UBC Faculty Research and Publications

Hf isotope constraints on mantle sources and shallow-level contaminants during Kerguelen hot spot activity.. Weis, Dominique 2003

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
Weis_AGU_2002GC000482.pdf [ 1.26MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0075976.json
JSON-LD: 1.0075976+ld.json
RDF/XML (Pretty): 1.0075976.xml
RDF/JSON: 1.0075976+rdf.json
Turtle: 1.0075976+rdf-turtle.txt
N-Triples: 1.0075976+rdf-ntriples.txt
Original Record: 1.0075976 +original-record.json
Full Text
1.0075976.txt
Citation
1.0075976.ris

Full Text

Hf isotope constraints on mantle sources and shallow-level contaminants during Kerguelen hot spot activity since 120 Ma Stephanie Ingle Department of Earth and Environmental Sciences, Faculte´ des Sciences, Universite´ Libre de Bruxelles, CP160/02, 50 Avenue F. D. Roosevelt, Brussels, Belgium, B-1050 (single@ulb.ac.be) Now at Department of Earth and Planetary Sciences, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8551, Japan. Dominique Weis Department of Earth and Environmental Sciences, Faculte´ des Sciences, Universite´ Libre de Bruxelles, CP160/02, 50 Avenue F. D. Roosevelt, Brussels, Belgium, B-1050 Now at Department of Earth and Ocean Sciences, University of British Columbia, 6339 Stores Road, Vancouver, British Columbia, Canada V6T 1Z4. Sonia Doucet Department of Earth and Environmental Sciences, Faculte´ des Sciences, Universite´ Libre de Bruxelles, CP160/02, 50 Avenue F. D. Roosevelt, Brussels, Belgium, B-1050 Now at Laboratory of Geochemistry and Cosmochemistry, I.P.G.P., 4 Place Jussieu, T14-15, 75252 Paris Cedex 05, France. Nadine Mattielli Department of Earth and Environmental Sciences, Faculte´ des Sciences, Universite´ Libre de Bruxelles, CP160/02, 50 Avenue F. D. Roosevelt, Brussels, Belgium, B-1050 [1] We report new Hf isotopic data for basalts from the Kerguelen large igneous province (LIP) obtained using high-precision multicollector inductively coupled mass spectrometry (MC-ICP-MS) analysis. All drill sites from the southern Indian Ocean Kerguelen Plateau–Broken Ridge, in addition to two volcanic suites from the Kerguelen Archipelago, are represented. These new data are integrated with other recently reported geochemical data for this LIP. We examine the geochemical signatures of the mantle sources and shallow-level contaminants present during the past 120 Ma history of Kerguelen hot spot activity. Our results highlight the contribution of distinct mantle source compositions during Cretaceous (Kerguelen Plateau and Broken Ridge) and Cenozoic (northernmost Kerguelen Plateau and Kerguelen Archipelago) magmatism arising from melting of the Kerguelen plume head and plume tail, respectively. The Cretaceous Kerguelen plume basalts have primitive mantle-like Pb, Sr, and Nd isotopic compositions and moderately depleted Hf isotopic compositions and are different from the Cenozoic plume basalts, which extend to more radiogenic Pb isotopic compositions. Neodymium and Hf isotopes are decoupled in Kerguelen plume-derived rocks, and this, in combination with their Pb isotopic compositions, implies that the Kerguelen plume contains small amounts of ancient pelagic sediment mixed with old, recycled enriched oceanic crust. Different contributions from pelagic sediments relative to oceanic crust are able to account for the distinction between the isotopic compositions of the Cretaceous and Cenozoic mantle plume sources. Assimilation of shallow-level continental crust, left stranded in the Indian Ocean during G3GeochemistryGeophysicsGeosystems Published by AGU and the Geochemical Society AN ELECTRONIC JOURNAL OF THE EARTH SCIENCES Article Volume 4, Number 8 8 August 2003 1068,? doi:10.1029/2002GC000482 ISSN: 1525-2027 Copyright 2003 by the American Geophysical Union 1 of 28Gondwana breakup, by plume-derived magmas was the dominant process recorded in Cretaceous Kerguelen Plateau basalts. During the Cenozoic, magmas from the Kerguelen plume mixed to varying degrees with local, Indian Ocean depleted upper mantle and assimilated the overlying Cretaceous Kerguelen Plateau lithosphere. Despite the geochemical heterogeneity of the Kerguelen LIP, we find evidence for a finite number of components involved in the genesis of Kerguelen Plateau–Broken Ridge and Kerguelen Archipelago rocks. Components: 15,187 words, 10 figures, 2 tables. Keywords: Kerguelen hot spot; large igneous provinces; Hf isotopes; Indian Ocean; ocean island basalt; Ocean Drilling Program. Index Terms: 1010 Geochemistry: Chemical evolution; 1025 Geochemistry: Composition of the mantle; 1040 Geochemistry: Isotopic composition/chemistry. Received 26 November 2002; Revised 9 May 2003; Accepted 18 June 2003; Published 8 August 2003. Ingle, S., D. Weis, S. Doucet, and N. Mattielli, Hf isotope constraints on mantle sources and shallow-level contaminants during Kerguelen? hot? spot? activity? since?120? Ma,?Geochem.? Geophys.? Geosyst.,?4(8),? 1068,? doi:10.1029/2002GC000482,? 2003. 1. Introduction 1.1. Objectives of the Hf Isotope Investigation [2] The Kerguelen hot spot is located in the southern Indian Ocean and has been active since at least 120 Ma [Duncan, 2002]. This long-lived volcanism has been attributed to a large mantle plume upwelling from deep in the mantle; the plume head created the Kerguelen Plateau–Bro- ken Ridge and the plume tail created the Nine- tyeast Ridge, Kerguelen Archipelago, Heard and McDonald Islands [Davies et al., 1989; Storey et al., 1989; Weis et al., 1989, 1991; Storey et al., 1996]. The Kerguelen LIP is extremely geo- chemically heterogeneous when compared with other oceanic plateaus [Arndt and Weis, 2002]. This complexity is probably largely the result of interaction between plume-derived magmas and continental fragments that are present in some parts of the plateau’s structure [e.g., Mahoney et al., 1995; Frey et al., 2000]. The presence of continental lithosphere leads to many ambiguities regarding the number and origin of mantle com- ponents and shallow-level contaminants involved in the genesis of this LIP. Fortunately, numerous recent investigations have provided a large back- ground dataset of major and trace elements and Sr, Nd and Pb isotopic compositions for lavas from the Kerguelen Plateau, Broken Ridge, 90E Ridge, Kerguelen Archipelago, Kerimis sea- mounts and Heard and McDonald Islands. These recent investigations have helped to identify basalts reflecting assimilation of continental crust by plume-derived magmas, and basalts where little or no role for continental crust is implicated [Mahoney et al., 1995; Weis et al., 1998, 2001, 2002a; Yang et al., 1998; Doucet et al., 2002; Frey et al., 2002a, 2002b; Ingle et al., 2002a, 2002b; Kieffer et al., 2002; Neal et al., 2002; Weis and Frey, 2002]. However, despite this abundant data, an integrated picture of the geo- chemical evolution of the Kerguelen hot spot has yet to be presented. We investigate whole rock Hf isotopic compositions in an attempt to further define the mantle sources and shallow-level con- taminants involved in the genesis of the Kergue- len Plateau, Broken Ridge and Kerguelen Archipelago rocks. Our study is intended to complement the work of Mattielli et al. [2002], who presented Hf isotopic compositions for many volcanic sections from the Kerguelen Archipelago. [3] Hafnium in the Earth is primarily sequestered in zircons that are abundant in continental crust and high-energy terrigenous sediments [Patchett et al., 1984]. Therefore Hf isotopes are a useful indicator of continental contamination and sediment recy- Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 2 of 28cling; non-continental (i.e., pelagic) sediments can develop radiogenic Hf isotopic signatures (because of the lack of zircons) whereas continental crust and terrigenous sediments develop unradiogenic Hf isotopic signatures over time [Patchett et al., 1984; Vervoort et al., 1999]. The Lu-Hf isotope system behaves geochemically similarly to the Sm-Nd isotope system but important differences exist. Partition coefficients are more disparate for Lu and Hf, relative to Sm and Nd, and this, coupled with the shorter half-life of 176Lu, relative to 147Sm, creates a range in radiogenic Hf isotopic compositions nearly double the variation in Nd isotopic compositions [Patchett and Tatsumoto, 1980]. Also, Hf isotopes may help to discriminate between continental crust and subcontinental litho- spheric mantle (SCLM) since initial investigations suggest that Hf isotopes may be strongly decoupled from Nd isotopes in the SCLM [Schmidberger et al., 2002; Simon et al., 2002; Ionov and Weis, 2002]. [4] Basalts from the Indian Ocean (both mid-ocean ridge basalts (MORB) and ocean island basalts (OIB)) are distinguished by their higher 87Sr/86Sr [Hedge et al., 1973], higher 207Pb/204Pb and 208Pb/204Pb relative to 206Pb/204Pb [Dupre´ and Alle`gre, 1983; Hart, 1984] and lower 143Nd/144Nd relative to 206Pb/204Pb [Hart, 1984;Mahoney et al., 1996] when compared to basalts from the northern hemisphere in the Pacific and Atlantic Oceans [Hart, 1984]. Hf isotopes have proved to be a powerful tool in investigating this distinctive isotopic feature of the Indian Ocean [Chauvel and Blichert-Toft, 2001; Kempton et al., 2002], the so-called ‘‘Dupal anomaly’’ [Hart, 1984]. These Hf isotopic studies have demonstrated that MORB from the Indian Ocean are displaced to lower 143Nd/144Nd for a given 176Hf/177Hf. Mattielli et al. [2002] also documented this feature in some sections of the Kerguelen Archipelago. The Kerguelen Archipelago and the Kerguelen Plateau share the other above mentioned geochemical char- acteristics with Indian Ocean basalts [Davies et al., 1989; Weis et al., 1989; Salters et al., 1992; Storey et al., 1996; Weis and Frey, 1996]. Therefore it seems likely that the process or processes that created the special isotopic signature in Indian Ocean basalts should be applicable to Kerguelen basalts and, in this respect, we may test the various hypotheses invoked to explain the existence of the Dupal anomaly. 1.2. Geological Setting and Tectonic Evolution [5] The currently submerged Kerguelen Plateau and Broken Ridge and a few volcanic, oceanic islands constructed upon the plateau, Kerguelen Archipelago, Heard and MacDonald Islands are believed to be the manifestation of the Kerguelen hot spot (Figure 1) [Coffin et al., 2002]. Conti- nental tholeiites on the margins of southwest Australia (Bunbury basalts, 132, 123 Ma) and east India (Rajmahal Traps, 118 Ma) may also be related to early Kerguelen hot spot activity [e.g., Mahoney et al., 1983; Frey et al., 1996; Kent et al., 1997]. The Kerguelen Plateau, together with the Broken Ridge, cover a geo- graphical area almost one-half the size of Aus- tralia, making it the world’s second biggest LIP [Coffin and Eldholm, 1994]. This LIP has been sampled during Ocean Drilling Program (ODP) Legs 119, 120 and 183; drilling at each of the 11 basement sites (Figure 2) has penetrated between a few tens of meters and a few hundred meters maximum into the upper Kerguelen Plateau and Broken Ridge crust. Argon-argon ages of plateau basalts show a rough, general progression from oldest in the south to youngest in the north [Duncan, 2002]. On the basis of these ages, Cretaceous Kerguelen plume activity appears to have created the Southern Kerguelen Plateau (SKP) and the Central Kerguelen Plateau (CKP) plus Broken Ridge in two distinct pulses at 120–110 Ma and 100–95 Ma, respectively, with less voluminous volcanism on Elan Bank, a western salient, at 108 Ma, occurring between the two major pulses [Duncan, 2002; Coffin et al., 2002]. Subsequent volcanic activity was volu- metrically minor, but aerially extensive and resulted in the formation of the 90E Ridge (80 to 40 Ma) [Duncan, 1978] and Skiff Bank (68 Ma) [Duncan, 2002], followed by the Northern Kerguelen Plateau (NKP; 34 Ma) [Duncan, 2002] and the Kerguelen Archipelago Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 3 of 28(constructed upon the NKP; <30 Ma to 0.1 Ma) [Weis et al., 1993, 1998; Nicolaysen et al., 2000; Doucet et al., 2002]. Additionally, Heard and Mac- Donald Islands are younger than the Kerguelen Archipelago but lie farther to the south, on the CKP [Barling et al., 1994]. The Kerimis seamounts also record recent volcanic activity, of intermediate age between the peaks of activity on Kerguelen and Heard Islands, and are located between these islands [Weis et al., 2002a]. Since Heard and McDonald Islands are volcanically active at the present time, they represent the most likely current location of the Kerguelen hot spot [Coffin et al., 2002]. [6] The Cretaceous Kerguelen Plateau was con- structed primarily within the newly formed eastern Indian Ocean, well after true oceanic crust had been created between Greater India and Australia (anomaly M11, 133 Ma [Lawver et al., 1992; Ramana et al., 1994] using magnetic polarity timescales of [Gradstein et al., 1994]). Seismic Figure 1. Physiographic map of the Indian Ocean and surrounding continents, showing the Kerguelen Plateau (including the various provinces: SKP, CKP and NKP are the Southern, Central and Northern Kerguelen Plateau, respectively, Elan Bank and Skiff Bank), Broken Ridge and 90E Ridge. Precise locations of drill sites for samples analyzed in this study may be found in Figure 2. KA and HI are the Kerguelen Archipelago and Heard and MacDonald Island, respectively (see Figure 3 for a detailed map of the archipelago). Locations for continental tholeiites on the margins of eastern India (Rajmahal Traps) and southwestern Australia (Bunbury basalts) are also noted; these may be related to Kerguelen plume activity [e.g., Frey et al., 1996; Kent et al., 1997]. