Crustal thickening prior to 38 Ma in southern Tibet: Evidence from lower crust-derived adakitic magmatism in the Gangdese Batholith

The petrogenesis and geodynamic implications of the Cenozoic adakites in southern Tibet remain topics of debate. Here we report geochronological and geochemical data for host granites and mafic enclaves from Wolong in the eastern Gangdese Batholith, southern Tibet. Zircon LA-ICP-MS dating indicates that the Wolong host granites and enclaves were synchronously emplaced at ca. 38 Ma. The host granites are medium- to high-K calc-alkaline, metaluminous (A/CNK = 0.93–0.96), with high Al2O3 (15.47–17.68%), low MgO (0.67–1.18%), very low abundances of compatible elements (e.g., Cr = 3.87–8.36 ppm, Ni = 3.04–5.71 ppm), and high Sr/Y ratios (127–217), similar to those typical of adakite. The mafic enclaves (SiO2 = 51.08–56.29%) have 3.83–5.02% MgO and an Mg# of 48–50, with negative Eu anomalies (δEu = 0.59–0.79). The Wolong host granites and enclaves have similar Sr–Nd isotopic compositions (initial 87Sr/86Sr = 0.7053–0.7055, εNd(t) = − 2.7 to − 1.4), with varying zircon εHf(t) values, ranging from + 6.0 to + 12.6. A comprehensive study of the data available for adakitic rocks from the Gangdese Batholith indicates that the Wolong adakitic host granites were derived from partial melting of a thickened lower crust, while the parental magmas of the mafic enclaves were most likely derived from lithospheric mantle beneath southern Tibet. The Wolong granitoids are interpreted as the result of mixing between the thickened lower crust-derived melts and lithospheric mantle-derived mafic melts, which are likely the protracted magmatic response to the break-off of the Neo-Tethyan oceanic slab at about 50 Ma. Our results suggest that the crustal thickening in southern Tibet occurred prior to ~ 38 Ma, and support the general view that the India–Asia collision must have occurred before 40 Ma.