Petrography and geochemistry of oceanic crust: provenance of sedimentary detritus, Macquarie Island

Macquarie Island, located 1500 km southeast of Australia, is a sub-aerial exposure of Miocene oceanic crust/upper mantle situated within the oceanic basin in which it formed. The seafloor spreading related ‘Finch – Langdon fault’ juxtaposes upper mantle serpentinised peridotites, gabbro and sheeted dolerite dykes in the northern quarter of the island with volcanic rocks (pillow basalts, tabular basalts, hyaloclastites and minor sedimentary rocks) that form the southern part of the island. Interbedded within the extensive volcanic sequences are volcaniclastic sedimentary units containing gabbro and dolerite clasts, indicative that lower oceanic crust was exposed near the mid-ocean ridge during active volcanism. Through a combination of petrography, major and trace-element clinopyroxene and zircon geochemistry, and U-Pb geochronology, this thesis investigates the provenance of the sedimentary units produced at an active mid-ocean ridge. The high abundance of hydrothermal monomineralic clasts including chlorite, epidote, prehnite, quartz, carbonates and zeolites, and clasts of fault rock within the sedimentary rocks is consistent with a fault-scarp derived provenance. The unusually large volume of sedimentary rocks exposed on Macquarie Island suggests a large plate boundary scale transform fault is likely in order to produce the large volume of sediment. The long offset transform that links the Miocene ‘A2/P2 – A4/P4’ seafloor spreading corridor to the Australia-Antarctic-Pacific triple junction is the most proximal and therefore most likely source. Detrital clinopyroxene major and trace-element geochemistry shows little overlap with clinopyroxene grains within the interbedded volcanic sequences, excluding the proximal volcanism as a likely source for the sediment. A lack of peridotite signature in clinopyroxene clasts suggests that it is likely that wasting of the nearby exposed or any other upper mantle provided very little (if any) input to the sediment. Overlapping clinopyroxene geochemical affinities between the gabbro and sheeted dolerite dyke sequences from Macquarie Island with the clinopyroxene clasts in the sedimentary rocks may suggest a lower crustal rock derived provenance. U-Pb geochronology of detrital zircons yielded ages of ~27 Ma, which is ca 20 Ma older than published ages of the island. This demonstrates a significantly more distal provenance than has previously been proposed for these sedimentary rocks, and excludes the island as the source of the zircons. Overlapping detrital zircon REE patterns indicate the zircons in the sedimentary rocks from across the island were all derived from a similar oceanic rock type. The most proximal crust of the appropriate age is crust that formed at the South East Indian Ridge (SEIR) on the western side of the Australia-Pacific plate boundary. Furthermore, it is unlikely that the clinopyroxene clasts in the volcaniclastic sedimentary rocks of this study were derived from the exposed lower crust on Macquarie Island, given the total lack of zircons of appropriate age. Therefore it is likely that the clinopyroxene clasts were also derived from a fault scarp along a long offset transform that uplifted and exposed SEIR crust. These results, combined with recently published paleogeographic reconstructions favour the SEIR crust as the most likely potential source area for the Macquarie Island volcaniclastic sandstones.