Mantle reservoirs and mafic magmatism associated with the break-up of Gondwana : the Balleny Plume and the Australian-Antarctic discordance : U-Pb zircon dating of a Proterozoic mafic dyke swarm in the Vestfold Hills, East Antarctica

Part 1: Within the recent literature, the isotopic heterogeneity of ocean island basalts (OIB) is generally ascribed to mixing between any two or more of four isotopically distinct mantle end-member components - DMM, HIMU, EMI and EMU - the origin and precise location of which are still the subject of much debate. An attempt is made here to constrain the geochemical characteristics of these end-member components using literature-derived endmember OIB data. This has confirmed the presence of consistent trace element differences and also suggested certain major element distinctions, which are interpreted to reflect diverse end-member mantle source compositions rather than differences in the pressure and temperature of melting involved in OIB production. EMI basalts, which extend to the lowest \\(^{143}\\)Nd/\\(^{144}\\)Nd values of all OIB, were found to possess the most distinctive major element characteristics, the latter evident as higher SiO\\(_2\\) and lower FeO*, CaO, TiO\\(_2\\), P\\(_2\\)O\\(_5\\) and K\\(_2\\)O values than both HIMU and EMU OIB. However, EMI basalts have the least well defined trace element systematics. In contrast, HIMU and EMU basalts, previously characterised by the highest Pb isotope ratios and the highest \\(^{87}\\)Sr/\\(^{86}\\)Sr values respectively, were found to have overlapping major element abundances but quite distinctive trace element systematics. HIMU basalts, have the highest CaO and generally lower LILE/HFSE and LREE/HFSE abundance ratios, intermediate Th/La and Th/Nb, lower Zr/Nb and higher Nb/Pb and Ce/Pb values than either of the EM components. EMU basalts extend to the highest K\\(_2\\)O and TiO\\(_2\\) values and are characterised by lower Ba/Th, higher Th/La and higher Th/Nb values than EMI OIB. Previous workers have recognised a regional HIMU radiogenic isotope and trace element signature within Cretaceous to Recent volcanics scattered throughout the dispersed continental fragments of eastern Gondwana and dredged from the TasmanSea - southwest Pacific Ocean seafloor. This study has ascribed the HIMU nature of continental volcanism in Tasmania, the South Island and offshore islands of New Zealand, and Marie Byrd Land and the McMurdo Volcanic Group of West Antarctica to localised and intermittent periods of tectonically-induced decompression and melting of underplated HIMU material emplaced at the base of the lithosphere by one or more upwelling plume heads prior to or coincident withthe onset of continental break-up in this region. Ocean floor HIMU volcanism in the southern Tasman Sea and southwest Pacific Ocean has been attributed to lithospheric plate movement over the persisting plume conduits. Previous plate-tectonic reconstructions have ascribed a 4000 km long cun/ed chain of seamounts and islands, extending from the western flank of Lord Howe Rise to the Balleny Islands close to the Ross Sea region of the southwest Pacific Ocean, to movement of the Indian-Australian and Antarctic Plates over the Balleny Plume with progressive similar latitude to the Balleny Islands, has an homologous isotopic and trace element signature and therefore appears to be the product of a second and parallel HIMU plume trace, here termed the Scott Plume. The latter is also held responsible for the enriched geochemical features of some Macquarie Island basalts. The Balleny and Scott Plumes appear to have been temporarily trapped by, and to have contaminated the eruptives of, the Southeast Indian and Pacific-Antarctic spreading ridges respectively. This has resuKed in the <28 Ma (between AS and the active spreading ridges) seafloor in this region bearing an EMORB to HIMU-DMM isotopic signature. The fact that samples dredged from virtually zero-age oceanic crust on the southern flank of the SEIR and northwest of the Balleny Islands, also bear this distinctive isotopic and trace element signature impliesthat the supply of HIMU material to this region of the SEIR has continued until very recently. Significant contamination of Balleny Plume HIMU volcanics by DMM material, particularly evident in Pb-Pb isotopic space, suggests that the Balleny Plume is a relatively weak plume which entrained large amounts of depleted upper mantle during upwelling. Progressive temporal dilution of the HIMU plume component by DMM material may imply either increasing amounts of upper mantle entrainment, suggesting a gradual weakening of the plume with time, or progressive preferential melting of entrained material within a heterogeneous plume. The Balleny Plume also appears to have slightly higher timeintegrated Th/U than other known HIMU sources, resulting in the most radiogenic Balleny Plume samples plotting above the HIMU field in \\(^{206}\\)/Pb/\\(^{204}\\).Pb\\(^{208}\\)Pb/\\(^{204}\\)Pb