The Petrology of Lavas and Associated Plutonic Inclusions of Ascension Island

Publisher Website
Google Scholar
Abstract
The lavas and pyroclastics of Ascension Island contain a suite of plutonic inclusions which range in composition from olivine-gabbro to peralkaline-granite thus paralleling, but extending beyond, the compositional range of the volcanics. The lavas range from alkali-basalt through hawaiite, trachybasalt, trachyandesite and trachyte to comendite. Basalt sensu stricto is relatively rare and there is a scarcity of analyses with 57 < SiO2 < 63. No high pressure mineral assemblages and hence no probable mantle fragments have been found. One suite of gabbros includes a small number containing cumulus and intercumulus orthopyroxene which is absent from the lavas and the remaining plutonic inclusions (except as a reaction product of olivine in some cases). These gabbros also have differing Pb isotopic composition thus suggesting derivation from a different mantle source. The remaining gabbroic-intermediate inclusions show features consistent with their being the residue of crystal fractionation of the evolving magmas now represented by the exposed lavas. The acid inclusions are peralkaline in character and chemically very similar to the acid lavas, the most evolved (granite) inclusions are unusual in that some contain vlasovite (Na2ZrSi4O11) or dalyite (K2ZrSi6O15), which are indicative of strong peralkalinity. A pegmatoid body exposed within the thick hawaiite sequence of Cricket Valley shows many features in common with the plutonic inclusions, yet is clearly relatable to its host lava. Relatively slow cooling of this body resulted in progressive changes in mineralogy which are similar to the range of mineralogies exhibited by the lavas and plutonic inclusions suggesting a similar origin, i.e. crystal fractionation and a liquid line of descent. Major element and minor element variations within the lavas suggest that crystal fractionation at low pressures is the prime differentiation process, and the very presence of cumulate blocks implies that crystal settling took place. Partial melting is seen in some plutonic inclusions but trace element evidence suggests that the acid lavas and plutonic inclusions did not evolve via partial melting of plutonic intermediate material. Mixing models for the major elements suggest that feldspar was the principal fractionating phase along with olivine, clinopyroxene and ilmenite and/or magnetite, in proportions which are in general agreement with phenocryst abundances in the lavas and modal mineralogies of the plutonic inclusions. Although there is continued depletion of FeOT with increasing SiO2, there is good evidence that an iron oxide phase is not continuously fractionating, allowing some short term FeO enrichment. Chemical analyses used to determine the density contrast between lavas and possible fractionating minerals show that settling of feldspar as well as pyroxene, olivine and an oxide phase could have occurred throughout the evolution of the suite. Coexisting magnetite and ilmenite compositions in the gabbroic inclusions and the assemblage aegirine + arfvedsonite ± aenigmatite ± fayalite in the granites suggest both crystallized under moderately reducing conditions, slightly to the HM side of the FMQ reaction curve.

This publication has 0 references indexed in Scilit: