Compositional Variations in the Sub-Continental Lithospheric Mantle and Implications for Its Role in Basalt Petrogenesis

Abstract

The role of the sub-continental lithospheric mantle (SCLM) as a source for continental basalts has been strenuously debated. On the one hand, the SCLM provides a convenient and compositionally distinct alternative to the continental crust as a reservoir in which to store the light rare earth element rich materials that will evolve low 143Nd/144Nd. On the other hand, major element, Re/Os isotope and density considerations suggest that the bulk of the lithospherie mantle is depleted in basaltic components calling into question its ability to generate large volume tholeiitic ~34rovinces. Coupled variat ions in initial 3Nd/144Nd and aS7OsflgSOs for Karoo picrite basalts (Ellam et al., 1992) indicated the involvement of a low Sm/Nd, low Re/Os source that has thus far only been encountered elsewhere in lithospheric mantle xenoliths (e.g., Walker et al, 1989) and lamproite (Lambert et aL, 1991) apparently confirming SCLM involvement. However, these data also suggested the involvement of asthenospheric or plume material with relatively high initial 143Nd/144Nd and lS7Os/lSSOs. A model in which small volume melts derived from the SCLM mix with sublithospheric magmas (MeKenzie, 1989; Ellam & Cox, 1991) is consistent with these conclusions. Here we investigate the validity of such a model for flood basalts elsewhere. At issue are two main points; firstly, is the lithospherie component typically a small volume melt or are some flood basalts derived from predominantly or wholly lithospheric sources; secondly, how typical is the well-studied Karoo example in which lithospherie involvement is now most clearly demonstrated?