Andean mountain building and magmatic arc migration driven by subduction-induced whole mantle flow

W P Schellart

doi:10.1038/s41467-017-01847-z

Abstract

Subduction along the western margin of South America has been active since the Jurassic, but Andean orogeny started in the middle Cretaceous and was preceded by backarc extension in the Jurassic-Early Cretaceous. The timing and sequence of these events has remained unexplained. Here I present a four-dimensional buoyancy-driven whole-mantle subduction model implying that the ~200 Myr geological evolution can be attributed to sinking of a wide slab into a layered mantle, where upper-mantle wide-slab subduction causes backarc extension, while whole-mantle (upper+lower) wide-slab subduction drives Andean orogeny. The model reproduces the maximum shortening and crustal thickness observed in the Central Andes and their progressive northward and southward decrease. The subduction evolution coincides with a 29° decrease in slab dip angle, explaining ~200 km of Jurassic-present eastward migration of the Central Andean magmatic arc. Such arc migration negates proposed long-term subduction erosion and continental destruction, but is consistent with long-term crustal growth.

Highlights

Subduction along the western margin of South America has been active since the Jurassic, but Andean orogeny started in the middle Cretaceous and was preceded by backarc extension in the Jurassic-Early Cretaceous
Andean mountain building and shortening have been active since the middle to Late Cretaceous (~120–70 Ma)[7, 11,12,13,14,15] during several orogenic phases, but were preceded by a period of Jurassic-Early Cretaceous backarc extension[7, 9, 13, 14] that was most pronounced in the north and south, and less extensive in the centre[9]
The cordilleran mountain range is characterized by magmatic arc migration and widening in the Central Andes since the Jurassic/Early Cretaceous[8, 11] (Fig. 1b, c), but a lack thereof in the north and south (Fig. 1a)

Summary

Introduction

Subduction along the western margin of South America has been active since the Jurassic, but Andean orogeny started in the middle Cretaceous and was preceded by backarc extension in the Jurassic-Early Cretaceous. I present a dynamic, buoyancy-driven, whole-mantle numerical subduction model to test a hypothesis in which the formation of the Andes is mostly a consequence of the long-term (~200 Myr) progressive evolution and large width (trench-parallel extent) of the subduction zone, focusing on the role of subduction-induced mantle flow in driving overriding plate deformation.

Results

Conclusion