CONSTRAINTS ON GEODYNAMIC MODELS OF SURFACE UPLIFT FROM HYDROGEN ISOTOPES OF VOLCANIC GLASS IN SOUTHERN PERU

The emerging picture in the Central Andean Plateau (CAP) is one of diachronous surface uplift with potentially disparate uplift mechanisms. For example, a late Miocene pulse of surface uplift in the Altiplano of Bolivia at 17°S contrasts with evidence for 2.5–3.5 km elevations in the Eastern Cordillera (EC) by ~17 Ma. At 19–23°S, compiled Δ47 paleothermometry data support a middle Miocene ~2 km pulse of surface uplift in the eastern Altiplano and EC. Farther south at 23–27°S, available evidence points to near-modern elevations (~4 km) in the Western Cordillera (WC) and Puna Plateau since the late Eocene while the EC remained at low elevations until the onset of local shortening in the middle–late Miocene.

The history of surface uplift in the CAP of southern Peru remains poorly documented compared to regions farther south. Basins such as the Condoroma basin in the WC and basins of the Cordillera Carabaya (Macusani, Crucero, and Picotani basins) in the EC include multiple intercalated volcaniclastic horizons. Miocene volcanic glass samples from the Condoroma basin show an abrupt decrease in δD values from a mean of -100‰ to -160‰ at 19–16 Ma and remain similarly negative through the Pliocene. Mean pre-shift reconstructed water δD values from the Condoroma basin (-70‰) are comparable to modern waters sampled at low elevations (-50‰). In contrast, mean reconstructed post-shift waters (-130‰) are consistent with modern waters near 4500 m (-140‰). We attribute this pattern to rapid uplift and attainment of high elevation in this location by ~16 Ma. In contrast, samples from the Cordillera Carabaya change from -105‰ to -130‰, and reconstructed-water δD values shift from an average of -75‰ to -100‰ between ~8 and 6 Ma. This suggests a delayed and lower magnitude episode of surface uplift in the EC relative to the WC of southern Peru.

These data are consistent with several geodynamic scenarios, including a west-to-east migrating hinge of lithospheric delamination. A viable alternative model involves early Miocene foundering of a dense, unstable lithospheric root beneath the WC followed by uplift via late Miocene shortening in the EC. Differentiating between these geodynamic models requires integration of paleoelevation data with currently unavailable histories of basin evolution, deformation, and exhumation.