppm) R b X R F 12.0 4.2 5.0 5.0 5.7 5.4 9.4 6.7 5.4 g ^ R F 226 392 345 354 368 621 665 372 378 B a , c p 79.2 88.5 101 187 121 Sc ' c p 25.2 33.2 35.7 30.5 36.4 yXRF 189 219 237 246 245 199 183 230 221 C o , o p 62.6 61.2 59.6 56.0 54.6 N f " 195 288 216 201 188 285 190 115 117 C u ' c p 64.5 98.3 98.0 80.2 104 Z n X R F 86 89 83 84 88 91 95 96 93 G a X R F 19 18 17 17 18 15 19 20 20 yXRF 19.8 20.5 22.9 23.4 24.6 20.4 21.5 25.4 24.1 Z r X R F 119 128 134 140 151 135 159 167 157 N b X R F 8.2 8.3 8.8 9.0 10.2 10.9 14.7 12.4 11.6 Hf 1 0 , 2.59 3.00 3.12 3.33 3.60 T a l c p 0.499 0.586 0.656 0.994 0.779 T h l c p 0.277 0.426 0.535 0.834 0.787 0.161 0.168 0.195 0.255 0.244 P B ! C P 0.781 0.761 0.862 1.08 0.992 La ' C P 10.9 10.5 11.3 17.3 15.9 C e l c p 26.1 27.1 30.4 40.4 36.5 P | J C P 3.68 3.71 4.60 5.51 5.06 N d i c p 15.5 16.0 18.8 21.4 20.4 S m l c p 4.13 4.12 4.47 4.87 5.09 E u I C P 1.48 1.51 1.66 1.79 1.80 G d , c p 4.81 4.73 4.85 4.92 5.48 T b I C P 0.681 0.693 0.809 0.760 0.830 D y , c p 4.08 4.35 5.08 4.31 4.69 H o l c p 0.785 0.857 1.06 0.888 0.988 Er.cp 2.09 2.30 2.57 2.35 2.68 Yb,cp 1.79 2.06 2.30 2.00 2.38 L u l c p 0.256 0.300 0.328 0.268 0.304 L i , c p 2.34 3.71 3.19 4.63 4.38 M o l c p 0.718 0.670 0.706 1.04 1.00 C d l c p 0.072 0.127 0.112 0.102 0.138 Sn ' c p 0.853 0.991 1.15 1.06 1.25 S b , c p <lod <lod <lod <lod <lod C s , C P 0.277 0.223 0.292 0.209 0.225 W I C P 0.891 1.50 1.46 0.766 1.36 B i l c p 0.211 0.154 0.213 0.198 0.207 *all Fe as Fe203 mg -number= Mg/tMg+Fe**). 2A.\. (Alkalinity Index) = (Na2O+K2O)-0.37'SiO2+14.43; mildly alkalic samples have positive Al values and transitionat-tholetittc samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020°C. <lod = below limit of detection. 118 Table 3.2 (continued): Major (wt. % oxides) and trace element abundances (ppm) in Marion Dufresne lavas. Rock type: Olivine-phyric high-MgO basalt Plagioclase-phyric basalt Sample: BOB-93-541 BOB-93-543 BOB-93-545 BOB-93-547 BOB-93-549 BOB-93-550 BOB-93-551 BOB-93-554 BOB-93-557 Height (m): 545 530 490 470 450 445 430 390 330 Major elements (wt%) S i 0 2 48.45 47.98 48.28 48.04 47.88 47.02 48.00 48.41 48.65 T i 0 2 1.76 2.09 1.87 1.84 1.81 2.25 1.89 2.83 3.04 A l 2 0 3 15.60 14.34 15.09 14.57 13.93 14.86 17.09 15.23 15.31 Fe:0 3* 12.10 11.98 12.68 12.72 12.81 13.08 12.03 13.44 13.52 MnO 0.18 0.17 0.18 0.18 0.19 0.20 0.17 0.19 .0.18 MgO 8.02 .9.77 8.77 9.74 10.93 8.61 7.10 4.78 4.69 CaO 10.48 9.90 9.41 9.47 9.08 10.45 9.82 10.13 10.21 NajO 2.15 2.37 2.22 2.09 2.02 2.55 2.39 3.11 2.74 K 2 0 0.56 1.05 0.88 0.72 0.82 0.57 0.77 1.19 0.77 P 2 O s 0.22 0.33 0.24 0.22 0.22 0.32 0.27 0.43 0.39 TOTAL 99.52 99.98 99.62 99.59 99.69 99.91 99.53 99.74 99.50 LOI 3.97 4.56 6.51 6.89 5.29 2.27 6.55 1.97 1.92 mg- number1 0.59 0.64 0.60 0.63 0.65 0.59 0.57 0.44 0.43 A.I. 2 -0.79 0.10 -0.33 -0.53 -0.45 0.15 -0.17 0.82 -0.06 Trace elements (ppm) R b X R F 13.0 19.8 17.1 7.8 15.3 6.9 12.7 16.5 7.3 S f X R F 308 547 293 229 297 394 464 409 403 Ba' C P 234 142 283 232 Sc 1 C P 23.6 22.5 31.4 31.3 yXRF 188 194 177 177 192 242 153 271 294 C o , C P 63.5 50.5 45.9 46.1 N i X R F 134 251 160 209 249 191 100 37 32 C u , C P 66.7 73.8 116 98.2 2 n X R F 95 88 101 106 105 100 96 114 125 Ga™ 21 18 24 18 18 20 21 22 23 yXRF. 20.7 23.0 19.8 19.0 19.3 25.4 19.3 27.4 29.3 2 r X R F 123 158 125 123 121 162 133 225 218 N b X R F 10.2 14.8 13.8 13.4 12.6 15.5 14.7 30.6 25.6 H f c p 2.82 3.87 4.76 4.76 T a I C P 0.828 0.993 1.76 1.45 T h l c p 1.59 0.769 3.10 2.58 U I C P 0.357 0.205 0.588 0.431 P B , C P 1.14 1.02 2.17 2.00 L a ' c p 13.6 16.7 28.4 25.1 C e l c p 28.7 38.6 57.6 52.4 3.7B 5.67 7.14 6.75 N d , c p 15.2 21.0 27.9 26.2 S m , c p 3.85 5.23 6.37 6.48 E u ' c p 1.31 1.78 2.08 2.13 G d l c p 4.36 5.58 6.54 6.93 T b ' c p 0.631 0.864 0.872 0.951 D y B P 3.93 4.96 5.34 5.79 H o , C P 0.740 1.03 0.977 1.06 E H C P 1.98 2.60 2.53 2.77 Y b l c p 1.67 2.37 2.12 2.30 L u ' c p 0.244 0.314 0.311 0.327 Liicp 4.24 3.26 7.32 4.46 M o , c p 1.14 1.19 1.60 1.46 C d l c p 0.072 0.044 0.143 0.161 S n l c p 1.03 1.12 1.63 1.59 S b l c p <lod <lod <lod 0.015 C s ' c p 0.263 <lod 0.449 0.183 W c p 0.843 0.396 0.838 0.970 B i l c p 0.207 <lod 0.212 r 0.156 'all Fe as F e ^ . mg -number= Mg/(Mg+Fe"). 'A. I. (Alkalinity Index) = (Na2O+K2O)-0.37* Si02+14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020°C. <lod = below limit of detection. 119 Table 3.2 (continued): Major (wt. % oxides) and trace element abundances (ppm) in Marion Dufresne lavas. Rock type: Sample: Height (m): Plagioclase-phyric basalt Plagiocla se-ultraphyric basalt Aphyric basalt B0B-93-558 330 BOB-93-563 BOB-93-571 290 165 B0B-93-572 150 B0B-93-556 B0B-93-564 360 240 B0B-93-565 215 BOB-93-552 400 BOB-93-553 390 Major elements (wt%) SiO z 49.70 48.83 48.37 48.48 48.37 47.78 48.62 49.36 48.95 TiOj 3.17 2.88 2.75 2.66 1.89 1.87 1.82 3.12 3.06 Al 2 0, 14.21 15.79 16.15 15.52 20.71 22.12 21.98 14.56 14.70 FejO,* 14.12 13.00 13.56 13.38 9.49 8.96 8.43 14.26 14.07 MnO 0.20 0.19 0.2 0.21 0.13 0.13 0.12 0.19 0.19 MgO 4.66 4.79 4.42 4.66 3.80 3.56 2.99 4.37 4.73 CaO 9.91 10.82 10.12 10.44 11.59 12.16 12.62 9.27 9.12 Na 20 2.67 2.54 2.96 3.04 2.97 2.27 2.36 3.08 2.89 K 2 0 0.54 0.43 1.1 1.12 0.51 0.69 0.70 1.41 1.41 P*Os 0.41 0.36 0.39 0.39 0.22 0.22 0.22 0.48 0.46 TOTAL 99.59 99.63 100.02 99.90 99.68 99.76 99.86 100.10 99.58 LOI 2.87 3.04 2.45 1.96 3.44 4.99 1.81 2.18 3.88 mg- number' 0.42 0.45 0.42 0.43 0.47 0.47 0.44 0.40 0.43 A. l . ! -0.75 -0.67 0.59 0.65 0.01 -0.29 -0.50 0.66 0.62 Trace elements (ppm) Rb"" 13.2 3.8 21.4 20.9 5.2 7.8 10.9 26.0 28.3 Sr™ 381 418 404 474 486 590 605 398 371 B a l c p 270 167 • 275 Sc I C P 34.4 20.3 27.0 yXRF 270 263 298 268 181 202 173 290 270 C o l c p 48.0 27.6 41.2 Ni X R F 32 35 19 30 32 25 16 23 26 C u l c p 55.9 56.6 67.0 Zn'" 126 112 107 ' 100 84 65 64 124 104 Ga* R F 23 22 23 22 22 21 21 23 24 yXRF 30.1 26.5 24.7 25.5 17.6 16.4 15.3 29.8 29.1 227 201 195 197 127 119 118 241 233 Ub™ 26.3 22.8 27.3 27.3 15.2 13.8 13.8 33.7 32.2 H f c p 4.38 2.50 4.87 T a ' c p 1.69 0.937 1.73 T h ' c p 2.94 1.41 3.29 ylCP 0.389 0.319 0.791 Pb l c p 1.91 1.15 2.45 La' c p 26.7 14.3 32.5 C e l c p 56.9 30.4 66.9 PH C P 7.57 3.88 8.15 N d l c p 28.2 16.0 32.1 S m , c p 6.17 3.77 7.25 E u i c p 2.07 1.47 2.37 G d , c p 6.01 3.94 7.60 Tb' c p 0.928 0.571 1.02 Dy i c p 5.13 3.35 6.18 Ho l c p 1.04 0.625 1.16 2.69 1.59 2.96 Y b , C P 2.28 1.34 2.50 L u l c p 0.310 0.193 0.361 L i i c p 4.56 3.19 4.57 Mo' c p 1.52 0.952 1.54 C d , c p 0.132 0.074 0.197 S n , c p 1.41 1.11 1.53 Sb , c p <lod <lod 0.023 C s i c p 0.235 0.265 0.216 W K P 0.829 1.11 1.67 B i l c p 0.209 0.222 0.218 *all Fe as Fe 20 3 , mg -number= Mg/(Mg+Fe2*). 2A.I. (Alkalinity Index) = (NajO+K2O)-0.37*SiO2+14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020°C. <lod = below limit of detection. 720 Table 3.2 (continued): Major (wt. % oxides) and trace element abundances (ppm) in Marion Dufresne lavas. Rock type: Aphyric basalt Sample: BOB-93-559 BOB-93-560 BOB-93-562 BOB-93-566 BOB-93-567 BOB-93-568 BOB-93-569 BOB-93-570 BOB-93-573 Height (m): 315 305 290 200 190 185 170 170 130 Major elements (wt%) S\02 50.17 46.92 49.75 48.60 48.51 48.97 48.27 48.87 47.84 T i 0 2 3.37 3.53 3.30 2.89 3.05 3.02 3.31 2.99 3.27 A I A 13.46 15.84 13.95 14.99 14.50 14.29 13.83 14.18 14.14 Fe,CV 14.92 15.6 14.46 13.96 14.46 14.19 15.25 14.66 15.68 MnO 0.21 0.26 0.21 0.21 0.22 0.20 0.22 0.21 0.23 MgO 4.28 4.99 4.38 5.22 5.19 4.63 4.33 4.48 4.52 CaO 8.78 9.26 9.30 10.15 9.69 10.14 9.23 9.47 9.24 Na 2 0 3.10 2.73 2.91 2.90 2.93 2.91 3.38 3.36 3.28 K 2 0 0.82 0.46 0.81 0.91 1.08 0.90 1.27 1.27 1.22 P 2 0 5 0.44 0.44 0.43 0.35 0.38 0.37 0.49 0.44 0.45 TOTAL 99.55 100.03 99.50 100.18 100.01 99.62 99.58 99.93 99.87 LOI 1.12 1.84 1.87 2.75 2.98 1.91 1.23 1.65 2.29 mg- number1 0.39 0.41 0.40 0.45 0.44 0.42 0.38 0.40 0.39 A.I.' -0.21 0.26 -0.26 0.26 0.49 0.12 1.22 0.98 1.23 Trace elements (ppm) R b x w 16.2 3.6 10.8 12.0 20.8 16.2 26.6 24.3 22.3 Sr"™1 378 557 392 382 355 407 382 395 363 B a l c p 250 232 28B 288 Sc I C P 32.2 34.2 30.6 34.1 y X R F 324 296 291 313 318 313 319 307 363 C o l c p 45.6 47.6 44.5 50.7 N i X R F 14 69 21 25 26 25 17 18 16 C u l c p 58.6 120 91.9 127 Z n X R F 148 147 132 111 121 117 131 122 134 G a X R P 23 24 23 24 28 22 23 23 25 yXRF 32.2 31.6 31.6 24.4 26.0 25.8 33.2 28.3 29.8 Z r * * 244 291 240 192 203 205 260 228 230 N b X R F 28.4 29.9 27.3 24.2 26.3 26.3 34.5 32.2 32.0 H f c p 5.36 4.31 5.60 4.96 T a l c p 1.70 1.55 2.12 1.90 T h l c p 3.00 2.62 3.28 3.36 ylCP 0.594 0.533 0.716 0.677 P b l c p 2.14 1.92 2.39 2.09 L a l c p 27.6 23.8 32.7 27.8 C e l c p 56.8 49.8 66.8 61.4 Pr 1 0" 7.52 6.86 8.36 7.63 N d , c p 29.6 25.5 33.1 30.0 S m , c p 7.06 5.66 7.49 6.53 E u l c p 2.35 1.99 2.40 2.26 G d , C P 7.43 5.63 7.98 6.78 T b , c p 1.06 0.855 1.07 0.953 Dy"* 6.16 4.74 6.53 5.70 H o , c p 1.23 0.938 1.24 1.04 E r , o 3.14 2.53 3.19 2.68 Y b ' c p 2.68 2.12 2.62 2.33 L u I C P 0.369 0.272 0.381 0.341 L J ,CP 4.69 4.71 6.25 4.78 M o l c p 1.73 1.24 2.09 1.80 C d , c p 0.197 0.177 0.234 0.156 S n I C P 1.78 1.46 1.80 1.68 S b , c p 0.020 0.005 0.016 <lod C s l c p 0.479 0.379 0.377 0.250 W I C P 1.60 0.926 1.22 1.02 B i , c p 0.212 0.214 0.226 0.212 *all Fe as Fe^Oa mg -number= Mg/fMg+Fe**). aA.I. (Alkalinity Index) = (Na2O+K2O)-0.37'SiO2+14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020°C. <lod = below limit of detection. 121 Table 3.2 (continued): Major (wt. % oxides) and trace element abundances (ppm) in Marion Dufresne lavas. Rock type: Aphyric basalt Quartz-bearing basaltic andesite Aphyric trachybasalt Sample: BOB-93-574 BOB-93-575 BOB-93-539 BOB-93-540 BOB-93-544 BOB-93-533 BOB-93-576 BOB-93-577 Height (m): 120 120 585 565 505 660 100 90 Major elements (wt%) Si0 2 48.06 48.42 54.37 55.08 55.86 48.25 48.93 48.64 T i 0 2 3.17 2.92 1.39 1.48 1.54 3.06 3.38 3.33 A l 2 0 3 13.94 14.74 15.02 15.12 15.26 17.59 14.17 14.15 Fe 2 CV 15.51 14.69 9.82 9.09 9.11 13.69 15.34 15.5 MnO 0.24 0.20 0.15 0.14 0.13 0.17 0.21 0.29 MgO 4.65 4.64 6.65 5.65 5.14 4.39 3.73 3.78 CaO 9.24 9.68 8.61 8.18 7.98 7.1 8.46 8.47 Na 2 0 3.22 3.06 2.52 2.71 2.92 3.58 3.65 3.83 K 2 0 1.24 1.06 1.2 1.77 1.67 1.81 1.46 1.49 P 2 0 s 0.43 0.38 0.19 0.22 0.29 0.63 0.55 0.55 TOTAL 99.70 99.79 99.92 99.44 99.9 100.27 99.88 100.03 LOI 2.09 2.34 2.46 2.10 1.47 2.49 1.79 1.26 mg- number1 0.40 0.41 0.60 0.58 0.55 0.41 0.35 0.35 A.I.' 1.11 0.63 -1.97 -1.47 -1.65 1.97 1.44 1.75 Trace elements (ppm) R b X B F 25.6 21.1 26.7 45.1 41.9 39.7 29.6 30.0 S r X R F 371 385 289 332 424 612 379 397 B a l c p 283 240 303 347 403 313 280 Sc ' c p 31.1 16.4 24.6 20.9 13.0 28.1 22.9 yXF,F 335 328 184 168 168 117 270 273 C o ' c p 46.2 40.4 34.0 33.4 38.2 44.2 38.5 N i X R f 18 23 154 98 74 16 5 5 C u ' c p 90.0 80.0 63.9 51.6 27.0 51.8 38.8 Z n X R F 120 119 83 75 81 102 145 144 G a " " 23 23 19 19 20 19 26 27 yXRF 28.5 25.5 23.7 25.0 22.3 25.0 33.6 33.4 Z f X R F 216 192 156 177 182 285 274 272 N b X R F 29.8 25.3 11.8 14.5 16.6 46.3 36.4 36.2 H f c p 5.63 3.88 4.38 4.38 6.49 5.97 5.20 T a l c p 1.67 1.03 1.24 1.31 2.31 2.17 1.91 T h ' c p 3.37 4.36 5.78 5.27 4.77 3.65 2.89 u l cp 0.587 0.854 0.987 0.962 0.885 0.727 0.682 P b 1 c p 1.80 2.72 3.11 3.14 2.83 2.45 2.50 L a I C P 29.9 19.5 22.6 26.2 37.2 35.2 31.6 C e ' c p 62.3 43.0 51.7 57.2 87.6 73.9 66.3 P r i c P 8.08 5.46 6.72 7.51 11.2 10.1 8.37 N d l c p 28.7 20.1 23.2 26.2 39.7 37.3 32.4 S m l c p 6.59 4.41 4.92 5.30 7.54 8.14 7.42 E u l c p 2.16 1.46 1.63 1.67 2.58 2.77 2.52 G d ' c p 6.77 4.55 4.73 5.01 6.56 8.20 7.80 T b l c p 0.996 0.718 0.810 0.784 1.00 1.19 1.05 Dy 1 0" 5.64 4.67 4.85 4.75 5.53 6.53 6.49 H o l c p 1.16 0.973 0.989 0.967 1.07 1.33 1.17 E r , c p 2.94 2.38 2.49 2.34 2.43 3.43 3.10 Y b l c p 2.48 2.26 2.29 2.19 2.12 3.05 2.56 L u l c p 0.353 0.301 0.311 0.300 0.279 0.395 0.364 L i ' c p 5.07 8.38 8.17 9.43 5.68 8.93 5.61 M o l c p 1.46 1.26 1.60 1.87 1.86 2.23 2.03 C d l c p 0.123 0.053 0.053 0.068 0.119 0.234 0.208 S n , c p 1.47 1.43 1.67 1.46 1.83 1.73 1.76 S b l c p <lod <lod <lod <lod <lod 0.072 <lod C s l c p <lod 0.249 0.346 0.346 0.232 0.300 0.343 W I C P 0.417 1.02 1.16 1.31 0.550 0.624 0.935 B i ' c p <lod <lod <lod <lod <lod 0.213 0.206 'all Fe as F^Oa. mg -number= Mg/fMg+Fe**). aA.I. (Alkalinity Index) = (Na2O+K2O)-0.37*SiO2+14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020"C. <lod = below limit of detection. 122 Table 3.2 (continued): Major (wt. % oxides) and trace element abundances (ppm) in Marion Dufresne lavas. Rock type: Mafic dikes and sills Sample; BOB-93-534 BOB-93-542 BOB-93-561 Height (m): 650 540 300 Major elements (wt%) SiO, 46.79 46.91 48.11 T i 0 2 2.60 2.08 2.43 Al,0, 15.05 14.85 14.90 Fe 20 3* 12.99 12.36 11.98 MnO 0.18 0.17 0.17 MgO 8.40 9.34 8.06 CaO 9.18 10.49 9.10 Na 20 2.80 2.51 2.85 K 2 0 1.03 0.96 1.29 P 2 O s 0.50 0.30 0.57 TOTAL 99.52 99.97 99.46 LOI 2.09 1.60 1.28 mg- number1 0.59 0.62 0.60 A.I.' 0.95 0.54 0.77 Trace elements (ppm) Rb X R F 13.6 16.3 16.3 Sr*R F 907 488 909 B a , c p 189 194 293 Sc 1 " 18.3 15.4 14.9 yXPF 203 228 199 C o , C P 56.1 53.2 47.3 Ni X R F 139 231 118 C u , c p 56.0 74.1 51.0 Z n X R F 98 91 99 G a X R F 20 20 21 yXRF 23.6 17.0 26.3 ZrXRF 291 149 295 Nb X R F 17.7 19.3 21.2 H f c p 5.52 3.69 6.10 T a i c p 1.11 1.30 1.37 T h l c p 1.22 1.70 1.76 ylCP 0.491 0.414 0.432 pbICP 2.37 1.45 2.19 L a i c p 24.6 19.9 27.2 C e l o p 57.2 44.7 68.0 pHcp 7.22 5.95 9.71 N d , c p 29.0 21.1 36.4 S m i c p 6.58 4.59 7.36 E u i c p 2.21 1.59 2.55 G d l c p 6.52 4.36 6.91 T b l c p 0.855 0.661 1.03 Dy' c p 4.88 3.57 5.73 Ho' c p 0.875 0.695 1.13 E r 1 " 2.18 1.76 2.59 Y b I C P 1.89 1.40 2.34 L u l c p 0.268 0.176 0.327 L | I C P 4.27 4.83 5.70 Mo I C F 1.39 1.39 1.28 C d i c p 0.227 0.050 0.158 S n i c p 1.68 1.04 1.68 S b i c p 0.039 <lod <lod C s l c p 0.287 0.106 0.164 W c p 0.779 0.685 0.673 B i l c p 0.186 <lod <lod "all Fe as Fe 3 0 3 . mg -number= Mg/(Mg+Fez*). 2A.I. (Alkalinity Index) = (NazO+K;O)-0.37*SiOz+14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020°C. <lod = below limit of detection. 123 plagioclase-ultraphyric basalts with 3.0-5.2 wt. % MgO and 46.9-50.2 wt. % Si0 2 . This group is overlain by a -400 m thick sequence of olivine-phyric high-MgO basalts (6.7-11.4 wt. %) with a relatively limited range of S i0 2 values (45.9-48.3 wt. %), hereafter referred to as the "high-MgO" group. Three distinct quartz-bearing basaltic andesites (5.1-6.7 wt. % MgO) with high S i0 2 contents (54.4-55.9 wt. %) lava flows are interlayered with the high-MgO group in the upper part of the section and reflect mixing of quartz-bearing high-MgO basaltic magma with more evolved, quartz-bearing magma (see Chapter 2), and therefore have distinct physical and geochemical characteristics. Alumina contents increase (from 13.5-17.6 wt. %) with increasing MgO in the aphyric low-MgO basalts, but systematically decrease (from 17.1-13.8 wt. %) with increasing MgO in lavas with >7 wt. % MgO (FIG. 3.7). The three plagioclase-ultraphyric samples (>50 vol. % plagioclase phenocrysts) in the Marion Dufresne section have significantly elevated A l 2 0 3 values (20.7-22.1 wt. %) consistent with plagioclase accumulation. These samples also have the highest CaO (>12 wt. %) observed in the section and low Fe 2 0 3 (8.4-9.5 wt. %; all Fe calculated as Fe 20 3). The quartz-bearing basaltic andesites have lower T i0 2 , CaO and Fe 2 0 3 contents than all basalts from the Marion Dufresne section (1.4-1.5, 8.0-8.6, and 9.1-9.8 wt. %, respectively). The mafic dikes and sills have 8.1-9.4 wt. % MgO and major element characteristics similar to the high-MgO basalts, but typically contain higher modal olivine contents than the lavas (-25 vol. %). The distinct compositions of the low-MgO and high-MgO lavas are emphasized by major element co-variation diagrams (FIG. 3.7). The low-MgO lavas, although characterized by a more limited range of MgO values, show higher variability with respect to other major elements (e.g. Na 2 0, T i 0 2 / Fe 2 0 3 , A l 2 0 3 and P 2 0 5 ) . All lavas have increasing Si0 2 , AI 2 0 3 /CaO, and Sr contents with decreasing MgO. Ni and Cr concentrations are strongly correlated with MgO content due to olivine fractionation. The extent of alteration of the samples from the Marion Dufresne section is reflected by LOI values (Table 3.2). Low-MgO lavas have low LOI values (1-3 wt. %), except for two plagioclase-ultraphyric basalts and one altered aphyric basalt (BOB93-560) that have <5 wt. % LOI. Higher, more variable LOI values (1 -7 wt. %) for the high-MgO basalts are consistent with petrographic observations 124 of more advanced alteration of groundmass ± phenocrysts and zeolite recrystallization in more vesicular samples. All basaltic andesites and mafic dikes and sills have relatively low LOI (1.5-2.5 and 1.3-2.1 wt. % respectively). 