Cordilleran granites and supercontinent break-up: the example of Patagonia

RAPELA, C.W.; PANKHURST, R.J.

Floarencia, Italia

Congreso; 32nd International Geological Congress,; 2004

IUGS

Precise geochronological, geochemical and isotope studies of I-type cordilleran granites can provide important constraints for models of supercontinent break-up. Patagonia is a key area in the break-up history of SW Gondwana as it preserves the largest acid magmatic province developed during rifting and, subsequently, protracted subduction-related magmatism on the Pacific margin. New geochemical data and SHRIMP isotopic dating (181-185 Ma) of the Subcordilleran belt in northwestern Patagonia indicate that this belt represents an early Jurassic subduction-related magmatic arc along the proto-Pacific margin of Gondwana. Thus subduction was synchronous with the initial phase of rhyolite volcanism in extra-Andean Patagonia, previously ascribed to the thermal effects of the Karoo mantle plume and heralding rifting of this part of the supercontinent. Overall, there is clear evidence that successive episodes of calc-alkaline arc magmatism from Late Triassic times (220-200 Ma) until establishment of the Andean Patagonian Batholith in the Late Jurassic (c. 140 Ma) involved westerly migration and clockwise rotation of the arc. This indicates a changing geodynamic regime during Gondwana break-up and suggests differential roll-back of the slab, with accretion of new crustal material and/or asymmetrical 'scissor-like' opening of back-arc basins. Nd and Sr initial isotope ratios indicate that simple mixing arrays between melts derived from the lithospheric mantle wedge and either upper or lower crustal contaminants could explain the isotopic trends observed in I-type granites. The change towards less evolved melts with time in the I-type granites culminate in Early Miocene plutons with ENdt values between -4 and -6 (Pankhurst et al., 1999, J. Geol. Soc. London, 156: 673-694), a significant depleted signature that is never observed in Triassic and Jurassic granites nor any of their coeval volcanic rocks. We suggest that melting occurs in progressively more LIL-depleted sources with time and that the relatively primitive signature of the abundant Mid Cretaceous-Tertiary granites is indicative of melting in the lithospheric mantle underlying the Patagonian batholith. The evolved composition of Late Triassic to Early Cretaceous granites could then be explained by variable amounts of upper and lower crustal contamination of melts derived from this effectively undepleted lithospheric source.