Miocene adakites in south Tibet: Partial melting of the thickened Lhasa juvenile mafic lower crust with the involvement of ancient Indian continental crust compositions

Although postcollisional adakitic rocks are widely distributed in the southern Lhasa subterrane, their petrogenesis remains controversial. Complex petrogenesis models, mainly including partial melting of subducted oceanic crust, partial melting of the Indian lower continental crust, and magma mixing, are pivotal in reconstruction of the postcol-lisional dynamic processes in south Tibet. In order to constrain the geodynamic processes, we present systemic geochronological and geochemical data for newly discovered adakitic dikes in the Xigaze area, southern Lhasa subterrane. Based on the K2O and Na2O contents, the Xigaze dikes can be divided into K-rich and Na-rich dikes. Zircon U-Pb dating for the Xigaze K- and Na-rich dikes yielded ages of ca. 10.31 Ma and 14.78-12.75 Ma, re-spectively. The K-rich dikes show porphyritic texture and are characterized by high SiO2 (68.91-69.59 wt %) and K2O (5.53-5.68 wt %) contents and low Na2O/K2O (0.48-0.60) ratios, with Al2O3/(CaO + Na2O + K2O) (=A/CNK) ratios of 1.07-1.23. They have lower MgO (0.63-0.64 wt %), Mg# (37-39), and Cr (18.56-26.62 ppm) and Ni (4.37-4.62) contents. In addition, the K-rich dikes display enriched (La/Yb], = 65-68) light rare earth elements (LREEs), low concentrations of heavy rare earth elements (HREEs) and Y (e.g., Yb = 0.83-0.86 ppm; Y = 10.56-11.55 ppm), and high Sr (841-923 ppm), with high Sr/Y (74-84) ratios, indicating geochemical characteristics of typical adakitic rocks. Compared with the K-rich dikes, the Na-rich dikes also display porphyritic texture, but they have lower SiO2 (59.14-64.87 wt %) and K2O (1.98-3.25 wt %) contents, and higher Na2O (4.43-5.64 wt %) and MgO (1.40-3.08 wt %) contents, Mg# (46-59), and Cr (22.62-82.93 ppm) and Ni (8.91-39.76 ppm) contents. The HREE abundances (e.g., Yb = 0.36-0.81 ppm; Y = 5.30-10.56 ppm) of the Na-rich dikes are generally lower than the K-rich dikes. These Na-rich dikes are also characterized by adakitic geochemical features with high Sr/Y (60-223) but low (La/Yb)(N) (15-40) ratios. Both the K-rich and Narich dikes display distinct whole-rock-element geochemistry and Sr-Nd isotopic composition, with (Sr-87/Sr-86)(i) = 0.7121, epsilon(Nd)(t) = -8.62 to -8.11 and (Sr-87/Sr-86)(i) = 0.7054-0.7086, epsilon(Nd)(t)= -7.55 to -1.23 for K-rich and Na-rich dikes, respectively, which indicate different magma sources for the two types of dikes. The K-rich dikes were most likely derived from partial melts of Lhasa juvenile mafic lower crust with significant involvement of Indian continental crust compositions, whereas the Na-rich dikes were generated in the same way with less input of Indian continental crust compositions. Moreover, the postcollisional adakites in the southern Lhasa subterrane display distinctive spatial variations in geochemistry along the strike of this subterrane, indicating that the magma sources were heterogeneous. In combination with previously published data, we therefore suggest that all these late Oligocene to Miocene adakitic rocks were most likely generated dominantly by partial melting of the Lhasa mafic lower crust with involvement of Indian continental crust components, which was probably triggered by the tearing of the subducting Indian plate.