3.5.3 Trace element variations Trace element abundances by XRF (46 samples) and HR-ICP-MS (26 samples) in the Marion Dufresne section are shown in Table 3.2 and Appendices II and III, and selected trace element co-variations are plotted in FIG. 3.8. The high-MgO basalts have lower Nb concentrations,(8-16 ppm) and other incompatible trace element abundances compared with the low-MgO lavas (23-46 ppm Nb). Trace elements that can be mobile during post-magmatic alteration of basalt (i.e. Ba, Rb, Ce, U and Pb) correlate with Nb and form positive linear trends with nearly constant ratios of Ba/Nb~ 10.5, Rb/Nb~ 0.8, Ce/Nb~ 2.2, U/Nb~ 0.02, and Pb/Nb~ 0.07 (FIG. 3.8). This indicates that the trace element abundances were not significantly affected by secondary alteration. Ba/Rb ratios span a wide range from -7-60 in all Marion Dufresne lavas, but are <20 in the samples chosen for trace element analysis by HR-ICP-MS, values near the 'magmatic' ratio observed in fresh MORB and OIB (~11.3; Hofmann & White, 1983). The limited variance of K/Rb values observed in Marion Dufresne lavas (163-530) is also within the variance of this ratio in fresh natural basalts (Hofmann & White 1983), which indicates that K mobility is limited and the observed K/Rb values closely represent magmatic values. Samples BOB93-549 and 551 are outliers in many trace element diagrams and have high LOI values (5.3- 6.6 wt. %), but were included for trace element analysis by HR-ICP-MS due to their high MgO contents (10.9 and 7.1 wt. %, respectively) and stratigraphic position at the base of the high-MgO succession. Abundances of Sr are highly variable (200-900 ppm) throughout the Marion Dufresne section and are not correlated with Nb concentrations, although Sr concentrations are relatively constant (-400 ppm) in all low-MgO basalts (FIG. 3.8). Plagioclase-ultraphyric basalts have low trace element abundances and generally overlap with the high-MgO basalts in trace element co-variation diagrams, except with respect to elements compatible in plagioclase (e.g. Sr = 486-605 ppm). Th and Nb concentrations correlate strongly and give Th/Nb values of -0.12. Values of Zr/Nb are lower in low-125 FIG. 3.8. Nb variation diagrams for samples from the Marion Dufresne section. Elements that can potentially be mobile during alteration of basaltic rocks, such as Ba, Rb, Ce, U and Pb, define positive linear correlations with Nb, which indicates that trace element concentrations have not been significantly affected by secondary alteration (except for possible Rb-loss in several aphyric and plagioclase-phyric basalts). The unusually high Rb, Th, U, and Pb contents in the quartz-bearing basaltic andesites reflect mixing of highly fractionated trachytic magma with olivine-phyric basaltic magma. Based on the observed relations, Th and Rb appear to have been more incompatible than Nb in the Marion Dufresne magmas. Legend as in FIG. 3.5. 127 MgO basalts (6.2-8.8) than in high-MgO basalts (9.1-15.6); the highest Zr/Nb observed in any Kerguelen Archipelago lava examined to date (15.6) is from the Marion Dufresne section (sample BOB93-531). The quartz-bearing basaltic andesites have significantly higher incompatible trace element concentrations (notably U, Pb, Th and Rb) for a given Nb concentration. The mildly alkalic dikes and sills have intermediate Nb concentrations (17.7-21.7 ppm) and higher Sr, Pb, Zr, and Ce concentrations (and Zr/Y and Ce/Y) than most lavas in the Marion Dufresne section. In FICs. 3.9 and 3.10, trace element abundances in the Marion Dufresne lavas are normalized to primitive mantle and C1 chondrite values (McDonough & Sun, 1995). The high-MgO basalts form relatively smooth patterns with negative Pb, Th, and U, and positive Ba anomalies. One sample (BOB93-536) of high-MgO basalt has a high positive Sr peak, and although no plagioclase phenocrysts occur in this sample, it contains abundant zeolite recrystallization. Most of the low-MgO basalts (e.g. aphyric and plagioclase-phyric groups) are characterized by higher incompatible element abundances with strong negative U and Pb anomalies and higher REE abundances than low-MgO basalts. The sample of plagioclase-ultraphyric basalt (BOB93-565) has a prominent Sr peak and distinctly lower trace element abundances than the other low-MgO basalts, which is consistent with significant accumulation of plagioclase and relative dilution of all elements that are not compatible in plagioclase. The aphyric trachybasalts have the highest trace element abundances of all low-MgO lavas with patterns characterized by small positive Ti and Zr anomalies, negative Pb and U anomalies, and distinctly higher light REE abundances (FIG. 3.10). One sample of trachybasalt (BOB93-533) has a prominent positive Nb anomaly, higher Rb, Ba, and Th than other low-MgO lavas, and a distinctive positive Sr anomaly. The mafic dikes and sills have positive Ba peaks and negative Th, U and Pb anomalies, similar to the high-MgO basalts, but have higher Sr and Zr peaks. Primitive mantle-normalized trace element abundance diagrams for Marion Dufresne lavas do not show patterns characteristic of continental contamination, such as positive Pb or negative Nb and Ta anomalies. The quartz-bearing basaltic andesites from Marion Dufresne have high positive Rb and Th peaks, and negative Ba, Nb, and Pb anomalies, which reflects mixing of a basaltic magma with small amounts of 128 100 80 60 40 20 100 80 60 40 20 100 80 60 c to 40 V > 20 E *-0. 100 1 u 80 c (0 •o 60 E 3 < 40 20 100 80 60 40 20 100 80 60 40 20 ( 1 1 - - i 1 1 — i 1 1 i r -Olivine-phyric high-MgO basalts n=7 o 528 - 536 -X-531 - • 5 3 7 - o - 5 3 2 —1—549 - * - 550 1 i Aphyric basalts n=6 552 —O—569 559 570 568 —I—574 t—¥—t- t —i— —i—t—i—i—i—i—i—+—5—5-Plagioclase-phyric basalts n=4 Quartz-bearing basaltic andesites n=3 O 539 <• 540 A 544 A A O * f * ° ft -i 1 1 1 I 1 1-S « • * • * * * * * * * * H I I 1 1 I I 1 I 1 1—4-Aphyric trachybasalts n=3 533 576 577 « 9 -H I 1 I H Dikes and sills n=3 -534 •542 -561 Rb Ba Th U Nb Ta La Ce Pb Pr Nd Sr Sm Zr Hf Eu Tl Tb Dy Ho Y Er Yb Lu FIGURE 3.9 (29 FIG. 3.9. Primitive mantle-normalized incompatible element abundances in Marion Dufresne samples. The shaded field indicates the range observed between all samples in the section. The low-MgO lavas are characterized by higher incompatible element abundances. The positive Nb and Ta and negative Pb anomalies shown here are characteristic of enriched ocean island basalts and are inconsistent with the effects of continental contamination. Element incompatibility increases to the right and normalizing values are from McDonough & Sun (1995). The y-axis scale is linear (not logarithmic) to allow for a better comparison between samples and groups. Plagioclase fractionation in the Marion Dufresne section was limited, as indicated by the absence a Eu anomalies and small positive Sr anomalies for a small number of samples (BOB93-533, 536, 564, and the mafic dikes and sills). 130 131 FIGURE 3.10 FIG. 3.10. Chondrite-normalized rare earth element abundances in the Marion Dufresne samples. The shaded field indicates the range observed between all samples. The olivine-phyric, aphyric, and plagioclase-phyric basalts, which are the three most volumetrically significant groups in the section, show varying degrees of light REE enrichment with the flattest patterns in the high-MgO basalts and the most light REE-enriched patterns in the aphyric trachybasalts. Element incompatibility increases to the right and normalizing values are from McDonough & Sun (1995). The y-axis scale is linear (not logarithmic) to allow for a better comparison between samples and groups. 132 evolved magma (e.g. trachyte), therefore these anomalies do not imply a continental component in their source. 3.5.4 Radiogenic Isotopic Variations Radiogenic isotopic compositions of Hf-Sr-Nd-Pb for 22 samples from the Marion Dufresne section are presented in Tables 3.3 and 3.4 and FIGs. 3.11 and 3.12. All isotopic compositions are age-corrected to represent initial values based on the 4 0 Ar/ 3 9 Ar ages of -25 Ma from samples from the base and top of the section. The low-MgO lavas have an extremely limited range of "enriched" (i.e. high Sr, Pb and low Hf, Nd) isotopic compositions ( 1 7 6 Hf/ 1 7 7 Hf = 0.28282-0.28287; 8 7Sr/ 8 6Sr = 0.7048-0.7050; 1 4 3 N d / 1 4 4 N d = 0.5126-0.5127; 2 0 7 Pb/ 2 0 4 Pb = 15.54-15.56) compared with the high-MgO lavas, which extend to more "depleted" values (i.e. low Sr, Pb and high Hf, Nd) ranging from 1 7 6 H f / , 7 7 H f = 0.28292-0.28300, 8 7Sr/ 8 6Sr = 0.7040-0.7046, 1 4 3 N d / 1 4 4 N d = 0.5127-0.5128, and 2 0 7 Pb/ 2 0 4 Pb = 15.49-15.52. The quartz-bearing basaltic andesites have compositions intermediate between the high- and low-MgO groups ( 1 7 6 Hf/ , 7 7 Hf = 0.28291-0.28292; 8 7Sr/ 8 6Sr = 0.7045-0.7047; , 4 3 N d / 1 4 4 N d = 0.5127; 207p b / 204 p b = 1 5.53 .15.54) a n f j typically plot along trends defined by the high-MgO basalts in FIG. 3.12. This is evidence that these samples represent high-MgO basaltic magma diluted with minor volumes of evolved magma, as suggested by phenocryst compositions and the intermediate MgO content of the quartz-bearing basaltic andesites (5.1-6.7 wt. % MgO; see Chapter 2). The mafic dikes and sills in the Marion Dufresne section have the least radiogenic 2 0 8 Pb/ 2 0 4 Pb observed in this study as well as low 2 0 7 Pb/ 2 0 4 Pb and 2 0 6 Pb/ 2 0 4 Pb (FIG. 3.12c,d). One dike and one sill have low 8 7Sr/ 8 6Sr and high 1 4 3 N d / 1 4 4 N d and , 7 6 Hf/ 1 7 7 Hf, whereas one dike (BOB93-542) is more "enriched" and overlaps the low-MgO lavas in many isotope co-variation diagrams (e.g. FIG. 3.12a,b). The low-MgO and high-MgO lavas from the Marion Dufresne section typically define distinct trends in isotope co-variation diagrams. All samples from this study are negatively correlated in a 8 7Sr/ 8 6Sr vs. l 4 3 N d / 1 4 4 N d diagram, spanning a wide range of relatively "depleted" to "enriched" values (FIG. 3.12a) that nearly cover the entire range of variations observed on the Kerguelen Archipelago. The high-MgO basalts primarily define this trend (87Sr/86Sr = 0.70401-0.70458; 1 4 3 N d / 1 4 4 N d = 0.51273-133 Table 3.3: Hf, Sr, and Nd Isotopic ratios in Marion Dufresne samples Sample Elevation (m) Rock type MgO Al "6Hf'"'Hfm 2a "W 'H f "W'Hf, eHfi , 7 S r / e 6 S r m 2o " R b ^ S r "Sr '^Sn u 3 N d / " " N d m 2a ""Sm/^Nd " 5Nd/'"Ndj cNd BOB-93-528 755 1 11.44 0.5 0.282977 4 0.0140 0.282971 7.6 0.704156 8 0.0310 0.70414 0.512841 6 0.1614 0.51281 4.1 BOB-93-531 730 1 9.77 0.0 0.283002 4 0.0142 0.282996 8.5 0.704026 7 0.0409 0.70401 0.512859 6 0.1552 0.51283 4.4 BOB-93-532 715 1 9.34 0.0 0.282997 5 0.0149 0.282990 8.3 0.704042 7 0.0448 0.70403 0.512851 5 0.1440 0.51283 4.3 BOB-93-537 600 1 7.43 -0.5 0.282969 5 0.0120 0.282963 7.3 0.704325 7 0.0521 0.70431 0.512795 6 0.1507 0.51277 3.2 BOB-93-550 445 1 8.61 0.2 0.282926 4 0.0115 0.282920 5.8 0.704599 7 0.0507 0.70458 0.512755 9 0.1504 0.51273 2.4 BOB-93-554 2 390 2 4.78 0.8 0.282834 4 0.0093 0.282830 2.6 0.704987 7 0.1168 0.70495 0.512641 7 0.1378 0.51262 0.2 BOB-93-557 330 2 4.69 -0.1 0.282859 4 0.0098 0.282855 3.5 0.704950 7 0.0524 0.70493 0.512672 7 0.1496 0.51265 0.8 BOB-93-572 150 2 4.66 0.7 0.282831 5 0.0100 0.282827 2.5 0.704954 8 0.1276 0.70491 0.512635 5 0.1324 0.51261 0.1 BOB-93-565 215 3 2.99 -0.5 0.282841 4 0.0110 0.282836 2.8 0.704905 7 0.0521 0.70489 0.512679 9 0.1422 0.51266 1.0 BOB-93-552 400 4 4.37 0.7 0.282820 4 0.0105 0.282815 2.1 0.705089 7 0.1891 0.70502 0.512629 7 0.1365 0.51261 0.0 BOB-93-559 315 4 4.28 -0.2 0.282873 4 0.0098 0.282868 4.0 0.704929 6 0.1240 0.70488 0.512687 7 0.1440 0.51266 1.1 BOB-93-568' 185 4 4.63 0.1 0.282857 4 0.0090 0.282855 3.5 0.704905 7 0.1152 0.70486 0.512678 5 0.1343 0.51266 1.0 BOB-93-569 170 4 4.33 1.2 0.282848 4 0.0097 0.282843 3.1 0.704957 7 0.2015 0.70489 0.512659 5 0.1369 0.51264 0.6 BOB-93-570 2 170 4 4.48 1.0 0.282844 , .4 0.0098 0.282840 3.0 0.704969 7 0.1780 0.70491 0.512646 9 0.1315 0.51262 0.4 BOB-93-574 120 4 4.65 1.1 0.282833 4 0.0089 0.282828 2.6 0.704970 8 0.1997 0.70490 0.512645 5 0.1388 0.51262 0.3 BOB-93-576 100 5 3.73 1.4 0.282861 4 0.0094 0.282857 3.6 0.704918 7 0.2260 0.70484 0.512662 7 0.1320 0.51264 0.7 BOB-93-577 90 5 3.78 1.8 0.282854 4 0.0099 0.282849 3.3 0.704912 9 0.2187 0.70483 0.512667 6 0.1386 0.51264 0.8 BOB-93-540 565 6 5.65 -1.5 0.282917 4 0.0101 0.282913 5.5 0.704674 7 0.3931 0.70453 0.512734 6 0.1285 0.51271 2.1 BOB-93-544 505 6 5.14 -1.6 0.282927 5 0.0097 0.282922 5.9 0.704796 7 0.2860 0.70469 0.512718 5 0.1225 0.51270 1.8 BOB-93-534 u 650 7 8.40 0.9 0.283049 5 0.0069 0.283048 10.3 0.703765 7 0.0434 0.70375 0.512911 5 0.1372 0.51289 5.5 BOB-93-542 540 7 9.34 0.5 0.282843 6 0.0068 0.282839 2.9 0.704797 6 0.0967 0.70476 0.512676 6 0.1317 0.51265 0.9 BOB-93-561 300 7 8.06 0.8 0.283016 5 0.0076 0.283012 9.1 0.704246 7 0.0519 0.70423 0.512836 7 0.1221 0.51282 4.1 All Hf ratios determined by MC-ICP-MS and all Sr and Nd ratios determined by TIMS at PCIGR, UBC. Initial ratios were calculated at 25 Ma using parent-daughter concentrations given in Table 3.2. Rock types: 1. high-MgO basalt; 2. plagioclase-phyric basalt; 3. plagioclase-ultraphyric basalt; 4. aphyric basalt; 5. aphyric trachybasalt; 6. quartz-bearing basaltic andesite; 7. mafic dikes and sills. 'Hf values for 534 and 568 are the average of full procedural duplicate analyses that agree within " 6 Hf/" 7 Hf m = 0.000002. 2 Sr and Nd values for 554 and 570 are the average of full procedural duplicate analyses that agree within "Sr/^Sr,, = 0.000007 and ' 4 3 Nd/ '"Nd m = 0.000007. 3 Sr and Nd values for 534 are the average of two aliquots of one leached sample (chemically processed and analyzed separately) that agree within "Sr/^Sr^ = 0.000007 and '"Nd/'^Nd™ = 0.000007. 2o errors apply to last digit(s) of measured values. 134 Table 3.4: Pb isotopic ratios in Marion Dufresne samples Sample Elevation (m) Rock type MgO Al 2 0 6 p b / 2 w p b m 2o 2 0 7 p b / 2 « p b m 2o zoe p b / 204p b m 2o 23 f l U / 204 p b ™{j/™pb Z 3 ? f h / 2 M P b ""Pb^Pbj ^ P b / ^ P b , 2 0 e P b / O 4 P b , BOB-93-528 755 1 11.44 0.5 18.3020 25 15.5049 22 38.4767 53 13.0551 0.0947 23.2425 18.251 15.503 38.448 BOB-93-531 730 1 9.77 0.0 18.2353 13 15.4980 12 38.3870 33 13.9408 0.1011 36.5295 18.181 15.496 38.342 BOB-93-532 715 1 9.34 0.0 18.2398 11 15.4932 11 38.3866 29 14.2791 0.1036 40.5322 18.184 15.491 38.336 BOB-93-537 600 1 7.43 -0.5 18.3029 14 15.5169 13 38.6025 35 15.6355 0.1134 52.0133 18.242 15.514 38.538 BOB-93-550 445 1 8.61 0.2 18.3043 19 15.5184 16 38.6628 43 12.7617 0.0926 49.4784 18.255 15.516 38.602 BOB-93-554 390 2 4.78 0.8 18.3910 15 15.5462 18 38.9811 45 17.3444 0.1258 94.4315 18.324 15.543 38.864 BOB-93-557 330 2 4.69 -0.1 18.4488 20 15.5563 19 38.9685 49 13.7897 0.1000 85.1623 18.395 15.554 38.863 BOB-93-572 150 2 4.66 0.7 18.4088 25 15.5505 21 39.0062 55 13.0141 0.0944 101.8481 18.358 15.548 38.880 BOB-93-565 215 3 2.99 -0.5 18.5259 20 15.5604 19 39.0691 46 17.7376 0.1286 81.0968 18.457 15.557 38.969 BOB-93-552 400 4 4.37 0.7 18.3646 16 15.5445 15 38.9638 38 20.6017 0.1494 88.4630 18.285 15.541 38.854 BOB-93-559 315 4 4.28 -0.2 18.4927 10 15.5626 9 38.9996 23 17.7433 0.1287 92.6227 18.424 15.559 38.885 BOB-93-568 185 4 4.63 0.1 18.5566 19 15.5675 17 39.1268 46 17.8084 0.1292 90.5079 18.487 15.564 39.015 BOB-93-569 170 4 4.33 1.2 18.5023 19 15.5556 17 39.0629 42 19.2076 0.1393 90.9318 18.428 15.552 38.950 BOB-93-570 170 4 4.48 1.0 18.4808 17 15.5538 17 39.0660 46 20.7685 0.1506 106.3421 18.400 15.550 38.934 BOB-93-574 120 4 4.65 1.1 18.4478 12 15.5520 11 39.0321 30 20.8707 0.1514 123.7494 18.367 15.548 38.879 BOB-93-576' 100 5 3.73 1.4 18.3852 16 15.5440 14 38.9228 39 18.9409 0.1374 98.3421 18.312 15.541 38.801 BOB-93-577 90 5 3.78 1.8 18.3893 19 15.5387 15 38.9219 43 17.4071 0.1262 76.1265 18.322 15.536 38.828 BOB-93-540 565 6 5.65 -1.5 18.3481 13 15.5351 12 38.8143 31 20.2180 0.1466 122.3306 18.270 15.531 38.663 BOB-93-544 505 6 5.14 -1.6 18.3808 12 15.5405 10 38.8327 27 19.5367 0.1417 110.5767 18.305 15.537 38.696 BOB-93-534 650 7 8.40 0.9 18.1721 15 15.4839 13 38.3014 35 13.0777 0.0948 33.6545 18.121 15.482 38.260 BOB-93-542 540 7 9.34 0.5 18.1427 15 15.5118 12 38.5531 36 18.0995 0.1313 77.0163 18.072 15.508 38.458 BOB-93-561 300 7 8.06 0.8 18.1877 25 15.4962 21 38.3914 58 12.4451 0.0903 52.3470 18.139 15.494 38.327 All Pb ratios determined by MC-ICP-MS at PCIGR, UBC. Initial ratios were calculated at 25 Ma using parent-daughter concentrations given in Table 3.2. Rock types: 1. high-MgO basalt; 2. plagioclase-phyric basalt; 3. plagioclase-ultraphyric basalt; 4. aphyric basalt; 5. aphyric trachybasalt; 6. quartz-bearing basaltic andesite; 7. mafic dikes and sills. 