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 4 of 28evidence suggests that continental crust exists beneath Elan Bank and parts of the SKP [Operto and Charvis, 1995, 1996; Borissova et al., 2003]. The isolation of microcontinents (e.g., Seychelles) may be accomplished by ridge jumps during the interaction between mantle plumes and newly formed continental margins [Mu¨ller et al., 2001]. Therefore, although the tectonic setting of the early Kerguelen Plateau is not well constrained [Coffin et al., 2002; Kent et al., 2002], it may be inferred that Figure 2. (a) Predicted bathymetry (in meters below seafloor) map of the Kerguelen Plateau depicting locations for drill site samples analyzed in this study. The locations for the Kerguelen Archipelago (KA) and Heard and MacDonald Islands (HI) are also indicated. (b) Predicted bathymetric map of the Broken Ridge showing the two drill site locations, Sites 1141 and 1142. Broken Ridge was conjugate with the eastern margin of the Central Kerguelen Plateau until spreading began on the Southeast Indian Ridge, ca. 40 Ma [Munschy et al., 1992]. Presently, Broken Ridge is located off the western coast of Australia (see Figure 1). Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 5 of 28it was most likely situated nearby or on a ridge and proximal to a continental margin [Coffin et al., 2002; Borissova et al., 2003]. The Kerguelen Plateau may have changed to an intraplate setting as India migrated northward and the 90E Ridge formed [Duncan, 1978]. Rifting along the South- east Indian Ridge began 40 Ma and this event separated the Broken Ridge from the Kerguelen Plateau [Munschy et al., 1992]. Relative migration between the Kerguelen hot spot (located on the Antarctic plate) and the Broken Ridge (located on the Australian plate) continues, since 40 Ma, to the present-day by spreading along the Southeast Indian Ridge; during this time, and with increasing distance to the ridge over time, the Kerguelen Archipelago, Kerimis Seamounts and Heard and MacDonald Islands were created. 2. Sampling Strategy [7] The goal of this work is to present the first comprehensive Hf isotope investigation of the Kerguelen LIP. In order to accomplish this objec- tive, we selected samples from every drill site on the Kerguelen Plateau and Broken Ridge, in order to cover its various stages of construction, as well as two volcanic sections on the Kerguelen Archi- pelago (Figures 2 and 3). We also included two rocks, possibly representative of contaminating continental crust, pebbles of a rhyolite and a garnet-biotite gneiss recovered at Site 1137 [Shipboard Scientific Party, 2000]. A large geochemical database exists for the Kerguelen Plateau and Archipelago, including major and trace elements and some radiogenic isotopes (Sr, Nd and Pb). This valuable data permitted selec- tion of the best samples, in terms of geographic distribution, geochemical variability and minimum extents of alteration. More samples were analyzed for Hf isotopic composition in sites that have particularly heterogeneous geochemistry (Sites 747, 749, 750, 1137 and 1140) whereas fewer samples were included for sites with more homo- geneous character (Sites 738, 1136, 1138, 1141, 1142). To complement the investigation of the Hf isotopic compositions on the archipelago from Mattielli et al. [2002], we also selected samples from two volcanic sections from the older part of the archipelago not included in their study, Mont des Ruches and Mont Fontaine. Table 1 lists the Figure 2. (continued) Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 6 of 28selected samples by their geographic province, drill site number or location and references for age and geochemical data. All sample locations are indicated on Figures 2 and 3. 3. Hf Isotopic Analytical Methods [8] Hf separation was carried out on whole rock powders. The general procedure followed that of Blichert-Toft et al. [1997]. Basalt powders were dissolved using sub-boiled HF and HNO3 in closed Savillex Teflon vials. The basalts were not leached with HCl prior to the dissolution because Lu and Hf are fairly immobile during post-eruptive alteration and previous studies have demonstrated leaching to be unnecessary [Kempton et al., 2002; Mattielli et al., 2002]. The rhyolite and gneiss were dissolved in Teflon bombs in an oven at 160C for one week, to ensure dissolution of refractory minerals such as zircon. All dissolved powders were leached in concentrated HF to separate the rare earth elements (which precipitate out as fluoride salts) from the remaining sample. The conditioned supernatant was passed through an anion ex- change column to separate high field strength elements from the bulk matrix. The high field strength element separate was passed through a final, cation exchange column to isolate Hf-Zr. This separate was then run on the Nu Plasma multicollector ICP-MS at the Universite´ Libre de Bruxelles. Both Lu and Yb beams were monitored during each run; the Yb beam was negligible (average of zero for 60 ratio measurements) and the Hf isotopic composition was corrected for Lu interference (although the necessary correction Figure 3. Map of the Kerguelen Archipelago with locations and ages indicated for all samples with Hf data from either this study or from Mattielli et al. [2002]. The Kerguelen Archipelago is located on the Northern Kerguelen Plateau (see Figures 1 and 2a). LMS and UMS are the Lower and Upper Miocene Series, respectively. Ages for the archipelago are from Weis et al. [1993, 1998]; Yang et al. [1998]; Nicolaysen et al. [2000]; Doucet et al. [2002]; and Frey et al. [2002b]. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 7 of 28was generally well within the error of the analysis). This suggests that the HF leaching step effectively removes the rare earth elements as reported by Blichert-Toft et al. [1997]. One hundred and twenty analyses of Hf standard JMC-475 were completed during the days the specific samples reported here were run. The JMC-475 Hf standard was run several times (generally 5 or 6) before beginning a day of analyses, once between each two samples, and a few times after completion of the day in order to accurately assess reproducibility and drift during the day. The average of these analyses for the 176Hf/177Hf Lu-corrected value was 0.282160 ± 22 (2 standard deviations with an external, 2 standard deviation precision of 9 ppm; Figure 4a), which is within the range of previously published values of this standard [Blichert- Toft et al., 1997; Kempton et al., 2000; Chauvel and Blichert-Toft, 2001; Woodhead et al., 2001]. Laboratory techniques and reproducibility may be assessed by our full-procedural duplicate analyses on seven samples, listed in addition to all samples analyzed in Table 2 and Figure 4b. Replicates, a re-run of the same sample solution, are also presented in Table 2 and Figure 4b. Procedural blanks were less than 23 pg of Hf and are negligible compared to the concentrations in the samples (>300 ng). 4. Hf Isotope Results [9] Previous workers have documented the extreme isotopic heterogeneity associated with the Cretaceous-age Kerguelen Plateau and Broken Ridge volcanic rocks [Salters et al., 1992; Mahoney et al., 1995; Frey et al., 2002a; Ingle et al., 2002a; Neal et al., 2002]. This heterogeneity is also present in the Hf isotopic compositions of these rocks that range in eHf(T) from 6.4 to +13 (Table 2; Figure 5). The Cenozoic Kerguelen hot spot products also cover a large isotopic range compared to other oceanic islands [e.g., Storey et al., 1988; Weis et al., 1993; Yang et al., 1998; Doucet et al., 2002] and Hf isotopic compositions do cover a comparably large range (3.3 < eHf (T) < +13), even when compared to that of the Creta- ceous plateau (Table 2; Figure 5). Nevertheless, on the basis of distinct differences in tectonic setting and geochemical characteristics during Cretaceous and Cenozoic times, we divide the results into these two periods. Table 1. Samples Selected for Hf Isotopic Analysis, Listed by Geographic Province of the Kerguelen Plateau, Broken Ridge, and Archipelago Locationa Area (105 km2)b Site Age Ma ± 1s Ref.c Number Lava Type Eruption Style Ref.c Southern KP 4.5 Site 1136 118.9 ± 1.5 1 1 Tholeiitic Subaerial 5 Site 750 112.4 ± 0.4 2 4 Tholeiitic Subaerial 6 Site 749 109.9 ± 1.0 2 5 Tholeiitic Subaerial 6 Site 738 >108 2 1 Tholeiitic Subaerial 7 Elan Bank 1.4 Site 1137 107.7 ± 0.5 1 9 Tholeiitic Subaerial 8 Site 1137 Uncertain 1 Tephrite clastd N/A 9 Site 1137 >109 3 1 Rhyolitic clastd N/A 9 Site 1137 >550 3 1 Gneiss clastd N/A 9 Central KP 4.3 Site 747 Uncertain 3 Tholeiitic Subaerial 6 Site 1138 100.4 ± 0.7 1 1 Tholeiitic Subaerial 5 Broken Ridge 4.8 Site 1141 95.1 ± 0.8 1 1 Transitional Subaerial 5 Site 1142 94.5 ± 0.6 1 1 Transitional Subaerial 5 Skiff Banke 0.53 Site 1139 68.3 ± 0.3 1 2 Alkalic (bimod.) Subaerial 10 Northern KP 3.64 Site 1140 34.3 ± 0.4 1 4 Tholeiitic Submarine 11 KA (incl. with NKP) Mt. des Ruches 28.3 ± 0.9 4 6 Thol. to trans. Subaerial 4 Mt. Fontaine 28.2 ± 0.7 4 5 Thol. to trans. Subaerial 4 aKP, Kerguelen Plateau; KA, Kerguelen Archipelago. bArea as calculated by Coffin et al. [2002]. cReferences for age, petrology, and geochemistry: 1, Duncan [2002]; 2, Coffin et al. [2002]; 3, Nicolaysen et al. [2001]; 4, Doucet et al. [2002]; 5, Neal et al. [2002]; 6, Frey et al. [2002a]; 7, Mahoney et al. [1995]; 8, Ingle et al. [2002a]; 9, Ingle et al. [2002b]; 10, Kieffer et al. [2002]; 11, Weis and Frey [2002]. dClasts were recovered from a fluvial conglomerate between basalt flows (deposited contemporaneously with the lavas). eSamples were selected from the mafic series only. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 8 of 284.1. Hf Isotope Results for the Cretaceous Kerguelen Plateau and Broken Ridge 4.1.1. Southern Kerguelen Plateau Sites 738, 749, 750, and 1136 (119 Ma to 110 Ma) [10] The SKP is believed to be the oldest part of the Kerguelen Plateau [Coffin et al., 2002]. However, the 4 basement drill sites on the SKP span almost 10 myr (Table 1), from 119 Ma (Site 1136) to 110 Ma (Site 749) [Duncan, 2002; Coffin et al., 2002]. The Hf isotopic compositions are highly variable between the different sites, but are also not homogeneous within each site, at least where more than one sample was studied. Site 1136, in the center of the SKP, has an eHf(T) = +5. This value is very different compared to the values for Site 750, Figure 4. (a) Values for Hf Standard JMC 475 analyzed during days on which samples reported in this study were measured. Also shown is the average 176Hf/177Hf value for this standard from our measurements, 0.28160, and the 2 standard deviations, ±22, indicated by the shaded blue area. (b) Measured values for duplicates and replicates analyzed during the course of our analyses. Reproducibility of our chemical procedures may be assessed by the analysis of full procedural duplicates, depicted on the left-hand side of each graph (two graphs are shown to maintain the same scale). On the right-hand side of each graph, replicates are shown to illustrate the reproducibility of the measurements on actual samples on the Nu Plasma (Nu 015) at Universite´ Libre de Bruxelles. