Melt–rock reaction in the lower oceanic crust and its implications for the genesis of mid-ocean ridge basalt

Abstract

Primitive cumulates from a 2–3 Ma old gabbro massif exposed in the Kane Megamullion (23°N, Mid-Atlantic Ridge) contain abundant clinopyroxene with high Mg# (86–91). Such magnesian clinopyroxenes have hitherto been taken to signify crystallization at elevated pressures. Kane clinopyroxenes, however, are dominantly oikocrysts that overgrow olivine and plagioclase, indicating crystallization occurred at low pressure. The oikocrysts have textures and compositions indicative of disequilibrium processes. First, many of the oikocrysts enclose resorbed plagioclase with lower anorthite contents than plagioclase in the host rock, and olivine is notably absent as a chadacryst despite being abundant in the host rock. Second, the oikocryst minor element compositions are inconsistent with equilibrium growth from a MORB melt. These data indicate that high-Mg# clinopyroxene in the Kane gabbros formed as a result of reaction between primitive cumulates and migrating melt in the lower oceanic crust, with clinopyroxene and secondary plagioclase growing at the expense of olivine and primary plagioclase. Thus high-Mg clinopyroxene does not result from high-pressure crystallization as has been inferred previously. Assimilation–fractional crystallization modeling indicates that melts undergoing such reactions are enriched in Al 2O 3 and MgO and depleted in CaO and SiO 2. This effect is similar to that expected for fractional crystallization of MORB at elevated pressures, and reacted melts yield higher calculated pressures than starting melts. This suggests that the CaO–Al 2O 3–MgO–SiO 2 relationships of MORB may result from melt–rock reaction, and that calculated pressures of MORB fractionation are overestimated as a result. Melt–rock reaction in the lower oceanic crust may thus account for both lines of evidence for high-pressure fractionation of MORB.