'Pb values for 576 are the average of full procedural duplicate analyses that agrees within ^ P b / ^ P b ™ = 0.005; ""Pb/^Pb™ = 0.003; 2 0 8 P b / ! ° 4 P b n , = 0 . 0 0 9 . 2o errors apply to last digit(s) of measured values. 135 O) o Q. 2 (0 5 800 700 $ : • A A A 600 O A 500 : A A A • A A • • • 400 L £ O 300 A « i : % A : * A 200 ' $ 100 1 - ^ 0 87 S r / 86g r i i i i °?03s °-^s O-tfy - 2 0 2 4 6 8 3 6 9 12 15 g> o o OL 2 O) 4-1 (fl 800 700 m 2 c t : 600 500 A A • A 400 • • • a 300 • A % • A % : • A • 200 100 i F 0 206p b y 2 04p b 207 p b / 204p b 208 p by204p b 18.0 18.2 18.4 7S-46 /S-S0 'St* 3 8 2 3 8 6 3 9 0 800 700 600 500 400 300 200 100 0 800 700 600 500 400 300 200 100 0 This study: • Aphyric trachybasalts O Aphyric basalts • Plagioclase-phyrtc basalts •j Plagioclase-ultraphyric b a s a l ? <^> Olivine-phyilc basalts A Quartz-bearing basaltic andesites A Dikes and sills 736 FIGURE 3.11 FIG. 3.11. Variations in Hf-Sr-Nd-Pb initial isotopic compositions (i.e. at 25 Ma) with stratigraphic height in the Marion Dufresne section. All 2o errors are smaller than symbol widths. The low-MgO lavas have a limited range of Hf-Sr-Nd isotopic compositions compared with the high-MgO lavas, which have more depleted compositions. However, the low-MgO lavas have a wide range of Pb isotopic compositions that increase from 2 0 6 Pb/ 2 0 4 Pb = 18.32 to 18.49 in the lower 185 m of the section, then systematically decrease with increasing stratigraphic height (to 2 0 6 Pb/ 2 0 4 Pb = 18.29 at 400 m elevation). The eN d and eHf values are normalized to CHUR ( , 7 6 Hf/ 1 7 7 Hf m = 0.282772 and 1 4 3 N d / 1 4 4 N d m = 0.512638; Wasserburg etal., 1981). 137 0.5125 1 0.7032 0.7036 0.7040 0 7044 0.7048 0.7052 0.7056 "Sr/^Sr 15.60 15.55 15.50 h 15.45 18.0 18.1 18.2 18.3 18.4 18.5 18.6 2oepby204pb FIGURE 3.12 0.7056 0.7048 h 0.7040 0.7032 18.0 18.1 18.2 18.3 18.4 18.5 18. 2 0 8Pb/ 2 MPb <^> Olivine-phyric basalts p i Plagioclase-phyric ™ basalts mt Plagioclase-ultraphyric • basalts O Aphyric basalts • Aphyric trachybasalts A Quartz-bearing t - i basaltic andesites A Dikes and sills 0.2831 0.2827 18.0 18.1 18.2 18.3 18.4 18.5 18.6 aNph/a^Pb FIG. 3.12 (continued) FIG. 3.12. Isotopic co-variation diagrams for Marion Dufresne lavas. A. 8 7Sr/ 8 6Sr vs. 1 4 3 Nd/ 1 4 4 Nd. The high-MgO basalts and quartz-bearing basaltic andesites (and mafic dikes and sills) define a linear trend with a wide range of values that extend from depleted to enriched signatures. The aphyric and plagioclase-phyric basalts cluster at the high Sr, low Nd end of the array. B. eN d vs. eHf. The Marion Dufresne lavas are positively correlated, but do not extend into negative values for either E N d or e H f . Values are normalized to CHUR ( 1 7 6 Hf/' 7 7 Hf m = 0.282772 and 1 4 3 N d / 1 4 4 N d m = 0.512638; Wasserburg et al., 1981). C. 2 0 7 Pb/ 2 0 4Pb vs. 2 0 6 Pb/ 2 0 4 Pb. Two distinct slopes are evident; the high-MgO basalts have lower 2 0 7 Pb/ 2 0 4 Pb and 2 0 6 Pb/ 2 0 4 Pb whereas the low-MgO lavas are characterized by more radiogenic values and more variable 2 0 7 Pb/ 2 0 4 Pb, which increases with increasing 2 0 6 Pb/ 2 0 4 Pb. D. 2 0 8 Pb/ 2 0 4 Pb vs. 2 0 6 Pb/ 2 0 4 Pb. Two distinct trends are also observed. A greater range of 2 0 8 Pb/ 2 0 4 Pb isotopic compositions is observed in the high-MgO basalts. E. 8 7Sr/ 8 6Sr vs. 2 0 6 Pb/ 2 0 4 Pb. The high-MgO lavas have more variable 8 7Sr/ 8 6Sr isotopic compositions that increase systematically with increasing 2 0 6 Pb/ 2 0 4 Pb, whereas the low-MgO lavas have nearly constant 8 7Sr/ 8 6Sr compositions that decrease very slightly with increasing 2 0 6 Pb/ 2 0 4 Pb. F. , 7 6 H f / 1 7 7 H f vs. 2 0 6 Pb/ 2 0 4 Pb. The trends defined by the high-MgO and low-MgO lavas have opposing slopes. 1 7 6 Hf / 1 7 7 Hf decreases with increasing 2 0 6 Pb/ 2 0 4 Pb in the high-MgO lavas, but trends toward slightly higher 1 7 6 H f / , 7 7 H f with increasing 2 0 6 Pb/ 2 0 4 Pb. In all isotopic co-variation diagrams (A-F), the quartz-bearing basaltic andesites have intermediate isotopic compositions between the high-MgO and low-MgO groups. The mafic dikes and sills have low (unradiogenic) Pb isotopic compositions. One dike and one sill (BOB93-534 and 561) have 8 7Sr/ 8 6Sr, , 4 3 N d / , 4 4 N d , 1 7 6 Hf / ' 7 7 Hf isotopic compositions similar to the high-MgO lavas, whereas one dike (BOB93-542) resembles the low-MgO lavas. 140 0.51283). Compositions of the low-MgO lavas form a tight cluster in a Sr-Nd isotopic co-variation diagram with a restricted range of 8 7Sr/ 8 6Sr values (0.70483-0.70502) and a small variation i n U 3 N d / 1 4 4 N d (0.51261-0.51266). Values for eHf and £ N d express the difference between 1 4 3 N d / , 4 4 N d and , 7 6 H f / , 7 7 H f isotopic compositions and the composition of a chondritic uniform reservoir (CHUR; 1 7 6 Hf/ , 7 7 Hf m = 0.282772 and 1 4 3 N d / ' 4 4 N d m = 0.512638; Wasserburg et al., 1981) at the time of crystallization, and range from eHf = 2.06-8.46 and E N d = 0.02-5.36. These values are positively correlated across the section with no change in slope observed between the high- and low-MgO lavas (FIG. 3.12b). The Marion Dufresne lavas do not extend into negative E N d values. Variations in Pb isotopic compositions for samples from the Marion Dufresne section reveal, as expected, more complex relationships. The high-MgO and low-MgO lavas define different trends with distinct slopes in plots of 2 0 8 Pb/ 2 0 4 Pb and 2 0 7 Pb/ 2 0 4 Pb vs. 2 0 6 Pb/ 2 0 4 Pb (i.e. FIG. 3.12c). The low-MgO lavas span a larger range of Pb isotopic values than the high-MgO lavas. A clear stratigraphic variation is observed in the low-MgO basalts on plots of all Pb isotopic compositions vs. stratigraphic elevation (FIG. 3.11). Pb isotopic compositions systematically increase with increasing elevation in the lower 100 m of the section, e.g. from 2 0 6 Pb/ 2 0 4 Pb = 18.3216 at 90 m elevation (sample BOB93-577) to 18.4874 at 185 m elevation (sample BOB93-568), then decrease with increasing stratigraphic height to 206 p b / 204p b = 18.2845 at 400 m elevation (sample BOB93-552). Sample BOB93-568, an aphyric low-MgO basalt with A.I.=0.12, has the most radiogenic Pb isotopic composition observed in this study { 2 0 6 p b / 2 0 4 p b = 18.4874; 2 0 7 Pb/ 2 0 4 Pb = 15.5642; 2 0 8 Pb/ 2 0 4 Pb = 39.0148). In diagrams of 1 7 6 Hf/ 1 7 7 Hf, 8 7Sr/ 8 6Sr, and , 4 3 N d / 1 4 4 N d vs. 2 0 6 Pb/ 2 0 4 Pb, the high-MgO and low-MgO lavas define trends with opposing slopes, although the high-MgO lavas show much larger variations in Sr, Nd and Hf. FIG. 3.12e shows that the high-MgO lavas are characterized by increasing 8 7Sr/ 8 6Sr with increasing 2 0 6 Pb/ 2 0 4 Pb (from 18.15 to 18.25), whereas the low MgO lavas have very slightly decreasing 8 7Sr/ 8 6Sr with increasing 2 0 6 Pb/ 2 0 4 Pb (from 18.25 to 15.5). 2 0 6 Pb/ 2 0 4 Pb decreases systematically with increasing , 7 6 H f / 1 7 7 H f in the high-MgO basalts, whereas the low MgO basalts show relatively constant, or very slightly increasing1 7 6Hf/1 7 7Hf with increasing 2 0 6 Pb/ 2 0 4 Pb (FIG. 3.12f). 141 3.6 DISCUSSION 3.6.1 Implications for the volcanic stratigraphy of the Plateau Central The geochemical variations observed in the Marion Dufresne section have important implications for the evolution over time of volcanism in the Plateau Central region of the Kerguelen Archipelago. The vast Plateau Central region (>1100 km2) lies between well-studied older (29-28 Ma) tholeiitic-transitional basalts to the north and younger (25-24 Ma) mildly alkalic basalts to trachyandesites to the east and southeast (FIG. 3.2). Relatively low relief and numerous northwest-southeast trending lakes up to 12 km long characterize the Plateau Central region. The highest peaks (>800 m) occur near the Cook Glacier, which forms the western boundary of the region, and near Mt. Ross to the south (FIG. 3.13a). Importantly, flood basalts in this region dip 2-5° to the southeast, with the result that older flows are exposed to the northwest and younger flows to the southeast. The Tourmente and Capitole sections (16 and 14 km north of Marion Dufresne, respectively) are similar in age to the Marion Dufresne section (Frey et al., 2002 and Xu et al., in prep) (FIG. 3.14), which indicates that the Plateau Central was formed over a relatively short period of time (1 -2 Myr). Assuming a maximum regional dip of 5° SE, the base of the Marion Dufresne section is stratigraphically -500 m higher than the top of the Tourmente section (FIG. 3.13b), which is consistent with the slightly younger age determined for the Marion Dufresne section. Given the close proximity and similar ages of these sections, can the stratigraphic sequence and temporal evolution of magmatism on the Plateau Central region be determined? Basalts from the Tourmente and Capitole sections have remarkably similar major and trace element characteristics and isotopic compositions (Frey et al., 2002; Xu et al., in prep), whereas the Marion Dufresne samples are geochemically distinct. The Tourmente and Capitole sections consist of basalts with 4.1-6.4 and 3.3-8.0 wt. % MgO, respectively, and lack the sequence of olivine-phyric high-MgO basalts (7-12 wt. %) observed at Marion Dufresne (FIG. 3.14). The Marion Dufresne section is generally characterized by both higher and more variable Alkalinity Index (A.l.) values and a greater proportion of mildly alkalic flows (63%, compared with 33% and 22% for the Tourmente and Capitole sections, respectively). Of these three sections, the highest A l 2 0 3 values and plagioclase phenocryst 142 FIG. 3.13 FIG. 3.13. A. Simplified geologic map of the Plateau Central region of the Kerguelen Archipelago, after Nougier (1970). The Plateau Central (>1100 km2) is dominantly covered by flood basalts that dip 2-5° to the southeast. It is bounded by the Cook Glacier to the west and the Courbet Peninsula to the east, and does not include Mt. Ross, which is a younger (2-0.1 Ma) volcanic edifice. The Marion Dufresne section and previously basaltic studied sections from the Plateau Central (Mt. Tourmente and Mt. Capitole) are indicated by black boxes. B. Stratigraphic reconstructions indicate that the base of the Marion Dufresne section lies -500 m above the top of the Tourmente section, if a dip of 5° SE is assumed. 144 Alkalinity Index (A.I.) •2 -1 o 1 z 900 800 E 700 £ 600 m £ 500 aph 400 ratigi 300 53 200 100 A. Marion Dufresne A J A • - 2 - 1 0 1 A I 2 O s (wt. %) 12 16 20 24 M g O (wt. %) Zr /Nb 10 12 6 6 10 12 14 16 16 soo 800 a 800 800 700 700 600 if* ° 600 500 500 400 • *i 400 300 o 300 200 200 100 0 | B. Capitole I - 100 n • A A * • A A 20 24 2 900 - 2 - 1 0 1 2 • A * A 10 12 6 8 10 12 14 16 19 300 800 700 500 600 400 300 200 100 0 f o 0 ricp= o c 0 30C 900 900 • 800 800 SOO C. Tourmente 700 700 700 600 fib 800 -25.3 ±0.7 Ma 600 i 500 4 IT 500 500 400 400 400 % 300 300 300 • 200 200 200 > 100 0 100 0 ^26 0 ± 1.0 Ma 100 0 • - 2 - 1 0 1 2 Alkalinity Index (A.I.) 12 16 20 A l 2 0 3 (wt %) 1 1 1 900 800 700 600 500 400 c o 300 % 200 100 0 6 8 10 12 14 16 18 300 SOO 700 500 400 300 200 100 0 4 6 6 10 12 6 8 10 12 14 16 18 M g O (wt. %) Zr /Nb This study: • Aphyric trachybasalts O Aphyric basalts • Plagioclase-phyric basalts • Plagioclase-ultraphyric basalts <3> Olivine-phyric basalts A Quartz-bearing ™ basaltic andesites Dikes and sills FIGURE 3.14 145 FIG. 3.14. Stratigraphic variations of Alkalinity Index (A.I.), A l 2 0 3 (wt. %), MgO (wt. %), and Nb/Zr in the Marion Dufresne (this study), Capitole (Xu et al., in prep), and Tourmente (Frey et al., 2002) sections. The locations of these three sections are indicated in FIG. 3.13. The Tourmente and Capitole sections appear to be geochemically similar, whereas the Marion Dufresne section is characterized by more variable and slightly higher A. l . values, and higher A l 2 0 3 contents, particularly in the plagioclase-ultraphyric basalts (>20 wt%). The distinctly higher MgO and Zr/Nb contents observed in the upper 400 m of the Marion Dufresne section are not observed in the Tourmente or Capitole sections. 4 0 Ar/ 3 9 Ar ages of the dated samples from the Marion Dufresne and Tourmente sections are indicated. The vertical grey lines (at A. l . = 0, 15 wt. % A l 2 0 3 , 6 wt. % MgO, and Zr/Nb = 10) are for reference only. A. l . = (Na 20+K 20) - 0.37*SiO2+ 14.43 (Macdonald & Katsura, 1964). 746 abundances are observed in the Marion Dufresne section (FIG. 3.14). All samples from Tourmente are aphyric to nearly aphyric (<5 vol. % phenocrysts), with the exception of two samples with 8-10 vol. % plagioclase and 3-4 vol. % clinopyroxene. Only three clinopyroxene-phyric (2-15 vol. %) samples are observed in the Capitole section, which is composed mostly of aphyric basalts (<1 vol. % phenocrysts), except for several plagioclase-rich flows (up to 40 vol. %) in the upper 180 m of the section. Similar strongly plagioclase-phyric basalts are observed in the lower part of the Marion Dufresne section, but these lavas have higher alkalinity (A.l. = 0.1 to 1.8) than those observed at Capitole (A.l. = -0.8 to 0.1). The Tourmente and Capitole sections lack the plagioclase-ultraphyric (>50 vol. %) basalts, quartz-bearing basaltic andesites, and trachybasalts observed in the younger Marion Dufresne section. The thick sequences of compositionally homogeneous, fractionated (low-MgO) and largely aphyric character of basalts from the Tourmente and Capitole sections appears to reflect the effects of a "steady-state" magmatic system where magma flux rates allowed for extensive fractionation and effective separation of crystallized phases prior to eruption relative to the Marion Dufresne section. Albarede et al. (1997) proposed that basalts from the Piton de la Fournaise volcano on Reunion (Indian Ocean) that show limited compositional variations are "steady state" basalts. These "steady state" basalts have compositions that are buffered in a solid-dominant environment, which is interpreted to be an ascending slurry zone of crystals (mainly olivine and clinopyroxene at deeper levels, with plagioclase joining the assemblage at shallower levels). A similar situation may have occurred during construction of the lower stratigraphic levels of the Plateau Central. With the emergence of plagioclase-phyric, the plagioclase-ultraphyric, and finally olivine-phyric basalts in the middle and upper parts of the Marion Dufresne section, all of which show significant geochemical variations, it is likely that a more liquid-dominant environment of crystallization was established. In this context, the presence of the plagioclase-ultraphyric basalts in the Marion Dufresne section is particularly important and appears to signal the breaking-down of the steady-state crystallization regime. The large (1-9 mm) and relatively calcic (An 7 5. 6 0) plagioclase phenocrysts grew from water-rich, fractionated magmas at a maximum depth of 5-6 km (see Chapter 2) and were effectively segregated from the co-crystallizing olivine and clinopyroxene (the magmas were multi-saturated). Could this be material from the top of 747 the slurry zone? The subsequent emplacement of 400 m of olivine-phyric basalts suggests a significant increase in eruptive rate, such that the relatively dense olivine was entrained from depth. A change in the processes of magma formation and evolution between the eruption of the Tourmente/Capitole and Marion Dufresne lavas is also indicated by isotopic differences between these three spatially and temporally close sections. The Tourmente and Capitole lavas are characterized by limited isotopic variability (FIG. 3.15), but do not have isotopic compositions typical of the enriched component of the Kerguelen mantle plume source (e.g. Weis et al., 1998; Mattielli et al., 2002; Doucet et al., 2005). They also show no direct evidence of the distinct depleted mantle component observed in the Marion Dufresne high-MgO basalts (i.e. nonradiogenic Sr and Pb, and radiogenic Hf and Nd isotopic compositions). The low-MgO basalts from Marion Dufresne are generally more enriched than lavas from Tourmente and Capitole and have 8 7Sr/ 8 6Sr and , 4 3 N d / ' 4 4 N d compositions similar to some enriched lavas from Capitole, but extend to higher 8 7Sr/ 8 6Sr and lower , 4 3 N d / , 4 4 N d (FIG. 3.15a). The Marion Dufresne low-MgO lavas also have higher 2 0 6 Pb/ 2 0 6 Pb and 2 0 8 Pb/ 2 0 6 Pb values compared with basalts from Capitole (FIG. 3.15c). The Tourmente and Capitole lavas may represent a compositionally distinct source, or more likely, more efficient mixing of the same source components present in Marion Dufresne lavas (i.e. both "enriched" and "depleted" components). Thus, the apparent stratigraphic, isotopic, and geochemical compositional relations for basaltic lavas forming this region of the Plateau Central suggest that the Marion Dufresne section represents a period of significant change in both source components and differentiation processes of Kerguelen Archipelago basalts. 3.6.2 The transition from tholeiitic to mildly alkalic volcanism on the Kerguelen Archipelago Although both mildly alkalic and tholeiitic-transitional basalts are observed in all Plateau Central sections, the Marion Dufresne section contains the greatest proportion of mildly alkalic flows and highest variability in A. l . values observed in a single section from the Kerguelen Archipelago to date. The results of this study allow for a more refined characterization of the transition from tholeiitic to mildly alkalic volcanism on the Kerguelen Archipelago. As shown in FIG. 3.14, most transitional-tholeiitic basalts at Tourmente and Capitole have A. l . values between 0 and -0.5 with a volumetrically 148 0.5129 0.5128 0.5127 0.5126 0.5125 T 1 1 1 r SEIR MORB Bureau Marion Dufresne: ^ Lava flows <> Dikes and sills I I [ Marion Dufresne Site 1140 34 M a Site 1140 submarine basalts 28-29 M a tholeiitic-transltional basalts 25-26 M a Plateau Central basalts 24-25 M a mildly alkalic basalts <10 Ma alkalic tlows and intrusions Capitole Baie Charrier SE/Charbon Crozier 0.7036 0.7040 0.7044 0.7048 0.7052 8 7 S R / 6 6 S R 0.7056 149 Q. 39.4 39.2 39.0 38.8 £• 38.6 38.4 38.2 38.0 C. UMS Ross / SEIR . MORB -Crozier Baie CharrieH LMS Marion Dufresne 17.8 18.0 18.2 18.4 ^ P b / ^ P b 18.6 18.8 0.708 0.707 0.706 </> i 0.705 CO oo 0.704 0.703 0.702 Ruches charbon Crozier LMS Capitole Tourmente Fontaine 1140 Bureau SEIR MORB 17.9 18.1 18.3 2 0 6 p b / 2 0 4 p b 18.5 18.7 FIGURE 3.15 (continued) 150 FIG. 3.15. Isotopic variations in basaltic sections from the Kerguelen Archipelago. All plotted compositions are measured values. Fields are coloured according to age and geographic location (see complete legend in A). Data sources are as follows: Weis et al., 1993 (Southeast Upper and Lower Miocene Series, or UMS and LMS); Mattielli etal., 2002 and D. Weis, unpublished data (Crozier); Yang etal., 1998 (Bureau and Rabouillere); Frey etal., 2000 (Southeast Ravin du Charbon & Ravin Jaune); Chauvel & Blichert-Toft, 2001 (Southeast Indian Ridge); Doucet etal., 2002 (Ruches and Fontaine); Frey etal., 2002 (Tourmente); Xu etal., in prep (Capitole); and Hanano, 2005 (Baie Charrier). A. 8 7Sr/ 8 6Sr vs. 1 4 3 N d / U 4 N d . The wide range of Sr and Nd isotopic compositions observed in the Marion Dufresne section overlaps ranges defined by the 29-28 Ma tholeiitic-transitional basalts and the 25-24 Ma mildly alkalic basalts. The Marion Dufresne samples show significantly larger Sr-Nd isotopic variations than previously studied Plateau Central sections. B. eHf vs. eNd. The Marion Dufresne samples form a trend parallel to fields for the 29-28 Ma tholeiitic-transitional basalts that extends through the Tourmente field toward SEIR MORB, but do not have the negative eN d values that are observed in most of the 25-24 Ma mildly alkalic basalts. C. 2 0 8 Pb/ 2 0 4 Pb vs. 2 0 6 Pb/ 2 0 4 Pb. The low-MgO lavas from the Marion Dufresne section overlap many of the younger mildly alkalic sections, whereas the high-MgO lavas are characterized by less radiogenic compositions, overlap the Bureau field, and form a linear trend toward lower 2oap b / 204p b a n d 206 p b / 204p^ w h j c h r e p r e s e n t s m j x i n g w j t h a SEIR-like component. D. 8 7Sr/ 8 6Sr vs. 2 0 6 Pb/ 2 0 4 Pb. The low-MgO lavas from Marion Dufresne have variable 2 0 6 Pb/ 2 0 4 Pb at a given 8 7Sr/ 8 6Sr and define a trend similar to those observed in younger mildly alkalic basalts. In contrast, the high-MgO basalts show greater Sr isotopic variability, similar to that observed in the older 29-28 Ma sections, and form a clear trend toward the SEIR MORB field. 75/ minor role for mildly alkalic volcanism, whereas 63% of all Marion Dufresne lavas have A. l . >0 and the basal 200 m of the section consists entirely of mildly alkalic lavas. The younger (25-24 Ma) Crozier section, in comparison, consists predominantly of mildly alkalic basalts to trachyandesites with A. l . values from 0 to 3.2, with the exception of only two samples of transitional-tholeiitic basalt (A.l.= -0.5 to -0.4) (Damasceno, 1996). The evolved, more mildly alkalic lava compositions (trachybasalts to trachyandesites) that are volumetrically significant in the Crozier section occur locally in the Marion Dufresne section (e.g. trachybasalts BOB93-533, 576 and 577), but are absent in the Tourmente and Capitole sections. The Plateau Central sections thus appear to represent the last predominantly transitional sequences of basalts before the definitive change to mildly alkalic basaltic volcanism as documented in the Crozier section on the Courbet Peninsula (Damasceno, 1996; Damasceno eta/., 2002) and in the Southeast Province (Weis et al., 1993; Frey et al., 2000). However, based on the results of this study, this change is not abrupt, but occurs over a relatively extended period of time (over one million years), and encompasses the eruptions of basalts from the Tourmente, Capitole, and Marion Dufresne sections, and possibly most other Plateau Central basalts. In particular, the progression toward more negative A. l . observed at the base of the Marion Dufresne section, where A.l . decreases sharply from 1.8 to 0.1 with increasing stratigraphic height over the lower 200 m of the section (FIG. 3.6), indicates that the abrupt shift to higher alkalinity observed by Frey et al. (2002) in the Tourmente section was not a terminal shift toward mildly alkalic volcanism. This is supported by trace element chemistry, particularly with respect to Zr/Nb concentrations, which change systematically over time in all Kerguelen Archipelago lavas studied to date (FIG. 3.16). The ratio of Zr/Nb is generally considered to be an index of source depletion in the melting regime for ocean island basalts (e.g. Iceland: Fitton etal., 1997; 2003) because Nb is more incompatible than Zr during partial melting of spinel and garnet peridotite. However, the range of Zr/Nb produced during extents of melting up to 20% is relatively small compared to natural variations in OIB (Fitton et al., 2003) , which suggests that source composition (incompatible-element depleted [higher Zr/Nb] vs. enriched [lower Zr/Nb]) plays an important role. On the Kerguelen Archipelago, older (29-28 Ma) tholeiitic-transitional basalts have more variable and generally higher Zr/Nb than the 26-25 Ma Plateau 152 700 600 500 E 400 a a N N 300 h 200 h 100 h 18 16 14 12 10 8 6 2 0 A . T Marion Dufresne: + Lava flows O Dikes and sills 34 Ma Submarine basalts 28-29 Ma Tholeiitic-transitional basalts 25-26 Ma Plateau Central basalts 24-25 Ma Mildly alkalic basalts Marion Dufresne low-MgO O O pVlarion Dufresne high-MgO 0 10 20 30 40 Nb (ppm) I I 20 40 60 Nb (ppm) 80 100 120 B. Fonta ine Rabou i l le re • Marion Dufresne high-MgO R u c h e s more depleted i6 source component 1 4 Marion Dufresne low-MgO • 1 0 1 2 Alkalinity Index C r o z i e r S o u t h e a s t L M S Increasing stratigraphic height In the Marion Dufresne section Tht. study (inset diagrams): P i a g l o c , a s e . o Aphyric basalts ^aphyric basalts D Aphyric trachybasalts , Plagioclase-phyric basalts •0 Olivine-phyric basalts A Quartz-bearing ' basaltic andesites Dikes and sills J— -2 1 2 Alkalinity Index 4 5 FIGURE 3.16 153 FIG. 3.16. Zr/Nb variations in Marion Dufresne samples compared with other Kerguelen Archipelago lavas. The inset diagrams show the range of compositions observed in the Marion Dufresne samples. A. Zr vs. Nb. The 29-28 Ma tholeiitic-transitional and 26-25 Ma Plateau Central basalts have relatively high Zr/Nb values compared to the 25-24 Ma mildly alkalic basalts. The relatively high Zr/Nb values observed in the Marion Dufresne high-MgO lavas are similar to those observed in the 29-28 Ma tholeiitic-transitional basalts. Some of the Marion'Dufresne low-MgO lavas have distinctly higher Zr/Nb values and overlap the Plateau Central and transitional-tholeiitic fields, whereas others plot along the trend of low Zr/Nb defined by the younger 25-24 Ma mildly alkalic lavas. B. Zr/Nb vs. Alkalinity Index. The 29-28 Ma tholeiitic-transitional basalts and Marion Dufresne high-MgO basalts show a wide range of Zr/Nb values over a relatively limited range of A.I. values. In contrast, the 25-24 Ma mildly alkalic basalts are characterized by Zr/Nb < 8 and show decreasing Zr/Nb with increasing alkalinity. The Marion Dufresne low-MgO lavas with higher Zr/Nb values are generally those with lower A.I. (<0) whereas those with lower Zr/Nb have A.I > 0 and overlap trends defined by the 25-24 Ma mildly alkalic lavas. 754 Central basalts and the 25-24 Ma mildly alkalic lavas (FIG. 3.16a). The Marion Dufresne high-MgO lavas largely overlap the field for the 29-28 Ma basalts. The low-MgO lavas fall into two groups: those with A.l.<0 have higher Zr/Nb (8.3-8.8) and overlap the fields defined for the 26-25 Ma Plateau Central and 29-28 Ma transitional-tholeiitic lavas, whereas the stratigraphically lower low-MgO lavas with higher alkalinity (A.l.>0) have lower Zr/Nb (6.2-7.9) and overlap the field for younger mildly alkalic (25-24 Ma) lavas (FIG. 3.16a). Importantly, the low-MgO mildly alkalic lavas from Marion Dufresne show progressively increasing Zr/Nb (from 6.2-8.6) with decreasing A. l . upsection, which may be related to the combined effects of changing proportions of source components (higher Zr/Nb = more depleted component) and increasing extents of melting (lower A.I.), which will be discussed in more detail below and in the following section. The temporal shift toward decreasing alkalinity with increasing stratigraphic height in the basal 200 m of the Marion Dufresne section contrasts with regional trends observed on the archipelago. This shift to negative A. l . values occurs within a 200 m thick sequence of intercalated aphyric and plagioclase-phyric to -ultraphyric lavas, which are then overlain by the olivine-phyric high-MgO basalts. Decreasing Ti/AI contents in clinopyroxene phenocrysts from the Marion Dufresne section indicate an increase in the depth of fractionation over time (see Chapter 2). Variations in alkalinity in ocean island basalts have been attributed to changes in the degree and pressure (depth) of melting (e.g. Hawaii: Chen etal., 1991; Kerguelen: Weis etal., 1998; Frey etal., 2000; 2002). In some ocean island settings, it appears that there may be a greater role for the generation of mildly alkalic basalt by differentiation of a tholeiitic parental magma through extensive fractionation of clinopyroxene at high pressures (depths) (e.g. Galapagos: Naumann & Geist, 1999; Reunion: Albarede etal., 1997; Canary Islands: Nikogosian etal., 2002). Changes in alkalinity have also been linked to mantle source heterogeneities (e.g. Chen & Frey, 1983). Results from an experimental study on the role of clinopyroxene fractionation on alkalinity variations in tholeiitic to mildly alkalic basalts from the Kerguelen Archipelago do not support the derivation of mildly alkalic basalts from tholeiitic parental magmas (Scoates et al., 2005b). Instead, these results indicate that the significant alkalinity variations in the archipelago flood basalts are related to melting conditions. The distance between the Kerguelen 155 hotspot and Southeast Indian Ridge (SEIR) increased over time, which resulted in an increase in the thickness of the crust beneath the archipelago and of the lithosphere beneath the Northern Kerguelen Plateau. As a result, the Kerguelen hotspot produced higher degree partial melts and tholeiitic basalts dominated by a depleted component while situated under relatively thin, hotter crust more proximal to the SEIR (e.g. ODP Leg 183, Site 1140 on the NKP; Weis & Frey, 2002), but over time produced lower degrees of melt and more alkalic basalts as it stalled at increasingly higher pressures and depths with increasing distance from the SEIR (e.g. Frey et al., 2000; Doucet et al., 2002; Scoates et al., 2005b). This is also supported by variations in the slope of chondrite-normalized REE plots, as indicated by (La/Yb)N values for basalts from across the archipelago. Values of (La/Yb)N show progressively increasing LREE-enrichment from the tholeiitic-transitional to the mildly alkalic basalts (consistent with lower extents of partial melting of upper mantle peridotite), and correlate with observed Zr/Nb variations (FIG. 3.17). The reverse trend towards more tholeiitic compositions observed in the lower 200 m of the Marion Dufresne section, combined with increasing Zr/Nb and decreasing (La/Yb)N, suggests the occurrence of a temporal and local change within the melting regime that resulted in enhanced degrees of melting of a multi-component source. This trend culminated in the eruption of a thick sequence (>400 m) of mainly transitional high-MgO basalts with progressively increasing Zr/Nb, indicating an increasing role for the depleted component in the source of Kerguelen Archipelago basalts at this time (see next section). 