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 9 of 28Table 2. Hf Isotopic Composition of Kerguelen Plateau and Archipelago Samplesa Location Sample Age, Ma Lu, ppm Hf, ppm (176Hf/177Hf)T 2s (176Hf/177Hf)T (Hf)T Refb Ref c Kerguelen Plateau Site 1136 18R-3-Piece 2 119 0.39 2.54 0.282887 5 0.28284 5 1 6 Site 750 15R-1-103-108 112 0.32 0.84 0.283172 10 0.28306 13 2 7 16R-2-50-56 112 0.32 0.78 0.283175 13 0.28305 12 2 7 16R-4-42-47 112 0.33 0.88 0.283181 13 0.28307 13 2 7 17R-3-23-27 112 0.41 1.43 0.283041 9 0.28296 8.9 2 7 Site 738 34R-1-85-86 112 0.44 3.89 0.282574 12 2 8 34R-1-85-86 rep 112 0.282564 12 2 8 34R-1-85-86 dupl 112 0.282554 10 0.28252 6.4 2 8 Site 749 12R-3-55-60 110 0.24 1.59 0.283013 8 0.28297 9.4 2 7 12R-5-76-79 110 0.35 2.37 0.282879 6 0.28284 4.7 2 7 15R-4-36-40 110 0.4 2.43 0.282864 6 0.28282 4 2 7 16R-1-22-29 110 0.43 2.75 0.283013 10 0.28297 9.3 2 7 16R-6-50-60 110 0.21 1.24 0.283065 10 0.28302 11 2 7 Site 1137-U 25R-1-64-72 108 0.31 5.25 0.28272 3 0.2827 0.1 1 9 26R-1-31-43 108 0.33 4.77 0.2827 15 1 9 26R-1-31-43 rep 108 0.282713 10 0.28269 0.4 1 9 27R-4-57-63 108 0.35 5.53 0.282724 8 0.28271 0 1 9 31R-3-5-10 108 0.27 3.96 0.282738 7 1 9 31R-3-5-10 dupl 108 0.282746 6 0.28273 0.8 1 9 33R-1-53-60 108 0.29 4.88 0.282723 8 1 9 33R-1-53-60 rep 108 0.282722 12 1 9 33R-1-53-60 108 0.28273 6 1 9 33R-1-53-60 rep 108 0.282738 6 1 9 33R-1-53-60 dupl 108 0.282738 6 0.28272 0.5 1 9 33R-1-53-60 dupl rep 108 0.28273 6 1 9 Site 1137 1137-34R-4 rhyolite 113 0.56 18.7 0.282505 4 0.2825 7.4 3 3 clasts 1137-35R-2-gneiss 550 0.43 9.8 0.282192 6 0.28218 19 4 3 1137-35R-2 rep 550 0.282197 7 4 3 1137-36R-2 tephrite – 0.41 11.1 0.282616 6 0.28261 3.5 - 3 Site 1137-L 38R-4-71-77 108 0.33 4.57 0.282722 6 1 9 38R-4-71-77 dupl 108 0.282716 5 0.2827 0.4 1 9 40R-3-117-124 108 0.35 4.64 0.282682 14 1 9 40R-3-117-124 dupl 108 0.282679 5 0.28266 1.7 1 9 41R-1-3-10 108 0.31 4.82 0.28269 7 0.28267 1.2 1 9 41R-1-3-10 rep 108 0.282677 8 1 9 46R-1-51-56 108 0.39 4.96 0.28277 17 1 9 46R-1-51-56 dupl 108 0.282752 6 0.28273 0.9 1 9 Site 1138 84R-5-Piece 3 100 0.61 4.97 0.282921 13 0.28289 6.3 1 6 Site 747 12R-4-53-56 100 0.36 2.69 0.282683 6 0.28265 2.3 2 7 14R-1-31-35 100 0.33 3.29 0.282651 6 0.28263 3.2 2 7 15R-1-15-19 100 0.27 2.61 0.282657 6 0.28263 3 2 7 Site 1141 21R-2-Piece 1 94.7 0.35 3.56 0.282875 4 0.28285 4.8 1 6 Site 1142 10R-3-Piece 1 94.7 0.57 5.44 0.282928 7 0.2829 6.7 1 6 Site 1139 64R-1-136-146 68.3 0.54 8.57 0.282703 8 0.28269 1.3 1 10 64R-4-103-115 68.3 0.48 6.99 0.282696 8 0.28268 1.6 1 10 64R-4-103-115 rep 68.3 0.282703 9 1 10 Site 1140 25R-6-51-56 34.3 0.46 2.39 0.283175 5 0.28316 14 1 11 25R-6-51-56 dupl 34.3 0.283185 6 1 11 31R-1-57-67 34.3 0.47 4.03 0.283033 6 0.28302 9.6 1 11 31R-1-57-67 rep 34.3 0.283027 6 1 11 32R-2-45-50 34.3 0.52 5.46 0.282995 6 0.28299 8.3 1 11 32R-2-45-50 rep 34.3 0.282999 6 1 11 34R-6-33-39 34.3 0.45 2.16 0.283126 6 0.28311 13 1 11 Kerguelen Archipelago M. d. Ruches BY96-24 28.2 0.27 2.45 0.282911 5 0.2829 5.2 5 5 BY96-27 28.2 0.23 3.1 0.282861 5 0.28286 3.6 5 5 BY96-31 28.2 0.4 5.89 0.282812 4 0.28281 1.8 5 5 BY96-31 rep 28.2 0.282836 5 5 5 BY96-34 28.2 0.23 3.55 0.282731 5 0.28273 1 5 5 Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 10 of 28on the eastern flank of the SKP, that are much more radiogenic (+8.9 > eHf(T) > +13). Site 749, on the western flank of the SKP, has eHf(T) intermediate to those of Sites 1136 and 750 (+4.0  +11). In strong contrast, a basalt from the southernmost site on the SKP, Site 738, has an eHf (T) of 6.4, much less radiogenic than the other sites from the SKP. Sites 749 and 750 have the highest documented eHf (T) and eNd(T) of all basalts studied from the Creta- ceous Kerguelen Plateau (Figures 5 and 6); how- ever, they have very different (87Sr/86Sr)T and are distinct in their Pb isotopic compositions as well (Figure 7) [Frey et al., 2002a]. Site 1136 lies at the unradiogenic end of the Site 749 range of eHf (T) values (Figure 5). In plots of eHf (T) vs. eNd(T) or (87Sr/86Sr)T, the Site 1136 sample is displaced slightly toward more radiogenic (87Sr/86Sr) and less radiogenic eNd(T) than the Site 749 samples (Figure 6); Site 1136 also has Pb isotopic compo- sitions quite comparable to, but slightly more radiogenic in 207Pb/204Pb and 208Pb/204Pb than those of Site 749 (Figure 7) [Neal et al., 2002]. 4.1.2. Elan Bank Site 1137 (108 Ma) [11] The 108 Ma basalts [Duncan, 2002] from Site 1137 have been divided into a lower and upper group distinguished by distinctive isotopic values reflecting more extensive contamination by upper continental crust in the lower basalt group [Weis et al., 2001; Ingle et al., 2002a]. Hf isotopes for these basalts cluster around chondritic compo- sitions, with the upper group basalts having eHf (T) from 0.4 to +0.8, similar to the lower group with eHf(T) from 1.7 to +0.9 (Table 2). Site 1137 basalts have eHf(T) values that do not overlap with those of any basalts from the SKP (Figure 5), and this is coherent with the fact that the Site 1137 basalts are completely out of the range of eNd(T) measured in basalts from Sites 749, 750 and 1136 (Figure 6). The Site 1137 lower basalts have (87Sr/86Sr)T comparable to those of the Site 750 basalts (Figure 6). One of the most characteristic features of Site 1137 basalts is a vertical trend in Pb-Pb isotope plots, with variable 207Pb/204Pb and 208Pb/204Pb for nearly constant 206Pb/204Pb (Figure 7) [Ingle et al., 2002a]. Site 749 basalts form a similar type of trend (Figure 7) [Frey et al., 2002a]. The three clasts from the fluvial conglomerate that were analyzed include a tephrite, a rhyolite and a gneiss which have eHf (T) of 3.5, 7.4 and 19, respectively. All values of eHf(T) for the clasts Table 2. (continued) Location Sample Age, Ma Lu, ppm Hf, ppm (176Hf/177Hf)T 2s (176Hf/177Hf)T (Hf)T Refb Ref c BY96-45 28.2 0.32 4 0.282965 6 0.28296 7.2 5 5 BY96-46 28.2 0.46 7.35 0.282905 6 5 5 BY96-46 rep 28.2 0.282882 6 5 5 BY96-46 rep 28.2 0.282891 5 5 5 BY96-46 rep 28.2 0.282891 5 0.28288 4.6 5 5 M. Fontaine BY96-80 28 0.26 3.1 0.282908 7 5 5 BY96-80 dupl 28 0.282927 6 0.28292 5.8 5 5 BY96-82 28 0.29 2.89 0.282974 5 0.28297 7.5 5 5 BY96-86 28 0.24 2.53 0.282873 4 0.28287 3.9 5 5 BY96-89 28 0.21 1.29 0.282987 7 0.28297 7.8 5 5 BY96-98 28 0.24 2.91 0.282974 5 0.28297 7.5 5 5 aLu and Hf concentration measurements were made by ICP-MS except for Sites 1137 and 738 (INAA). Approximate ages for the clasts (recovered from a conglomerate deposited between lava flows at Site 1137) are listed, but these samples are age corrected only to the age of the basalt flows (108 Ma) recovered at Site 1137. Reported 2s applies to the sixth decimal place. Site 1137-U indicates ‘‘upper group,’’ and -L is ‘‘lower group.’’ A sample with ‘‘dupl’’ listed next to the sample name is a full procedural duplicate, while a sample listed as ‘‘rep’’ is a repeat analysis on the MC-ICP-MS using the same solution as the sample listed above. Age corrections are only applied to the analysis with the lowest reported error in the case of duplicates and replicates. Hf(T) = [(176Hf/177Hf)sample/(176Hf/177Hf)CHUR)  1]*104, where both the sample and CHUR values are corrected to time T; CHUR values ((176Hf/177Hf)Today = 0.282772, 147Sm/144Nd = 0.0332) from Blichert-Toft and Albare`de [1997] and Hf decay constant from Scherer et al. [2001]. bThe references for the listed age are as follows: 1, Duncan [2002]; 2, Coffin et al. [2002]; 3, Ingle et al. [2002b]; 4, Nicolaysen et al. [2001]; 5, Doucet et al. [2002]. cThe references for the Lu and Hf concentrations are as follows: 1, Duncan [2002]; 2, Coffin et al. [2002]; 3, Ingle et al. [2002b]; 4, Nicolaysen et al. [2001]; 5, Doucet et al. [2002]; 6, Neal et al. [2002]; 7, Frey et al. [2002a]; 8, Mahoney et al. [1995]; 9, Ingle et al. [2002a]; 10, Kieffer et al. [2002]; 11, Weis and Frey [2002]. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 11 of 28are much less radiogenic than those of the basalt flows (Table 2). 4.1.3. Central Kerguelen Plateau Sites 747 and 1138 (100 Ma) and Broken Ridge Sites 1141 and 1142 (95 Ma) [12] The CKP and Broken Ridge formed between 100–95 Ma and were conjugate until around 40 Ma [Munschy et al., 1992]. The CKP has been sampled at Site 747 in the south and at Site 1138 slightly farther to the north. The single sample from Site 1138 has eHf(T) of +6.3, this is a very different value from those obtained for the 3 samples from Site 747 which range in eHf(T) from 3.2 to 2.3 (Table 2, Figure 5). The Site 1138 sample has an eHf(T), eNd(T) and (87Sr/86Sr)T similar to those values found in the Site 1136 sample (Figure 6). The 206Pb/204Pb for the Site 1138 basalts is 18, similar to the values for basalts from Sites 749, 1136 and 1137; however, Site 1136 basalts do not form a vertical distribution in Pb-Pb plots. On the Broken Ridge, both Sites 1141 and 1142 were drilled from the southeastern margin. Despite their proximity, these sites have slightly different Hf isotopic compositions: Site 1141 has eHf(T) = +4.8 and Site 1142 has eHf (T) = +6.7 (Table 2, Figure 5). The two samples from Sites 1141 and 1142 are also slightly different from each other in eNd(T) and (87Sr/86Sr)T but Pb isotopic composi- tions overlap in some Site 1141 and 1142 samples and are similar to Pb isotopes in Site 1137 basalts (Figure 7) [Neal et al., 2002]. 4.1.4. Skiff Bank Site 1139 (68 Ma) [13] The small topographic high on the western part of the NKP, Skiff Bank, was drilled at Site 1139. It Figure 5. eHf(T) vs. relative stratigraphic age (true age noted next to each locale) for all Kerguelen Plateau and Kerguelen Archipelago samples analyzed during this study or during the study of Mattielli et al. [2002] (archipelago only). Hf isotopic composition is given in epsilon units eHf and age-corrected (T) for the Ar-Ar ages (references given in Table 1 and Figure 3 caption). The horizontal line divides samples from the Cretaceous Kerguelen Plateau (older than 68 Ma) and the Cenozoic Kerguelen Plateau and Kerguelen Archipelago. Note that there is no trend of either increasing or decreasing eHf (T) in the Cretaceous samples but that a general trend of decreasing eHf (T) is present in the Cenozoic samples. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 12 of 28has been dated as Late Cretaceous in age, 68 Ma [Duncan, 2002], and it therefore erupted during the timing of the 90E Ridge construction (40 to 80 Ma) [Duncan, 1978]. Its relationship to the remainder of the NKP is unknown as the northern part of the NKP (Site 1140) appears to be signifi- cantly younger (34 Ma; see discussion below) [Duncan, 2002]. The two Site 1139 samples have eHf (T) =1.6 and1.3 (Table 2, Figure 5). Basalts from Site 1139 have similar isotopic characteristics to basalts from Site 747 from the CKP, either overlapping in some plots or lying very close to each other in other plots (Figures 5, 6, and 7) [Kieffer et al., 2002; Frey et al., 2002a; this study]. 