B. C. Storey

The Chon Aike province of Patagonia and related rocks in West Antarctica: A silicic large igneous province

The field occurrence, age, classification and geochemistry of the Mesozoic volcanic rocks of Patagonia and West Antarctica are reviewed, using published and new information. Dominated by rhyolitic ignimbrites, which form a bimodal association with minor mafic and intermediate lavas, these constitute one of the largest silicic igneous provinces known, equivalent in size to many mafic LIPs. Diachronism is recognized between the Early–Middle Jurassic volcanism of eastern Patagonia (Marifil and Chon Aike formations) and the Middle Jurassic–earliest Cretaceous volcanism of the Andean Cordillera (El Quemado, Ibanez and Tobifera formations). This is accompanied by a change in geochemical characteristics, from relatively high-Zr and -Nb types in the east to subalkaline arc-related rocks in the west, although the predominance of rhyolites remains a constant factor. All of the associated mafic rocks are well fractionated compared to direct mantle derivatives. Petrogenetic models favour partial melting of immature lower crust as a result of the intrusion of basaltic magmas, possibly with some hybridisation of the liquids and subsequent fractionation by crystal settling or solidification and remelting. The formation of large amounts of intracrustal silicic melt acted as a density barrier against the further rise of mafic magmas, which are thus rare in the province.

Antarctica-New Zealand rifting and Marie Byrd Land lithospheric magmatism linked to ridge subduction and mantle plume activity

Mid-Cretaceous igneous rocks of central Marie Byrd Land, Antarctica record a rapid change from subduction-related to rift-related magmatism. This correlates with the final stages of subduction of the Phoenix plate and the subsequent rifting of New Zealand from West Antarctica, prior to the opening of the Southern Ocean. Rift magmatism produced diverse A-type granitoids and mafic intrusive rocks of continental flood-basalt affinity that were derived ultimately from lithospheric mantle sources. Rifting was caused by changes in plate boundary forces; however, mantle plume activity may have begun in mid-Cretaceous time, triggering melting of the lithosphere and controlling the locus of rifting.

Geochronology and geochemistry of pre-Jurassic superterranes in Marie Byrd Land, Antarctica

Marie Byrd Land, Antarctica, is a major part of the proto-Pacific supercontinental margin. On the basis of new geochronological and geochemical data relating to its pre-Jurassic evolution, Marie Byrd Land is subdivided into western or interior (“Ross”) and eastern or exterior (“Amundsen”) provinces, equivalent to two superterranes in New Zealand. The Ross province is characterized by Cambrian? metagraywackes and I-type orthogneiss dated at 505±5 Ma by U-Pb SHRIMP (Sensitive High Resolution Ion Microprobe). Its magmatic record consists of Devonian-Carboniferous (375±5 Ma and circa 339±6 Ma), predominantly I-type granitoids, and further minor granitic magmatism in Permo-Triassic times. This Paleozoic history is comparable to that of the Gondwana margin in northern Victoria Land, western New Zealand, and SE Australia. The Amundsen province has no observed Paleozoic graywacke succession; evidence from Rb-Sr and U-Pb SHRIMP dating supports calc-alkaline granitoid events in Ordovician/Silurian (450–420 Ma) and Permian (276±2 Ma) times. The latter may be the previously unknown source of Permian volcanic detritus in the Ellsworth and Transantarctic mountains. The Amundsen province is considered to be the equivalent of the Median Tectonic Zone of New Zealand, and arc magmatism of comparable ages is found in the Antarctic Peninsula and Thurston Island. The underlying lithosphere of the two provinces may be distinguished by Nd isotope data; granitoids and metasedimentary rocks of the Ross province have Meso-Proterozoic Nd model ages, generally 1300–1500 Ma, compared to 1000–1300 Ma for the Amundsen province. On the basis of published palaeomagnetic data, the two provinces amalgamated to form Marie Byrd Land in mid-Cretaceous times, only shortly before rifting of the New Zealand continental block away from Antarctica.

