Markers of the pyroxenite contribution in the major-element compositions of oceanic basalts: Review of the experimental constraints

International audienceBased on previous and new results on partial melting experiments of pyroxenites at high pressure, we attempt to identify the major element signature of pyroxenite partial melts and to evaluate to what extent this signature can be transmitted to the basalts erupted at oceanic islands and mid-ocean ridges. Although peridotite is the dominant source lithology in the Earth's upper mantle, the ubiquity of pyroxenites in mantle xenoliths and in ultramafic massifs, and the isotopic and trace elements variability of oceanic basalts suggest that these lithologies could significantly contribute to the generation of basaltic magmas. The question is how and to what degree the melting of pyroxenites can impact the major-element composition of oceanic basalts. The review of experimental phase equilibria of pyroxenites shows that the thermal divide, defined by the aluminous pyroxene plane, separates silica-excess pyroxenites (SE pyroxenites) on the right side and silica-deficient pyroxenites (SD pyroxenites) on the left side. It therefore controls the melting phase relations of pyroxenites at high pressure but, the pressure at which the thermal divide becomes effective, depends on the bulk composition; partial melt compositions of pyroxenites are strongly influenced by non-CMAS elements (especially FeO, TiO2, Na2O and K2O) and show a progressive transition from the liquids derived from the most silica-deficient compositions to those derived from the most silica-excess compositions. Another important aspect for the identification of source lithology is that, at identical pressure and temperature conditions, many pyroxenites produce melts that are quite similar to peridotite-derived melts, making the determination of the presence of pyroxenite in the source regions of oceanic basalts difficult; only pyroxenites able to produce melts with low SiO2 and high FeO contents can be identified on the basis of the major-element compositions of basalts. In the case of oceanic island basalts, high CaO/Al2O3 ratios can also reveal the presence of pyroxenite in the source-regions. Experimental and thermodynamical observations also suggest that the interactions between pyroxenite-derived melts and host peridotites play a crucial role in the genesis of oceanic basalts by generating a wide range of pyroxenites in the upper mantle: partial melting of such secondary pyroxenites is able to reproduce the features of primitive basalts, especially their high MgO contents, and to impart, at least in some cases, the major-element signature of the original pyroxenite melt to the oceanic basalts. Finally, we highlight that the fact the very silica depleted compositions (SiO2 < 42 wt.%) and high TiO2 contents of some ocean island basalts seem to require the contribution of fluids (CO2 or H2O) through melting of either carbonated lithologies (peridotite or pyroxenite) or amphibole-rich veins