Paleomagnetic constraints on deformation of superfast-spread oceanic crust exposed at Pito Deep Rift

Abstract

[1] The uppermost oceanic crust produced at the superfast spreading (∼142 km Ma −1 , full-spreading rate) southern East Pacific Rise (EPR) during the Gauss Chron is exposed in a tectonic window along the northeastern wall of the Pito Deep Rift. Paleomagnetic analysis of fully oriented dike (62) and gabbro (5) samples from two adjacent study areas yield bootstrapped mean remanence directions of 38.9°± 8.1°, −16.7°± 15.6°, n = 23 (Area A) and 30.4°± 8.0°, −25.1°± 12.9°, n = 44 (Area B), both are significantly distinct from the Geocentric Axial Dipole expected direction at 23°S. Regional tectonics and outcrop-scale structural data combined with bootstrapped remanence directions constrain models that involve a sequence of three rotations that result in dikes restored to subvertical orientations related to (1) inward-tilting of crustal blocks during spreading (Area A = 11°, Area B = 22°), (2) clockwise, vertical-axis rotation of the Easter Microplate (A = 46°, B = 44°), and (3) block tilting at Pito Deep Rift (A = 21°, B = 10°). These data support a structural model for accretion at the southern EPR in which outcrop-scale faulting and block rotation accommodates spreading-related subaxial subsidence that is generally less than that observed in crust generated at a fast spreading rate exposed at Hess Deep Rift. These data also support previous estimates for the clockwise rotation of crust adjacent to the Easter Microplate. Dike sample natural remanent magnetization (NRM) has an arithmetic mean of 5.96 A/m ± 3.76, which suggests that they significantly contribute to observed magnetic anomalies from fast-to superfast-spread crust.

Diverse magma flow directions during construction of sheeted dike complexes at fast- to superfast-spreading centers

“…For Hess Deep data, we follow a two-stage reverse sequence model outlined by Varga et al (2004) involving (1) (1), and determined from mean dike attitude (2) (Varga et al, 2004). The data set from two study areas, A and B, at Pito Deep is consistent with a model involving a three-stage reverse sequence of rotations about (1) a horizontal axis (150 • /00 • ) parallel to Pito Deep Rift normal faults, (2) a vertical axis (000 • /90 • ) to account for Easter Microplate rotation, and (3) a horizontal axis (010 • /00 • ) parallel to southern EPR normal faults (Horst et al, 2011). Rotation magnitudes for each part of the sequence are determined from best-fit models that restore both mean remanence directions and dikes to near expected initial orientations (geocentric axial dipole direction for 21 • S latitude: D = 000 • , I = −40.2 • ; initial dike strike/dip = 010 • /90 • ).…”

“…Sheeted dikes at both exposures typically dip away from the ridge axis with strikes parallel to adjacent abyssal hills. On average, a wider range in dike orientations measured at Hess Deep and qualitative observations of more cataclastic deformation and related faulting (Varga et al, 2004;Hayman and Karson, 2007), suggests slightly more tilting and subaxial subsidence occurred near the ridge axis during the formation of Hess Deep crust than that at Pito Deep (Horst et al, 2011). The crust exposed at each of the two areas has a complicated tectonic history involving initial formation at the EPR, thermal subsidence during spreading, and more recent faulting and tilting due to rifting.…”

“…Although part of the Nazca Plate since its formation at the EPR, the crust northeast of Pito Deep has been subjected to deformation and clockwise rotation related to coupling with the adjacent Easter Microplate (Hey et al, 1985;Francheteau et al, 1988;Naar and Hey, 1991;Rusby and Searle, 1995). Attempts to resolve the post-spreading deformation have been made through paleomagnetic studies of fully-oriented samples (Varga et al, 2004;Horst et al, 2011). Both of these studies combine structural and paleomagnetic data to create structural rotation models that best constrain the tectonic history and restore the dike samples to their assumed initial vertical and rift-parallel orientation.…”

“…Amounts of rotation for the reverse sequence model of area B are 10 • , 44 • , and 22 • (Horst et al, 2011). We use the two-stage rotation model for Hess Deep Rift and three-stage rotation model for Pito Deep Rift to restore the AMS eigenvectors to their initial orientation relative to the EPRspreading geometry.…”

“…However, collecting fully-oriented block samples from seafloor escarpments prohibited us from directly adhering to this strategy. Block samples of basalt dikes were fully oriented in situ by either submersible (Alvin) or remote-operated vehicle (Jason II) using the Geocompass, a device designed to measure planar features on the seafloor (Hurst et al, 1994). Geocompass measurements on two or more surfaces on individual blocks, predominantly dike margins and orthogonal fractures, allowed the unique in situ orientation of blocks to be reconstructed.…”