3.6.3 Significance of the depleted mantle component It has long been known that Kerguelen Archipelago basalts have isotopically variable compositions (e.g. Gautier et al., 1990). The significance of the current study is that the isotopic compositions of the Marion Dufresne lavas cover much of the compositional range determined for all archipelago flood basalts studied to date (FIG. 3.15). The Sr-Nd isotopic compositions define a nearly continuous linear trend, from depleted (low 8 7Sr/8 6Sr, high U 3 Nd/ 1 4 4 Nd) to enriched (high 8 7Sr/ 8 6Sr, low l 4 3 Nd/ 1 4 4 Nd) signatures, that has not previously been documented in a single section from the Kerguelen Archipelago. The occurrence of distinct depleted and enriched isotopic signatures was previously 156 FIG. 3.17. (La/Yb)N vs Zr/Nb. Kerguelen Archipelago lavas show progressive LREE-enrichment from the tholeiitic-transitional to the mildly alkalic basalts with decreasing Zr/Nb. The Marion Dufresne low-MgO and 25-24 Ma mildly alkalic lavas are characterized by higher (La/Yb)N concentrations, which is consistent with lower degrees of partial melting of a mantle peridotite source. Older lavas, including the Plateau Central basalts, are characterized by more shallowly dipping trends toward lower (La/Yb)N with increasing Zr/Nb, likely due to higher extents of partial melting. The inset diagram shows the range of compositions observed in the Marion Dufresne samples. Kerguelen Archipelago: ( ] 2 8 - 2 9 M a Tholei i t ic- transi t ional basa l ts Um 25 -26 M a P la teau Cen t ra l basal ts ( ) 2 4 - 2 5 M a Mi ld ly a lka l ic basa l ts Mar ion D u f r e s n e : * L a v a f lows o D i k e s a n d si l ls This study (inset): • Aphyric trachybasalts O Aphyric basalts • Plagioclase-phyric basalts m Plagioclase-ultraphyric basalts <Q> Olh/ine-phyric basalts A Quartz-bearing basaltic andesites 1. Dikes and sills 157 observed in studies of older 29-28 Ma transitional-tholeiitic lavas on the Kerguelen Archipelago (Yang et al., 1998; Doucet et al., 2002). In the Marion Dufresne section, the depleted isotopic compositions that occur only in the stratigraphically higher olivine-phyric high-MgO basalts (7.4-11.4 wt. % MgO) resemble the compositional characteristics of the 29-28 Ma depleted tholeiitic-transitional lavas, but mark the first occurrence of this depleted signature in basalts of this age and alkalinity. All of the high-MgO depleted samples have A. l . between -0.8 and 0.5, which demonstrates that the depleted mantle signature is not limited to tholeiitic basalts as previously documented, but is found in mildly alkalic basalts as well. Whereas the isotopically depleted basalts in the 29-28 Ma Bureau, Rabouillere, Ruches, and Fontaine sections typically occur at the base of a sequence of lavas that trend toward enriched isotopic compositions with increasing stratigraphic height (Yang et al., 1998; Doucet et al., 2002), the reverse occurs at Marion Dufresne. The low-MgO lavas from the Marion Dufresne section and isotopically enriched lavas from the 29-28 Ma tholeiitic-transitional sections are characterized by a limited range of 1 7 6 Hf/ 1 7 7 Hf, 8 7Sr/8 6Sr, and , 4 3 N d / , 4 4 N d isotopic compositions, coupled with more radiogenic, but also more variable, Pb isotopic compositions. In FIG. 3.15, it is evident that these values approach and partly overlap the compositional ranges of the younger (25-24 Ma) mildly alkalic lavas from the Crozier section in the eastern Kerguelen Archipelago, which are interpreted to represent the composition of the enriched component of the Kerguelen mantle plume source (Weis etal., 1998; Mattielli et al., 2002; Doucet et al., 2005). The Pb isotopic ratios of the low-MgO lavas from Marion Dufresne are slightly less radiogenic than the Crozier lavas (FIG. 3.15c). The Marion Dufresne low-MgO lavas form a trend of increasing 2 0 8 Pb/ 2 0 4 Pb with increasing 2 0 6 Pb/ 2 0 4 Pb that is parallel to trends defined by the 25-24 Ma mildly alkalic lavas, in which a depleted component is interpreted to be either absent or volumetrically minor (Mattielli et al., 2002; Doucet et al., 2005). Although the most isotopically enriched samples from Marion Dufresne have 2 0 6 Pb/ 2 0 4 Pb comparable to the Kerguelen plume average (~18.533; Weis & Frey, 2002), they have lower 8 7Sr/8 6Sr, 2 0 7 Pb/ 2 0 4 Pb, and 2 0 8 Pb/ 2 0 4 Pb and higher 1 4 3 N d / 1 4 4 N d and , 7 6 H f / 1 7 7 H f isotopic compositions. Thus, the low-MgO basalts from the Marion Dufresne section do not represent a pure expression of the enriched component of the Kerguelen mantle plume source. 158 The near continuum observed between relatively depleted and enriched isotopic compositions in Kerguelen Archipelago basalts was originally interpreted to represent variable mixing between an enriched, plume-like and a more depleted, MORB-like endmember (Storey et al., 1988; Gautier et al., 1990). This depleted component has been attributed to the contribution of melts from a Southeast Indian Ridge (SEIR) MORB source for sections erupted when the Kerguelen hotspot was situated -50 km from the ridge (Weis & Frey, 2002), and to sublithospheric mixing of SEIR asthenospheric melts with the plume when this distance had increased to -300 km (Doucet et al., 2002; 2005). However, the existence of similarly depleted isotopic compositions in 25 Ma basalts that were likely erupted about 400 km from the SEIR (assuming spreading rates of 35 mm/yr after Royer & Sandwell, 1989) following the eruption of relatively thick older (26 Ma) sequences in which this depleted component is relatively minor (i.e. Tourmente, Capitole) is problematic. At 25 Ma, could the ridge and plume sources still communicate through sublithospheric channels as proposed by Doucet et al. (2002)? Importantly, the Pb isotopic compositions of the Marion Dufresne high-MgO lavas overlap the Bureau trend toward less radiogenic Pb isotopic compositions, approaching those of the SEIR ( 2 0 6Pb/ 2 0 4Pb= 16.9-18.4, 207 p b / 204 p b = 1 5 4.15 6 a n c j 208 p b / 204 p b = 37 3,33 j . c h a u v e | & Blichert-Toft, 2001) (FIG. 3.15c), indicating that the depleted components observed in both the 25 Ma Marion Dufresne and the 29-28 Ma tholeiitic-transitional basalts have similar compositions, suggesting a similar origin. Another important line of evidence is the strong correlation between basalts with isotopic evidence for a depleted component and both source composition (Zr/Nb, Nb/Y, and Zr/Y) and relative melting indices, such as (La/Yb)N, A.I., and Nb-Zr-Y systematics. The older (29-28 Ma) tholeiitic-transitional basalts and Marion Dufresne high-MgO basalts are characterized by higher and considerably more variable Zr/Nb values (up to 15.6), which are coupled with lower (La/Yb)N, lower alkalinity, and isotopically depleted compositions (FIG. 3.18). These characteristics can be further illustrated on a logarithmic plot of Nb/Y vs. Zr/Y (FIG. 3.19), such as those used to discriminate between plume and N-MORB (normal, depleted MORB) sources in Icelandic basalts (e.g. Fitton etal., 1997; 2003). This diagram shows the effects of source depletion (low Nb/Y and Zr/Y in the source) as well as relative melting, since higher degrees of melting result in lower Nb/Y and Zr/Y. An important 159 E 0.5132 0.5130 r ^ 0.5128 z 3 0.5126 b 0.5124 F o n t a i n e L M S C h a r b o n S E I R M O R B T o u r m e n t e C a p i t o l e R u c h e s B u r e a u R a b o u i l l e r e S i t e 1 1 4 0 0.5129 •o 0.5128 ^ 0.5127 05126 O OS 4 6 8 10 12 14 16 18 Zr/Nb 10 20 30 Z r / N b 40 50 FIG. 3.18. 1 4 3 N d / 1 4 4 N d vs. Zr/Nb. The high Zr/Nb values observed in the Marion Dufresne section and all >26 Ma sections are correlated with increasingly depleted isotopic compositions (e.g. higher 1 4 3 N d / 1 4 4 N d ) . Al l Kerguelen Archipelago lavas define a curvilinear trend of increasing Zr/Nb with decreasing 1 4 3 N d / 1 4 4 N d , which clearly indicates that the observed compositions and isotopic signatures resulted from mixing between a depleted SEIR M O R B component and an enriched component, interpreted to be that of the Kerguelen plume. The compositions of 34 Ma submarine basalts recovered during O D P Leg 183, Site 1140 (Weis & Frey, 2002) are shown here for comparison. The inset diagram shows the range of compositions observed in Marion Dufresne samples. F i e l d s : 34 M a Site 1140 submarine basalts 28-29 M a Tholeiitic-transitional basalts 25-26 M a Plateau Central basalts 24-25 M a Mildly alkalic basalts ^ | <10 M a ^mmW Alkal ic flows & intrusions I "i Southeast Indian _ _ J Ridge M O R B Mar ion Du f resne : • L a v a f lows > D i k e s a n d si l ls • • T h i s s t u d y (inset): • Aphyric trachybasalts O Aphyric basalts n Plagioclase-phyric basalts m Plagioclase-ultraphyric basalts •()> Olivine-phyric basalts A Quartz-bearing basaltic andesites Dikes and sills 160 FIG. 3.19. Nb/Y vs Zr/Y variations in Kerguelen Archipelago samples. Nb-Zr-Y systematics can be used illustrate the effects of relative extents of melting (higher extents = lower Nb/Y and Zr/Y values) and source composition. The ANb parameter expresses relative excesses (ANb > 0) or deficiencies (ANb < 0) in Nb which have been interpreted to represent plume and ridge source components, respectively, in Iceland (Fitton et al., 1997). The Marion Dufresne low-MgO basalts, 25-24 Ma alkalic lavas, and <10 Ma alkalic flows and intrusions are characterized by lower degrees of melting and more plume-like sources, whereas the Marion Dufresne high-MgO basalts, 29-28 Ma tholeiitic-transitional basalts, 26-25 Ma Plateau Central basalts, and 34 Ma site 1140 submarine basalts represent higher degrees of melting and a greater contribution of a ridge-like source component. The inset diagram shows the range of compositions observed in Marion Dufresne samples. F i e l d s : M | 34 M a W Site 1140 submarine basalts 28-29 M a Tholeiitic-transitional basalts 25-26 M a Plateau Central basalts 24-25 M a Mildly alkalic basalts "1 < 1 0 M a Alkalic flows & intrusions ~i Southeast Indian . ; Ridge M O R B Mar ion Du f resne : • L a v a f lows o D i k e s a n d si l ls Th is s tudy (Inset): • Aphyric trachybasalts Q Aphyric basalts • Plagioclase-phyric basalts fj Plagioclase-ultraphyric basalts <D> Olivine- phyric basalts A Quartz-bearing basaltic andesites Dikes and sills parameter in a Nb/Y vs. Zr/Y diagram is ANb, which expresses a relative excess (ANb > 0) or deficiency (ANb < 0) in Nb that is indicated by the lower limit of the Iceland Array (FIG. 3.19). All Icelandic basalts are defined by ANb > 0, whereas MORB has ANb < 0, which suggests that ANb is a function of mantle source composition. In FIG. 3.19, the high Nb/Y and Zr/Y values observed in the Marion Dufresne low-MgO basalts, the 25-24 Ma alkalic lavas, and the <10 Ma alkalic flows and intrusions from the Kerguelen Archipelago are again consistent with generation by lower degrees of melting compared to the Marion Dufresne high-MgO basalts, the 29-28 Ma tholeiitic-transitional basalts, the 26-25 Ma Plateau Central basalts, and the 34 Ma Site 1140 submarine basalts. In addition, although all groups contain samples with both positive and negative ANb values, source differences between these groups are evident, and the Marion Dufresne high-MgO basalts in particular have negative ANb. Thus, higher degrees of melting result in progressively depleted and more MORB-like isotopic signatures. The Site 1140 submarine basalts have negative ANb and low Nb/Y and Zr/Y values that overlap with the field for SEIR MORB, which strongly indicates that they were derived by mixing between ridge-sourced and plume-sourced melts. In flood basalts from across the Kerguelen Archipelago, these depleted source characteristics become less apparent with decreasing eruption age, indicating that the depleted component likely ceased over time as suggested by Doucet et al. (2002), possibly due to the decreasing extent of melting as distance between the SEIR and the Kerguelen hotspot increased and the plume became situated beneath the thick oceanic crust of the Northern Kerguelen Plateau (Scoates et al., 2005b). 3.7 CONCLUSIONS This study of the -700 m high Marion Dufresne basaltic section in the southern Plateau Central region of the Kerguelen Archipelago (southern Indian Ocean) indicates that -25 Ma magmatism on the archipelago was physically more variable (aphyric, plagioclase-phyric, plagioclase-ultraphyric, olivine phyric) and chemically more variable (based on major and trace element and radiogenic isotopic compositions) than was previously observed. In particular, the wide range in phenocryst contents, alkalinity index, Zr/Nb, (La/Yb)N, and Hf-Sr-Nd-Pb isotopic compositions observed in lavas from the 162 Marion Dufresne section contrasts with the very uniform compositions observed in previously studied sections from the Plateau Central (Tourmente and Capitole) that were erupted earlier and likely crystallized from melts that were stored and homogenized in "steady-state" crystallization regimes. The occurrence of both transitional-tholeiitic and mildly alkalic basalts in all three sections from the Plateau Central suggests that the transition from tholeiitic to mildly alkalic volcanism on the archipelago occurred at a protracted rate between -26-25 Ma, which was largely due to increasing depths of melt segregation, decreasing magma supply rates, and decreasing extents of melting over this time interval. The geochemical characteristics of lavas from the Marion Dufresne section are consistent with a temporal increase in magma flux and increasing extents of partial melting during eruption of the 700 m high sequence of lavas, which contrast with regional trends on the archipelago. The occurrence of 400 m of olivine-phyric high-MgO basalts with isotopically depleted compositions in the upper part of the section probably represents enhanced melting in the melting regime beneath Kerguelen at 25 Ma. The depleted component is most likely depleted asthenospheric mantle that became mixed with plume-derived melts due to the relative proximity of the Kerguelen hotspot to the Southeast Indian Ridge and their continued interaction from -43 Ma to -25 Ma. The cessation of flood basalt volcanism at -24 Ma on the Kerguelen Archipelago occurred shortly after the eruption of the Marion Dufresne basaltic section, and suggests that after -24 Ma, (1) extents of melting beneath the archipelago decreased below a critical amount for effective segregation and ascent, (2) the depth of melting, as influenced by the cooling and thickening lithosphere, increased and effectively shut off melting, and/or (3) hotspot volcanism related to the Kerguelen mantle plume shifted to the south relative to the archipelago, resulting in the formation of seamounts between the archipelago and Heard and McDonald Islands on the Central Kerguelen Plateau (e.g. Weis et al., 2002). 163 3.8 ACKNOWLEDGEMENTS We are grateful to the Institut Polaire Francais Paul Emile Victor (IPEV) for supporting field work on the Kerguelen Archipelago and to Olivier Brisse for his careful sampling of the Marion Dufresne section. The first author thanks the PCIGR team at UBC (Wilma Pretorius, Bruno Kieffer, Jane Barling, Bert Mueller, Diane Hanano and Gwen Williams), as well as Claude Maerschalk at ULB, for their assistance in sample preparation and analysis. ULB students Maud Ibanez and Laurent Deraymaeker are thanked for the crushing of the Marion Dufresne samples. J. Michael Rhodes and Michael Vollinger at the University of Massachusetts are gratefully acknowledged for bulk chemical analyses by XRF. H. Annell was supported by an NSERC Discovery Grant to D. Weis and an NSERC PGS-M. Sample preparation and analytical costs for this study were provided by grants from the Actions de Recherche Concertee, Belgium (ARC 98/03-233 to D. Weis and J. Scoates) and NSERC Discovery Grants to D. Weis and J. Scoates. 164 3.9 REFERENCES Albarede, F., Luais, B., Fitton, G., Semet, M. , Kaminski, E., Upton, B.G.J., Bachelery, P. & Cheminee, J.L. (1997): The geochemical regimes of Piton de la Fournaise volcano (Reunion) during the last 530 000 years. Journal of Petrology 38, 171 -201. Blichert-Toft, J., Chauvel, C. & Albarede, F. (1997): Separation of Hf and Lu for high-precision isotope analysis of rock samples by magnetic sector-multiple collector ICP-MS. Contributions to Mineralogy and Petrology 127, 248-260. Charvis, P., Recq, M. , Operto, S. & Brefort, D. (1995): Deep structure of the northern Kerguelen Plateau and hotspot-related activity. Geophysical Journal International 122, 899-924. Charvis, P., & Operto, S. (1999): Structure of the Cretaceous Kerguelen Volcanic Province (southern Indian Ocean) from wide-angle seismic data. Journal of Geodynamics 28, 51-71. Chauvel, C. & Blichert-Toft, J. (2001): A hafnium isotope and trace element perspective on melting of the depleted mantle. Earth and Planetary Science Letters 190, 137-151. Chen, C.-Y., Frey, F.A., Garcia, M.O., Dalrymple, G.B. & Hart, S.R. (1991): The tholeiite to alkalic basalt transition at Haleakala Volcano, Maui, Hawaii. Contributions to Mineralogy and Petrology 106, 183-200. Chen, C.