HR: 08:00h
AN: T31B-01 INVITED     [PDF]
TI: The Production and Destruction of Chemical and Lithological Heterogeneities in the Mantle
AU: * Hart, S R
EM: shart@whoi.edu
AF: Woods Hole Oceanographic Inst., Mail Stop 22, Woods Hole, MA 02543 United States
AU: Workman, R
EM: rworkman@whoi.edu
AF: Woods Hole Oceanographic Inst., Mail Stop 22, Woods Hole, MA 02543 United States
AB: Gast, Tilton and Hedge first showed that Earth's mantle was isotopically heterogeneous in 1964. Since that time, work on the isotopic taxonomy of Earth's mantle has delineated 5 chemical species (domains): DMM (depleted MORB mantle), FOZO (Focus zone depleted mantle), HIMU (high U/Pb mantle), EM1 and EM2 (enriched mantles). Despite decades of debate, the mode of formation and placement of these domains in the mantle is still contentious. DMM is certainly the upper mantle, and FOZO is likely the deep mantle. The "Standard Model" of Hofmann and White ascribes the other three domains to subduction and long-term storage of the oceanic lithosphere, followed by recycling to surface hotspots via plumes. The HIMU and EM domains reflect mantle containing ocean crust and/or sediment components. The challenge is to translate the arguably miniscule isotopic heterogeneities of this isotopic zoo into a macroscopic view of the mantle. What is the variability in major element composition? What are the scale-lengths of these heterogeneities? Are there mafic lithologies as well as peridotitic lithologies? Do these heterogeneities imply bulk chemical and thermal variations that are important in seismological and dynamical studies of the mantle? The Standard Model creates heterogeneities in both lithology (eclogite and peridotite) and chemistry (mafic and ultramafic) on only 10-km scale lengths. Can these survive mixing, stirring and storage for billions of years? Can they be imaged tomographically? Do they provide dynamically significant chemical or thermal buoyancy? Is the Standard Model even operative? One weakness of this model is its requirement for arbitrary and ad hoc chemical processing during subduction of the lithosphere, in order to provide the right protoliths for evolution of the mantle zoo. We have probed this weakness using the classic example of an EM2 mantle domain, the Samoa hotspot. Basalt with the most extreme Sr (0.7089) in the oceans shows a very smooth trace element pattern with only slight negative anomalies at Ti and Ba. This spidergram is inconsistent with the standard model that invokes a sediment component to explain the enrichment of the EM2 mantle source. Furthermore, it is highly unlikely that any chemical processing during subduction would "smooth out" the typically jagged spidergram of oceanic sediment. We propose instead that EM2 mantle represents the bottom, not the top, of the oceanic lithosphere. The enriched character results from a metasomatic melt/rock reaction process involving small-degree partial melts percolating up from underlying partially molten asthenosphere. This causes the lower regions (tens of kilometers?) of all oceanic lithosphere to become trace-element enriched as a matter of course; subduction and long-term storage then generates the EM2 isotope signature, and recycling in plumes/hotspots provides us the witness. The model EM2 source will have a major element composition inconsequentially different from that of DMM mantle, but with a heat production some 10 times higher (11 pW/kg versus 0.9 pW/kg).
DE: 8105 Continental margins and sedimentary basins
DE: 8120 Dynamics of lithosphere and mantle--general
DE: 8125 Evolution of the Earth
SC: Tectonophysics [T]
MN: 2003 Fall Meeting