4.2. Hf Isotopic Results for the Cenozoic Kerguelen Plateau and Kerguelen Archipelago [14] The NKP was drilled on its northernmost margin at Site 1140, and these basalts are the youngest yet recovered from the Kerguelen Plateau (34 Ma) [Duncan, 2002]. The 4 samples ana- lyzed from this site range in eHf(T) from +8.3 to +14 (Table 2; Figure 5). Some Site 1140 basalts extend into the Indian MORB field in eHf(T), eNd(T) and Pb isotopes and only have (87Sr/86Sr)T slightly outside of the reported field for Indian MORB (Figures 6 and 7) [Weis and Frey, 2002; this study]. We have plotted the new Site 1140 and the new archipelago data from Mont Fontaine and Mont des Ruches as individual points but have divided the previously reported Kerguelen Archi- pelago data into three groups (plotted as fields in the various diagrams): basalts >26 Ma, 24–25 Ma basalts and the <10 Ma lavas. [15] The Kerguelen Archipelago was constructed upon the NKP and there is a general age progres- sion from the northwest (30 Ma) to the southeast (24 Ma) but some younger lavas and dikes (>0.1 Ma) are also present in the southeast [e.g., Figure 6. (a) eHf (T) vs. eNd(T) and (b) eHf (T) vs. (87Sr/86Sr)T for samples from the Kerguelen Plateau and Archipelago. Samples analyzed during this study are shown as the individual points and are included in the key. Kerguelen Archipelago (KA) samples analyzed by Mattielli et al. [2002] are grouped in the different green-shaded fields by age and composition. The Indian MORB field is data from Chauvel and Blichert-Toft [2001], and references therein. The new data on the Kerguelen Plateau samples analyzed in this study greatly extends the range of eHf (T) associated with the archipelago. The Site 738 sample does not fit in the plotted area but has eHf (T) = 6.4, eNd(T) = 7.6 and (87Sr/86Sr)T = 0.70910 (Sr, Nd and Pb isotopes from [Mahoney et al., 1995]). The new data for the Kerguelen Archipelago mostly fall in the field of the archipelago corresponding to their age, but one sample from Mont des Ruches falls in the field associated with the youngest lavas on the archipelago, and near plateau samples from Sites 1137, 747 and 1139. The juvenile rock array from Vervoort and Blichert-Toft [1999] is shown for reference; note that most Kerguelen samples, except for those most contaminated by continental crust, lie to the left of this array. Nd and Sr isotopic data for the Kerguelen Plateau sites: Mahoney et al. [1995]; Frey et al. [2002a]; Ingle et al. [2002a]; Kieffer et al. [2002]; Neal et al. [2002]; Weis and Frey [2002] and Kerguelen Archipelago: Weis et al. [1993, 1998]; D. Weis, unpublished data (2001); Yang et al. [1998]; Doucet et al. [2002]; Frey et al. [2002b]. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 13 of 28Storey et al., 1988; Weis et al., 1993, 1998; Yang et al., 1998; Nicolaysen et al., 2000; Doucet et al., 2002]. Previous workers have noted an isotopic distinction in the Kerguelen Archipelago lavas between the >26 Ma sections, that are tholeiitic to transitional in composition, the 24–25 Ma mildly alkalic sections and the <10 Ma lavas, that are highly alkalic [e.g., Gautier et al., 1990;Weis et al., 1998; Damasceno et al., 2002]. Mattielli et al. [2002] reported Hf isotopes for 39 samples from the Figure 7. (a) Measured 208Pb/204Pb vs. 206Pb/204Pb and (b) Measured 207Pb/204Pb vs. 206Pb/204Pb for rocks from the Kerguelen Plateau and Kerguelen Archipelago. All sites from the plateau are grouped and labeled in individual fields for clarity. Kerguelen Archipelago samples are divided into the fields as presented in Figure 6 and individual Mont des Ruches (blue X) and Mont Fontaine (pink asterisk) samples are shown as their Hf isotopic compositions were analyzed during this study (all Pb isotopic compositions from Doucet et al. [2002]). The large range covered by Kerguelen Plateau and Archipelago data is reinforced by the plotting of the mantle end-members (gray fields) [Hart, 1988]. Note the displacement of the Indian Ocean MORB and Kerguelen basalts to the left of the northern hemisphere reference line (NHRL as defined by Hart [1984]). All Pb isotopic references for Kerguelen samples are as in Figure 6. Indian MORB data sources: Hamelin et al. [1985];Michard et al. [1986]; Price et al. [1986]; Dosso et al. [1988]; Pyle et al. [1992]; Klein et al. [1988]; Klein et al. [1991];Mahoney et al. [1989, 1992]; White [1993]; Schiano et al. [1997]. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 14 of 28Kerguelen Archipelago from all 3 groups. We analyzed 11 additional samples for Hf isotopes from 2 previously unreported volcanic sections of the >26 Ma group, the Mont des Ruches and Mont Fontaine basaltic sections from the northwestern part of the archipelago (age = 28 Ma and other isotopic values from [Doucet et al., 2002]). The Mont des Ruches samples are quite variable and range in eHf(T) from 1.0 to +7.2 and the Mont Fontaine samples range in eHf (T) from +3.9 to +7.8 (Table 2; Figure 5). Although the majority of the studied samples fall within the reported range for the >26 Ma lavas from the Kerguelen Archi- pelago, one sample from Mont des Ruches lies well out of this field in eHf(T) vs. eNd(T) or (87Sr/86Sr)T (Figure 6) and closer to the <10 Ma lava values. Additionally, one Mont Fontaine sample is out of the range for the other similarly aged Kerguelen Archipelago samples, having slightly higher (87Sr/86Sr)T (Figure 6). 5. Identification of the Mantle Sources and Continental Contaminants Recorded in the Kerguelen Plateau– Broken Ridge and Kerguelen Archipelago Rocks 5.1. Cretaceous Kerguelen Plateau–Broken Ridge Sources and Contaminants [16] The studied samples from the Cretaceous Kerguelen Plateau–Broken Ridge alone, excluding the continental clasts from Site 1137, span >20 eHf units (Table 2; Figure 5). The large differences in the geochemical characteristics between basalts from the same geographic area and age leads us to group the samples on the basis of geochemical characteristics. Sites sharing some geochemical characteristics may then be grouped together and possible mechanisms to account for the diversity of the Hf isotopic compositions examined. We group the sites on the basis of their Pb isotopic compo- sitions as the Pb-Pb plots provide a useful tool to discriminate the various Kerguelen Plateau–Bro- ken Ridge sites that have similar isotopic character- istics but that are widely dispersed geographically. Then, we incorporate our Hf isotopic data to yield new insights to the origin of Kerguelen Plateau– Broken Ridge mantle sources and shallow-level interactions. 5.1.1. Sites With Basalts Having Limited 206Pb/204Pb Variation From 17.9 to 18.1 [17] This group includes basalts from various areas of the Kerguelen Plateau including Sites 749 and 1136 (SKP), Site 1137 (Elan Bank), Site 1138 (CKP) and Sites 1141 and 1142 (Broken Ridge). We use measured Pb isotopic ratios because U, Th and Pb concentration data are not available for all samples and because the total shift in Pb isotopic compositions resulting from age-correction is not significant for most of these samples. These sites have a narrow range in 206Pb/204Pb (17.9–18.1), cover a wide range in all other isotopic systems and form a diffuse, vertical trend in Pb-Pb isotope diagrams (Figures 6, 7, and 8) [Frey et al., 2002a; Ingle et al., 2002a; Neal et al., 2002; this study]. Binary mixing in Pb-Pb diagrams results in linear correlations and so the Pb compositions of these samples appear to originate from a common source having 206Pb/204Pb  18 and relatively low 207Pb/204Pb and 208Pb/204Pb that has been mixed with a contaminant also having 206Pb/204Pb  18 but high 207Pb/204Pb and 208Pb/204Pb. Site 1138 may be the best representation of the mantle component during Cretaceous time, as suggested by Neal et al. [2002]. Site 1138 basalts have homogeneous Sr, Nd and Pb isotopic values and these values are characteristic of the primitive mantle [Neal et al., 2002]. The eHf(T) for the Site 1138 sample is well within the range of many Kerguelen Plateau, Broken Ridge and Kerguelen Archipelago rocks. These basalts also have no trace element characteristics which reflect a continental lithosphere signature (e.g., La/Nb  1, La/Ta  1) [Neal et al., 2002]. The Site 1138 sample has a moderately depleted signature in Hf isotopes, with an eHf(T) = +6.4, despite the primitive mantle-like Sr-Nd-Pb isotopic values (Figures 6 and 8). A low eNd(T) for a given eHf(T) results in displacement to the left in the eHf(T) vs. eNd(T) diagram relative to the mantle array [Vervoort and Blichert-Toft, 1999]. This quality in Hf-Nd isotope space is also characteristic of most Kerguelen Archipelago lavas [Mattielli et al., 2002] and of Indian MORB [Chauvel and Blichert-Toft, 2001]. In summary, Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 15 of 28Site 1138 samples have (1) homogeneous Sr and Nd isotopic compositions typical of the values at many sites on the Kerguelen Plateau–Broken Ridge, (2) 206Pb/204Pb values of 18, (3) high eHf(T) for their eNd(T) relative to non-Indian Ocean basalts samples, (4) trace element characteristics consistent with derivation from an enriched reser- voir and (5) no signature indicative of crustal contamination [Neal et al., 2002; this study]. Therefore, at the present time, these samples ap- pear to be the best representation of the mantle source available for generating most Cretaceous Kerguelen Plateau–Broken Ridge basalts and may be inferred to represent the composition available in the Kerguelen plume head. [18] The contaminant that would best account for the full isotopic variation in the other basalts from this group is upper continental crust [Ingle et al., 2002a]. This hypothesis also is consistent with the Hf isotopic data: assimilation of upper crust having 206Pb/204Pb  18, high 207Pb/204Pb and 208Pb/204Pb, subchondritic eHf and eNd and radiogenic 87Sr/86Sr. Minor input from a compo- nent more depleted than that of the Site 1138 basalts is required to explain the Sr, Nd, Hf and Pb isotopic compositions of Site 749 (SKP) basalts [Frey et al., 2002a; Ingle et al., 2002a; this study]. It is possible that this component could be the depleted, upper Indian MORB mantle (Figures 6–8). Figure 8. eHf(T) vs. measured 206Pb/204Pb (a, d), 207Pb/204Pb (b) and 208Pb/204Pb (c) for rocks from the Kerguelen Plateau and Archipelago. The field for Indian Ocean MORB is also shown: [Salters, 1996; Chauvel and Blichert-Toft, 2001, and references therein]. Kerguelen Plateau sites are listed in the key individually (squares = SKP; diamonds = Elan Bank and CKP; circles = Broken Ridge; triangles = Skiff Bank and NKP) and archipelago samples are divided into age and composition groups with the exception of the new data for Mont des Ruches and Fontaine which are plotted individually. Kerimis volcanic rocks, recovered from seamounts between the Kerguelen Archipelago and Heard Island are also plotted [Weis et al., 2002a]. In (a) fields are drawn around sites with multiple samples analyzed for Hf isotopes. Figure 8a is extended in (d) in order to show the mantle end-members (Pb from Hart, [1988]; Hf from Salters and White [1998]) for comparison to the Kerguelen samples, and to highlight the absence of any HIMU-like component in any Kerguelen rocks. Note the extreme isotopic values in the gneiss clast from Site 1137 on Elan Banks. References for Kerguelen isotopic data are as in Figures 6 and 7. