Magmatic and Seismic Evidence for the Neogene Evolution of the Subducting Slab and Crustal and Mantle Lithosphere under the Central Andes

Geophysical models coupled with the distribution, chemistry and age of magmatic rocks provide powerful tools for reconstructing the thermal and material balance and deformational history of the Central Andean crust and lithosphere in time and space. Two examples are given. In the first, a model for changing slab geometry, delamination (foundering) of the crust and mantle and forearc subduction erosion beneath the southern Puna plateau comes from studies of Miocene to Recent magmatic rocks linked with seismic studies. The distribution and chemistry (e.g., Sm/Yb, La/Ta, Ba/La, isotopes) of the volcanic rocks support an 18-7 Ma period of slab shallowing, followed by slab steepening and forearc subduction erosion linked with backarc crustal and lithospheric delamination and eruption of large ignimbrites. Support for delamination comes from seismic attenuation and Vs tomographic images that reveal an 100 km wide high velocity anomaly associated with an irregular shear wave splitting pattern, which is interpreted as a delaminated block above a nearly aseismic segment of the subducting slab at a depth of 150-200 km (Calixto et al., 2013, 2014; Liang et al. 2014). This block underlies the < 7 Ma giant Cerro Galan dacitic ignimbrites and bordering mafic flows and glassy andesites and dacites to the east. The characteristics of the flows support equilibration of basaltic magmas at > 1350°C at 2 Gpa followed by fractionation and mixing with melts of garnet-pyroxene-amphibole bearing crust (Risse et al., 2013). In accord, the lavas are over a region where receiver functions indicate a lithosphere-asthenosphere boundary at 60-80 km and a regionally thin 45-55 km thick crust with a low Vp/Vs (< 1.70) ratio (Heit et al., 2014). Calculations of crustal loss and gain allow up to 10% of the southern Puna lower crust to have been lost in the last 10 Ma. A second region where the characteristics of the magmatic rocks provide clues to the timing of slab shallowing and proposed slab tears (e.g., Lynner et al, 2017) is over and on the margins of the Chilean flat-slab). In this case, shallowing of the slab as the trench normal portion of the Juan Fernandez Ridge began to subduct at 11-10 Ma correlates well with the magmatic and deformational history. If the magmatism on the margins of the flat slab corresponds to slab tears, these tears also began at 10 Ma.