Summary of Main Tectonic Stages of Gondwana Breakup

The first major tectonic stage in the break-up of Gondwana corresponds to an initial rifting phase that started in the Weddell Sea, initially as a back-arc basin, in the Late Jurassic (Fig. 7.4) (Lawver et al., 1991). This stage involved right-lateral transtension as East Gondwana (Antarctica, Australia, India and New Zealand) and West Gondwana (South America and Africa) moved apart with stretching beginning in the north and propagating southward (Lawver et al., 1992). Initial break-up involved complex geodynamic evolution characterized by the rotation and translation of several microplates, such as the Ellsworth Mountains block, a displaced part of the Gondwanide fold belt. The original position of the various microplates is still controversial, but the Ellsworth-Whitmore Mountains crustal block most likely originated from near south-eastern Africa (e.g. Curtis and Storey, 1996). Rotation of West Antarctic microplates must have been accomplished by c. 165 Ma (the time of opening of the Weddell Sea), and rotation of the Ellsworth-Whitmore Mountains crustal block was finished before translation into its present position by 175 Ma (Grunow et al., 1987). Rotation of microplates did not apparently involve the production of oceanic crust

(Marshall, 1994) but may have occurred as block rotation with controlling faults concealed beneath Mesozoic sedimentary basins in the Weddell Sea (Storey, 1996).

The break-up has been explained by several authors as the result of a hot mega-plume, which, according to Storey and Kyle (1999), could have promoted domal uplift and formation of a triple junction. According to Dalziel et al. (1999), the Gondwana plume may have also caused or expedited formation of the early Mesozoic Gondwanide fold belt, due to the buoyancy of a hot plume acting on the downgoing slab and causing it to flatten. In both cases, plume activity led to the production of magma batches reflecting different degrees of plume-lithosphere interaction, which migrated along crustal shear zones to ultimately form the various large igneous provinces.

The plume-related magmatic products are represented by huge within-plate mafic and felsic magmatic provinces in many Gondwana continents as well as Antarctica (Cox, 1988; White and MacKenzie, 1989; Storey, 1996).

In the Transantarctic Mountains (LeMasurier and Thomson, 1990), Jurassic mafic rocks are generally known as the Ferrar Supergroup or the Jurassic Ferrar Large Igneous Province (FLIP). They are divided into a volcanic component, the Kirkpatrick Basalt - preceded by extensive phreatomagmatic volcanoclastic rocks (Elliot et al., 2006; Viereck-Gotte et al., 2007) - and the intrusive dolerite (diabase) sills and dikes of the Ferrar Dolerite. Geochemi-cally, the FLIP is unusual with upper crustal-like characteristics such as high large ion lithophile element concentrations and enriched isotopic signatures (87Sr/86Sr> 0.709). The crust-like signature suggests derivation from an enriched lithosphere, possibly connected to subduction along the Pacific margin of Gondwana during the Palaeozoic and Mesozoic.

The mafic rocks are concentrated within a long linear belt exposed in Tasmania, Antarctica and South Africa. Cox (1988) considered that the linear pattern of the Ferrar and Tasman provinces could not be compatible with classic circular plumes and proposed a hot line rather than a hot spot. A number of rifts that intersected at the Dufek Massif (Elliot, 1992) could have favoured the development of zones of weakened lithosphere, which acted as pathways for lateral migration of magmas derived from lithospheric sources. In spite of the still not completely understood tectonic setting, it is important to note that, similarly to other continental flood basalt provinces, all the mafic products formed during a short period of eruption (Tasman province: 175718Ma; Ferrar Supergroup: 180-183Ma; Dronning Maud Land province: 177 7 2Ma; Karoo province: 18272Ma) (Hergt et al., 1989; Hooper et al., 1993; Heimann et al., 1994). Coeval felsic intrusions considered to be rift-related are known in West Antarctica (Storey et al., 1988) and southern South America (Chon-Aike or Tobifera province) (Gust et al., 1985).

The second major geodynamic stage occurred in the Early Cretaceous (Fig. 7.4), as a consequence of the change of the Gondwana break-up stress regime from dominantly north-south between East and West Gondwana to dominantly east-west with the two-plate system being replaced by a multiple-plate system (e.g. Lawver et al., 1992). This modification in stress regime is thought to have induced large-scale ductile deformation concentrated along shear zones in the Antarctic Peninsula (Storey et al., 1996) and thin-skinned deformation and inversion of existing sedimentary basins such as the Latady Basin (Kellogg and Rowley, 1989).