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 16 of 285.1.2. Sites With Basalts Having Unusually Low 206Pb/204Pb (17.3 to 17.8) [19] This group includes Sites 738 and 750 (SKP), Site 747 (CKP) and Site 1139 (Skiff Bank). Basalts from these sites have low 206Pb/204Pb (<17.8) but have variable isotopic compositions in all other systems. Sites 747 (CKP) and 1139 (Skiff Bank) have generally comparable Sr, Nd, Pb and Hf isotopic values [Frey et al., 2002a; Kieffer et al., 2002; this study]. Basalts from Site 750 (SKP) overlap with Sites 747 and 1139 in all Pb isotopic composi- tions but have other isotopic values quite distinct from those characterizing Sites 747 and 1139, and fall much closer to the Indian MORB field (Figures 6 and 8). Site 738 basalts have signifi- cantly higher 208Pb/204Pb (>38.9) and 207Pb/204Pb (>15.7) than Site 747, 750 and 1139 basalts [Mahoney et al., 1995; Frey et al., 2002a; Kieffer et al., 2002]. The diversity of isotopic composi- tions in this low 206Pb/204Pb group suggests that either multiple mantle components or multiple continental contaminants (or both) are necessary to explain their variation. [20] The origin of the low-206Pb/204Pb component in Site 747 and 750 basalts has been explained by incorporation of delaminated lower continental crust in the Kerguelen plume, perhaps as a result of a ridge-centered tectonic environment, where the lithosphere would be thin and exposed to the hot mantle plume [Frey et al., 2002a]. Basalts from Sites 747 and 1139 share comparable Sr, Nd, Pb and Hf isotopic compositions and it is reasonable to assume that their mantle sources and continental contaminants were similar. If Site 747 and 1139 basalts originated from melting in the plume head, the continental contaminant must have had low Pb, Nd and Hf and high Sr isotopic compositions [Frey et al., 2002a; this study]. These basalts could therefore be mixtures between the plume head source and a lower crustal con- taminant, comparable to the upper crust contam- inant discussed in the previous section in all isotopic values except for Pb. Site 1139 basalts appear to be less contaminated than Site 747 basalts, since the Site 747 basalts extend to lower Pb isotopic compositions. [21] Site 738 and 750 basalts must have developed their low-206Pb/204Pb isotopic compositions for different reasons than those applicable to Site 747 and 1139 samples. This is clear because Site 750 basalts are very different from Site 747 and 1139 basalts in eHf(T) and eNd(T), having values consid- erably more MORB-like than those ratios in Site 747 basalts [Frey et al., 2002a; this study]. Frey et al. [2002a] argued that Site 750 basalts may have been generated by the same mantle component as that of Site 747 basalts (and by inference, Site 1139) but that Site 747 and 750 basalts experienced contamination by different types of lower continental crust, to account for the discrepancies between the Sr and Nd isotopic compositions in basalts from these two sites. However, Site 747 and 1139 basalts are deflected toward higher 87Sr/86Sr, and lower eHf and eNd, similar to those values in Site 1137 basalts, relative to the plume head composition (Site 1138). Lower crust may generally have lower Pb isotopic ratios than upper crust because of preferential removal of U, Th and Rb relative to their daughter products during partial melting [e.g., Doe and Zartman, 1979]. However, Nd and Hf isotopic compositions in the lower crust should not be significantly different than the those found in the upper crust [Vervoort et al., 2000]. Thus the presence of the high Hf and high Nd isotopic values in Site 750 is inconsistent with contami- nation of enriched, plume-derived magmas by any likely lower continental crust component. An additional constraint is that the high (87Sr/86Sr)T in Site 750 basalts (0.705) requires that a con- taminant with sufficiently high Sr concentrations was involved. [22] An alternative explanation is that the Site 738 and 750 basalts result from magmas derived from a depleted source having extremely unradiogenic Pb isotopic compositions, that subsequently assimilate upper continental crust, characterized by high Sr concentrations, high 87Sr/86Sr, 207Pb/204Pb and 208Pb/204Pb, 206Pb/204Pb  18 and unradiogenic eHf(T) and eNd(T). This upper crust contaminant could be quite comparable to that necessary in generating the Site 1137 basalts. Binary mixing between a depleted source with low-206Pb/204Pb by Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 17 of 28up to 10% of this upper crust could explain the distinctive isotopic characteristics of Site 750 basalts (Figure 9). One Site 750 sample falls below the mixing curve in 176Hf/177Hf vs. 87Sr/86Sr but this may be a result of alteration in this sample (>7% loss on ignition; [Frey et al., 2002a]). Site 738 basalts reflect extensive contamination [Mahoney et al., 1995]; their isotopic compositions are probably nearly equivalent to those of the contaminant. Although there is no sample to represent the proposed crustal contaminant, it most likely has isotopic characteristics slightly more extreme than those of the Site 738 basalts. 5.2. Cenozoic Kerguelen Plateau and Kerguelen Archipelago Mantle Sources and Possible Contaminants [23] During the Cenozoic, spreading on the South- east Indian Ridge changed the tectonic setting of the Kerguelen plume from ridge-centered (or nearby), around 40 Ma, to intraplate [Munschy et al., 1992]. Of all the Cenozoic Kerguelen samples, an ‘‘extreme’’ in Pb isotopic compositions is present in the 24–25 Ma mildly alkalic basalts from Mont Crozier [Mattielli et al., 2002; Weis et al., 2002b]. The age of Mont Crozier samples also coincides with a change in depth of melting (alkali index increases) and style of magmatism (flood basalts to plugs and dikes) between the older, >26 Ma tholei- itic to transitional, and the younger,10 Ma, highly alkalic groups on the archipelago [Damasceno et al., 2002; J. S. Scoates et al., manuscript in prepa- ration, 2003]. The Mont Crozier basalts, in addition to having relatively high 206Pb/204Pb (18.6), are characterized by Sr, Nd and Hf isotopic composi- tions intermediate to those found in the >26 Ma group and the 10 Ma group [Weis et al., 2002b; Mattielli et al., 2002]. We discuss the other two groups from the Cenozoic period beginning with this apparent source composition, represented by the 24–25 Ma Mont Crozier basalts. 5.2.1. Tholeiitic to Transitional >26 Ma Basalts From the Kerguelen Plateau and Kerguelen Archipelago [24] Site 1140 basalts from the NKP have Sr, Nd, Hf and Pb isotopic compositions previously reported to reflect binary mixing between the Kerguelen plume and the ambient depleted Indian MORB source [Mattielli et al., 2000; Weis and Frey, 2002]. Most of the older Kerguelen Archi- pelago basalts have also been attributed to mixing between a depleted mantle component and the enriched Kerguelen plume source, with the depleted component variably represented by (1) normal Southeast Indian Ocean depleted upper mantle [e.g., Storey et al., 1988; Gautier et al., 1990; Doucet et al., 2002], (2) assimilation of depleted gabbroic cumulates [Yang et al., 1998] or (3) an intrinsic depleted component in the Kerguelen plume source [Frey et al., 2002b]. Recently, Fitton et al. [2003] have discussed the depleted component in the Iceland volcanics, particularly whether this component is intrinsic to the plume or results from interaction with the surrounding North Atlantic MORB source mantle [Hanan et al., 2000]. Using high-precision Hf-Nd isotopic analyses only, Fitton et al. [2003] were able to delineate separate fields for the normal Atlantic MORB and the samples from Iceland containing a depleted component, suggesting that the depleted component was not the same as the Atlantic MORB source. High-precision Hf-Nd isotopic data for the Indian MORB from Chauvel and Blichert-Toft [2001] do, however, overlap with our data for the most depleted samples from Kerguelen Site 1140 (Figure 6). This may suggest that the depleted component originates from the depleted, upper Indian Ocean mantle. An addi- tional argument in favor of the depleted compo- nent originating from the ambient upper mantle, is that the depleted signature in Cenozoic Kerguelen basalts decreases over time (decreasing Nd isoto- pic composition and increasing Sr and Pb isotopic compositions) [e.g., Storey et al., 1988; Gautier et al., 1990; Doucet et al., 2002]. Furthermore, unlike Iceland, which is situated on the ridge axis, the Kerguelen Archipelago has changed from ridge-centered to intraplate with increasing distance between the Kerguelen Archipelago and the Southeast Indian Ridge from 40 Ma to the present [Munschy et al., 1992; Royer and Sandwell, 1989; Coffin et al., 2002]. In summary, (1) the growing distance between the ridge and the archi- pelago correlates with the apparently decreasing amount of necessary depleted component and Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 18 of 28Figure 9. Mixing diagrams for SKP Site 750 and NKP Site 1140 in the isotopic systems of (176Hf/177Hf)T vs. (87Sr/86Sr)T (a), (176Hf/177Hf)T vs. (143Nd/144Nd)T (b), 206Pb/204Pb vs. 208Pb/204Pb (c) and (176Hf/177Hf)T vs. 206Pb/204Pb (d). Site 750 samples have unusual characteristics including very unradiogenic Pb isotopic compositions and radiogenic Hf, Sr and Nd isotopic compositions and generally do not fall near the field of other Kerguelen Plateau rocks [Frey et al., 2002a; this study]. We calculate bulk mixing (red line, small yellow squares) between an unradiogenic depleted mantle component, chosen from within the Indian MORB field, and an upper crust-like component, here represented by the Site 738 basalts from the SKP, thought to be highly contaminated by continental material [Mahoney et al., 1995]. Although the Site 738 sample could not be representative of the precise continental end-member involved in such a mix, and thus the proportions of each are not accurate, it should lie near the position of the crustal end-member. Site 1140 samples have been previously interpreted as representing bulk mixing between a MORB component from the southeast Indian Ridge and the Mont Crozier samples, which form an end-member in Pb and Hf isotopic compositions for the Kerguelen Archipelago rocks [Mattielli et al., 2000; Weis and Frey, 2002]. In most isotopic systems, this mixing equation works rather well, with the possible exception of the 206Pb/204Pb vs. 208Pb/204Pb diagram where the Site 1140 samples lie somewhat below the projected mix (further discussion of this mixture may be found in [Weis and Frey, 2002]. It is important to note that, particularly in (c) and (d), the mixture between the Mont Crozier samples and the depleted component is insufficient to fully explain most of the archipelago data including the new reported Hf isotopic data in this study. For the young archipelago lavas (the <10 Ma group) this has been previously explained by either possible contamination by plateau material or a distinct heterogeneity in the Kerguelen plume [Mattielli et al., 2002]. References are the same as given in Figures 6 and 7. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 19 of 28(2) high-precision Hf-Nd isotopic compositions for Indian Ocean MORB overlap with those from the most depleted Kerguelen basalts (i.e., Site 1140). Therefore the simplest explanation for the depleted component in the Kerguelen Archipelago lavas is that it is derived from local depleted MORB mantle, the same source as that available in generating the Southeast Indian Ridge MORB [Mattielli et al., 2000; Doucet et al., 2002; Weis and Frey, 2002]. [25] All of the sampled >26 Ma sections on the Kerguelen Archipelago have some basalts that reflect input from the depleted component in Hf isotopes [Mattielli et al., 2002; this study], as has been also suggested from studies of the Sr-Nd-Pb isotopes [e.g., Gautier et al., 1990; Yang et al., 1998; Doucet et al., 2002]. In these basalts, an additional, minor component is necessary to explain the full range of both Hf and Pb isotopic compositions (Figures 7 and 8). Binary mixing trends, between the Kerguelen plume and the depleted MORB mantle, account for the Site 1140 basalts but some of the basalts from the >26 Ma group on the archipelago fall in a field deflected toward the Kerguelen Plateau basalts that have 206Pb/204Pb  18. It is possible that magmas derived from variable proportions of Kerguelen plume tail and MORB mantle assimi- lated small amounts of the overlying Cretaceous Kerguelen Plateau lithosphere prior to eruption. This additional but minor plateau component more completely accounts for the range of isotopic compositions in the Kerguelen Archipelago basalts than binary mixing between plume tail and depleted components alone. 