Enriched Asthenosphere and Depleted Plumes

Abstract

“Lithosphere” and “asthenosphere” are mechanical concepts; “depleted mantle” (DM), “enriched mantle” (EM), and “primitive mantle” (PM) are chemical concepts. Upper mantle, lower mantle, and D″ are seismological subdivisions. Geochemistry provides few constraints on the locations of mantle reservoirs, but it is generally assumed that the MORB reservoir (DM) lies immediately below the lithosphere, i.e., DM is the asthenosphere or in the asthenosphere. This leads to various paradoxes involving the temporal and spatial distributions of non-depleted basalts. To accommodate evidence for a widespread, shallow, enriched (relative to MORB) layer, it has been proposed that giant plume heads arise, as needed, from the core-mantle boundary, or that continental lithosphere delaminates and contaminates the shallow mantle. “Fossil plume heads” have been proposed to accommodate the need for long-lived, enriched, shallow mantle. Evidence is presented that a global, shallow, enriched layer is a permanent part of the upper mantle, and is absent or attenuated only when long-sustained spreading has replaced it by upwelling depleted mantle. It is the absence of EM, rather than its presence, that needs to be explained. I call this shallow, enriched (or metasomatized) region the “perisphere” (for “all around”). It is constantly refreshed by fluids from subducting slabs and by residual melts trapped beneath the lithosphere. There is no one-to-one correspondence of the perisphere with the lithosphere or asthenosphere, since it can be either hot or cold, weak or strong, fertile or infertile. The perisphere protects the underlying depleted mantle from recycling and contamination. Recognition that noble gases and high ^3He/^4He material collect in pelagic sediments increases the power of the recycling hypothesis. The shallow mantle is the likely sink for recycled sediments and slab-derived fluids, and the likely source of the enriched components of island-arc, ocean-island, and continental flood basalts. I argue that enriched magmas are caused neither by melting of continental lithosphere in continental domains nor by channeling in narrow hoses from point-source hotspots in the oceanic domain. Large-scale enriched domains exist in the shallow mantle. Narrow point-like or line-like eruptive centers are the result of lithospheric control, and are not the dimensions of active plume-like upwellings. I attribute the homogeneity of the MORB source to isolation from the effects of recycling rather than to efficient stirring. The best candidates for hot plume material are high-temperature magmas, picrites, and komatiites. Geochemically these are more similar to MORB than to enriched magmas thought to be plume derived. The tectonic context of many socalled hotspot magmas implies that it is the lithospheric dynamics of the upper Earth that control eruption, not the fluid dynamics of the deep Earth or lower mantle.