We report geochronologic, whole‐rock geochemical and Sr–Nd isotopic analyses of the granites that are exposed to both the north and the inside of the Bangong–Nujiang Suture (BNS) zone as well as the implications for the Mesozoic history of Tibet. To the north of the BNS, the Larelaxin pluton consists of I‐type quartz diorite and highly fractionated I‐type biotite granite. The Larelaxin pluton is enriched in large‐ion lithophile elements (LILE) but depleted in high‐field‐strength elements (HFSE); therefore, it exhibits the features of volcanic arc rocks. The initial Sr (0.7102 to 0.7215) and negative εNd (t) (−2.91 to −5.20) values imply a mixture of depleted mantle and continental crust. The mean 206Pb/238U zircon age is 168 Ma; we therefore propose that the Bangong–Nujiang Ocean (BNO) had already been subducted beneath the Qiangtang terrain by the middle Jurassic. Inside the BNS, the Rutog granites intruded into the Lagongtang and Duoren formations, which show a continental margin and a forearc basin sedimentary facies, respectively. The mean 206Pb/238U zircon age is 101 Ma. The Rutog granites are monzogranites with a high Na/K ratio (Na2O/K2O > 1) and a high LILE/HFSE ratio, and A/CNK < 1.1. The high Sr/Y ratio (22 to 56) implies that these granites are adakitic. The low initial Sr (0.7044 to 0.7055) and positive εNd (t) (+1.46 to +2.70) values indicate that the protolith of the Rutog granites originated mainly from a depleted source. We attribute the Rutog plutonism to the development of an oceanic arc during the continuing northward subduction of the BNO and propose that the Rutog adakitic granites were formed by melting of the subducted BNO crust with limited crustal contamination. Copyright © 2013 John Wiley & Sons, Ltd.

The subduction and close of the Mesozoic Bangong–Nujiang Ocean (BNO) led to a collision of the Lhasa and Qiangtang blocks, which formed the backbone of the Tibetan Plateau (the largest and highest plateau on Earth). However, the detailed subduction processes (in particular, the oceanic subduction processes) within the BNO are still not clear. Here, we focus on the plutonic complex of the oceanic arc in the Bangong–Nujiang suture (BNS) and report field observations on zircon U–Pb ages, Lu–Hf isotopes, and the Al-in-hornblende barometry of quartz diorites from the Lameila pluton in western Tibet. Zircon from the quartz diorites yielded a LA-ICP-MS U–Pb age of 164 Ma. The zircon showed very positive εHf(t) values from 10.5 to 13.9, suggesting the Lameila pluton was likely sourced from the depleted-mantle wedge, which is in contrast with contemporary (164–161 Ma) volcanic rocks in the region that had negative εHf(t) values of −7.4 to −16.2 and a magma source from partial melting of subducted sediments. The Lameila pluton showed a temperature-corrected Al-in-hornblende pressure of 3.9 ± 0.8 kbar, corresponding to an emplacement depth of 13 ± 3 km. Therefore, the thickness of the Jurassic oceanic arc crust must have doubled since the initial growth of the oceanic arc on the BNO crust, with a crustal thickness of 6.5 km during the Middle Jurassic. In combination with previous works on volcanic rocks, this study further supports a two-subduction zone model in association with the BNO during the Middle Jurassic, namely, a north-dipping BNO–Qiangtang subduction zone and an oceanic subduction zone within the BNO. The latter oceanic subduction zone produced the depleted-mantle-derived Lameila pluton and the subducted sediment-derived volcanic rocks in the fore arc.