Key points are not available for this paper at this time.
Abstract Post‐collisional potassic‐ultrapotassic rocks can provide key clues to the change of the recycled material type and/or tectonic transition in subduction‐related zones. Despite continental materials widely recognized in their sources, it remains unclear whether such continental materials were contributed by former oceanic subduction or recent continental subduction. Here we address this issue by systematically investigating previously reported and our new chemical and Sr–Nd–Pb isotopic compositions of the post‐collisional K‐rich rocks in Southeast Tibet. Kink‐like compositional variations provide solid evidence for a primary control of fractional crystallization on the evolution of these K‐rich magmas. Their primary melts are demonstrated to have been produced by partial melting of phlogopite‐bearing peridotites in subcontinental lithospheric mantle (SCLM). The trace element and Sr–Nd–Pb isotopic signatures argue against involvement of the deeply subducted Indian continent but suggest a great contribution from sediments in oceanic slabs. A thinned (∼70–100 km) and hot (∼55–70 mW/m 2 ) lithosphere is also unraveled beneath Southeast Tibet during the potassic‐ultrapotassic magmatism. Together with geophysical data, here we suggest a two‐stage geodynamic model for post‐collisional potassic‐ultrapotassic magmatism in Southeast Tibet: (a) Before the Indian‐Asia continental collision, phlogopite/K‐richterite‐bearing SCLM sources were formed through oceanic subduction‐related metasomatism; (b) After the Indian‐Asia continental collision, asthenosphere upwelling induced by post‐collisional tectonic extension or deep subduction of the Indian continental slab caused lithospheric thinning, partial melting of pre‐existing phlogopite/K‐richterite‐rich SCLM and thus K‐rich magmatism. This study provides new insights into the role of oceanic subduction and continental collision in post‐collisional potassic‐ultrapotassic magmatism.
Geng et al. (Thu,) studied this question.