DUCTILE FLOW WITHIN AN OROGENIC WEDGE, AND ITS IMPLICATIONS FOR SEISMIC ANISTROPY

A commonly held view is that seismic anisotropy at convergent plate boundaries is related to shearing on localized ductile shear zones. This idea is problematic given that lattice preferred orientation (LPO) “saturates” to maximum values at relatively low shear strains (3 to 5). Thus, if convergence is accommodated mainly by localized shearing, then how can we explain strong seismic anisotropy, which seem to require the generation of widespread LPO (mainly mica) by penetrative deformation? I consider here an alternative idea, based on the deformation associated with flow in orogenic wedges. An orogenic wedge deforms primarily by the flux of material through the wedge. The shape of the wedge is constrained by its rheology, ranging from triangle tapers for frictional wedges to parabolic tapers for viscous wedges. The distribution and rates of accretionary and erosional fluxes is thus the main factor that controls the velocity field within the wedge. The important difference from the shear zone idea is that ductile fabrics will form throughout the entire thickness of the wedge, which can locally reach a thickness of 35 km or more beneath the wedge high. The intensity of these ductile fabrics should increase rearward within the wedge, given that materials in these locations have moved farther through the wedge and have acquired more strain. These predictions are supported by strain studies at a number of exhumed subduction wedges. In general, one finds that the strongest fabrics are located beneath the wedge high. In these areas, the symmetry of the deformation is typically uniaxial oblate at the regional scale. Linear fabrics may be present locally but tend to be unimportant when the deformation is averaged to the regional scale. Recent papers have argued that the dip of this planar deformational fabric beneath a wedge high can vary from steeply dipping where erosion and frontal accretion dominate, to flat lying where underplating dominates. The most important feature linking these fabrics to seismic anisotropy is mica LPO. The geometry, magnitude, and distribution of mica LPO can be reasonably predicted using existing models. Thus, seismic anisotropy represents an intriguing way to test these ideas.