Phase-equilibria constraints on the genesis and magmatic evolution of oceanic basalts

Abstract

This review is concerned with the genesis and subsequent evolution of the magmas of oceanic spreading centres, here termed ridge basalts, and of ocean islands, as exemplified by the Kilauea volcano of Hawaii. A summary of geophysical data emphasizes the limitations to the long-standing postulate, based on ophiolite studies, that spreading centres are underlain by extensive, permanent shallow-level magma chambers. Against this background, the published data from phase-equilibria studies of basic and ultrabasic rocks, and their synthetic analogues, are used to constrain a discussion of the relative importance of the following processes in oceanic magmatism: polybaric fractional crystallisation, under either closed- or open-system conditions; magma mixing; variable degrees of partial fusion of mantle with diverse compositions, under isobaric or polybaric conditions. The controversy as to whether the primary magmas of ridge basalts are picritic or basaltic is debated, with emphasis on some of the difficulties caused by the widespread use of ophiolite analogies in discussions of this problem. It is concluded that the hypothesis most compatible with all the foregoing constraints is that ridge basalts originate by small-to-moderate degrees of essentially-anhydrous mantle fusion (up to about 25%) over a pressure range between about 5 and 25 kb, leaving a lherzolite residuum. Over a rising convective mantle plume or jet, such as the Azores region, there is evidence that the total amount of mantle fusion is increased, so that the residue is harzburgite. The results of recent fluid-dynamical analyses of decompression melting within mantle upwelling beneath spreading-centres are shown to agree very well with the phase-equilibria-based model. There is also close convergence between the phase-equilibria constraints and published fluid-dynamical and geochemical approaches to estimating the composition of Kilauean primary magma. This appears to be a picrite, formed by about 15% of decompression melting (lherzolite residuum) in an upwelling hot plume from sources probably several hundred km deep. As the melt leaving the rising jet collects at about 70 km depth, at the base of the sub-Hawaiian lithosphere, its major-element composition converges with those postulated in alternative models which invoke shear-melting at this depth.