Adakite-like porphyries from the southern Tibetan continental collision zones: evidence for slab melt metasomatism
Tóm tắt
We present new whole rock trace element and Pb-isotope data for a suite of Neogene adakitic rocks that formed during the post-collisional stage of the India-Asia collision in an east-west- trending array along the Yalu Tsangpo suture. Compared to classic ‘adakites’ that form along certain active convergent plate margins, the Tibetan adakitic rocks show even stronger enrichment in incompatible elements (i.e. Rb, Ba, Th, K and LREEs) and even larger variation in radiogenic (Pb, Sr, Nd) isotope ratios. Tibetan adakitic rocks have extraordinarily low HREE (Yb: 0.34–0.61 ppm) and Y (3.71–6.79 ppm), high Sr/Y (66–196), high Dyn/Ybn and Lan/Ybn. They show strong evidence of binary mixing both in isotopic space (Sr-Nd, common Pb, thorogenic Pb) and trace element systematics. The majority of the adakitic rocks in south Tibet, including published and our new data, have variational Mg# (0.32–0.70), clear Nb (and HFSE) enrichment, the lowest initial 87Sr/86Sr and 206Pb/204Pb ratios, and the highest 144Nd/143Nd ratios of all Neogene volcanic rocks in south Tibet. These results indicate an involvement of slab melts in petrogenesis. Major and trace element characteristics of the isotopically more enriched adakites are compatible with derivation from subducted sediment but not with assimilation of crustal material. Thus, the south Tibetan adakitic magmas are inferred to have been derived from an upper mantle source metasomatised by slab-derived melts. An interesting observation is that temporally coeval and spatially related lamproites could be genetically related to the adakitic rocks in representing partial melts of distinct mantle domains metasomatised by subducted sediment. Our favoured geodynamic interpretation is that along-strike variation in south Tibetan post-collisional magma compositions may be related to release of slab melts and fluids along the former subduction zone resulting in compositionally distinct mantle domains.