5.2.2. Mantle Sources and Contamination in Alkalic Lavas From the <10 Ma Part of the Kerguelen Archipelago [26] As mentioned above, limited assimilation of Cretaceous Kerguelen Plateau lithosphere by plume (± depleted MORB) derived magmas accounts for much of the variation in isotopic data on the Kerguelen Archipelago. However, in the young (<10 Ma), evolved alkalic lavas from the Archipelago, the isotopic characteristics do not simply trend toward the plateau, but for many samples are almost indistinguishable from the plateau basalts, especially those from Site 1137 [Weis et al., 1998; Mattielli et al., 2002; this study]. Either the Kerguelen plume head source was avail- able for the generation of these lavas or the plume tail magmas assimilated significant quantities of the overlying plateau lithosphere. Some of the <10 Ma archipelago lavas have slightly higher Sr and lower Nd isotopic compositions than either the Kerguelen plume head or plume tail compositions; all <10 Ma lavas have lower Hf isotopic values than those from either ‘‘head’’ or ‘‘tail’’ values. In Pb isotopic diagrams, these lavas trend between the Mont Crozier samples and values for the basalts from the plateau with the higher extents of conti- nental crust contamination. Although we cannot rule out a distinct mantle source origin for the <10 Ma lavas, as suggested by Mattielli et al. [2002], the isotopic data are consistent with assim- ilation of the Cretaceous Kerguelen Plateau litho- sphere by magmas derived from the plume tail source. It is probably also significant that the <10 Ma lavas are highly alkalic and fractionated, indicating that they come from small amounts of deep melting beneath the archipelago that have undergone extensive crystal fractionation; these conditions would facilitate assimilation of the thick, overlying plateau since the ascending magmas are (1) not voluminous and therefore more readily affected by smaller amounts of assimilate and (2) are in contact with the plateau’s lithosphere for a longer time, both since the melting takes place at greater depths and because the magmas need time to extensively fractionate. 6. Summary of Mantle Sources and Contaminants Involved in the 120 Ma History of the Kerguelen Hot Spot [27] During Cretaceous time, two mantle sources were important in the generation of basalts recov- ered from the Kerguelen Plateau–Broken Ridge (Figure 10). The primary mantle source may have been the dominant component available in the Kerguelen plume head and is best represented by basalts from Site 1138, that have 206Pb/204Pb  18, primitive mantle-like Sr and Nd isotopic compo- sitions and Hf isotopic compositions moderately Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 20 of 28depleted compared to the primitive mantle (Figure 10) [Neal et al., 2002; this study]. Basalts derived from this source whose magmas interacted with an upper crust contaminant can be recognized by their vertical trend in Pb-Pb isotope diagrams (Figure 7) and include those from Sites 749, 1136, 1137, 1141 and 1142. At Site 749, the basalts also record minor input from a normal-206Pb/204Pb (of about 18) Indian MORB mantle source, needed to fully account for their isotopic characteristics (Figure 10) [Frey et al., 2002a; Ingle et al., 2002a; this study]. Some magmas derived from the Ker- guelen plume head assimilated lower continental crust [Frey et al., 2002a]; this lower crust must have had Sr, Nd and Hf isotopic compositions broadly consistent with the aforementioned upper crust contaminant but with very unradiogenic Pb isotopic values (Figure 10). Site 747 and 1139 basalts fall into this group. A secondary mantle source, in addition to the plume head source, is necessary to account for the remaining Cretaceous plateau basalts from Sites 738 and 750.We suggest that this Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 21 of 28component had radiogenic Nd and Hf isotopic compositions, unradiogenic Pb isotopic values and is possibly represented by low-206Pb/204Pb depleted Indian MORB (Figure 10). Magmas derived from this source assimilated upper continental crust, either to very small extents (Site 750) or to very large extents (Site 738). [28] The post-Cretaceous Kerguelen Plateau and Kerguelen Archipelago samples also reflect two mantle sources, one potentially derived from the plume tail and the other from the normal, depleted Indian upper mantle, best represented by basalts from the Southeast Indian Ridge [e.g., Storey et al., 1988; Doucet et al., 2002; Weis and Frey, 2002]. The plume tail source, during the Cenozoic, is characterized by moderately radiogenic Pb isotopic compositions (206Pb/204Pb  18.6), nearly primi- tive mantle-like Hf isotopic compositions and Sr and Nd isotopic values within the field defined by the older Kerguelen Archipelago basalts and is best represented by basalts from the Mont Crozier section on the archipelago (Figure 10) [Mattielli et al., 2002; Weis et al., 2002b]. Radiogenic ingrowth cannot account for the shift from the plume head to the plume tail isotopic composition. Minor extents of assimilation of the Cretaceous Kerguelen Plateau by magmas reflecting mixtures between the depleted MORB component and the Kerguelen plume component can explain the remaining isotopic variation in some of the older Kerguelen Archipelago basalts (Figure 10). More extensive assimilation of the Kerguelen Plateau lithosphere by plume-derived basalts is required to fully account for the younger than 10 Ma lavas on the archipelago (Figure 10). 7. Origin of the Mantle Sources in the Kerguelen Hot Spot [29] The two Kerguelen plume sources can be described relative to each other by noting that the plume tail component has similar eNd(T), marginally lower eHf(T), higher (87Sr/86Sr)T and significantly more radiogenic Pb isotopic values than the plume head component. Both components fall between the fields associated with the enriched mantle end-members (EM I and EM II) [Hart, 1988], except that Site 1138 (plume head) basalts are slightly less radiogenic in 207Pb/204Pb and 208Pb/204Pb than either EM I or EM II. The most intriguing feature of both plume head and tail components is the low eNd(T) for a given eHf(T) [Mattielli et al., 2002; this study] compared to the mantle array of Vervoort and Blichert-Toft [1999]. In fact, this is also a characteristic of Indian MORB [Chauvel and Blichert-Toft, 2001; Kempton et al., 2002]. To generate the decoupling Figure 10. (opposite) Schematic representation of the possible mantle components (squares) and contaminants (ovals) involved in the genesis of Kerguelen Plateau and Kerguelen Archipelago rocks as depicted in skeletal form in plots of 208Pb/204Pb vs. 206Pb/204Pb (a) and (176Hf/177Hf)T vs. 206Pb/204Pb (b). Primary bulk mixing trajectories are shown in the thick, vectorless lines whereas tertiary contamination trends are shown in the thinner lines with vectors. During Cretaceous time, the mantle sources available in generating the Kerguelen Plateau rocks include (1) the low 206Pb/204Pb depleted component (as discussed in the caption for the previous diagram) important in generating basalts at Sites 750 and 738 on the SKP, and (2) a ‘‘plume head’’ component with general geochemical characteristics consistent with those found in Site 1138 basalts from the CKP [Neal et al., 2002] important for the genesis of Sites 747, 749 and 1136 (SKP), Site 1137 (Elan Bank), Site 1138 (CKP), Sites 1141 and 1142 (Broken Ridge) and Site 1139 (Skiff Bank). Also during Cretaceous times continental crust played a large role in generating the highly diverse isotopic compositions of the Kerguelen Plateau rocks. A minimum of two crustal contaminants may be accountable for the isotopic compositions in some Kerguelen Plateau basalts including lower crust, important at Site 747 (CKP) and Site 1139 (Skiff Bank) and upper crust, important for Sites 738, 749 and 750 (SKP), Site 1137 (Elan Bank) and Sites 1141 and 1142 (Broken Ridge). Minor contamination by a normal-type MORB may account for some ‘‘depleted’’ characteristics of Site 749 basalts. Cenozoic mantle sources include a ‘‘plume tail’’ component with general geochemical characteristics consistent with those found in the Mont Crozier basalts from the archipelago, and that is important for the generation of all post-Cretaceous lavas at Site 1140 (NKP) and on the archipelago. Large scale contamination of plume-tail-derived magmas with plateau-type material, as represented by the Site 1137 basalts (Elan Bank) can account for the isotopic characteristics of the Kerguelen Archipelago <10 Ma group. Smaller extents of contamination of mixtures between the normal-MORB and plume-tail sources with plateau- type lithosphere fully accounts for the total range of isotopic compositions in the Kerguelen Archipelago samples. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 22 of 28between eHf and eNd that results in samples plotting to the left of the mantle array requires an aged mantle source with a high Lu/Hf relative to Sm/Nd. Recycled continental sediments can be excluded on the basis that even small amounts of zircon in terrigenous sediments, incorporated in the mantle source, tends to move samples to below the mantle array in the Hf-Nd isotopic diagram [e.g., Eisele et al., 2002]. Two possibil- ities exist that might carry a high Lu/Hf (‘‘high’’ (eHf)T) relative to Sm/Nd (‘‘normal’’ (eNd)T): recycled pelagic sediment plus oceanic crust [e.g., Patchett et al., 1984; Chauvel and Blichert-Toft, 2001] or delaminated subcontinental mantle lithosphere (SCML) [Schmidberger et al., 2002]. A third possibility has been discussed by Kempton et al. [2002] to account for a high time- integrated Lu/Hf in Indian Ocean MORB. They propose an upper mantle wedge that has been subjected to several depletion events during ancient subduction-generated melting events. Although this model may be appropriate for the upper Indian Ocean mantle, it probably cannot account for the enriched geochemical signatures associated with the Kerguelen hot spot, the volcanic products of which are presumably generated primarily from melting within a mantle plume and do not always reflect mixing with the ambient, depleted upper mantle. Therefore, of the crust + pelagic sediment or delaminated SCML model, which is more appropriate for the Kerguelen plume sources? [30] It is difficult to fully evaluate the geochem- ical composition of ‘‘typical’’ SCML because it cannot be directly sampled and its composition is therefore determined by studying mantle xeno- liths. Nevertheless, some general characteristics may characterize the SCML in terms of its isotopic compositions. Recent studies on xeno- liths from the SCML document high eHf(T) relative to eNd(T) [Schmidberger et al., 2002; Simon et al., 2002; Ionov and Weis, 2002]. Pb isotopic values, although variable, are distinct from modern-day oceanic basalts and Sr and Nd isotopic compositions of SCML xenoliths are similar to present-day upper oceanic mantle [Walker et al., 1989]. Isotopic compositions of Os are also variable but are generally unradiogenic [Walker et al., 1989; Pearson, 1999]. The SCML is mostly comprised of highly depleted restite [e.g., Pearson, 1999; Lee et al., 2001] and trace element contents in unmetasomatized xenoliths may be insufficient for SCML to be a major factor in contaminating any mantle source as large volumes would be required [McDonough, 1990]. Some isotopic characteristics of inferred SCML are consistent with those of the Kerguelen plume sources, specifically a high eHf relative to eNd but others are not, specifically low Os iso- topic compositions are not characteristic of those associated with the Kerguelen plume tail (Mont Crozier), which has Os isotopic values overlap- ping with those of modern oceanic basalts [Weis et al., 2000]. Therefore, unless metasomatism is ubiquitous in the SCML and unless large quan- tities of metsomatized SCML are involved, the SCML is not a good candidate for the origin of the Kerguelen plume mantle sources. [31] The arguments for and against evidence favor- ing small additions of pelagic sediment to recycled oceanic crust as the cause of the distinctive isotopic signature of the Indian Ocean MORB have been discussed by Rehka¨mper and Hofmann [1997] and Chauvel and Blichert-Toft [2001] and so this po- tential source will be evaluated here, strictly as it applies to the Kerguelen plume. Patchett et al. [1984] and Vervoort et al. [1999] have described the fractionation of Hf from the REEs in pelagic sediments as arising from the lack of zircons in pelagic sediments. Pelagic sediments, injected back into the mantle along with oceanic crust, should be characterized by low U/Pb, leading to low-206Pb/204Pb signatures over time [e.g., Weaver, 1991] and, as stated above, show decoupled Nd and Hf isotopes resulting in a low 143Nd/144Nd relative to a given 176Hf/177Hf [Patchett et al., 1984]. Recycled oceanic crust alone should normally re- sult in HIMU-like isotopic compositions as the crust contains high U/Pb and subchondritic Nd resulting, over the long term, in high 206Pb/204Pb [Hofmann and White, 1982] and 176Hf/177Hf and 143Nd/144Nd below the ‘‘juvenile rock array’’ [Salters and White, 1998]. Thus the Pb and Hf isotopic compositions of the Kerguelen plume Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 23 of 28strongly support a source that is a mixture of recycled oceanic crust and pelagic sediments. The 235U and Th contents in this recycled crust must have been somewhat elevated above those consid- ered by Rehka¨mper and Hofmann [1997], because the Kerguelen mantle sources have 207Pb/204Pb and 208Pb/204Pb relative to 206Pb/204Pb that are well above the Indian Ocean MORB values. A possible way to account for this would be to invoke old oceanic crust characterized by higher 235U and Th concentrations, perhaps significantly older than 1.5 Ga and more fractionated than the crust proposed by Rehka¨mper and Hofmann [1997] in order to account for the present-day isotopic compositions. The Sr, Nd and Hf isotopic compo- sitions of this oceanic crust end-member could still be broadly consistent with those chosen by Rehka¨mper and Hofmann [1997] with Lu/Hf and 176Hf/177Hf of the oceanic crust represented by juvenile 1.8 Ga basalts from Vervoort and Bli- chert-Toft et al. [1999]. The ratio of the pelagic sediment to oceanic crust mixture would necessar- ily be different during the Cretaceous and Cenozo- ic, with the proportion of pelagic sediment required reduced in the Cenozoic, in order to account for the more radiogenic Pb isotopic compositions. Melting extents would have been higher during the plume head stage (with sediments melting more readily than the crustal component) relative to the plume tail stage, and this could account for the decreased proportion of sediment signature in the plume tail-derived lavas. 