Mantle processes during Gondwana break-up and dispersal

Abstract This paper reviews the Mesozoic continental flood basalts (CFBs) associated with the break-up and dispersal of Gondwana from 185-60 Ma, the conditions for melt generation in mantle plumes and within the continental mantle lithosphere, and possible causes for lithospheric extension. The number of CFB provinces within Gondwana is much less than the number of mantle plumes that are likely to have been emplaced beneath it in the 300 Ma prior to its initial break-up. Also, the difference between the age of the peak of CFB volcanism and that of the oldest adjacent ocean crust decreases with the age of volcanism during the break-up and dispersal of Gondwana. The older CFBs of Karoo and Ferrar appear to have been derived largely from source regions within the mantle lithosphere. It is only in the younger Paranâ-Etendeka and Deccan CFBs that there are igneous rocks with major, trace element and radiogenic isotope ratios indicative of melting within a mantle plume. These younger CFBs are also clearly associated with hot spot traces on the adjacent ocean floor. The widespread 180 Ma magmatic event is attributed to partial melting within the lithosphere in response to thermal incubation over 300 Ma. In the case of the Ferrar (Antarctica) this was focussed by regional plate margin forces. The implication is that supercontinents effectively self-destruct in response to the build up of heat and resultant magmatism, since these effects significantly weaken the lithosphere and make it more susceptible to break-up in response to regional tectonics. The younger CFB of Paranâ-Etendeka was generated, at least in part, because the continental lithosphere had been thinned in response to regional tectonics. While magmatism in the Deccan was triggered by the emplacement of the plume, that too may have been beneath slightly thinned lithosphere.

Mantle plumes and Antarctica-New Zealand rifting: evidence from mid-Cretaceous mafic dykes

Ocean floor magnetic anomalies show that New Zealand was the last continental fragment to separate from Antarctica during Gondwana break-up, drifting from Marie Byrd Land, West Antarctica, about 84 Ma ago. Prior to continental drift, a voluminous suite of mafic dykes (dated by Ar–Ar laser stepped heating at 107 ± 5 Ma) and anorogenic silicic rocks, including syenites and peralkaline granitoids (95–102 Ma), were emplaced in Marie Byrd Land during a rifting event. The mafic dyke suite includes both high- and low-Ti basalts. Trace element and Sr and Nd isotope compositions of the mafic dykes may be modelled by mixing between tholeiitic OIB (asthenosphere-derived) and alkaline high- to low-Ti alkaline magmas (lithospheric mantle derived). Pb isotopes indicate that the OIB component had a HIMU composition. We suggest that the rift-related magmatism was generated in the vicinity of a mantle plume. The plume helped to control the position of continental separation within the very wide region of continental extension that developed when the Pacific–Phoenix spreading ridge approached the subduction zone. Separation of New Zealand from Antarctica occurred when the Pacific–Phoenix spreading centre propagated into the Antarctic continent. Sea floor spreading in the region of the mantle plume may have caused an outburst of volcanism along the spreading ridge generating an oceanic plateau, now represented by the 10–15 km thick Hikurangi Plateau situated alongside the Chatham Rise, New Zealand. The plateau consists of tholeiitic OIB-MORB basalt, regarded as Cretaceous in age, and similar in composition to the putative tholeiitic end-member in the Marie Byrd Land dykes. The mantle plume is proposed to now underlie the western Ross Sea, centred beneath Mount Erebus, where it was largely responsible for the very voluminous, intraplate, alkaline McMurdo Volcanic Group. A second mantle plume beneath Marie Byrd Land formed the Cenozoic alkaline volcanic province.

