The continental lithospheric mantle: characteristics and significance as a mantle reservoir.

TLDR: Observations lead us to question the requirement for CLM participation in the source of oceanic magmas and to promote consideration of a mantle that is chemically heterogeneous on all scales.

Abstract: The continental lithospheric mantle (CLM) is a small-volumed (ca. 2.5% of the total mantle), chemically distinct mantle reservoir that has been suggested to play a role in the source of continental and oceanic magmatism. It is our most easily identifiable reservoir for preserving chemical heterogeneity in the mantle. Petrological and geophysical constraints indicate that the maximum depth of the CLM is ca. 250 km. There is a clear secular variation of CLM composition, such that CLM formed in the last 2 Gyr is less depleted and therefore less dynamically stable than ancient CLM formed in the Archean. We present new trace-element data for kimberlite-hosted lithospheric peridotites and metasomites. These data, combined with other data for spinel peridotites from non-cratonic regions, show that neither hydrous nor anhydrous lithospheric mantle xenoliths make suitable sources for continental or oceanic basalts. Addition of a hydrous phase, either amphibole or phlogopite, to depleted peridotite results in positive Nb and Ti anomalies that are the opposite of those predicted for some flood-basalt sources on the basis of their trace-element abundances. Overall, the Sr and Nd isotopic composition of cratonic and non-cratonic CLM is close to bulk Earth, with cratonic CLM showing small numbers of extreme compositions. Thus, while the CLM is certainly ancient in many locations, its average composition is not significantly 'enriched' over primitive upper mantle, in terms of either radiogenic isotopes or trace elements. These characteristics, plus a change in lithospheric chemistry with depth, indicate that the elemental and isotopic composition of lithospheric mantle likely to be re-incorporated into convecting mantle via delamination/thermal erosion processes is probably not very distinct from that of the convecting mantle. These observations lead us to question the requirement for CLM participation in the source of oceanic magmas and to promote consideration of a mantle that is chemically heterogeneous on all scales. Hf and Nd isotope compositions identify a distinctive source component in deeply derived alkaline volcanics associated with continents. This component cannot be constrained to the CLM but may originate from a deeper reservoir of ancient, subducted oceanic crust stored in the mantle.