8. Conclusions [32] We present high-precision Hf isotopes for basalts from all stages of construction of the Kerguelen LIP. Integration of these data with other radiogenic isotopic systems results in several new conclusions regarding the geochemical origins of this large igneous province. During Cretaceous time, the primary mantle source for the Kerguelen Plateau–Broken Ridge basalts had primitive man- tle-like Sr, Nd and Pb isotopic compositions but moderately depleted Hf isotopic compositions. This source, probably located in the head of the Kerguelen plume, produced voluminous magmas that interacted with stranded fragments of conti- nental crust (both upper and lower) at shallow- levels beneath the Indian Ocean. An additional, minor mantle component with unradiogenic Pb isotopic compositions, similar to some Indian MORB today, is the source of basalts at two sites on the SKP. These magmas also assimilated upper continental crust. The origin of this mantle com- ponent (in the ambient upper mantle, or entrained from a depleted, mid-mantle boundary) cannot be determined at this time. In the post-Cretaceous evolution of the Kerguelen plume, the magmas derived from the plume tail source, having higher Sr and Pb, lower Nd and similar Hf isotopic compositions to the plume head source, mixed readily with depleted ambient Indian MORB man- tle. These magmas erupted either with or without assimilation of the overlying, thick Cretaceous Kerguelen Plateau lithosphere. The extreme of this interaction is present in the youngest lavas from the Kerguelen Archipelago, which reflect extensive contamination by plateau lithosphere. The geo- chemical characteristics of the Kerguelen plume head and plume tail compositions may be accounted for by ancient, recycled oceanic crust that had high 235U and Th contents mixed with ancient recycled pelagic sediments. The propor- tions of each must change from larger amounts of pelagic sediments in the plume head during the Cretaceous, compared to the plume tail during the Cenozoic. Acknowledgments [33] We are very appreciative of discussions, clarifications and references provided by G. Bru¨gmann, C. Chauvel, M. F. Coffin, C. Devey, J. G. Fitton, J. Hertogen, P. Kempton, J. J. Mahoney, M. Menzies, F. Nauret and J. S. Scoates. M. F. Coffin and R. Miura generously provided Figures 1 and 2. J. DeJong is thanked for guidance with and maintenance of the Nu Plasma at ULB. We are most grateful for thorough and constructive reviews by J. Barling and J. J. Mahoney and editorial comments from C. Chauvel that have significantly improved the concepts and contents discussed in this manu- script. N. Arndt, W. Chazey, F. A. Frey, C. R. Neal and J. J. Mahoney supplied some of the samples. This study used samples provided by the Ocean Drilling Program (ODP). The ODP is funded by the U. S. National Science Foundation and member countries and is under management from the Joint Oceanographic Institute. The Fonds National de la Recherche Scientifique (F.N.R.S.) funds Belgian participation in ODP. The first author is supported by an ARC grant (#98/03-233) Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 24 of 28from the Communaute´ Franc¸aise de Belgique (C.F.B.). The F.N.R.S. and the C.F.B. jointly funded this research. D. Weis is the F.N.R.S. Director of Research. References Arndt, N., and D. Weis, Oceanic plateaus as windows to the Earth’s interior: An ODP success story, Joides J., 28, spec. issue, 79–84, 2002. Barling, J., S. L. Goldstein, and I. A. Nicholls, Geochemistry of Heard Island (southern Indian Ocean): Characterization of an enriched mantle component and implications for enrich- ment of the sub-Indian Ocean mantle, J. Petrol., 35(4), 1017–1053, 1994. Blichert-Toft, J., and F. Albare`de, The Lu-Hf isotope geo- chemistry of chondrites and the evolution of the mantle- crust system, Earth Planet. Sci. Lett., 148, 243–258, 1997. Blichert-Toft, J., C. Chauvel, and F. Albare`de, Separation of Hf and Lu for high-precision isotope analysis of rock samples by magnetic sector-multiple collector ICP-MS, Contrib. Mineral. Petrol., 127, 248–260, 1997. Borissova, I., M. Coffin, P. Charvis, and S. Operton, Structure and development of a microcontinent: Elan Bank in the southern Indian Ocean, Geochem. Geophys. Geosyst., 4, doi:10.1029/2003GC000535, in press, 2003. Chauvel, C., and J. Blichert-Toft, A hafnium isotope and trace element perspective on melting of the depleted mantle, Earth Planet. Sci. Lett., 190, 137–151, 2001. Coffin, M. F., and O. Eldholm, Large igneous provinces: Crus- tal structure, dimensions, and external consequences, Rev. Geophys., 32, 1–36, 1994. Coffin, M. F., M. S. Pringle, R. A. Duncan, T. P. Gladczenko, M. Storey, R. D. Mu¨ller, and L. A. Gahagan, Kerguelen hotspot magma output since 130 Ma, J. Petrol., 43, 1121– 1139, 2002. Damasceno, D., J. S. Scoates, D. Weis, F. A. Frey, and A. Giret, Mineral chemistry of mildly alkalic basalts from the 25 Ma Mont Crozier Section, Kerguelen Archipelago: Constraints on phenocryst crystallization environments, J. Petrol., 43, 1389–1413, 2002. Davies, H. L., S.-S. Sun, F. A. Frey, I. Gautier, M. T. McCulloch, R. C. Price, Y. Bassias, C. T. Klootwijk, and L. Leclaire, Basalt basement from the Kerguelen Plateau and the trail of a Dupal plume, Contrib. Mineral. Petrol., 103, 457–469, 1989. Doe, B. R., and R. E. Zartman, Plumbotectonics: The Pha- nerozoic, in Geochemistry of Hydrothermal Ore Deposits, edited by H. Barnes, pp. 22–70, John Wiley, New York, 1979. Dosso, L., H. Bougault, P. Beuzart, J.-Y. Calvez, and J.-L. Joron, The geochemical structure of the South-East Indian Ridge, Earth Planet. Sci. Lett., 88, 47–59, 1988. Doucet, S., D. Weis, J. S. Scoates, K. Nicolaysen, F. A. Frey, and A. Giret, The depleted mantle component in Kerguelen Archipelago basalts: Petrogenesis of the tholeiitic-transi- tional basalts from the Loranchet Peninsula, J. Petrol., 43, 1341–1366, 2002. Duncan, R. A., Geochronology of basalts from the Ninetyeast Ridge and continental dispersion in the eastern Indian Ocean, J. Volcanol. Geotherm. Res., 4, 283–305, 1978. Duncan, R. A., A timeframe for construction of the Kerguelen Plateau and Broken Ridge, J. Petrol., 43, 1109–1119, 2002. Dupre´, B., and C. J. Alle`gre, Pb-Sr isotope variation in Indian Ocean basalts and mixing phenomena, Nature, 303, 142– 146, 1983. Eisele, J., M. Sharma, S. J. G. Galer, J. Blichert-Toft, C. W. Devey, and A. W. Hofmann, The role of sediment recycling in EM-1 inferred from Os, Pb, Hf, Nd, Sr isotope and trace element systematics of the Pitcairn hotspot, Earth Planet. Sci. Lett., 196, 197–212, 2002. Fitton, J. G., A. D. Saunders, P. D. Kempton, and B. S. Hardarson, Does depleted mantle form an intrinsic part of the Iceland plume?, Geochem. Geophys. Geosyst., 4(3), 1032, doi:10.1029/2002GC000424, 2003. Frey, F. A., N. J. McNaughton, D. R. Nelson, J. R. deLaeter, and R. A. Duncan, Petrogenesis of the Bunbury Basalt, Wes- tern Australia: Interaction between the Kerguelen plume and Gondwana lithosphere?, Earth Planet. Sci. Lett., 144, 163– 183, 1996. Frey, F. A., et al., Origin and evolution of a submarine large igneous province: The Kerguelen Plateau and Broken Ridge, southern Indian Ocean, Earth Planet. Sci. Lett., 176, 73–89, 2000. Frey, F. A., D. Weis, A. Borissova, and G. Xu, Involvement of continental crust in formation of the Kerguelen Plateau: New perspectives from ODP Leg 120 Sites, J. Petrol., 43, 1207– 1239, 2002a. Frey, F. A., K. Nicolaysen, B. K. Kubit, D. Weis, and A. Giret, Flood basalt from Mont Tourmente in the Central Kerguelen Archipelago: The change from transitional to alkalic basalt at 25 Ma, J. Petrol., 43, 1367–1387, 2002b. Gautier, I., D. Weis, J.-P. Mennessier, P. Vidal, A. Giret, and M. Loubet, Petrology and geochemistry of the Kerguelen Archipelago basalts (south Indian Ocean): Evolution of the mantle sources to intraplate position, Earth Planet. Sci. Lett., 100, 59–76, 1990. Gradstein, F. M., F. P. Agterberg, J. G. Ogg, J. Hardenbol, P. van Veen, J. Thierry, and Z. Huang, A Mesozoic time scale, J. Geophys. Res., 99, 24,051–24,074, 1994. Hamelin, B., B. Dupre´, and C. J. Alle`gre, Pb-Sr-Nd isotopic data of Indian Ocean ridges: New evidence of large-scale mapping of mantle heterogeneities, Earth Planet. Sci. Lett., 76, 288–298, 1985. Hanan, B. B., J. Blichert-Toft, R. Kingsley, and J.-G. Schilling, Depleted Iceland mantle plume geochemical signature: Artifact of multicomponent mixing?, Geochem. Geophys. Geosyst., 1, Paper number 1999GC000009, 2000. Hart, S. R., A large-scale isotope anomaly in the Southern Hemisphere mantle, Nature, 309, 753–757, 1984. Hart, S. R., Heterogeneous mantle domains: Signatures, genesis and mixing chronologies, Earth Planet. Sci. Lett., 90, 273–296, 1988. Hedge, C. E., N. D. Watkins, R. A. Hildreth, and W. P. Doering, 87Sr/86Sr ratios in basalts from islands in the Indian Ocean, Earth Planet. Sci. Lett., 21, 29–34, 1973. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 25 of 28Hofmann, A. W., and W. M. White, Mantle plumes from ancient oceanic crust, Earth Planet. Sci. Lett., 57, 421– 436, 1982. Ingle, S., D. Weis, J. S. Scoates, and F. A. Frey, Relationship between the early Kerguelen plume and continental flood basalts of the paleo-Eastern Gondwanan margins, Earth Pla- net. Sci. Lett., 197, 35–50, 2002a. Ingle, S., D. Weis, and F. A. Frey, Indian continental crust recovered from Elan Bank, Kerguelen Plateau (ODP Leg 183, Site 1137), J. Petrol., 43, 1241–1257, 2002b. Ionov, D., and D. Weis, Hf-Nd-Sr isotope relationships in spinel and garnet facies peridotite xenoliths: Inferences for the age and evolution of the lithospheric mantle, Geochim. Cosmochim Acta, 66, spec. issue, A356, 2002. Kempton, P. D., J. G. Fitton, A. D. Saunders, G. M. Nowell, R. N. Taylor, B. S. Hardarson, and G. Pearson, The Iceland plume in space and time: A Sr-Nd-Pb-Hf study of the North Atlantic rifted margin, Earth Planet. Sci. Lett., 177, 255– 271, 2000. Kempton, P. D., J. A. Pearce, T. L. Barry, J. G. Fitton, C. Langmuir, and D. M. Christie, Sr-Nd-Pb-Hf isotope results from ODP Leg 187: Evidence for mantle dynamics of the Australian-Antarctic Discordance and origin of the Indian MORB source, Geochem. Geophys. Geosyst., 3(12), 1074, doi:10.1029/2002GC000320, 2002. Kent, R. W., A. D. Saunders, P. D. Kempton, and N. C. Ghose, Rajmahal Basalts, eastern India: Mantle sources and melt distribution at a volcanic rifted margin, in Large Igneous Provinces: Continental, Oceanic, and Planetary Flood Volcanism, Geophys. Monogr. Ser., vol. 100, edited by M. Coffin and J. Mahoney, pp. 145–182, AGU, Washington, D.C., 1997. Kent, R. W., M. S. Pringle, R. D. Mu¨ller, A. D. Saunders, and N. C. Ghose, 40Ar/39Ar geochronology of the Rajmahal basalts, India, and their relationship to the Kerguelen Plateau, J. Petrol., 43, 1141–1153, 2002. Kieffer, B., N. T. Arndt, and D. Weis, A bimodal alkalic shield volcano on Skiff Bank: Its place in the evolution of the Kerguelen Plateau, J. Petrol., 43, 1259–1286, 2002. Klein, E. M., C. H. Langmuir, A. Zindler, H. Staudigel, and B. Hamelin, Isotope evidence of a mantle convection boundary at the Australian-Antarctic discordance, Nature, 333, 623–629, 1988. Klein, E. M., C. H. Langmuir, and H. Staudigel, Geochemistry of basalts from the Southeast Indian Ridge, 115E–138E, J. Geophys. Res., 96, 2089–2107, 1991. Lawver, L. A., L. A. Gahagan, and M. F. Coffin, The devel- opment of paleoseaways around Antarctica, in The Antarctic Paleoenvironment: A Perspective on Global Change, Ant- arct. Res. Ser., 56, 7–30, 1992. Lee, D.