We have conducted a comparative study of the tectonic morphology of young seafloor using SeaMARC II side scan sonar Surveys of the intermediate spreading Ecuador Rift, the fast spreading East Pacific Rise (EPR) (8¿30'--10¿N), and the super fast spreading EPR (18¿--19¿S). We find that characteristics of fault populations are not only a function of spreading rate but also vary along axis within individual ridge segments (i.e., with proximity to large- and short-offset discontinuities). We also find that fault azimuths can be used to examine plate kinematics on a finer scale than can be obtained using magnetic data alone. Most of the variation in fault populations with spreading rate can be explained by an inverse relationship between spreading rate and thickness of the brittle layer. For example, regions of super fast spreading are characterized by the largest numbers of short faults, the smallest average fault spacing and throw, and the highest fault density. In addition, clusters of short, closely spaced antithetic faults subsidiary to long master inward dipping faults are common within the super fast spreading area, presumably the result of a thinner, weaker brittle layer. Faults facing away from the ridge axis occur in increasing numbers with increasing spreading rate such that few outward facing faults are found at slow to intermediate rates and approximately equal numbers of inward and outward facing faults are observed at the fastest rates. Rapid thickening of the brittle layer with distance from the ridge may account for the predominance of inward facing faults at slower spreading rates. Outward facing faults at all spreading rates have shorter mean lengths and lower vertical offsets. These differences may reflect the shorter time outward facing faults are active owing to increasing strength of the lithosphere with distance from the ridge. Fault lengths and spacings in all areas approximate exponential distributions.

The extensional strain represented by fault populations is calculated from the displacement and length distributions of faults, and strain estimates of ~4% are obtained for each area. Assuming that fault spacing reflects fracture depth extent where faults initiate, we infer a brittle layer thickness of ~1 km when faulting begins. Fault populations are examined for ridge segment scale variations in amagmatic extension. We see evidence for greater amagmatic extension associated with long-term reduced magma supply along the eastern third of the Ecuador Rift. Evidence for local increased brittle extension is also found within 15 km of transform faults. Discordant zones left by overlapping spreading centers (OSCs) are characterized by low fault abundances. At OSCs, discrete events of ridge tip propagation may accommodate extension taken up elsewhere along the ridge by normal faulting. Fault azimuthsdo appear to be useful indicators of plate motion. Within the EPR 8¿30'--10¿N area, fault trends record a recent change in Pacific-Cocos plate motion (3¿--6¿ at ~1 m.y.) consistent with magnetic anomaly and fault lineation data from elsewhere along the northern EPR. Within the Ecuador Rift, fault azimuths scatter within 3¿ of predicted trends and are consistent with constant spreading about one pole for the past 1.5 m.y. ¿ American Geophysical Union 1994