|
The composition of subcontinental lithospheric mantle (SCLM) varies broadly with the age of the last major tectonothermal event in the overlying crust. Archean SCLM is highly depleted, Phanerozoic terrains are generally underlain by fertile mantle, and most Proterozoic SCLM is intermediate. This secular evolution in SCLM composition implies quasi-contemporaneous formation (or modification) of the crust and its mantle root. Worldwide, there is a strong correlation between lithosphere composition and lithosphere thickness, interpreted in terms of evolution in the processes that produce SCLM.
Typical Archean mantle used in most geochemical/geophysical modelling is a garnet lherzolite composition (from peridotite xenoliths in kimberlites, mainly from the SW Kaapvaal Craton). However, most such xenoliths show repeated metasomatism, with a progression from dunite/harzburgite to depleted then fertile lherzolite. Similar refertilisation can be studied in situ in peridotite massifs (eg W Norway, Lherz). Such depleted rocks are rare in the xenolith record; the bias partly reflects the collecting of rocks useful for P-T studies, but also has a geological basis.
Seismic tomography of Archean cratons shows high-Vs volumes surrounded and dissected by zones of lower Vs. In detail, kimberlites preferentially sample the low-Vs zones, biasing our mantle sample toward the metasomatic products. The low-Vs parts can be modelled using the typical lherzolite compositions, but the higher-Vs volumes require strongly depleted dunite/harzburgite, which apparently underlies most Archean cratons to depths of ∼150 km, but is poorly sampled by kimberlites. Some of these high-velocity "roots" extend to 300-400km, and the geochemically defined LAB lies within them. What do these roots represent, and what and where is the LAB in such upper mantle domains?
Relict SCLM is also imaged as high-Vp blobs at shallow levels (0-150 km) in oceanic regions, providing a likely source for "recycled" geochemical signatures in ocean island basalts, and insights into the mechanisms of continental breakup. What is the nature of the LAB in such oceanic upper mantle domains? Hf-isotope data on zircons show that much Proterozoic crust, especially in shield areas, has Archean protoliths, suggesting that the underlying SCLM was originally Archean. The intermediate composition of much Proterozoic SCLM may reflect progressive metasomatism of Archean SCLM by asthenospheric melts, rather than a change in the nature of lithosphere-forming processes since the Archean. This refertilisation would move the LAB (defined as base of the depleted SCLM) upward through time, helping to explain the correlation between SCLM composition and depth to the LAB.
Fertile Phanerozoic SCLM, cooled to a conductive geotherm, will tend to delaminate; accretion of upwelling asthenosphere will produce new fertile SCLM, which can in turn delaminate. This cyclic process would ensure that fertile SCLM is thin SCLM.
|