By c. 110 Ma, the microplates of West Antarctica had nearly reached their present location with respect to East Antarctica. In the same time period, separation also began between India and Antarctica (Lawver et al., 1991). Initial stretching between Australia and Antarctica began as early as 125 Ma (Stagg and Willcox, 1992), but sea-floor spreading was delayed to c. 95 Ma (Cande and Mutter, 1982; Veevers et al., 1990; Royer and Rollet, 1997), in the Ross embayment (Elliot, 1992), as well as extension between the Lord Howe Rise and northern New Zealand (Lawver et al., 1992). By the Late Cretaceous, Antarctica had reached its final polar location and configuration, and the final stage of break-up was completed when New Zealand (Campbell Plateau) rifted from Marie Byrd Land at 84 Ma (e.g. Stock and Molnar, 1987; Lawver et al., 1991).

In this geodynamic context, four particularly large and conspicuous intraplate fracture zones have been investigated and all related to extensive and prolonged extensional regimes spanning in time from Mesozoic to present (Fig. 7.2). These major extensional zones include:

• the Lambert Graben or Lambert Rift (East Antarctica);

• the WARS and its main part, the Ross Sea Rift (Pacific sector);

• the graben of Jutulstraumen and Penckmulde (occasionally called Jutul Penck Graben; Atlantic sector);

• the Rennick Graben as the main element of a strike-slip fault system in Victoria and Oates Lands.

The Lambert Graben developed in the East Antarctic Craton and is filled by sediments of the Permo-Triassic Beacon Supergroup. Faulting started already during the early Palaeozoic, reached its peak in the Permian and continued to the Early Cretaceous (Hofmann, 1996). Possibly, subglacial Lake Vostok belongs to the same rift system, but somewhat offset. The continuation of the Lambert Graben is the Indian Mahanadi Rift south-west of Calcutta in the state of Orissa (Hofmann, 1996), filled with sediments of the same type and age as the Lambert Graben. The reconstruction of the Gondwanan India-Antarctica fit by these graben systems coincides with reconstructions by Archean to Early Proterozoic elements for Rodinia. That means, interestingly enough, that the relationships of Antarctica and India during Rodinian and Gondwanan times do not differ substantially.

The Ross Sea Rift is extremely wide (about 1,000 km). Its subsidence started during the late Mesozoic (about 140 Ma ago), reached its main activity in the Early Tertiary (about 40 Ma ago) and produced an enormous relief at its western shoulder. The difference in altitude between the tops of the Transantarctic Mountains and the floor of the adjacent Ross Sea exceeds 14 km. The crustal extension is combined with alkaline intra-continental volcanism, which is still active at Mt. Erebus (3,794 m) and at Mt. Melbourne (2,732m), both located in Victoria Land (Kyle and Cole, 1974; Worner et al.. 1989; Kyle, 1990a,b; Tessensohn and Worner, 1991).

The Jutul Penck Graben of western Dronning Maud Land originated probably around 140 Ma ago or a little bit later (Jacobs and Lisker, 1999). The graben marks the boundary of the Grunehogna Craton towards the south-east and thus it follows a much older geological structure. A possible continuation of the Jutul Penck Graben into the still active East African rift system is under discussion (Grantham and Hunter, 1991). By no means is this out of the question, because parts of the East African rift system were already active in the Jurassic (Ring and Betzler, 1993).

The strike-slip fault system of Victoria and Oates Lands runs obliquely to the Ross Sea Rift and is cut by it. The principle element of the system is the Rennick Graben, which is presently active, as demonstrated by earthquakes in 1952, 1974 and 1998. The graben contains downfaulted Ferrar volcanics and sediments of the Beacon Supergroup, which have been spectacularly folded and squeezed onto the graben shoulders (Rossetti et al., 2003). This demonstrates alternating dextral transpression and transtension (with formation of pull-apart basins), being parts of a complicated strike-slip system in detail (Rossetti et al., 2003). The 1974 quake happened some 120 km to the west at the parallel structure of the Matusevich Glacier. There is an ongoing discussion, as to whether this strike-slip fault system constitutes the continuation of oceanic fracture zones between Australia and Antarctica (e.g. the Tasman Fracture Zone) into the continental crust of Antarctica (Salvini et al., 1997; Salvini and Storti, 2003; Kleinschmidt and Laufer, 2006).

0 0

Post a comment