-C., R. L. Rudnick, and G. H. J. Brimhall, Deep lithospheric dynamics beneath the Sierra Nevada during the Mesozoic and Cenozoic as inferred from xenolith petrology, Geochem. Geophys. Geosyst., 2, doi:10.1029/ 2002GC000482, 2001. Mahoney, J. J., J. D. Macdougall, G. W. Lugmair, and K. Gopalan, Kerguelen hotspot source for Rajmahal Traps and Ninetyeast Ridge?, Nature, 303, 385–389, 1983. Mahoney, J. J., J. H. Natland, W. M. White, R. Poreda, S. H. Bloomer, R. L. Fishe, and A. N. Baxter, Isotopic and geo- chemical provinces of the western Indian Ocean spreading centers, J. Geophys. Res., 94, 4033–4052, 1989. Mahoney, J., A. P. LeRoex, Z. Peng, R. L. Fisher, and J. H. Natland, Southwestern limits of Indian Ocean ridge mantle and the origin of low 206Pb/204Pb mid-ocean ridge basalt: Isotope systematics of the central Southwest Indian Ridge (17–50E), J. Geophys. Res., 97, 19,771–19,790, 1992. Mahoney, J. J., W. B. Jones, F. A. Frey, V. J. M. Salters, D. G. Pyle, and H. L. Davies, Geochemical characteristics of lavas from Broken Ridge, the Naturaliste Plateau and southern- most Kerguelen Plateau: Cretaceous plateau volcanism in the southeast Indian Ocean, Chem. Geol., 120, 315–345, 1995. Mahoney, J. J., W. M. White, B. G. J. Upton, C. R. Neal, and R. A. Scrutton, Beyond EM-1: Lavas from Afanasy-Nikitin Rise and the Crozet Archipelago, Indian Ocean, Geology, 24, 615–618, 1996. Mattielli, N., D. Weis, S. Ingle, J. Blichert-Toft, J. Barling, and F. A. Frey, Hafnium isotopic constraints on the source of Kerguelen Plateau lavas (Leg 183 Sites 1140 and 1137), Eos Trans. AGU, 81(19), 430–431, 2000. Mattielli, N., D. Weis, J. Blichert-Toft, and F. Albare`de, Hf isotope evidence for a Miocene change in the Kerguelen mantle plume composition, J. Petrol., 43, 1327–1339, 2002. McDonough, W. F., Constraints on the composition of the continental lithospheric mantle, Earth Planet. Sci. Lett., 101, 1–18, 1990. Michard, A., R. Montigny, and R. Schlich, Geochemistry of the mantle beneath the Rodriguez Triple Junction and the South-East Indian Ridge, Earth Planet. Sci. Lett., 78, 104– 114, 1986. Mu¨ller, R. D., C. Gaina, W. R. Roest, and D. L. Hansen, A recipe for microcontinent formation, Geology, 29, 203–206, 2001. Munschy, M., J. Dyment, M. O. Boulanger, D. Boulanger, J. D. Tissot, R. Schlich, Y. Rotstein, and M. F. Coffin, Breakup and sea floor spreading between the Kerguelen Plateau- Labuan Basin and Broken Ridge-Diamantina Zone, Proc. Ocean Drill. Program Sci. Results, 120, 931–944, 1992. Neal, C. R., J. Mahoney, and W. Chazey, 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, J. Petrol., 43, 1177– 1205, 2002. Nicolaysen, K., F. A. Frey, K. V. Hodges, D. Weis, and A. Giret, 40Ar/39Ar geochronology of flood basalts from the Kerguelen Archipelago, southern Indian Ocean: Implica- tions for Cenozoic eruption rates of the Kerguelen plume, Earth Planet. Sci. Lett., 174, 313–328, 2000. Nicolaysen, K., S. Bowring, F. A. Frey, D. Weis, S. Ingle, M. S. Pringle, and M. F. Coffin, Provenance of Proterozoic garnet-biotite gneiss recovered from Elan Bank, Kerguelen Plateau, southern Indian Ocean, Geology, 29, 235–238, 2001. Operto, S., and P. Charvis, Kerguelen Plateau: A volcanic passive margin fragment?, Geology, 23, 137–140, 1995. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 26 of 28Operto, S., and P. Charvis, Deep structure of the southern Kerguelen Plateau (southern Indian Ocean) from ocean bot- tom seismometer wide-angle seismic data, J. Geophys. Res., 101, 25,077–25,103, 1996. Patchett, P. J., and M. Tatsumoto, Hafnium isotope variations in oceanic basalts, Geophys. Res. Lett., 7, 1077–1080, 1980. Patchett, P. J., W. M. White, H. Feldmann, S. Kielinczuk, and A. W. Hofmann, Hafnium/rare earth element fractionation in the sedimentary system and crustal recycling into the Earth’s mantle, Earth Planet. Sci. Lett., 69, 365–378, 1984. Pearson, D. G., The age of continental roots, Lithos, 48, 171– 194, 1999. Price, R. C., A. K. Kennedy, M. Riggs-Sneeringer, and F. A. Frey, Geochemistry of basalts from the Indian Ocean triple junction: Implications for the generation and evolution of Indian Ocean ridge basalts, Earth Planet. Sci. Lett., 78, 379–386, 1986. Pyle, D. G., D. M. Christie, and J. J. Mahoney, Resolving an isotopic boundary within the Australian-Antarctic Discor- dance, Earth Planet. Sci. Lett., 112, 161–178, 1992. Ramana, M. V., et al., Mesozoic anomalies in the Bay of Bengal, Earth Planet. Sci. Lett., 121, 469–475, 1994. Rehka¨mper, M., and A. W. Hofmann, Recycled ocean crust and sediment in Indian Ocean MORB, Earth Planet. Sci. Lett., 147, 93–106, 1997. Royer, J.-Y., and D. T. Sandwell, Evolution of the eastern Indian Ocean since the Late Cretaceous: Constraints from Geosat altimetry, J. Geophys. Res., 94, 13,755–13,782, 1989. Salters, V. J. M., The generation of mid-ocean ridge basalts from the Hf and Nd isotope perspective, Earth Planet. Sci. Lett., 141, 109–123, 1996. Salters, V. J. M., and W. M. White, Hf isotope constraints on mantle evolution, Chem. Geol., 145, 447–460, 1998. Salters, V. J. M., M. Storey, J. H. Sevigny, and H. Whitechurch, Trace element and isotopic characteristics of Kerguelen- Heard Plateau basalts, Proc. Ocean Drill. Program Sci. Results, 120, 55–62, 1992. Scherer, E., C. Mu¨nker, and K. Mezger, Calibration of the Lutetium-Hafnium Clock, Science, 293, 683–687, 2001. Schiano, P., J.-L. Birck, and C. J. Alle`gre, Osmium-strontium- neodymium-lead isotopic covariations in mid-ocean ridge basalt glasses and the heterogeneity of the upper mantle, Earth Planet. Sci. Lett., 150, 363–379, 1997. Schmidberger, S. S., A. Simonetti, D. Francis, and C. Garie´py, Probing Archean lithosphere using the Lu-Hf isotope systematics of peridotite xenoliths from Somerset Island kimberlites, Canada, Earth Planet. Sci. Lett., 197, 245– 259, 2002. Shipboard Scientific Party, Site 1137, Proc. Ocean Drill. Pro- gram Initial Rep. [CD-ROM], 183, 2000. Simon, N. S. C., R. W. Carlson, D. G. Pearson, and G. F. Davies, The Lu-Hf isotope composition of cratonic litho- sphere: Disequilibrium between garnet and clinopyroxene in kimberlite xenoliths, Geochim. Cosmochim Acta, 66, spec. issue, A717, 2002. Storey, M., A. D. Saunders, J. Tarney, P. Leat, M. F. Thirlwall, R. N. Thompson, M. A. Menzies, and G. F. Marriner, Geo- chemical evidence for plume-mantle interactions beneath Kerguelen and Heard Islands, Indian Ocean, Nature, 336, 336–371, 1988. Storey, M., A. D. Saunders, J. Tarney, I. L. Gibson, M. J. Norry, M. F. Thirwall, P. Leat, R. N. Thompson, and M. A. Menzies, Contamination of Indian Ocean asthenosphere by the Kerguelen-Heard mantle plume, Nature, 338, 574–576, 1989. Storey, M., M. S. Pringle, M. F. Coffin, and J. Wijbrans, Geo- chemistry and geochronology of Kerguelen Plateau basalts: Results from ODP Legs 119 and 120, Eos Trans. AGU, 77, 123, 1996. Vervoort, J. D., and J. Blichert-Toft, Evolution of the depleted mantle: Hf isotope evidence from juvenile rocks through time, Geochim. Cosmochim. Acta, 63(3/4), 533–556, 1999. Vervoort, J. D., P. J. Patchett, J. Blichert-Toft, and F. Albare`de, Relationships between Lu-Hf and Sm-Nd isotopic systems in the global sedimentary system, Earth Planet. Sci. Lett., 168, 79–99, 1999. Vervoort, J. D., P. J. Patchett, F. Albare`de, J. Blichert-Toft, R. L. Rudnick, and H. Downes, Hf-Nd isotopic evolution of the lower crust, Earth Planet. Sci. Lett., 181, 115–129, 2000. Walker, R. J., R. L. Carlson, S. B. Shirey, and F. R. Boyd, Os, Sr, Nd, and Pb isotope systematics of southern African peri- dotite xenoliths: Implications for the chemical evolution of subcontinental mantle, Geochim. Cosmochim. Acta, 53, 1583–1595, 1989. Weaver, B. L., The origin of ocean island basalt end-member compositions: Trace element and isotopic constraints, Earth Planet. Sci. Lett., 104, 381–397, 1991. Weis, D., and F. A. Frey, Role of the Kerguelen Plume in generating the eastern Indian Ocean seafloor, J. Geophys. Res., 101, 13,831–13,849, 1996. Weis, D., and F. A. Frey, Submarine basalts of the Northern Kerguelen Plateau: Interaction between the Kerguelen plume and the Southeast Indian Ridge revealed at ODP Site 1140, J. Petrol., 43, 1287–1309, 2002. Weis, D., Y. Bassias, I. Gautier, and J. P. Mennessier, Dupal anomaly in existence 115 Ma ago: Evidence from isotopic study of the Kerguelen Plateau (South Indian Ocean), Geochim. Cosmochim. Acta, 53, 2125–2131, 1989. Weis, D., F. A. Frey, A. Saunders, and I. L. Gibson, Ninetyeast Ridge (Indian Ocean): A 5000 km record of a Dupal mantle plume, Geology, 19, 99–102, 1991. Weis, D., F. A. Frey, H. Leyrit, and I. Gautier, Kerguelen Archipelago revisited: Geochemical and isotopic study of the Southeast Province lavas, Earth Planet. Sci. Lett., 118, 101–119, 1993. Weis, D., F. A. Frey, A. Giret, and J.-M. Cantagrel, Geochem- ical Characteristics of the youngest volcano (Mount Ross) in the Kerguelen Archipelago: Inferences for magma flux, lithosphere assimilation and composition of the Kerguelen Plume, J. Petrol., 39, 973–994, 1998. Weis, D., S. B. Shirey, and F. A. Frey, Re-Os systematics of Kerguelen plume basalts: Enriched components and lower mantle sources, Eos Trans. AGU, 81(48), Fall Meet. Suppl., abstract V11C-09, 2000. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 27 of 28Weis, D., S. Ingle, D. Damasceno, F. A. Frey, K. Nicolaysen, and J. Barling, Origin of continental components in Indian Ocean basalts: Evidence from Elan Bank (Kerguelen Pla- teau, ODP Leg 183, Site 1137), Geology, 29, 147–150, 2001. Weis, D., F. A. Frey, R. Schlich, R. Montigny, D. Damasceno, N. Mattielli, K. E. Nicolaysen, and J. S. Scoates, Trace of the Kerguelen mantle plume: Evidence from seamounts between Kerguelen Archipelago and Heard Island, Indian Ocean, Geochem. Geophys. Geosyst., 3(6), 1033, doi:10.1029/ 2001GC000251, 2002a. Weis, D., J. de Jong, and C. Maerschalk, High-precision Pb isotope systematics of basalts from the Kerguelen Archipe- lago: New insights on the Kerguelen plume components, Eos Trans. AGU, 83(47), Fall Meet. Suppl., F1426, 2002b. White, W. M., 238U/204Pb in MORB and open system evolu- tion of the depleted mantle, Earth Planet. Sci. Lett., 115, 211–226, 1993. Woodhead, J. D., J. M. Hergt, J. P. Davidson, and S. M. Eggins, Hf isotope evidence for ‘conservative’ element mo- bility during subduction zone processes, Earth Planet. Sci. Lett., 192, 331–346, 2001. Yang, H.-J., F. A. Frey, D. Weis, A. Giret, D. Pyle, and G. Michon, Petrogenesis of the flood basalts forming the northern Kerguelen Archipelago: Implications for the Kerguelen plume, J. Petrol., 39, 711–748, 1998. Geochemistry Geophysics Geosystems G3 ingle et al.: hf isotope constraints on mantle sources 10.1029/2002GC000482 28 of 28

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
France 20 0
United States 7 4
China 5 14
Ukraine 1 0
City Views Downloads
Unknown 20 1
Ashburn 6 0
Shenzhen 3 14
Fuzhou 2 0
Columbus 1 0
Toulouse 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

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.38318.1-0075976/manifest

Comment

Related Items