Episodic Neogene Southward Growth of the Andean Subduction Orogen between 30°S and 40°S — Plate Motions, Mantle Flow, Climate, and Upper-Plate Structure

Abstract

The tectonic shortening of the South American upper plate that formed the Andean subduction orogen shows significant along-strike variations in magnitude and timing. Shortening in the Central Andes started in the Eocene, whereas the contraction of the Southern Central Andes at 30$\,^{\circ}$ S started only in the Early Miocene and then migrated southwards. Upper-plate shortening reached 38$\,^{\circ}$ S in the Late Miocene and tapered off at around 40$\,^{\circ}$ S. At the beginning of the Pliocene, contraction in the Southern Central Andes significantly slowed and eventually changed to transtensional conditions south of about 33.5$\,^{\circ}$ S, where the intersection of the Juan Fernandez Ridge results in a radical change in subduction mechanics. A southward decrease of both the duration and the average rate of contraction explains the observed decrease of the absolute Neogene shortening from 160 km at 30$\,^{\circ}$ S to 15 km at 38$\,^{\circ}$ S. The southward growth of the subduction orogen in the Mid-Late Miocene is not related to an increase of the relative Nazca-South America convergence but takes place during an acceleration of the westward motion of the South American Plate (overriding) with respect to the hot-spot reference frame. This suggests that overriding is the main plate kinematic driving mechanism for subduction orogeny. Kinematical considerations show that overriding is converted either into plate boundary deformation or into subduction-zone rollback. From the correlation of the deformation time-series with climatic, plate kinematic and geological data we suggest that the latitudinal and temporal variations of three factors can explain the varying proportions of rollback and upper plate shortening along the orogen. These are (1) the distance to the southern edge of the subducting Nazca slab, (2) the strength of the plate interface and (3) the strength of the upper-plate. From the fact that subduction orogeny expands southward during an acceleration of overriding we conclude that contraction of the upper plate in the more southerly regions required faster overriding velocities than in the north. This is most likely the result of easier rollback in the south which is enabled by the proximity of the Drake Passage or --- after initiation of the Chile Rise triple junction at 14 Ma --- the opening of slab windows under Patagonia that facilitate the transfer of mantle material from the sub-slab region to the slab wedge. The almost simultaneous post-Miocene slowing of upper-plate shortening south of the Juan Fernandez Ridge intersection can be best explained by Late Cenozoic increased sediment flux to the trench that ultimately reduced the strength of the plate interface. The initial Mid-Miocene exhumation of the main source area for trench sediments in the Patagonian Andes was tectonically controlled by the northward migrating Chile rise collision and not related to global cooling. However, the rapid spreading of Patagonian ice sheets after 7 Ma on the previously uplifted topography most likely led to a sudden increase of the average erosion rates and sediment flux into the trench. This increase and the blocking of the sediment dispersal within the trench by the Juan Fernandez Ridge enabled the accumulation of a thick trench fill. Hence, the pronounced Pliocene slowing of upper-plate contraction by sediment induced lubrication and weakening of the plate interface in this region is most likely the combined result of global climate change, tectonic uplift of the sediment source region and limited sediment dispersal in the trench. Once contraction started, deformation in the upper plate became focused in pre-Andean sedimentary basins and shortening rates in these basins were twice those in thick-skinned belts. Hence, we confirm that inherited soft regions in the upper-plate accelerate orogenic shortening. Subduction zone shallowing as a result of convergence-parallel ridge subduction weakens the upper plate thermally and has a similar accelerating effect on subduction orogeny.