Constraints on mantle evolution from Ce-Nd-Hf isotope systematics. (1st March 2020)

• Record Type: Journal Article
• Authors: Willig, Michael; Stracke, Andreas; Beier, Christoph; Salters, Vincent J.M.
• Is Part Of: Geochimica et cosmochimica acta. Volume 272(2020)
• Page Start: 36
• Page End: 53
• Publication Date: 2020-03-01
• Subjects: OIB -- MORB -- Cerium isotopes -- REE -- Mantle heterogeneity -- Depleted mantle; Geochemistry -- Periodicals; Meteorites -- Periodicals; Géochimie -- Périodiques; Météorites -- Périodiques; Geochemie; Astrochemie; Electronic journals; 551.905
• DOI: 10.1016/j.gca.2019.12.029
• Languages: English
• ISSNs: 0016-7037

Abstract

Abstract: Mantle evolution is governed by continuous depletion by partial melting and replenishment by recycling oceanic and continental crust. Several important unknowns remain, however, such as the extent of compositional variability of the residual depleted mantle, the timescale, mass flux and composition of recycled oceanic and continental crust. Here, we investigate the Ce-Nd-Hf isotope systematics in a globally representative spectrum of mid ocean ridge and ocean island basalts. Using a Monte Carlo approach for reproducing the observed Ce-Nd-Hf isotope variation shows that the type and age of depleted mantle and recycled crust have the dominant influence on the slope, scatter, and extent of the modeled Ce-Nd-Hf isotope array. The model results suggest a relatively young (<1.5 Ga) average depletion age of the depleted mantle, consistent with Nd and Os isotope model ages of abyssal peridotites, and an apparent moderate extent of incompatible element depletion. The latter, however, is deceiving, because it reflects a natural sampling bias, resulting from melting an inherently heterogeneous depleted mantle. In principal, recycling of oceanic crust can explain most of the isotopic range of the isotopically enriched end of the Ce-Nd-Hf mantle array, but only if the entire compositional variability of the recycled crust is preserved during recycling, residence in the mantle, and re-melting. The latter is unlikely, however, because many sources of internal chemical variance average out on the scale of the bulk oceanic crust, during residence in the mantle, and subsequent sampling by partial melting. Moreover, both the slope and limited scatter of the observed Ce-Nd-Hf mantle array show that recycling of bulk oceanic crust, that is, both the extrusive basalts and intrusive gabbros of the lower oceanic crust must be considered, and are key to better understand crust-mantle cycling in general. The Monte-Carlo simulation also indicates that the return flux from the continental crust into the mantle mainly derives from the lower continental crust, consistent with current models of continental crust evolution, which all require that a substantial amount of the mafic lower continental crust must be recycled into the mantle to maintain the average andesitic composition of the continental crust.