Source and pressure effects in the genesis of the Late Triassic high Sr/Y granites from the Songpan-Ganzi Fold Belt, eastern Tibetan Plateau

The petrogenesis of granites with high Sr/Y signatures similar to adakitic rocks in continental settings is much debated. It is especially controversial whether these rocks are indicative of high-pressure magmatism (i.e., partial melting or crystallization fractionation) related to a thickened crust, or their high Sr/Y signatures are merely inherited from a high Sr/Y source. To address this, new chronological and geochemical data are presented for high Sr/Y granites from the Riluku batholith in eastern Tibetan Plateau. Zircon U-Pb dating and amphibole barometry suggest a final granite crystallization stage of ca. 207 Ma and an emplacement depth of ~14 km (~ 4 kbar), respectively. They are geochemically characterized by high Sr and low Y contents (481–1195 ppm and 6.62–20.6 ppm, respectively) with high Sr/Y and (La/Yb)_N ratios (35–112 and 15–113, respectively). Enriched isotope character of the high Sr/Y granites, in combination with regional tectonics, indicates that these rocks are unlikely to be derived from partial melting of subducting oceanic crust. Low Cr (0.8–1.0 ppm), Ni (0.9–3.2 ppm), and Mg^# (24–41) suggest that derivation from the partial melting of delaminated lower crust is also unlikely. Water-fluxed partial melting of crustal rocks at low pressure, which preferentially consumes plagioclase over micas, is inconsistent with the high contents of Rb (mostly >150 ppm) and low Sr/Rb ratios (mostly <5) of the samples. Reverse geochemical modeling, which simulated source composition under different pressures, indicates that evaluation of source compositions is critical to constrain the derivation pressure of the high Sr/Y granites. Our study emphasizes that high Sr/Y granites cannot be simply used to deduce a thickened crust at the time of magmatism without a priori knowledge of the nature of the source.