Origin of the Australian–Antarctic Discordance from an ancient slab and mantle wedge
A variety of evidence has suggested that the enigmatic Australian–Antarctic Discordance is caused by a cold zone beneath the present-day South East Indian Ridge south of Australia. We show that the present position of the subduction zone which surrounded Gondwanaland until the Mesozoic is within a few hundred kilometres of the Australian–Antarctic Discordance. Beneath the Australian–Antarctic Discordance, tomographic inversions show a north-south-trending seismic anomaly with a higher than average shear velocity in the lower mantle and a prominent, nearly circular, high-velocity anomaly within the transition zone. These seismic inversions are consistent with the predictions of three-dimensional models of mantle convection with imposed plate tectonics. However, these earlier models, incorporating only a thermal slab, resulted in a circular topographic depression on the present South East Indian Ridge and are inconsistent with the observed residual depth anomaly which is continuous from the South East Indian Ridge nearly to the Australian and Antarctic margins. To resolve this discrepancy, we propose that the Australian–Antarctic Discordance results from the sampling of both an ancient mantle wedge, depleted by prolonged melting, and mantle cooled by the subduction system. Dynamic models show that the new ridge forming between Australia and Antarctica would have sampled the wedge first, resulting in a V-shaped structure, which matches the observed residual depth anomaly. The models show that the South East Indian Ridge would later sample the cool mantle which became trapped within the transition zone at 20 Ma, consistent with the observed increase in fracture zone density at the Australian–Antarctic Discordance in the Neogene. We propose that along the trace of the residual depth anomaly, a fundamental change occurred in the dominant mechanism causing the topography, first by sampling of refractory mantle from the old wedge and later by the sampling of cold mantle. This hypothesis can be tested with ocean floor recovered from drillholes along the depth anomaly.