-Y. & Frey, F.A. (1983): Origin of Hawaiian tholeiite and alkalic basalt. Nature 302, 785-790. Coffin, M.F. & Eldholm, O. (1994): Large igneous provinces: crustal structure, dimensions, and external consequences. Reviews of Geophysics 32, 1-36. Coffin, M.F., Pringle, M.S., Duncan, R.A., Gladczenko, T.P., Storey, M. , Muller, R.D. & Gahagan, L.A. (2002): Kerguelen hotspot output since 130 Ma. Journal of Petrology 43, 1121-1140. Damasceno, D. (1996): Evolution magmatique de la serie basaltique du Mont Crozier (Archipel de Kerguelen, ocean Indien): etude petrologique et chimique. Unpublished B.Sc. thesis, Universite Libre de Bruxelles, Brussels, Belgium, 47 p. Damasceno, D., Scoates, J.S., Weis, D., Frey, F.A., & Giret, A. (2002): Mineral chemistry of mildly alkalic basalts from the 25 Ma Mont Crozier section, Kerguelen Archipelago: constraints on phenocryst crystallization environments. Journal of Petrology 43, 1389-1413. Doucet, S., Weis, D., Scoates, J.S., Nicolaysen, K., Frey, F.A. & Giret, A. (2002): The depleted mantle component in Kerguelen Archipelago basalts: pedogenesis of tholeiitic-transitional basalts from the Loranchet Peninsula. Journal of Petrology 43, 1341 -1366. Doucet, S., Scoates, J.S., Weis, D. & Giret, A. (2005): Constraining the components of the Kerguelen mantle plume: a Hf-Pb-Sr-Nd isotopic study of picrites and high-MgO basalts from the Kerguelen Archipelago. Geochemistry Geophysics Geosystems (in press). Duncan, R.A. (2002): A time frame for construction of the Kerguelen Plateau and Broken Ridge. Journal of Petrology 43, 1109- 1120. Duncan, R.A., Hooper, P.R., Rehacek, J., Marsh, J.S. & Duncan, A.R. (1997): The timing and duration of the Karoo igneous event, southern Gondwana. Journal of Geophysical Research 102, 18127-18138. Fitton, J.G., Saunders, A.D., Norry, M.J., Hardarson, B.J. & Taylor, R.N. (1997): Thermal and chemical structure of the Iceland plume. Earth and Planetary Science Letters 153, 197-208. Fitton, J.G., Saunders, A.D., Kempton, P.D. & Hardarson, B.J. (2003): Does depleted mantle form an intrinsic part of the Iceland plume? Geochemistry Geophysics Geosystems 4, 10.1029/2002GC000424. Frey, F.A., Weis, D., Yang, H.-J., Nicolaysen, K., Leyrit, H. & Giret, A (2000): Temporal geochemical trends in Kerguelen Archipelago basalts: evidence for decreasing magma supply from the Kerguelen plume. Chemical Geology 164, 61-80. Frey, F.A., Nicolaysen, K., Kubit, B.K., Weis, D. & Giret, A. (2002): Flood basalt from Mont Tourmente in the Central Kerguelen Archipelago: the change from transitional to alkalic basalt at -25 Ma. Journal of Petrology 43, 1367-1387. Frey, F.A., Huang, S., Blichert-Toft, J., Regelous, M . & Boyet, M . (2005): Origin of depleted components in basalt related to the Hawaiian hot spot: Evidence from isotopic and incompatible element ratios. Geochemistry Geophysics Geosystems 6, 10.1029/2004GC000757. 166 Galer, S.J.G. & Abouchami, W. (1998): Practical application of lead triple spiking for correction of instrumental mass discrimination. Mineralogical Magazine 62a, 491-492. Gautier, I., Weis, D., Mennessier, J.-P., Vidal., P., Giret, A. & Loubet, M . (1990): Petrology and geochemistry of Kerguelen basalts (South Indian Ocean): evolution of the mantle sources from ridge to an intraplate position. Earth and Planetary Science Letters 100, 59-76. Giret, A. (1983): Le plutonisme oceanique intraplaque, exemple de I'archipel Kerguelen. These doctorat d'Etat, Universite Pierre et Marie Curie, Paris, 290 pp. Giret, A. & Lameyre, J., (1983): A study of Kerguelen plutonism: petrology, geochronology and geological implications. In: Oliver, R.L., James, P.R. & Jago, J.B. (eds.) Antarctic Earth Science. Australia: Australian Academy of Sciences, pp. 646-651. Goolaerts, A., Mattielli, N. , de Jong, J., Weis, D. & Scoates, J. S. (2004): Hf and Lu isotopic reference values for the zircon standard 91500 by MC-ICP-MS. Chemical Geology 206, 1 -9. Hanan, B.B., Blichert-Toft, J., Kingsley, R & Schilling, J.-G. (2000): Depleted Iceland mantle plume geochemical signature: Artifact of multicomponent mixing? Geochemistry Geophysics Geosystems 1, 10.1029/1999GC000009. Hanano, D. (2004): Geochemistry of the Baie Charrier basaltic section, Courbet Peninsula, Kerguelen Archipelago: Implications for the composition of the Kerguelen mantle plume. Unpublished B.Sc. thesis, University of British Columbia, Vancouver, Canada, 92 p. Harpp, K.S. & White, W.M. (2001): Tracing a mantle plume: Isotopic and trace element variations of Galapagos seamounts. Geochemistry Geophysics Geosystems 2, 10.1029/2000GC000137. Hofmann, A. & White, W. (1983): Ba, Rb and Cs in the Earth's mantle. Zeitschrift fur Naturforschung Section A-A, Journal of Physical Sciences 38, 256-266. Kempton, P.D., Fitton, J.G., Saunders, A.D., Nowell, G.M., Taylor, R.N., Hardarson, B.S. & Pearson, G. (2000): The Iceland plume in space and time: a Sr-Nd-Pb-Hf study of the North Atlantic rifted margin. Earth and Planetary Science Letters 177, 255-271. Le Bas, M.J., Le Maitre, R.W., Streckeisen, A., & Zanettin, B. (1986): A chemical classification of volcanic rocks based on the total alkali-silica diagram. Journal of Petrology 27, 745-750. Leitch, A . M . & Davies, G.F. (2001): Mantle plumes and flood basalts: enhanced melting from plume ascent and an eclogite component. Journal of Geophysical Research 106, 2047- 2059. Ludwig, K.R. (2003): Isoplot 3.09: A Geochronological Toolkit for Microsoft Excel. Berkeley Geochronology Center, Special Publication No. 4. Macdonald, G.A. & Katsura, T. (1964): Chemical composition of Hawaiian lavas. Journal of Petrology 5, 82-133. McDonough, W., F. & Sun, S.,-s. (1995): The composition of the Earth. Chemical Geology 120, 223-253. McDougall, I. & Harrison, T.M. (1999): Geochronology and Thermochronology by the 4 0 Ar/ 3 9 Ar Method. New York: Oxford University Press, 269 p. Mattielli, N. , Weis, D., Blichert-Toft, J. & Albarede, F. (2002): Hf isotope evidence for a Miocene change in the Kerguelen mantle plume composition. Journal of Petrology 43, 1327-1339. Nativel, P., Verdier, O. & Giret, A. (1994): Nature et diversite des zeolites de Kerguelen. Memoire de la Societe Geologique, France 166, 31-45. Naumann, T.R. & Geist, D.J. (1999): Generation of alkaline basalt by crystal fractionation of tholeiitic magma. Geology 27, 423-426. Nicolaysen, K., Frey. F.A., Hodges, K.V., Weis, D & Giret, A. (2000): 4 0 Ar/ 3 9 Ar geochronology of flood basalts from the Kerguelen Archipelago, southern Indian Ocean: implications for Cenozoic eruption rates of the Kerguelen plume. Earth and Planetary Science Letters 174, 313-328. Nikogosian, I.K., Elliott, T. & Touret, J.L.R (2002): Melt evolution beneath thick lithosphere: a magmatic inclusion study of La Palma, Canary Islands. Chemical Geology 183, 169-193. Nougier, J. (1970): Carte geologique au 1/200 000 de I'archipel des Kerguelen. Paris: Institut Geographique National. Recq, M. , Le Roy, I., Charvis, P., Goslin, J. & Brefort, D. (1994): Structure profonde du mont Ross d'apres la refraction sismique (Ties Kerguelen, ocean Indien austral). Canadian Journal of Earth 'Sciences 31, 1806-1821. Renne, P.R., Swisher, C C , III, Deino, A.L., Karner, D.B., Owens, T. & DePaolo, D.J. (1998): Intercalibration of standards, absolute ages and uncertainties in 4 0 Ar/ 3 9 Ar dating. Chemical Geology 145, 117-152. Rhodes, J.M. (1996). Geochemical stratigraphy of lava flow samples by the Hawaii Scientific Drilling Project, journal of Geophysical Research 101, 11729-11746. Rosman, J. R. & Taylor, P. D. P. (1998): Isotopic compositions of the elements. Pure and Applied Chemistry 70, 217-236. Royer, J.-Y. & Sandwell, D.T. (1989): Evolution of the eastern Indian Ocean since the late Cretaceous: constraints from Geosat altimetry. Journal of Geophysical Research 94, 13755-13782. Scoates, J.S., Lo Cascio, M. , Weis, D., & Lindsley, D.H. (2005a): Experimental constraints on the origin and evolution of mildly alkalic basalts from the Kerguelen Archipelago, southeast Indian Ocean. Submitted to Contributions to Mineralogy and Petrology. Scoates, J.S., Weis, D., Franssens, M. , Mattielli, N. , Annell, H., Frey, F.A., Nicolaysen, K. & Giret, A. (2005b): The Val gabbro plutonic suite, Kerguelen Archipelago: evolution and significance of a volcanic feeder system in an ocean island. To be submitted to Journal of Petrology. Sharp, W.D., Turrin, B.D. & Renne, P.R. (1996): The 4 0 Ar/ 3 9 Ar and K/Ar dating of lavas from the Hilo 1-km core hole, Hawaii Scientific Drilling Project. Journal of Geophysical Research 101, 11607-11616. Smith, W.H.F. & Sandwell, D.T. (1997): Global seafloor topography from satellite altimetry and ship depth soundings. Science 277, 1956-1962. Storey, M. , Saunders, A.D., Tarney, J., Leat, P., Thirwall, M.F., Thompson, R.N., Menzies, M.A. & Marriner, G.F. (1988): Geochemical evidence from plume-mantle interactions beneath Kerguelen and Heard Islands, Indian Ocean. Nature 336, 371-374. Tikku, A.A. & Cande, S.C. (2000): On the fit of Broken Ridge and Kerguelen plateau. Earth and Planetary Science Letters 180, 117-132. 169 Wasserburg, G.J., Jacobsen, S.B., DePaolo, D.J., McCulloch, M.T. & Wen, T. (1981): Precise determination of Sm/Nd ratios, Sm and Nd isotopic abundances in standard solutions. Geochimica et Cosmochimica Acta 45, 2311 -2323. Weis, D. & Frey, F.A. (1991). Isotope geochemistry of Ninetyeast Ridge basalts: Sr, Nd, and Pb evidence for the involvement of the Kerguelen hot spot. In: Weissel, J., Peirce, J., Taylor, E., Alt, J., et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 121. College Station, TX: Ocean Drilling Program, pp. 591-610. Weis, D., Frey, F.A., Leyrit, H. & Gautier, I. (1993): Kerguelen Archipelago revisited: geochemical and isotopic study of the SE Province lavas. Earth and Planetary Science Letters 118, 101-119. Weis, D. & Giret, A. (1994): Kerguelen plutonic complexes: Sr, Nd, Pb isotopic study and inferences about their sources, ages and geodynamic setting. Bulletin de la Societe Geologique, France, 166, 47-59. Weis, D., Frey, F.A., Giret, A. & Cantagrel, J.-M. (1998): Geochemical characteristics of the youngest volcano (Mount Ross) in the Kerguelen Archipelago: inferences for magma flux, lithosphere assimilation, and composition of the Kerguelen Plume, journal of Petrology 39, 973-994. Weis, D. & Frey, F.A. (2002): Submarine basalts of the Northern Kerguelen Plateau: interaction between the Kerguelen plume and the Southeast Indian ridge revealed at ODP Site 1140. Journal of Petrology 43, 1287-1309. Weis, D., Frey, F.A., Schlich, R., Schaming, M. , Montigny, R., Damasceno, D., Mattielli, N. , Nicolaysen, K.E. & Scoates, J.S. (2002): Trace of the Kerguelen mantle plume: evidence from seamounts between the Kerguelen Archipelago and Heard Island, Indian Ocean. Geochemistry Geophysics Geosystems 3, 10.1029/2001GC000251. Weis, D., Kieffer, 8., Maerschalk, C , Pretorius, W. & Barling, J. (2005): High-precision Pb-Sr-Nd-Hf isotopic characterization of USGS BHVO-1 and BHVO-2 reference materials. Geochemistry Geophysics Geosystems 6, 10.1029/2004GC000852. Xu, C , Frey, F., Weis, D., Scoates, J. S. & Giret, A. (2005): Geochemical characteristics of flood basalts from Mont Capitole in the central Kerguelen Archipelago: further constraints on the Kerguelen plume. In preparation. Yale, M . M . & Phipps Morgan, J. (1998): Asthenosphere flow model of hotspot-ridge interactions: a comparison of Iceland and Kerguelen. Earth and Planetary Science Letters 161, 45-56. Yang, H.-J., Frey, F.A., Weis, D., Giret, A., Pyle, D. & Michon, G. (1998): Petrogenesis of the flood basalts forming the Northern Kerguelen Archipelago: implications for the Kerguelen plume. Journal of Petrology 39, 711 -748. 171 CHAPTER 4 IMPLICATIONS FOR THE MAGMATIC EVOLUTION OF THE KERGUELEN ARCHIPELAGO 4.1. SUMMARY This detailed study of the 700 m high Marion Dufresne basaltic section in the southern part of the Plateau Central region of the Kerguelen Archipelago (Indian Ocean) encompasses petrology, mineral chemistry, geochronology, and major and trace element and radiogenic isotope (Hf-Sr-Nd-Pb).geochemistry in order to develop a detailed characterization of Kerguelen hotspot magmatism at -25 Ma. The diversity of phenocryst types (olivine, clinopyroxene, and plagioclase), the relative abundance of phenocrysts, and the variety of lava compositions observed in the Marion Dufresne section clearly distinguishes this section from the aphyric basalts that dominate previously studied sections on the Kerguelen Archipelago, particularly those from the Plateau Central region. Basalts from the base and top of the Marion Dufresne section have been dated at 25 ± 1 Ma using the 4 0 Ar/ 3 9 Ar geochronologic method. This thesis expands the known geographic extent of mildly alkalic volcanism on the Kerguelen Archipelago to include parts of the Plateau Central and also documents the first occurrence of a depleted mantle component in mildly alkalic -25 Ma basalts on the archipelago. This contrasts with previous observations of basaltic sections on the Kerguelen Archipelago that indicated that this depleted mantle component (and thus chemical interactions between the Southeast Indian Ridge and the Kerguelen mantle plume) effectively ceased prior to 25 Ma. Importantly, the processes affecting basalt petrogenesis and alkalinity at Marion Dufresne show lateral variations from the northeast-southwest regional trends observed in previously studied basaltic sections on the archipelago. At Marion Dufresne, the superposition of basalts with depleted isotopic compositions over a series of comparatively enriched lavas contrasts with regional trends on the archipelago, where previously studied sections indicate a progression toward more enriched isotopic compositions over time. Results from the Marion Dufresne section indicate that the original proposals by Weis et al. (1998) and Frey et al. (2000) that mildly alkalic magmas of the archipelago were produced by progressively A) lower degrees of partial melting, B) increasing depths of melting, 173 and C) decreasing magma supply from the plume over time do not necessarily hold when shorter timescales are considered, as temporal deviations from these trends can be significant. Higher degrees of partial melting in the Marion Dufresne section are demonstrated specifically by the occurrence of the olivine-phyric high-MgO basalts in the upper 400 m of the section. Olivine phenocryst-whole rock Fe-Mg equilibria indicate that these high-MgO basalts were derived from relatively unfractionated parental magmas with 8-10 wt. % MgO. Higher degrees of partial melting will result in magmas with lower abundances of the most incompatible elements compared to low-degree partial melts. Although the absolute abundances of incompatible trace elements can be influenced by fractionation and other differentiation processes, higher-degree partial melts of the same starting material will result in flatter patterns (less enriched in large ion lithophile elements such as Rb and Ba) when normalized to chondrite or primitive mantle values. The high-MgO lavas from Marion Dufresne show lower (La/Yb)N values and lower degrees of REE enrichment relative to chondritic values than all other lavas in the section, which may partly reflect olivine accumulation. The high-MgO lavas also have lower trace element abundance ratios (e.g. Zr/Y, Nb/Y), which are not affected by fractionation, than lavas from the basal 300 m of the Marion Dufresne section, which is consistent with derivation by higher degrees of partial melting of a mantle peridotite and a signifies a temporal increase in the degree of melting at -25 Ma. Increasing depths of magma fractionation are exemplified by changes in the phenocryst assemblage with increasing stratigraphic height in the Marion Dufresne section. Plagioclase-phyric and plagioclase-ultraphyric basalts at the base of the section reflect significant low-pressure magma fractionation and accumulation, and the concentrations of non-quadrilateral components of clinopyroxene phenocrysts and microphenocrysts from these samples reflect crystallization from plagioclase-saturated magmas (i.e. high Ti/AI). In addition, the occurrence of highly calcic plagioclase with >An 8 0 indicates crystallization at <2 kbar, or 5-6 km depth. However, with increasing stratigraphic height, the crystallizing assemblage is dominated by olivine and clinopyroxene. Clinopyroxene compositions from the upper 400 m of the section have lower Ti/AI and likely crystallized from plagioclase-undersaturated magmas at greater depths, which signifies a 174 temporal increase in the depth of magma fractionation over time. This is consistent with the proposal for an enhanced role for high-pressure high-AI clinopyroxene fractionation in the slightly younger 25-24 Ma Crozier basalts to the east (Damasceno et al., 2002). Increasing