Corundum-bearing Garnet Clinopyroxenites at Beni Bousera (Morocco): Original Plagioclase-rich Gabbros Recrystallized at Depth within the Mantle?

Publisher Website
Google Scholar
Abstract
The Beni Bousera ultramafic massif, Morocco, is composed of peridotite with subordinate garnet pyroxenitc units which belong to two different families: (1) the Type I pyroxenites, which are characterized by an Fe-enrichment trend; and (2) the Type II pyroxenites, which are characterized by high but nearly constant Mg/Fe ratios and highly variable concentrations of Ca and Al; the latter family includes corundum-bearing garnet pyroxenites which resemble the peraluminous eclogites and grospydites described as xenoliths in kimberlite diatremes. The Type II pyroxenites appear as layered sheets in the peridotite, and have granuloblastic metamorphic texture. They contain a primary association of a coarse-grained assemblage (cpx + gt; cpx + gt + sp; cpx + gt + co), and a variety of secondary and tertiary associations includ ng clinopyrox-ene, orthopyroxene, olivine, spinel, corundum, sapphirine, plagioclase, and amphibole. The primary assemblage in the corundum-bearing pyroxenite is characterized by clinopyroxene rich in A12O3 (up to 20 wt%), and poor in Na2O (generally less than 2 wt.%). The clinopyroxene phase is therefore richer in the Ca-Ts molecule than in the jadeite molecule. On the other hand, the composition of the primary and secondary clinopyroxene and garnet phases shows strong variation across the pyroxenite sheets. These variations express compositional variations of the rock system across the sheets. The cpx-gt associations indicate high temperatures (1200–1350 °C) in the central parts of the sheets. The crystallization pressure may have reached at least 20 kb in the corundum-bearing assemblages. The bulk-rock composition and the compatible element's behaviour in the Type II pyroxenite sheets suggest that the modal and cryptic layering mainly resulted from igneous fractionation processes. The REE patterns of corundum-bearing Type II pyroxenite are characterized by low concentrations of HREE and by significant Eu anomalies. These, together with the high bulk-rock Sr/Nd ratios, suggest that plagioclase segregation may have played a significant part in the rock genesis. These geochemical features are similar to those described, in the literature, in some low-pressure, plagioclase-bearing adcumulates (e.g., in the crustal sequence of the Oman ophiolite). They are quite different from those observed in the Type I pyroxenite sheets in the Beni Bousera massif, whose geochemistry suggests that plagioclase played no part in the fractionation process, whereas garnet probably fractionated as an early igneous phase. The Type II pyroxenite sheets have a primary isotopic signature similar to MORB, based on the composition of leached clinopyroxene. It is concluded that the Mg-rich Type II pyroxenite sheets resulted ultimately from the fractionation of a basaltic melt at low pressure, and from the accumulation of olivine, clinopyroxene, and plagioclase along dykes cross-cutting the surrounding peridotite. The close similarities with the geochemical features in the Oman ophiolite lead us to suggest that these processes may have been operative in an oceanic crustal environment. The high-pressure and high-temperature crystallization of the ‘primary’ cpx+gt + co assemblage was achieved deep in the mantle, after subduction and/or dragging down in convection currents of this particular piece of the (oceanic?) lithosphere. Further ascent may have resulted in partial melting of peridotite and/or pyroxenite, and in the emplacement of the Type I pyroxenite sheets.

This publication has 0 references indexed in Scilit: