Peridotite versus pyroxenite input in Mongolian Mesozoic-Cenozoic lavas, and dykes

Abstract

Here we test for peridotite versus pyroxenite input in Mongolian Mesozoic and Cenozoic magmatism. A combination of new 40Ar/39Ar radiometric dating results, whole-rock major- and trace-element, SrNd isotope, and mineral phenocryst geochemical data is used to decipher the petrogenesis of Cretaceous lavas (Tsagaan Nuur and Khukh Tolgoi) and dykes (Samaan Damba) from the Argalant Range, Gobi-Altai (Mongolia). This magmatism is compared to Cretaceous asthenospheric mantle-derived basalts from Tsost Magmatic Field and Cenozoic volcanism from the Gobi-Altai, Khangai Range Watershed, Tariat and Togo to assess changing source conditions. We also compare this magmatism to Cenozoic magmatism from the North China Craton. The Argalant Range magmatism has geochemical signatures consistent with the involvement of both peridotite and pyroxenite-like components, and we suggest that this pyroxenite-like component was obtained through the melting of metasomatized subcontinental lithospheric mantle (SCLM). Mineral-liquid thermobarometer results for samples from Khukh Tolgoi and Samaan Damba indicate that upwelling magma stalled at ~30 km depth, before finally traversing further to surface. A model to explain Mesozoic magmatic genesis is presented here, whereby piecemeal delamination and convective erosion of a metasomatized SCLM drives magmatism. The Cenozoic volcanism also has geochemical signatures consistent with the melting of non-peridotite components, and the presence of samples with >9 wt% MgO from Khangai Range Watershed, Tariat and Togo enabled assessments on the relative contribution of non-peridotite melt input. We suggest that magmatism from Togo contains the greatest amount of non-peridotite melt input, followed by Tariat and Khangai Range Watershed localities. We hypothesize that intermittent Cenozoic magmatism is the result of a slab graveyard under East Asia foundering into the upper mantle.