magma supply from the plume is inferred from comparisons of the phenocryst abundances and major and trace element geochemical characteristics between the Marion Dufresne section and previously studied sections from the Plateau Central region (Tourmente and Capitole). The limited compositional variations observed in other Plateau Central sections, as well as in the aphyric basalts observed in the lower 200 m of the Marion Dufresne section, appear to represent a "steady-state" magmatic system where melt compositions were buffered by equilibration with cumulate assemblages in a solid-dominated environment, as melts percolated through a crystal slurry. With increasing stratigraphic height in the Marion Dufresne section, lava compositions are more variable and some lavas contain high phenocryst abundances (e.g. plagioclase-ultraphyric basalts). This may indicate a significant disruption of the steady-state system resulting from a progressive increase in either magma flux rate or magma volume due to an overall increase in magma supply from the plume. The influx of a geochemically distinct magma during the eruption of the Marion Dufresne section is also evident from the radiogenic isotopic compositions (Hf-Sr-Nd-Pb), which change systematically with increasing stratigraphic height in the section. As another possible result of this increase in magma flux, ascending magma may have exploited previously abandoned channels and intersected pockets of highly differentiated magma in equilibrium with quartz and sodic plagioclase (e.g. trachytes), which were mixed with olivine-phyric high-MgO basaltic magmas to yield the rare quartz-bearing basaltic andesite flows that are interlayered with olivine-phyric high-MgO basalts in the upper 400 m of the section. The olivine-phyric high-MgO basalts (7-12 wt. % MgO with olivine cores of Fo 6 9to Fo88) certainly represent lower degrees of fractionation compared to the underlying lavas, which have <5.2 wt% MgO, and they may reflect the breaking down of the steady-state system and transition to a more liquid-dominated crystallization regime. Samples with unsupportedly high whole-rock mg-numbers 175 relative to the forsterite content of olivine phenocrysts (i.e. those with >10 wt% MgO) likely entrained and erupted accumulated olivine from the dense lower parts of the crystal slurry. The temporal increase in the degree of partial melting, magma supply from the plume, and increasing magma flux coupled with changes in the depths of magma fractionation observed at -25 Ma strongly affected the nature of the transition from tholeiitic to mildly alkalic volcanism on the Kerguelen Archipelago, which was relatively protracted and occurred over a span of >1 Myr. Trachybasalt flows at the base of the Marion Dufresne section are the most alkalic lava flows in this section (A.I. = 1.8-1.4) and the earliest evidence of the evolved magmatism that is typically associated with the terminal stages of basaltic magmatism on the Kerguelen Archipelago observed in the Crozier, Ravin du Charbon, and Ravin Jaune sections from the eastern and southeastern parts of the archipelago, which contain basaltic to trachytic lavas. The occurrence of isotopically depleted compositions in mildly alkalic lavas within the same basaltic section as evolved lava compositions (trachybasalts) at Marion Dufresne is a case that has not been observed elsewhere on the archipelago. Similarities between the radiogenic isotopic characteristics of the depleted components observed in both the Marion Dufresne lavas and the tholeiitic-transitional lavas from the northern part of the Kerguelen Archipelago strongly suggest that this depleted component is depleted upper mantle, similar to that generating basaltic magmas at the Southeast Indian Ridge, which interacted with the Kerguelen mantle plume source subsequent to the intersection of the Kerguelen Plateau by the Southeast Indian Ridge at -43 Ma. Although this depleted mantle signature is interpreted to result from binary mixing at -34 Ma in submarine basalts from ODP Site 1140, which were erupted when SEIR axis was -50 km from the Kerguelen hotspot (Weis & Frey, 2002), the character of the depleted mantle component in flood basalts on the archipelago remains enigmatic and requires further study. This depleted component is not strongly expressed in other sections from the Plateau Central (-26 Ma), but is observed in the younger (-25 Ma) Marion Dufresne section, which was erupted when the Kerguelen hotspot was situated approximately 400 km from the Southeast Indian Ridge. This indicates that prolonged interactions (from -43 Ma to -25 Ma) between the Kerguelen 176 mantle plume and the SEIR, and a magmatic interval characterized by enhanced melting and higher magma flux, ultimately resulted in a temporal increase in the amount of the depleted upper mantle component in mildly alkalic flood basalts on the Kerguelen Archipelago at -25 Ma. 777 4.2. REFERENCES Damasceno, D., Scoates, J.S., Weis, D., Frey, F.A., & Giret, A. (2002): Mineral chemistry of mildly alkalic basalts from the 25 Ma Mont Crozier section, Kerguelen Archipelago: constraints on phenocryst crystallization environments. Journal of Petrology 43, 1389-1413. Doucet, S., Scoates, J.S., Weis, D. & Giret, A. (2005): Constraining the components of the Kerguelen mantle plume: a Hf-Pb-Sr-Nd isotopic study of picrites and high-MgO basalts from the Kerguelen Archipelago. Geochemistry Geophysics Geosystems (in press). Frey, F.A., Weis, D., Yang, H.-J., Nicolaysen, K., Leyrit, H. & Giret, A (2000): Temporal geochemical trends in Kerguelen Archipelago basalts: evidence for decreasing magma supply from the Kerguelen plume. Chemical Geology 164, 61-80. Weis, D., Frey, F.A., Giret, A. & Cantagrel, J.-M. (1998): Geochemical characteristics of the youngest volcano (Mount Ross) in the Kerguelen Archipelago: inferences for magma flux, lithosphere assimilation, and composition of the Kerguelen Plume. Journal of Petrology 39, 973-994. Weis, D. & Frey, F.A. (2002): Submarine basalts of the Northern Kerguelen Plateau: interaction between the Kerguelen plume and the Southeast Indian ridge revealed at ODP Site 1140. Journal of Petrology 43, 1287-1309. 178 APPENDICES 179 Appendix I: Example of a mass balance calculation for correction of olivine accumulation In a high-MgO basalt Sample SI02 TiQ2 AIA Fe;Q3 FeO' MnQ MgO CaO Na;0 K2Q P ;Q5 Total mg*4 KD* Fo' Whole-rock composition (XRF): 527' 46.81 1.83 14.36 2.10 1 0.83 0.19 10.74 9.92 2.38 0.62 0.23 100.01 0.64 0.37 Olivine composition (EPMA): 527-04' 39.29 0.04 0 0 16.22 0.23 43.82 0.30 0 0 0 99.90 0.83 82.81 Fraction olivine removed (wt. % 6.39 Melt composition after removal: 44.30 1.83 14.36 2.10 9.79 0.18 7.94 9.90 2.38 0.62 0.23 93.62 Normalized melt composition: 47.32 1.95 15.34 2.24 10.46 0.19 8.48 10.58 2.54 0.66 0.25 100.00 0.59 0.30 All oxides in wt. %. Abbreviations used: XRF: X-ray fluorescence; EPMA: electron-probe microanalysis 'Sample BOB93-527 is an olivine-phyric high-MgO basalt at 760 m elevation 'Analysis 527-04 is the olivine core analysis that most resembles the average core composition of olivine in this sample 3A.I. (Alkalinity Index) = (Na2O+K!O)-0.37*SiO!+14.43 ' mg # is the mg-number, = Mg/(Mg+Fe2*) 5KD is the Fe-Mg exchange coefficient between olivine and basaltic magma (KD= 0.30 ± 0.03; Roeder & Emslie, 1970) 'assuming 10% Fe present as Fe3' Fig. A1: Plot of forsterite content of olivine vs. whole rock mg-number. The three curves indicate, within error, the equilibrium field for Fe-Mg exchange between olivine and basaltic magma (KD= 0.30 ± 0.03). The olivine composition used is that of 574-04, which is representative of the average olivine composition in this sample. A. The black square falls well below the equilibrium field, which indicates that olivine in this sample (Fo e 2 8) is not in equilibrium with the whole rock (mg-number = 0.64). B . At KD=0.3, the whole rock composition in equilibrium with olivine from this sample (Fo8 / „) has a mg -number of 0.59, which is equivalent to the 5 [ i I removal of 6.4% olivine. 0.56 0.58 0.60 0 62 0.64 0.66 0 68 180 Appendix II: Major (wt. % oxides) and trace element abundances in Marion Dufresne samples by XRF Rock type: Olivine-phyric high-MgO basalt Sample: BOB-93-527 BOB-93-528 BOB-93-530 BOB-93-53T BOB-93-532 BOB-93-535 BOB-93-536 BOB-93-537 BOB-93-538 BOB-93-541 BOB-93-543 BOB-93-545 BOB-93-547 Height (m): 760 755 740 730 715 635 620 600 595 545 530 490 470 Major elements (wt%) Si0 2 46.81 45.89 46.51 46.14 46.31 47.21 47.47 48.29 47.90 48.45 47.98 48.28 48.04 T i 0 2 1.83 1.82 1.83 1.83 1.92 1.74 1.94 1.99 1.93 1.76 2.09 1.87 1.84 Al ,0 3 14.36 14.01 15.23 15.08 15.13 13.83 14.98 15.60 15.81 15.60 14.34 15.09 14.57 Fe 2 0 3 " 12.74 12.99 12.74 12.58 12.60 12.81 12.39 12.37 12.20 12.10 11.98 12.68 12.72 MnO 0.19 0.19 0.20 0.20 0.20 0.19 0.18 0.18 0.18 0.18 0.17 0.18 0.18 MgO 10.74 11.44 9.77 9.77 9.34 11.43 9.66 7.43 7.61 8.02 9.77 8.77 9.74 CaO 9.92 9.96 10.72 11.05 10.95 10.10 9.79 10.85 10.80 10.48 9.90 9.41 9.47 Na 2 0 2.38 2.63 1.92 2.29 2.30 1.86 2.20 2.44 2.42 2.15 2.37 2.22 2.09 K 2 0 0.62 0.39 0.40 0.40 0.42 0.49 0.84 0.48 0.48 0.56 1.05 0.88 0.72 P 2 O s 0.23 0.23 0.24 0.25 0.27 0.27 0.35 0.31 0.31 0.22 0.33 0.24 0.22 TOTAL 99.82 99.55 99.56 99.59 99.44 99.93 99.80 99.94 99.64 99.52 99.98 99.62 99.59 LOI 6.79 2.20 3.90 1.65 0.88 4.09 5.00 1.13 3.15 3.97 4.56 6.51 6.89 mg- number1 0.65 0.66 0.63 0.63 0.62 0.66 0.63 0.57 0.58 0.59 0.64 0.60 0.63 A.I. 2 0.11 0.47 -0.46 0.05 0.02 -0.69 -0.09 -0.52 -0.39 -0.79 0.10 -0.33 -0.53 Trace elements (ppm) Rb 12.0 4.2 5.0 5.0 5.7 5.4 9.4 6.7 5.4 13.0 19.8 17.1 7.8 Sr 226 392 345 354 368 621 665 372 378 308 547 293 229 Ba 110 106 113 98 113 193 215 133 127 163 271 246 169 V 189 219 237 246 245 199 183 230 221 188 194 177 177 Cr 544 525 454 443 396 478 377 260 266 284 422 302 339 Ni 195 288 216 201 188 285 190 115 117 134 251 160 209 Zn 86 89 83 84 88 91 95 96 93 95 88 101 106 Ga 19 18 17 17 18 15 19 20 20 21 18 24 18 Y 19.8 20.5 22.9 23.4 24.6 20.4 21.5 25.4 24.1 20.7 23.0 19.8 19.0 Zr 119 128 134 140 151 135 159 167 157 123 158 125 123 Nb 8.2 8.3 8.8 9.0 10.2 10.9 14.7 12.4 11.6 10.2 14.8 13.8 13.4 Th 1 1 1 1 1 1 1 1 1 1 2 2 2 Pb 1 1 1 1 1 2 2 1 1 1 2 1 2 La 8 11 20 9 10 11 15 16 12 11 17 12 15 Ce 25 27 31 30 33 32 39 37 34 29 42 30 34 *all Fe as Fe 2 0 3 . mg -number= Mg/(Mg+Fe2*). 2A.I. (Alkalinity Index) = (Na2O+K2O)-0.37*SiO2+14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020°C. 181 Appendix II (continued): Major (wt. % oxides) and trace element abundances in Marion Dufresne samples by XRF Rock type: Olivine-phyric high-MgO basalt Plagioclase-phyric basalt Plagioclase-ultraphyric basalt Sample: BOB-93-549 BOB-93-550 BOB-93-551 BOB-93-554 BOB-93-557 BOB-93-558 BOB-93-563 BOB-93-571 BOB-93-572 B0B-93-556 BOB-93-564 B0B-93-565 Height (m): 450 445 430 390 330 330 290 165 150 360 240 215 Major elements (wt%) Si0 2 47.88 47.02 48.00 48.41 48.65 49.70 48.83 48.37 48.48 48.37 47.78 48.62 TiOj 1.81 2.25 1.89 2.83 3.04 3.17 2.88 2.75 2.66 1.89 1.87 1.82 A l 2 0 3 13.93 14.86 17.09 15.23 15.31 14.21 15.79 16.15 15.52 20.71 22.12 21.98 Fe 2 0 3 * 12.81 13.08 12.03 13.44 13.52 14.12 13.00 13.56 13.38 9.49 8.96 8.43 MnO 0.19 0.20 0.17 0.19 0.18 0.20 0.19 0.2 0.21 0.13 0.13 0.12 MgO 10.93 8.61 7.10 4.78 4.69 4.66 4.79 4.42 4.66 3.80 3.56 2.99 CaO 9.08 10.45 9.82 10.13 10.21 9.91 10.82 10.12 10.44 11.59 12.16 12.62 NajO 2.02 2.55 2.39 3.11 2.74 2.67 2.54 2.96 3.04 2.97 2.27 2.36 KJO 0.82 0.57 0.77 1.19 0.77 0.54 0.43 1.1 1.12 0.51 0.69 0.70 P 2 O s 0.22 0.32 0.27 0.43 0.39 0.41 0.36 0.39 0.39 0.22 0.22 0.22 TOTAL 99.69 99.91 99.53 99.74 99.50 99.59 99.63 100.02 99.90 99.68 99.76 99.86 LOI 5.29 2.27 6.55 1.97 1.92 2.87 3.04 2.45 1.96 3.44 4.99 1.81 mg- number1 0.65 0.59 0.57 0.44 0.43 0.42 0.45 0.42 0.43 0.47 0.47 0.44 A.I. 2 -0.45 0.15 -0.17 0.82 -0.06 -0.75 -0.67 0.59 0.65 0.01 -0.29 -0.50 Trace elements (ppm) Rb 15.3 6.9 12.7 16.5 7.3 13.2 3.8 21.4 20.9 5.2 7.8 10.9 Sr 297 394 464 409 403 381 418 404 474 486 590 605 Ba 286 168 178 321 264 251 218 270 313 159 211 186 V 192 • 242 153 271 294 270 263 298 268 181 202 173 Cr 451 416 273 29 64 44 49 23 44 54 33 20 Ni 249 191 100 37 32 32 35 19 30 32 25 16 Zn 105 100 96 114 125 126 112 107 100 84 65 64 Ga 18 20 21 22 23 23 22 23 22 22 21 21 Y 19.3 25.4 19.3 27.4 29.3 30.1 26.5 24.7 25.5 17.6 16.4 15.3 Zr 121 162 133 225 218 227 201 195 197 127 119 118 Nb 12.6 15.5 14.7 30.6 25.6 26.3 22.8 27.3 27.3 15.2 13.8 13.8 Th 2 1 2 4 3 3 3 3 3 2 2 1 Pb 1 1 2 3 2 2 2 3 2 1 1 1 La 12 14 16 28 23 26 21 25 25 15 16 12 Ce 27 37 33 64 54 61 51 56 59 33 34 31 *all Fe as Fe 2 0 3 , mg -number= Mg/(Mg+Fe2+). !A.I. (Alkalinity Index) = (Na2O+K2O)-0.37*SiO2+14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020"C. 182 Appendix II (continued): Major (wt % oxides) and trace element abundances in Marion Dufresne samples by XRF Rock type: Sample: Height (m): Mafic dikes and sills Aphyric basalt BOB-93-534 650 BOB-93-542 540 BOB-93-561 300 BOB-93-552 400 BOB-93-553 390 BOB-93-559 315 BOB-93-560 305 BOB-93-562 290 BOB-93-566 200 BOB-93-567 190 BOB-93-568 185 BOB-9 170 Major elements (wt%) Si0 2 46.79 46.91 48.11 49.36 48.95 50.17 46.92 49.75 48.60 48.51 48.97 48.27 TiQ, 2.60 2.08 2.43 3.12 3.06 3.37 3.53 3.30 2.89 3.05 3.02 3.31 Al:0 3 15.05 14.85 14.90 14.56 14.70 13.46 15.84 13.95 14.99 14.50 14.29 13.83 Fe 2 0 3 " 12.99 12.36 11.98 14.26 14.07 14.92 15.6 14.46 13.96 14.46 14.19 15.25 MnO 0.18 0.17 0.17 0.19 0.19 0.21 0.26 0.21 0.21 0.22 0.20 0.22 MgO 8.40 9.34 8.06 4.37 4.73 4.28 4.99 4.38 5.22 5.19 4.63 4.33 C a O 9.18 10.49 9.10 9.27 9.12 8.78 9.26 9.30 10.15 9.69 10.14 9.23 Na 2 0 2.80 2.51 2.85 3.08 2.89 3.10 2.73 2.91 2.90 2.93 2.91 3.38 K 2 0 1.03 0.96 1.29 1.41 1.41 0.82 0.46 0.81 0.91 1.08 0.90 1.27 P 2 0 5 0.50 0.30 0.57 0.48 0.46 0.44 0.44 0.43 0.35 0.38 0.37 0.49 TOTAL 99.52 99.97 99.46 100.10 99.58 99.55 100.03 99.50 100.18 100.01 99.62 99.58 LOI 2.09 1.60 1.28 2.18 3.88 1.12 1.84 1.87 2.75 2.98 1.91 1.23 mg- number1 0.59 0.62 0.60 0.40 0.43 0.39 0.41 0.40 0.45 0.44 0.42 0.38 A . l . 