The Paradox of the Axial Melt Lens: Petrology and Geochemistry of the Upper Plutonics at Hess Deep

Abstract:

The axial melt lens (AML) is a steady-state magma-rich body located at the dyke-gabbro transition at intermediate- and fast-spreading ridges. It is widely believed to be the reservoir from which mid-ocean ridge basalt (MORB) is erupted. The paradox of the axial melt lens is that the plutonic rocks that occur at this level are far too evolved to be in equilibrium with MORB, which is basaltic by definition; hence, the plutonic and volcanic records do not match. We explore this paradox by study of the first comprehensive sample suite of the uppermost plutonics of a fast-spreading ridge, taken by remotely-operated vehicle from the Hess Deep rift during cruise JC21. 23 samples (8 dolerites, 14 gabbronorites, and 1 gabbro) were collected from a section containing the transition from the uppermost gabbroic section into sheeted dykes. We present the results of a detailed petrographic and microanalytical investigation of these samples. They are dominated by evolved, varitextured (both in hand sample and thin section) oxide gabbronorites; olivine occurs in only one sample. A preponderance of the samples have positive Eu/Eu* and Sr/Sr*, indicating a cumulate origin. However, the minerals have evolved compositions, and are in equilibrium with melts significantly more evolved than East Pacific Rise MORB. Furthermore, the trace element contents of clinopyroxene differ significantly from clinopyroxene in equilibrium with MORB, being more enriched in incompatible elements.

To account for both the evidence of derivation of MORB from the AML and the evolved nature of its rock record, we posit that the AML must be fed by melts on two different timescales: continual low-volume feeding by evolved interstitial melt from the cumulus pile below is augmented episodically by delivery of high volumes of more primitive melt. The latter episodes may trigger eruptions; hence the primitive melts are held in the magma chamber for only short periods, and erupt on the seafloor before significant crystallisation in the AML has taken place. This model for the feeding of the AML provides ample opportunity for mixing between the relatively primitive melts and the evolved, trace-element-rich melt, and accounts for the observed over-enrichment in incompatible elements of MORB.