Mafic dykes of the East Antarctic shield : experimental, geochemical and petrological studies focusing on the proterozoic evolution of the crust and mantle

The Vestfold Hills block of the East Antarctic Shield is an Archean granulite facies complex crossed by hundreds of mafic dykes. The dykes compose five chemically distinct suites emplaced in three separate magmatic episodes during the Early-Middle Proterozoic. Sampling of Chilled margins coupled with detailed documentation of cross-cutting relationships form the basis for studying the chemical and temporal evolution of each dyke suite as well as providing insight into the chemical evolution of the subcontinental mantle during the Proterozoic. Furthermore, the emplacement of the swarms is bracketed by tectonothermal metamorphic events at ca.2500 Ma and ca.1100 Ma. Previous studies of the metamorphic assemblages which developed during these events have provided constraints on the P,T, time path of the Shield. A principal purpose of the present study has been to determine emplacement depths of the dyke suites in order to further restrict the possible uplift path of the shield during the -1400 Ma interval separating, the final two major deformations. High pressure experimental techniques were used to determine the load pressure at the emplacement depth of two samples representing dyke suites intruded at ca.2400 Ma and ca.1360 Ma. This was accomplished by experimentally reproducing the chilled margin phenocryst assemblage and the phase compositions after first rigorously evaluating crystal/liquid KD's to ensure the studied material represented liquid compositions: The results of the experimental studies were also applied to compositionally similar dykes from the Napier Complex of Enderby Land. The studies indicate the Napier Complex and the Vestfold Hills are characterized by -2000 Ma of crustal stability following the peak metamorphic event at ca .3100 Ma. Both terrains followed identical isobaric cooling paths, to ca.2500 Ma, After which the Napier Complex remained at deep crustal levels ( -28 km) until ca.1000 Ma, while the Vestfold Hills was exhumed at a rate of -1cm/1000 a until ca. 1100 Ma. This long period of crustal stability experienced by both terrains was terminated by a Himalayan-style tectonic event which resulted in isothermal decompression at a minimum net rate of 11cm/1000 a in the Napier Complex, while crustal loading, depressed the Vestfold Hills from depths of -16 km to -22 km resulting in the formation of garnet in the mafic dykes. In the absence of more recent deformational events, it is assumed that these terrains experienced slow, erosion controlled uplift since ca.1000 Ma. The oldest undeformed mafic dyke swarm in the Vestfold Hills was emplaced at ca.2400 Ma and is composed of two contemporaneous suites, a Si02 -rich, high-Mg tholeiite suite and a Fe-rich tholeiite suite, which compare favourably to basaltic komatiite -Fe tholeiite associations of Archean greenstone belts. Normalized plots of incompatible trace elements, phase compositions, and major element ratios divide the high-Mg suite into three distinct subgroups. The trace element characteristics and isotopic studies prohibit these subgroups from being related by crystal fractionation and previous isotopic studies preclude crustal contamination as a source of chemical variations. Major and trace element evaluation also indicates that two of the chemically distinct subgroups were derived from a primitive liquid extracted from "chondritic" mantle sources leaving an olivine + orthopyroxene residue. Superimposed upon the chondritic characteristics is evidence for selective trace element contamination through a "wall-rock" reaction-type process, possibly with a plagioclasebearing lherzolite. Comparison of an estimated parental liquid composition with experimental melting studies indicates magma extraction took place at pressures of -10 kbar (35 km), consistent with the geochemical signature indicating partial melting with a plagioclase-bearing mantle. The more differentiated Fe-rich tholeiites are depleted in LIL elements compared to the high-Mg tholeiites which excludes any simple relationship between the suites by a fractional crystallization process. Three petrographically and chemically distinct groups can be identified within the Fe-rich suite. Two groups can be related in both major and trace element variations by crystal fractionation, but field relationships preclude this simple interpretation. All the groups have REE patterns indicating derivation from an unfractionated mantle, which precludes a relationship with the high-Mg suite by polybaric melt extraction of a single source. The two closely related groups display enrichment in P and Zr in excess of that due to crystal fractionation. This may represent a relatively deep level wall-rock reaction process as the third group displays elemental enrichments identical to that of the high-14g tholeiites (formed at -10 kbar). The youngest and volumetrically largest dyke swarm was emplaced in the Vestfold Hills at ca.1360 Ma and is dominated by Fe-rich tholeiites with subsidiary, though widespread alkaline dykes. The major and trace element diversity of the Fe-rich suite can be accounted for by a crystal fractionation dominated process. However, the geochemical trends are defined by clusters of analyses which in themselves define differentiation trends that cannot be related to fractionation of the phenocryst phases. Furthermore, the ordering of the groups does not relate to a simple, monotonic emplacement sequence. The nature of the small-scale geochemical trends and the complex emplacement style may be the result of "open system" magma chamber processes. Rare earth element profiles, as well as the incompatible element plots indicate the parental liquids to this suite were derived from a primitive source which had not experienced previous melting events