2 0.95 0.54 0.77 0.66 0.62 -0.21 0.26 -0.26 0.26 0.49 0.12 1.22 Trace elements (ppm) Rb 13.6 16.3 16.3 26.0 28.3 16.2 3.6 10.8 12.0 20.8 16.2 26.6 Sr 907 488 909 398 371 378 557 392 382 355 407 382 Ba 234 223 344 343 342 273 212 285 254 252 263 346 V 203 228 199 290 270 324 296 291 313 318 313 319 Cr 239 398 177 17 51 10 87 22 38 52 38 12 Ni 139 231 118 23 26 14 69 21 25 26 25 17 Zn 98 91 99 124 104 148 147 132 111 121 117 131 Ga 20 20 21 23 24 23 24 23 24 28 22 23 Y 23.6 17.0 26.3 29.8 29.1 32.2 31.6 31.6 24.4 26.0 25.8 33.2 Zr 291 149 295 241 233 244 291 240 192 203 205 260 Nb 17.7 19.3 21.2 33.7 32.2 28.4 29.9 27.3 24.2 26.3 26.3 34.5 Th 2 2 2 4 4 3 3 3 3 3 3 3 Pb 3 2 3 2 3 3 2 3 2 2 3 3 La 24 20 30 30 31 26 22 28 22 23 23 34 Ce 64 46 71 71 69 59 60 62 52 50 56 72 *all Fe as Fe 2 0 3 . mg -number= Mg/(Mg+Fe2*). !A.I. (Alkalinity Index) = (Na 2 0+K£)-0.37*SiOV<'14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020X. 183 Appendix II (continued): Major (wt. % oxides) and trace element abundances in Marion Dufresne samples by XRF Rock type: Sample: Height (m): Aphyric basalt Quartz-bearing basaltic andesite Aphyric trachybasalt BOB-93-570 170 BOB-93-573 130 BOB-93-574 120 BOB-93-575 120 BOB-93-539 585 BOB-93-540 565 BOB-93-544 505 BOB-93-533 660 BOB-93-576 100 BOB-93-577 90 Major elements (wt%) S i 0 2 48.87 47.84 48.06 48.42 54.37 55.08 55.86 48.25 48.93 48.64 TiQ, 2.99 3.27 3.17 2.92 1.39 1.48 1.54 3.06 3.38 3.33 A I 2 O 3 14.18 14.14 13.94 14.74 15.02 15.12 15.26 17.59 14.17 14.15 Fe 2 0 3 * 14.66 15.68 15.51 14.69 9.82 9.09 9.11 13.69 15.34 15.5 MnO 0.21 0.23 0.24 0.20 0.15 0.14 0.13 0.17 0.21 0.29 MgO 4.48 4.52 4.65 4.64 6.65 5.65 5.14 4.39 3.73 3.78 C a O 9.47 9.24 9.24 9.68 8.61 8.18 7.98 7.1 8.46 8.47 Na 2 0 3.36 3.28 3.22 3.06 2.52 2.71 2.92 3.58 3.65 3.83 K 2 0 1.27 1.22 1.24 1.06 1.2 1.77 1.67 1.81 1.46 1.49 P 2 O s 0.44 0.45 0.43 0.38 0.19 0.22 0.29 0.63 0.55 0.55 TOTAL 99.93 99.87 99.70 99.79 99.92 99.44 99.90 100.27 99.88 100.03 LOI 1.65 2.29 2.09 2.34 2.46 2.10 1.47 2.49 1.79 1.26 mg- number1 0.40 0.39 0.40 0.41 0.60 0.58 0.55 0.41 0.35 0.35 A.I. 2 0.98 1.23 1.11 0.63 -1.97 -1.47 -1.65 1.97 1.44 1.75 Trace elements (ppm) Rb 24.3 22.3 25.6 21.1 26.7 45.1 41.9 39.7 29.6 30.0 Sr 395 363 371 385 289 332 424 612 379 397 Ba 318 282 319 260 246 323 352 441 342 343 V 307 363 335 328 184 168 168 117 270 273 Cr 3 7 11 24 306 246 188 3 0 0 Ni 18 16 18 23 154 98 74 16 5 5 Zn 122 134 120 119 83 75 81 102 145 144 Ga 23 25 23 23 19 19 20 19 26 27 Y 28.3 29.8 28.5 25.5 23.7 25.0 22.3 25.0 33.6 33.4 Zr 228 230 216 192 156 177 182 285 274 272 Nb 32.2 32.0 29.8 25.3 11.8 14.5 16.6 46.3 36.4 36.2 Th 3 4 3 3 5 6 6 6 4 4 Pb 3 2 2 3 3 4 3 3 3 3 La 28 29 30 23 21 25 26 39 34 32 Ce 67 66 63 54 42 53 ' 57 85 76 78 'all Fe as Fe 2 O a . mg -number= Mg/(Mg+Fe3*)-2A.I. (Alkalinity Index) = (Na2O+K2O)-0.37*SiO2+14.43; mildly alkalic samples have positive Al values and transitional-tholeiitic samples have negative values. LOI= Loss on Ignition (wt%) is the sample weight lost after 30 minutes at 1020°C. 184 A p p e n d i x III: T race e lement a b u n d a n c e s (ppm) in Mar ion Duf resne s a m p l e s by H R - I C P - M S . Rock type: High-MgO basalt Sample: BOB-93-528 BOB-93-531 BOB-93-532 BOB-93-536 BOB-93-537 B O B - 9 Height (m): 755 7 3 0 715 620 600 450 Trace elements (ppm) L i ' C P 2.69 3.71 3.19 4.63 4.38 4.24 S c l c p 18.5 33.2 35.7 30.5 36.4 23.6 V I C P 286 331 321 267 318 276 C o ' c p 61.7 61.2 59.6 56.0 54.6 63.5 N i l c p 297 227 216 222 138 297 C u , c p 65.0 98.3 98.0 80.2 104 66.7 Z n l c p 85.5 80.1 84.1 91.6 96.4 104 G a l c p 113 15.5 227 233 249 15.4 R b l c p 3.70 5.92 7.11 11.8 8.37 18.4 S r l 0 P 355 381 411 740 429 300 y l C P 17.9 20.6 24.1 22.3 25.7 18.2 ZrCP 128 143 152 167 174 129 N b ' c p 8.56 ' 9.78 11.0 16.3 14.1 13.6 M o ' c p 0.778 0.670 0.706 1.04 1.00 1.14 C d l o p 0.044 0.127 0.112 0.102 0.138 0.072 S n l c p 0.882 0.991 1.15 1.06 1.25 1.03 S b l c p <lod <lod <lod <lod <lod <lod C s , C P 0.120 0.223 0.292 0.209 0.225 0.263 B a l c p 83.5 88.5 101 187 121 234 L a I C P 10.5 10.5 11.3 17.3 15.9 13.6 C e l c p 26.3 27.1 30.4 40.4 36.5 28.7 p R . C P 3.74 3.71 4.60 5.51 5.06 3.78 N d l c p 16.0 16.0 18.8 21.4 20.4 15.2 S m ' c p 4.01 4.12 4.47 4.87 5.09 3.85 E u I C P 1.47 1.51 1.66 1.79 1.80 1.31 G d l c p 4.51 4.73 4.85 4.92 5.48 4.36 T b l c p 0.677 0.693 0.809 0.760 0.830 0.631 D y , c p 4.12 4.35 5.08 4.31 4.69 3.93 H o I C P 0.819 0.857 1.06 0.888 0.988 0.740 E R , C P 2.08 2.30 2.57 2.35 2.68 1.98 Y b I C P 1.81 2.06 2.30 2.00 2.38 1.67 L u I C P 0.255 0.300 0.328 0.268 0.304 0.244 H f C P 2.85 3.00 3.12 3.33 3.60 ' 2.82 T a l c p 0.529 0.586 0.656 0.994 0.779 0.828 W I C P 0.683 1.50 1.46 0.766 1.36 0.843 P B , C P 0.737 0.761 0.862 1.08 0.992 1.14 B i l c p 0.046 0.154 0.213 0.198 0.207 0.207 T h l c p 0.296 0.426 0.535 0.834 0.787 1.59 U I C P 0.162 0.168 0.195 0.255 0.244 0.357 All analyses by H R - I C P - M S at P C I G R , U B C . <lod = below limit of detection. 185 A p p e n d i x III (cont inued) : T race e lement a b u n d a n c e s (ppm) in Mar ion Du f resne s a m p l e s by H R - I C P - M S . Rock type: H igh-MgO basalt Plagioclase-phyric basalt P . -U .B . Sample: BOB-93-550 BOB-93-554 BOB-93-557 BOB-93-572 BOB-93-565 Height (m): 445 390 3 3 0 150 215 Trace elements (ppm) L i l c p 3.26 7.32 4.46 4.56 3.19 S c ' c p 22.5 31.4 31.3 34.4 20.3 V I C P 297 338 359 342 216 C o ' 0 " 50.5 45.9 46.1 48.0 27.6 N i I C P 184 43.8 36.8 35.2 19.4 C u ' c p 73.8 116 98.2 55.9 56.6 Z n I C P 102 117 118 111 66.1 G a I C P 227 20.7 21.3 286 16.9 R b l c p 5.70 21.2 8.47 28.3 14.8 S r , c p 381 473 462 563 668 y l C P 23.5 27.8 26.2 26.8 14.6 Z [ J C P 166 237 225 212 119 N b , c p 16.8 32.7 • 27.8 29.9 14.9 M o l c p 1.19 1.60 1.46 1.52 0.952 C d I C P 0.044 0.143 0.161 0.132 0.074 S n ' c p 1.12 1.63 1.59 1.41 1.11 S b l c p <lod <lod 0.015 <lod <lod C s , c p <lod 0.449 0.183 0.235 0.265 B a , c p 142 283 232 270 167 L a l c p 16.7 28.4 25.1 26.7 14.3 C e I C P 38.6 57.6 52.4 56.9 30.4 P r , c p 5.67 7.14 6.75 7.57 3.88 N d l c p 21.0 27.9 26.2 28.2 16.0 S m , c p 5.23 6.37 6.48 6.17 3.77 E u ' c p 1.78 2.08 2.13 2.07 1.47 G d l c p 5.58 6.54 6.93 6.01 3.94 T b l c p 0.864 0.872 0.951 0.928 0.571 D y l c p 4.96 5.34 5.79 5.13 3.35 H o l c p 1.03 0.977 1.06 1.04 0.625 E r , c p 2.60 2.53 2.77 2.69 1.59 Y b l c p 2.37 2.12 2.30 2.28 1.34 L u I C P 0.314 0.311 0.327 0.310 0.193 H f c p 3.87 4.76 4.76 4.38 2.50 T a l c p 0.993 1.76 1.45 1.69 0.937 W I C P 0.396 0.838 0.970 0.829 1.11 P B , C P 1.02 2.17 2.00 1.91 1.15 B i l c p <lod 0.212 0.156 0.209 0.222 T h l c p 0.769 3.10 2.58 2.94 1.41 U I 0 P 0.205 0.588 0.431 0.389 0.319 Rock type P. -U.B indicates plagioclase-ultraphyric basalt. Al l analyses by H R - I C P - M S at P C I G R , U B C . <lod = below limit of detection. 186 A p p e n d i x III (cont inued) : T race e lement a b u n d a n c e s (ppm) in Mar ion Du f resne s a m p l e s by H R - I C P - M S . Rock type: Mafic dikes and sills Aphyr ic basalt Sample : BOB-93-534 BOB-93-542 BOB-93-561 BOB-93-552 BOB-93-559 B O B - 9 Height (m): 650 540 300 400 315 185 Trace elements (ppm) L J , C P 4.27 4.83 5.82 4.42 4.69 4.71 S c , c p 18.3 15.4 19.8 27.8 32.2 34.2 V I C P 270 283 245 344 376 371 C o l c p 56.1 53.2 46.7 42.6 45.6 47.6 N i l c p 170 228 131 25.2 17.5 26.9 C u ' c p 56.0 74.1 50.6 72.3 58.6 120 Z n l c p 95.7 99.0 103 114 135 108 G a l c p 17.9 246 248 154 295 268 R b I C P 12.7 14.8 14.5 30.9 20.3 20.4 S r l c p 835 487 887 409 424 449 y l C P 20.2 14.9 24.0 27.5 32.1 25.6 Z r , c P 302 155 299 231 249 200 N b l c p 18.7 21.8 22.5 33.1 27.9 25.9 M o l c p 1.39 1.39 1.34 1.74 1.73 1.24 C d l c p 0.227 0.050. 0.174 0.201 0.197 0.177 S n , c p 1.68 1.04 1.69 1.63 1.78 1.46 S b l c p 0.039 <lod <lod 0.016 0.020 0.005 C s l c p 0.287 0.106 0.173 0.271 0.479 0.379 B a l c p 189 194 297 283 250 232 L a ' c p 24.6 19.9 28.3 31.5 27.6 23.8 C e ' c p 57.2 44.7 69.8 66.8 56.8 49.8 p R I C P 7.22 5.95 9.51 8.14 7.52 6.86 N d l c p 29.0 21.1 35.1 31.9 29.6 25.5 S m l c p 6.58 4.59 7.40 7.01 7.06 5.66 E u , Q P 2.21 1.59 2.52 2.34 2.35 1.99 G d l 0 p 6.52 4.36 6.93 7.31 7.43 5.63 T b l c p 0.855 0.661 1.01 1.00 1.06 0.855 D y l c p 4.88 3.57 5.65 5.94 6.16 4.74 H o l c p 0.875 0.695 1.12 1.10 1.23 0.938 Er,cp 2.18 1.76 2.67 2.86 3.14 2.53 Y B , C P 1.89 1.40 2.33 2.40 2.68 2.12 L u , C P 0.268 0.176 0.321 0.347 0.369 0.272 H f c p 5.52 3.69 6.18 4.92 5.36 4.31 T a l c p 1.11 1.30 1.32 1.85 1.70 1.55 W ' c p 0.779 0.685 0.669 1.34 1.60 0.926 P B , C P 2.37 1.45 2.22 2.36 2.14 1.92 B i ' c p 0.186 <lod <lod 0.215 0.212 0.214 T h l c p 1.222 1.70 1.87 3.31 3.00 2.62 U I C P 0.491 0.414 0.440 0.756 0.594 0.533 All analyses by H R - I C P - M S at P C I G R , U B C . <lod = below limit of detection. 187 A p p e n d i x III (cont inued) : T race e lement a b u n d a n c e s (ppm) in Mar ion Du f resne s a m p l e s by H R - I C P - M S . Rock type: Aphyr ic basalt Quartz-bearing basaltic andesite Sample : BOB-93-569 BOB-93-570 BOB-93-574 BOB-93-539 BOB-93-540 B O B - 9 Height (m): 170 170 120 585 565 505 Trace elements (ppm) L i , 0 P 6.25 4.78 5.07 8.38 8.17 9.43 S c l c p 30.6 34.1 31.1 16.4 24.6 20.9 y l C P 384 376 387 215 223 207 C o l c p 44.5 50.7 46.2 40.4 34.0 33.4 N i l c p 20.6 21.6 17.9 148 97.2 77.3 C u l c p 91.9 127 . 90.0 80.0 63.9 51.6 Z n I O P 131 120 133 82.4 76.7 85.0 G a l c p 22.2 20.9 272 235 240 242 R b ' c p 33.4 32.5 24.0 25.6 46.1 41.2 S r ' 0 " 415 480 370 298 340 439 y l C P 32.8 29.2 27.1 21.7 24.0 21.6 Z r , C P 272 238 224 161 179 184 N b l c p 36.2 34.8 28.7 13.1 15.8 18.6 M o , c p 2.09 1.80 1.46 1.26 1.60 1.87 C d l c p 0.234 0.156 0.123 0.053 0.053 0.068 S n l c p 1.80 1.68 1.47 1.43 1.67 1.46 S b l c p 0.016 <lod <lod <lod <lod <lod C s l c p 0.377 0.250 <lod 0.249 0.346 0.346 B a ' c p 288 288 283 240 303 347 L a l 0 p 32.7 27.8 29.9 19.5 22.6 26.2 G e I C P 66.8 61.4 62.3 43.0 51.7 57.2 P r l c p 8.36 7.63 8.08 5.46 6.72 7.51 N d l c p 33.1 30.0 28.7 20.1 23.2 26.2 S m l c p 7.49 6.53 6.59 4.41 4.92 5.30 E u l c p 2.40 2.26 2.16 1.46 1.63 1.67 G d , c p 7.98 6.78 6.77 4.55 4.73 5.01 T b , c p 1.07 0.953 1.00 0.718 0.810 0.784 D y l c p 6.53 5.70 5.64 4.67 4.85 4.75 H o l c p 1.24 1.04 1.16 0.973 0.989 0.967 E r l c p 3.19 2.68 2.94 2.38 2.49 2.34 Y b l c p 2.62 2.33 2.48 2.26 2.29 2.19 L u l c p 0.381 0.341 0.353 0.301 0.311 0.300 H f c p 5.60 4.96 5.63 3.88 4.38 4.38 T a I C P 2.12 1.90 1.67 1.03 1.24 1.31 W I C P 1.22 1.02 0.417 1.02 1.16 1.31 p b ( C P 2.39 2.09 1.80 2.72 3.11 3.14 B i l c p 0.226 0.212 <lod <lod <lod <lod T h , c p 3.28 3.36 3.37 4.36 5.78 5.27 U ' C P 0.716 0.677 0.587 0.854 0.987 0.962 All analyses by H R - I C P - M S at P C I G R , U B C . <lod = below limit of detection. 188 A p p e n d i x III (cont inued) : T race e lement a b u n d a n c e s (ppm) in Mar ion Duf resne s a m p l e s by H R - I C P - M S . Rock type: Aphyric trachybasalt Sample: BOB-93-533 BOB-93-576 BOB-93-577 Height (m): 660 100 90 Trace elements (ppm) L J , C P 5.68 8.93 5.61 S c l c p 13.0 28.1 22.9 y l C P 143 327 291 C o l c p 38.2 44.2 38.5 N i l c p 16.1 7.63 6.76 C u I C P 27.0 51.8 38.8 Z n ' c p 113 140 125 G a I O P 251 336 24.9 R b l c p 38.0 39.7 27.4 S r l c p 604 410 370 y l C P 23.0 34.9 29.7 Z I . . C P 288 291 257 N b l o p 44.1 36.9 34.5 M o l c p 1.86 2.23 2.03 C d l c p 0.119 0.234 0.208 S n l c p 1.83 1.73 1.76 S b I C P <lod 0.072 <lod C s , c p 0.232 0.300 0.343 B a l c p 403 313 280 L a ' c p 37.2 35.2 31.6 C e I O P 87.6 73.9 66.3 p r I C P 11.2 10.1 8.37 N d , c p 39.7 37.3 32.4 S m l c p 7.54 8.14 7.42 E u ' c p 2.58 2.77 2.52 G d l c p 6.56 8.20 7.80 T b I C P 1.00 1.19 1.05 D y I C P 5.53 6.53 6.49 H o l c p 1.07 1.33 1.17 E r , C P 2.43 3.43 3.10 y b , C P 2.12 3.05 2.56 L u l c p 0.279 0.395 0.364 H f c p 6.49 5.97 5.20 T a ' c p 2.31 2.17 1.91 W I C P 0.550 0.624 0.935 p b , C P 2.83 2.45 2.50 B i l o p <lod 0.213 0.206 T h l 0 p 4.77 3.65 2.89 U I C P 0.885 0.727 0.682 All analyses by H R - I C P - M S at P C I G R , U B C . <lod = below limit of detection. 189 Appendix IV: Duplicate analysis of trace element abundances by HR-ICP-MS Sample: BOB-93-528 BOB-93-528-R Average a %RSD Li 2.34 3.05 2.69 0.506 18.8 Sc 25.2 11.8 18.5 9.51 51.3 V 283 288 286 3.39 1.19 Co 62.6 60.9 61.7 1.16 1.88 Ni 304 290 297 9.73 3.27 Cu 64.5 65.5 65.0 0.712 1.09 Zn 80.2 90.9 85.5 7.61 8.90 Ga 15.7 21.0 18.3 3.78 20.6 Rb 4.25 3.15 3.70 0.779 21.0 Sr 332 379 355 33.2 9.36 Y 15.57 14.53 15.0 0.732 4.86 Zr 126 129 128 2.72 2.13 Nb 8.501 8.61 8.56 0.076 0.889 Mo 0.718 0.837 0.778 0.084 10.8 Cd 0.072 0.017 0.044 0.039 87.0 Sn 0.853 0.910 0.882 0.041 4.60 Sb <lod <lod Cs 0.277 <lod Ba 79.2 87.9 83.5 6.17 7.39 La 10.9 10.1 10.5 0.527 5.01 Ce 26.1 26.5 26.3 0.282 1.07 Pr 3.68 3.80 3.74 0.085 2.28 Nd 15.5 16.4 16.0 0.680 4.27 Sm 4.13 3.89 4.01 0.168 4.18 Eu 1.48 1.46 1.47 0.016 1.07 Gd 4.81 4.22 4.51 0.420 9.31 Tb 0.681 0.672 0.677 0.006 0.938 Dy 4.08 4.17 4.12 0.058 1.40 Ho 0.785 0.854 0.819 0.048 5.92 Er 2.09 2.07 2.08 0.011 0.535 Yb 1.79 1.84 1.81 0.035 1.91 Lu 0.256 0.253 0.255 0.002 0.828 Hf 2.59 3.11 2.85 0.363 12.7 Ta 0.499 0.558 0.529 0.042 7.91 W 0.891 0.475 0.683 0.294 43.0 Pb 0.781 0.693 0.737 0.062 8.44 Bi 0.211 <lod Th 0.277 0.314 0.296 0.026 8.65 U 0.161 0.163 .0.162 0.001 0.632 "R" after sample number indicates a complete procedural duplicate. "<lod" indicates that abundance was below the level of detection. 190 Appendix IV (continued): Duplicate analysis of trace element abundances by HR-ICP-MS Sample: BOB-93-552 BOB-552-R Average a %RSD Li 4.57 4.28 4.42 0.208 4.70 Sc 27.0 28.7 27.8 1.19 4.26 V 341 346 344 3.57 • 1.04 Co 41.2 44.0 42.6 1.99 4.67 Ni 23.5 26.9 25.2 2.34 9.29 Cu 67.0 77.7 72.3 7.57 10.5 Zn 105 123 114 12.3 10.7 Ga 20.5 28.8 24.6 5.83 23.7 Rb 29.6 32.1 30.9 1.77 5.75 Sr 398 419 409 14.8 3.61 Y 26.8 28.3 27.5 1.04 3.77 Zr 220 241 231 14.9 6.45 Nb 32.3 33.9 33.1 1.13 3.41 Mo 1.54 1.93 1.74 0.273 15.7 Cd 0.197 0.205 0.201 0.005 2.65 Sn 1.5 1.7 1.6 0.131 8.03 Sb 0.023 0.010 0.016 0.009 53.6 Cs 0.216 0.327 0.271 0.079 29.1 Ba 275 290 283 10.1 3.57 La 32.5 30.5 31.5 1.46 4.64 Ce 66.9 66.7 66.8 0.135 0.202 Pr 8.15 8.12 8.14 0.018 0.222 Nd 32.1 31.7 31.9 0.270 0.847 Sm 7.25 6.77 7.01 0.338 4.82 Eu 2.37 2.30 2.34 0.048 2.03 Gd 7.60 7.03 7.31 0.402 5.49 Tb 1.02 0.975 0.999 0.034 3.40 Dy 6.18 5.70 5.94 0.341 5.73 Ho 1.16 1.05 1.10 0.074 6.72 Er 2.96 2.75 2.86 0.150 5.26 Yb 2.50 2.30 2.40 0.140 5.82 Lu 0.361 0.332 0.347 0.020 5.82 Hf 4.87 4.97 4.92 0.074 1.50 Ta 1.73 1.97 1.85 0.176 9.51 W 1.67 1.02 1.34 0.463 34.4 Pb 2.45 2.26 2.36 0.137 5.81 Bi 0.218 0.212 0.215 0.004 2.02 Th 3.29 3.32 3.31 0.025 0.769 U 0.791 0.721 0.756 0.050 6.60 "R" after sample number indicates a complete procedural duplicate. "<lod" indicates that abundance was below the level of detection. 191 Appendix IV (continued): Duplicate analysis of trace element abundances by HR-ICP-MS Sample: BOB-93-561 BOB-93-561-R Average a %RSD Li 5.70 5.94 5.82 0.176 3.02 Sc 14.9 24.7 19.8 6.91 34.9 V 247 243 245 2.53 1.03 Co 47.3 46.0 46.7 0.885 1.90 Ni 133 128 131 3.70 2.83 Cu 51.0 50.2 50.6 0.608 1.20 Zn 103 103 103 0.151 0.146 Ga 247 248 248 0.879 0.355 Rb 13.6 15.4 14.5 1.33 9.19 Sr 874 900 887 17.9 2.02 Y 22.9 25.0 24.0 1.42 5.92 Zr 302 297 299 3.80 1.27 Nb 23.5 21.5 22.5 1.38 6.14 Mo 1.28 1.39 1.34 0.077 5.74 Cd 0.158 0.190 0.174 0.023 13.0 Sn 1.68 1.69 1.69 0.004 0.242 Sb <lod <lod Cs 0.164 0.182 0.173 0.013 7.55 Ba 293 301 297 5.70 1.92 La 27.2 29.4 28.3 1.56 5.50 Ce 68.0 71.6 69.8 2.55 3.65 Pr 9.71 9.30 9.51 0.285 3.00 Nd 36.4 33.8 35.1 1.88 5.36 Sm 7.36 7.44 7.40 0.062 0.841 Eu 2.55 2.49 2.52 0.041 1.64 Gd 6.91 6.94 6.93 0.020 0.296 Tb 1.03 0.979 1.01 0.039 3.83 Dy 5.73 5.57 5.65 0.114 2.02 Ho 1.13 1.10 1.12 0.019 1.69 Er 2.59 2.75 2.67 0.114 4.26 Yb 2.34 2.32 2.33 0.010 0.426 Lu 0.327 0.315 0.321 0.009 2.65 Hf 6.10 6.25 6.18 0.109 1.76 Ta 1.37 1.27 1.32 0.072 5.50 W 0.673 0.665 0.669 0.006 0.851 Pb 2.19 2.24 2.22 0.036 1.63 Bi <lod <lod Th 1.76 1.98 1.87 0.16 8.46 U 0.432 0.449 0.440 0.01 2.77 "R" after sample number indicates a complete procedural duplicate. "<lod" indicates that abundance was below the level of detection. 192 Appendix V: Comparison of trace element abundances (ppm) in sample 1140-31R-1 Weis & Frey, 2002 Doucet et al., 20021 This study2 (XRF, ICP-MS, NA) (ICP-MS) a (HR-ICP-MS) a Ba 100.0 100.43 2.83 103 2.67 Bi 0.17 0.126 Cd 0.19 0.038 Ce 31. 17 31.3 0.74 32.4 1.48 Co 48 48.0 2.31 Cs 0.25 0.108 Cu 102 5.56 Dy 6.19 6.23 0.11 5.80 0.234 Er 3.41 3.34 0.03 3.21 0.162 Eu 1.79 1.78 0.05 1.81 0.050 Ga 20 18.35 0.28 25.0 0.588 Gd 6.18 6.12 0.278 Hf 3.75 3.89 0.21 4.09 0.217 Ho 1.30 1.27 0.01 1.20 0.091 La 13.0 13.08 0.41 13.6 0.374 Li 9.01 0.562 Lu 0.47 0.43 0 0.422 0.023 Mo 1.02 0.119 Nb 13.3 12.07 0.15 13.3 0.402 Nd 20.4 20.25 0.49 18.9 0.735 Ni 68 70.7 3.04 Pb 1.52 1.45 0.87 1.12 0.060 Pr 4.40 4.42 0.13 4.51 0.291 Rb 6.23 5.96 5.96 6.36 0.980 Sc 36.5 36.61 0.96 37.6 1.28 Sm 5.51 5.35 0.10 5.13 0.142 Sn 1.36 0.018 Sr 262 191.36 3.37 272 10.3 Ta 0.87 0.82 0.07 0.89 0.034 Tb 1.03 1.02 0.01 0.937 0.057 Th 1.64 1.51 0.05 1.49 0.105 Tm 0.53 0.52 0.01 U 0.37 0.34 0.34 0.35 0.018 V 300 348 13.0 W 1.24 0.187 Y 31.9 33.97 0.3 30.2 0.789 Yb 2.98 3.01 0.04 2.89 0.189 Zn 118 101 4.34 Zr 167 154.25 1.85 166 3.28 1Average of eight full duplicate analyses Values for REE, U, and Pb are the average of 8 replicate analyses of 2 separate digestions; average values for all other elements represent 6 replicate analyses of 2 separate digestions. 193 2 350 a. O £ 250 E 150 a a. 50 50 | 40 I a O 30 CC X 20 ¥ a. 10 Q. Vanadium y= 1.1X + 44 R2= 0.92 150 250 350 ppm (XRF) 50 100 150 200 250 300 350 ppm (XRF) 60 80 100 120 140 160 ppm (XRF) Rubidium • y= 1 .Ox + 1.9 o / & R2= 0.92 - y 325 275 225 175 125 — 1 — 1 — 1 — 1 — 1 — • Zirconium 1 1 1 r-yr - y= 1 .Ox + 3.2 . R2= 0.98 ' • i • i • / T O • i • 10 20 30 40 ppm (XRF) 75 125 175 225 275 ppm (XRF) 200 400 600 800 ppm (XRF) 5 Cerium 80 • y= 0.92x + 2.3 ICP-I 60 Rz= 0.97 (HR-40 ppm 20 20 40 60 80 ppm (XRF) 100 200 300 400 ppm (XRF) 20 30 40 ppm (XRF) 10 20 30 ppm (XRF) 10 20 30 ppm (XRF) 8 10 12 14 ppm (XRF) Appendix VI. Comparison of trace element abundances determined by XRF and HR-ICP-MS. All samples are included in 1:1 plots; see Appendices II and III for values used. R2 values are between 0.9 and 0.99 for all trace elements analyzed by both methods. 194 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0052331/manifest

Comment

Related Items