Layering of Oceanic Crust

Abstract

Magma generated by decompression melting of the upwelling mantle beneath mid-ocean ridges (MORs) rises buoyantly and accumulates in crustal magma chambers (e.g., Forsyth, 1992). The long-held view is that oceanic crust is built from in situ crystallization of melts in these reservoirs, as well as from melts extracted from the magma chamber(s) in the form of dikes that eventually break the seafloor during submarine eruptions (e.g., Sinton and Detrick, 1992). These processes result in oceanic crust that is magmatically layered, with a lower section typically 4–5 km thick composed of mafic plutons (gabbros), a 1–2 km thick layer of sheeted basaltic dikes above it, and a carapace of a few to several hundred meters thick of extrusive basalts (pillow lavas and sheet flows), known as the Penrose model of oceanic crust (Penrose Conference participants, 1972). Although these units all have common parental magmas, they exhibit distinct elastic properties because of differences in mineralogical textures produced by crystallization processes and different porosity structure (i.e., the shallower extrusive section is highly porous, with porosity decreasing with depth throughout the crust). As a result, the speed of seismic waves propagating through the crust increases with depth while at the same time the rate at which seismic velocities increase with depth (velocity gradient) diminishes with depth. These seismic layers have been historically known as seismic Layer 2 (upper and mid-crust, subdivided in 2A and 2B based on changes in velocity gradient) and Layer 3 (lower crust) (e.g., White et al., 1992). To a large degree they correspond to the lithological layering of the oceanic crust, although deviations from this direct correlation are known to exist (Christeson et al., 2007; Detrick et al., 1994; Wilson et al., 2006). The lithological and seismic layering described above has successfully explained many observations made at ophiolites (sections of oceanic crust exposed on land) as well as observations at crust formed at MORs with fast-spreading rates (>80 mm/year full rate), where magmatism is the dominant mode of lithospheric accretion. However, it is becoming increasingly accepted that the Penrose model fails to explain many features observed at MORs where plates spread apart at slower rates. Along slowand ultra-slow-spreading MORs, the lithosphere is generally thicker and colder than at fast-spreading MORS, tectonic activity is a dominant process accommodating plate separation, and magmatism is highly variable both in time and space (e.g., Dick et al., 2010; Michael et al., 2003; Schroeder et al., 2007). In these settings, the interplay between magmatic and tectonic extension and their space-time variability result in oceanic lithosphere in which the crust is no longer stratified but disrupted, forming a heterogeneous mixture of magmatic and mantle rocks (Cannat, 1993).