space, and thereby potentially expanding the known range of HIMU isotopic values. Part 2: The Southern Ocean between the southern margin of Australia and Wilkes Land in Antarctica hosts an important section of the global mid-ocean ridge system. Other researchers have concluded that the Southeast Indian Ridge (SEIR) in this region experienced a dramatic change in spreading rate at -44.5 Ma, coincident with a major period of global plate reorganisation. Comparative studies, within the recent literature, of individual fast and slow spreading ridges have revealed distinct morphological and geochemical parameters which have been attributed to different underlying magma chamber processes. An attempt is made here to discern if the geochemical parameters correlated with spreading rate in the literature are also evident with a change in spreading rate along a single segment of mid-ocean ridge. This effort has been concentrated on the chemistry of basalts collected from four Southern Ocean dredge sites north of the SEIR. Two of these sites were located on oceanic crust formed during a period of relatively slow seafloor spreading prior to 49 Ma, whereas the other two correspond to a period of more rapid seafloor spreading after 44.5 Ma. The lack of recognisable and systematic geochemical differences between the SEIR slow and intermediate spreading rate eruptives suggests that the change in spreading rate at -44.5 Ma, and the inferred change in underlying mantle processes, may not have been great enough to result in the significant geochemical variations cited in the literature. Alternatively, the off-axis Southern Ocean basalt database may be insufficient to enable the detection of such differences. The Southern Ocean south of the Great Australian Bight is also host to a globally anomalous region of mid-ocean ridge, known as the Australian-Antarctic Discordance (AAD). In addition to its recognised morphological and geophysical anomalies, the AAD represents the current on-axis location of a proposed isotopic boundary between Indian and Pacific Ocean upper mantle convective regimes, previously defined from analyses of ‚Äöv¢¬ß4 Ma MORB dredged from within and adjacent to the AAD. An investigation is undertaken here into the off-axis location of this isotopic boundary, concentrating on radiogenic Pb, Sr and Nd isotope data for -36-66 Ma seafloor dredged from either side of the northward extrapolation of the AAD. The results show that ‚Äöv¢‚Ä¢36 Ma seafloor east of the AAD has an Indian Ocean MORB isotopic signature, thereby implying that the proposed Indian-Pacific Ocean isotopic boundary does not extend directly north of the ridge towards the southern margin of Australia. Progressive westward migration of an arcuate-shaped front of Pacific Ocean upper mantle therefore appears to be a consequence of Australian-Antarctic rifting and Southern Ocean opening, favouring models of active mantle flow outlined in the recent literature such as Pacific Ocean basin shrinkage, convergence of hotspot-driven along-axis asthenospheric flow and/or direct Indian and Pacific Ocean upper mantle convergence. Part 3: The Vestfold Hills, one of several Archaean cratonic blocks within the East Antarctic Shield, comprises a high-grade metamorphic basement complex intruded by at least nine generations of Early to Middle Proterozoic mafic dykes. Extensive U-Pb ion microprobe (SHRIMP) analyses of zircons, derived predominantly from late-stage felsic differentiates of the mafic dykes, provide precise crystallisation ages for several dyke generations. These new ages enable constraints to be placed on both the history of mafic magmatism in the Vestfold Hills and the timing of the various interspersed, and already documented, Proterozoic deformation events. In addition to demonstrating the utility of zircons derived from felsic late-stage differentiates for the dating of co-genetic mafic dykes, this study also places doubt on previous whole-rock Rb-Sr dating of mafic dyke suites in this and other areas of East Antarctica. \\(^{207}\\)Pb/\\(^{206}\\)Pb zircon ages of 2241¬¨¬±4 Ma and 2238¬¨¬±7 Ma for the Homogeneous and Mottled Norites, respectively, provide a younger emplacement age for associated Group 2 High-Mg tholeiite dykes than the whole-rock Rb-Sr date (2424¬¨¬±72 Ma) originally interpreted as the age of all high-Mg intrusives in the Vestfold Hills. Zircon ages of 1754¬¨¬±16 Ma and 1832¬¨¬±72 Ma confirm the previously defined Rb-Sr age ofthe Group 2 Fe-rich tholeiites. Two later dyke generations, the Group 3 and 4 Fe-rich tholeiites, distinguished on the basis of field orientations and cross-cutting relationships, yield zircon emplacement ages of 1380¬¨¬±7 Maand 1241¬¨¬±5 Ma,which also defineminimum ages fortwo suites of lamprophyre dykes. Xenocrystic zircons within both felsic segregations and mafic dykes yield zircon ages of 2478¬¨¬±5 Ma to -2740 Ma, indicating the presence of Archaeancrustal source rocks of this antiquity beneath the Vestfold Hills.