Crustal growth of the Antarctic Peninsula by accretion, magmatism and extension

A subduction-accretion model incorporating new geophysical data is presented to explain the geology of the Antarctic Peninsula from late Palaeozoic to Cenozoic time. According to the model, the peninsula consists of overlapping accretionary, magmatic and extensional regimes that were diachronous across the peninsula and have built the crust to its present form. The crust, which contains a small proportion of sialic basement, was mainly formed by accretionary and magmatic processes and modified to its present shape by extension. The Gondwanide Orogeny for the Antarctic Peninsula is interpreted in terms of the accretionary processes.

The eastern Palmer Land shear zone: a new terrane accretion model for the Mesozoic development of the Antarctic Peninsula

A major ductile fault zone, the eastern Palmer Land shear zone, has been identified east of the spine of the southern Antarctic Peninsula. This shear zone separates newly identified geological domains, and indicates that during Late Jurassic terrane accretion and collision, two and possibly three separate terranes collided, resulting in the Palmer Land orogeny. The orogeny is best developed in eastern Palmer Land and eastern Ellsworth Land. There, shallow-marine sedimentary rocks of the Latady Formation, and a metamorphic and igneous basement complex of possible Lower Palaeozoic to pre-Early Jurassic age, are thrust and folded. This forms an arcuate, east-directed, foreland, fold and thrust belt up to 100 km wide and 750 km long, parallel to the axis of the Antarctic Peninsula. The newly identified Antarctic Peninsula domains include: (1) a parautochthonous Eastern Domain that represents part of the margin of the Gondwana continent, comparable to the Western Province of New Zealand, the Ross Province superterrane of Marie Byrd Land, the Eastern Series of south-central Chile, the Pampa de Agnía and Tepuel rocks of north Patagonia, and the Cordillera Darwin rocks of Tierra del Fuego, (2) a suspect Central Domain that represents an allochthonous, microcontinental, magmatic arc terrane, comparable to the Median Tectonic Zone of New Zealand, the Amundsen Province superterrane of Marie Byrd Land, and Coastal Cordillera of north Chile and (3) a suspect Western Domain, with strong similarities to the Eastern Province of New Zealand, Western Series of south-central Chile, and Chonos metamorphic complex of north Patagonia, that represents either a subduction–accretion complex to the Central Domain, or another separate crustal fragment. Although an allochthonous terrane hypothesis for the Antarctic Peninsula remains to be fully tested, this has much in common with models for the New Zealand and South American parts of the Pacific margin of Gondwana. The identification of a potential allochthonous terrane–continent collision zone allows us to define the edge of the Gondwana continent in the Antarctic Peninsula sector of the supercontinent margin, which has implications for Mesozoic reconstructions of Gondwana.

Pb, Nd, and Sr isotope mapping of Grenville‐age crustal provinces in Rodinia

New Pb, Nd, and Sr isotope data are presented for geochemically similar, ∼1.1–1.2 Ga, granitoids and tonalitic‐granitic orthogneisses from Antarctica, southern Africa, and the Falkland Islands and adjoining plateau, areas originally within the supercontinents of Rodinia and Gondwana. These data support contentions for the presence of a Mesoproterozoic (∼1.2 Ga) destructive plate margin running from Namaqua‐Natal (southern Africa), through the displaced microplates of the Falkland Plateau and Falkland Islands, the Haag Nunatak crustal block (West Antarctica) and into western Dronning Maud Land (East Antarctica). The bulk of these granitoids represent juvenile Mesoproterozoic additions to the crust, except for in parts of East Antarctica (i.e., the Sverdrupfjella) where older (Paleoproterozoic or Archean) crust was involved in granitoid generation. Our isotope data permit plate reconstructions in which southern Africa, East Antarctica, and the Falkland Islands and plateau were adjacent within Rodinia.

