RHEOMANTLE · Evaluation of mantle rheology in exhumed strike-slip faults

Evaluation of mantle rheology in exhumed strike-slip faults

The main objective of the RHEOMANTLE project is to understand what governs the cycle of earthquakes that occur in strike-slip tectonic plate boundaries. Strike-slip faults, which result in horizontal motion, are thought to continue downward through the entire tectonic plate (i.e. lithosphere). Consequently, there is a debate about which part of the tectonic plate controls deformation: Is it the upper crust that contains discrete faults or the upper mantle where distributed flow occurs in shear zones? Our work shows that mantle-crust interactions affect deformation and mechanical properties of tectonic plates in strike-slip plate boundaries, and therefore, can control the cycle of earthquakes. To study the deformation processes and mechanical properties of the deeper sections of the lithosphere in strike-slip plate boundaries, we used naturally deformed rocks brought up to the Earth’s surface from the upper mantle and lower crust during volcanic eruptions (such rocks are known as xenoliths) or as parts of oceanic lithosphere, which have been uplifted and exposed at the Earth’s surface. Rocks from four major strike-slip fault zones were studied, providing direct evidence for deep tectonic processes. In particular, we analyzed upper mantle xenoliths from beneath the San Andreas Fault system in western California (USA), as well as xenoliths extracted from the lower crust and upper mantle in the vicinity of the Baja California Shear Zone, in western Baja California (Mexico). The rocks studied from these two fault zones provide unprecedented insights into the strength of different lithospheric layers (variation of strength as a function of depth) and locations (lateral strength variations) along the Pacific-North American Plate boundary. Moreover, we studied exhumed upper mantle rocks in the Bogota Peninsula shear zone (New Caledonia), which comprised the boundary between two oceanic tectonic plates, and the Mavrovouni shear zone (Greece), which served as a major zone of localized deformation in the oceanic lithosphere. In both the Bogota Peninsula and the Mavrovouni shear zones, we studied lateral variations in the strength of the upper mantle rocks and the processes leading to localization of deformation in the deeper parts of strike-slip faults. The upper mantle xenoliths from the San Andreas Fault system and the upper mantle and lower crustal xenoliths from the Baja California Shear Zone have low water contents and record shear stresses that are nearly identical to the low upper crustal strength of the strike-slip faults in the San Andreas Fault system. Thus, strength of the lithosphere in the Pacific-North American plate boundary remains constant, and low, with depth. This result indicates that deformation is not controlled by a single lithospheric layer (upper crust, lower crust, or upper mantle). At the grain scale, deformation in both the upper mantle and lower crust involves a combination of movement of crystal defects within the grains and sliding along the grain boundaries (dislocation accommodated grain boundary sliding). The upper mantle rocks in the Bogota Peninsula shear zone are hydrous and record a lateral increase in both strength and change in strain with respect to time toward the high strain parts of the shear zone. Such lateral variations induce a decrease in the resistance of rock to deformation, which promotes strain localization. The upper mantle rocks in the Bogota Peninsula shear zone have low strength, comparable to the strength of the upper mantle in the San Andreas Fault system and the Baja California Shear Zone. The observed localization of deformation in the upper mantle shear zones could be imposed by earthquake rupturing at shallower lithospheric levels. To explain these observations, we developed the “lithospheric feedback” model, in which the behavior of major strike-slip fault systems is controlled by the interaction of the faulting upper crust and flowing upper mantle acting as an integrated system. The interaction of the lithospheric layers – rather than their relative strengths – may be the critical driver of the seismic cycle in strike-slip plate boundaries.

Data: CORDIS, © European Union