The geometry of oceanic spreading centers can change by rift propagation, in which a new ridge segment propagates and preempts a preexisting one. As the locus of relative motion between the two plates shifts, the dying and growing ridge segments are joined by a boundary, usually a transform, which migrates with time, causing seafloor to be transferred from one plate to the other. As a result, originally ridge-parallel isochrons in a zone between the growing and failed rifts can be reoriented to trends oblique to both the ridge and spreading direction. An attractive feature of this model is that the reorientation occurs by strictly rigid plate tectonics, since relative motion occurs only between points on different plates. Recent surveys of the Galapagos propagating rift system at 95¿W show seafloor fabric curving away from the growing rift and toward the failed rift. These observations have been interpreted as requiring a zone of distributed simple shear between the two rifts. Such behavior would represent a local deviation from rigid plate tectonics, since relative motion would be distributed over a finite region not part of either of the two plates.

Here we explore possible rigid plate models for the formation of curved seafloor lineaments by rift propagation. We find that such fabric can be generated by changes either in the velocity (rate or direction) of rift propagation, or in the spreading rate during rift propagation. The curvature of various tectonic elements is diagnostic of the different possibilities. The geometry observed at the Galapagos can result from either rift propagation acceleration, or a spreading rate decrease, during the last few hundred thousand years. The reverse curvature could result from either deceleration of rift propagation or an increase in spreading rate. Propagator acceleration and spreading rate slowing, which produce similar structural trends, could in principle be distinguished using spreading rate data. At Galapagos, however, since either would have occurred within the present period of normal polarity, both are consistent with the magnetic anomaly data. We thus conclude that the data interpreted as requiring a shear zone are equally consistent with two distinct models based on rigid plate tectonics. It may, however, be possible to discriminate between rigid and shear models using fabric observations at other propagators. In contrast, the rigid plate tectonic model allows for fabric with a variety of orientations, including reverse curvature formed by either propagation deceleration or increasing spreading rate and linear fabric formed by constant propagation and spreading velocities. ¿ American Geophysical Union 1988