Role of subduction-plate boundary forces during the initial stages of Gondwana break-up: evidence from the proto-Pacific margin of Antarctica

Abstract In the West Antarctic sector of Gondwana, early stages of break-up are associated with the large Antarctic-Karoo-Tasman basalt province. Formation of this within-plate province was synchronous with active margin tectonics and development of both a proto-Pacific margin magmatic suite along the Antarctic Peninsula and the extensive Tobífera volcanic suite associated with the Rocas Verdes marginal basin system of southern South America and South Georgia. Extension, concurrent with subduction and oceanward migration of the magmatic focus, resulted in a broad extensional province in a back-arc and intra-arc-setting. High geothermal gradients and basalt underplating caused crustal melting on the east coast of the Antarctic Peninsula and formation of bimodal basalt-rhyolite suites. Large-ion lithophile element enriched initial rifting magmas were succeeded, at least in part of the Rocas Verdes basin, by early drift magmas of transitional chemistry and then by entirely asthenospheric MORB magmas representing lithospheric rupture and sea-floor spreading. A plate interaction model is proposed for the initial stages of Gondwana break-up relating the broad zone of mantle melting to a reduction in subduction-plate boundary forces. The change from Gondwanide compression to lithospheric extension in the Jurassic may be linked to a change from shallow to steeply dipping subduction, and to a slowing of subduction rates caused by a change in plate boundary zone parameters. A possible reduction of compressive boundary stresses may have enabled unconfined, overthickened Permo-Triassic crust to extend because of gravitational instability, thus facilitating break-up. We suggest that break-up was not plume-related, but was due to variations in the regional stress field associated with changing plate-boundary forces. The continental crust was placed under tension with substantial lithospheric thinning and decompression melting of an enriched mantle source forming the broad linear zone of within-plate magmatism. The presence of a plume beneath the Karoo province may have thermally weakened the lithosphere and induced local rifting, contributing to, but not causing the eventual separation of East and West Gondwana.

Coats Land dolerites and the generation of Antarctic continental flood basalts

Abstract On the basis of geochemical signatures, Mesozoic magmatism in Antarctica is divided into the Ferrar Magmatic Province and the Dronning Maud Land Province. The tholeiitic magmatism of the Ferrar Magmatic Province is distinguished by such features as low Ti/Y (< 200) and Zr/Y (< 5.0) ratios, negative εNd values (< −3) and relatively enriched initial 87Sr/86Sr ratios (> 0.709). All of these geochemical features indicate a major contribution from the continental mantle lithosphere in the generation of these magmas. In contrast, the Dronning Maud Land magmatism has elevated trace element ratios and εNd values (Ti/Y 250–600; Zr/Y 3.0–9.0; εNd −2 to +3) and lower initial 87Sr/86Sr ratios (< 0.707) relative to the Ferrar Magmatic Province. The trace element and isotopic correlations suggest that these magmas were derived by the mixing of an OIB like asthenospheric component with a continental lithosphere component. The transition between these two geochemical provinces is located in Coats Land. In Coats Land, the Mesozoic tholeiitic magmatism is represented by doleritic sills and minor dykes which intrude Permo-Triassic sedimentary rocks. The dolerites can be subdivided into two series based on their TiO2 contents. Series 1 dolerites (TiO2 < 1.5%) can be further subdivided into three groups, which give Ar/Ar ages of 171±6 Ma (Group 1) and 193±7 Ma (Groups 2 and 3). It is only Group 2 magmas which have trace element and isotopic signatures akin to the Ferrar Magmatic Province. Group 1 dolerites have geochemical signatures which are transitional between the Ferrar Magmatic Province and Dronning Maud Land magma types. The Ferrar Magmatic Province signature in Coats Land is confined to the early magmatic episode (193±7 Ma) and this appears to mark the initiation of rift related magmatism in this region. It is argued that extension was limited and that most of the melt was derived from the continental mantle lithosphere. In contrast, the younger rocks (176±5 Ma) have relatively lower initial 87Sr/86Sr and higher trace element ratios relative to the Ferrar Magmatic Province, and this appears to be associated with the later stages of rifting and relatively enhanced crustal extension which allowed for the encorporation of a small asthenosphere component.