Segment-scale variations in the crustal structure of 150–300 kyr old fast spreading oceanic crust (East Pacific Rise, 8◦15′N–10◦5′N) from wide-angle seismic refraction profiles

S U M M A R Y We have simultaneously inverted seismic refraction and wide-angle Moho reflection traveltimes for the 2-D crustal thickness and velocity structure of 150–300 kyr old crust along the East Pacific Rise (EPR) between the Siqueiros and Clipperton fracture zones (FZs). Our results show a strong correlation between ridge segmentation and upperand mid-crustal seismic velocities, with higher velocities near segment centres and lower velocities near segment ends. Low crustal velocities at the Clipperton and Siqueiros FZs are interpreted as fracturing resulting from brittle deformation of the crust in the transform domain. A relict overlap basin left on the Pacific Plate by the 9◦03′N overlapping spreading centre (OSC) as it propagated southward is associated with a large (∼1 km s−1), negative upperand mid-crustal velocity anomaly. This anomaly is consistent with the presence of an unusually thick extrusive section within the basin and with tectonic alteration, fracturing and shearing arising from rotation of the basin as it was formed. The discordant zone left by this OSC on the Cocos Plate is characterized by moderately low crustal velocities, probably because of crustal fracturing as the OSC propagated into older crust. Higher crustal velocities near segment centres may reflect a higher ratio of dikes to extrusives in the upper crust, and lower-intensity tectonic alteration of the crust, than near segment ends. The mean crustal thickness along the EPR between the Siqueiros and Clipperton FZs is 6.7–6.8 km. The thickest crust is found beneath the Lamont seamounts (∼9 km), and in a southward-pointing, V-shaped band located just north of the off-axis trace of the 9◦03′N OSC (7.3–7.8 km). The thinnest crust (<6 km) is found proximal to the Clipperton and Siqueiros FZs. The crust associated with the off-axis trace of the 9◦03′N OSC is not anomalously thin, suggesting that magma supply beneath the OSC is similar to that of the northern and southern segments. We see a similar pattern of crustal thickness variation to that determined using multichannel reflection data, including a gradual thickening of the crust from north to south along the northern ridge segment, and the location of the thickest crust just north of the 9◦03′N OSC. However, the magnitude of the along-axis crustal thickness variation we observe along the northern ridge segment between 9◦50′N and 9◦15′N (∼1.3–1.8 km, excluding the Lamont seamounts) is significantly less than the 2.3 km of variation previously reported, weakening the case for the existence of a low-density mantle diapir at 9◦50′N inferred from gravity data. The band of thick crust located just north of the off-axis trace of the 9◦03′N OSC suggests a close genetic link between this feature and the OSC. Thus we attribute the pattern of crustal thickness variations along the northern segment to the kinematics of the southward-propagating 9◦03′N OSC over the past 0.5 Myr, and not to along-axis melt migration away from a mantle diapir as previously proposed.