T34A-01
Oceanic corrugated surfaces and the strength of the axial lithosphere at slow spreading ridges
We analyse the topography and gravity signature of 39 corrugated surfaces formed over the past 26 myrs in the footwall of axial detachment faults at the easternmost Southwest Indian Ridge. We show that these corrugated surfaces formed at intermediate melt supply and that their presently elevated topography, relative to adjacent non-corrugated seafloor, was mostly acquired at the end of their formation, at the "termination stage". This configuration, which also characterizes many off-axis corrugated surfaces in other oceans, suggests that the plate's flexural rigidity is very low during the formation of the corrugated surface, and increases significantly at the termination stage. We present a numerical model in which these variations in the axial plate's rigidity are explained by variable mechanical accommodation within the footwall of axial detachment faults, due to partial hydration of exhumed mantle material. We further propose that weak talc- coated shear zones developped in mantle-derived ultramafics due to interactions with Si-rich hydrous fluids derived from the alteration of gabbro intrusions drastically reduce the strength of the upper part of the brittle lithosphere in the detachment footwall. Based on this conceptual model, we predict that corrugated surfaces, with subdued dynamic topography, form preferentially when such talc-coated weak shear zones are frequent in the exhumed ultramafics. This requires that gabbro intrusions are sufficiently abundant to provide Si-rich metasomatic fluids but do not form a coherent rigid framework in the detachment's footwall. This hypothesis, which may be tested by drilling, is compatible with the Buck et al. (2005) model which links for the formation of long-lasting detachments at mid-ocean ridges to an intermediate supply of melt to the axial crust.
T34A-02
High-Precision TIMS Dating of Oceanic Crustal Accretion: Vema Lithospheric Section, Mid- Atlantic Ridge
We present new high-precision thermal ionization mass spectrometry (TIMS) U-Pb zircon dates from the Vema Lithospheric Section (VLS; 11°N, Mid-Atlantic Ridge). The VLS exposes a complete, intact section of slow-spreading oceanic lithosphere that includes mantle peridotites, lower crustal gabbros and basaltic upper crust. The precision of single-crystal 206Pb/238U dates attained by the TIMS analyses generally ranges from 0.07% to 0.75% (ca. 10,000 to 100,000 years; 2σ) and are controlled by zircon U and Pb concentrations (U = 2 to 134 ppm; total Pb load = 0.9 to 9.2 pg). The single grain dates are over an order of magnitude more precise than previous ion probe dates on oceanic gabbros and provide the first opportunity to study magmatic processes occurring on timescales of 104 to 105 years. Zircon dates range from ~13 to 14 Ma, and show a good correlation with distance from the ridge axis, defining a half spreading rate of 16 mm/yr. The systematic progression of zircon dates, combined with the continuous nature of the lower crust and scarcity of gabbroic plutons in the Vema mantle section, are consistent with shallow, ridge-centered melt delivery. Our results contrast with previous zircon studies of oceanic gabbros, which focused on oceanic core complexes and document up to a 2.5 Ma range in zircon dates from individual segments of oceanic crust. Zircons separated from individual samples from Vema show an age range of up to 200 ka, which we interpret to reflect the timescale of zircon crystallization within differentiating gabbroic plutons.
T34A-03 INVITED
Emplacement of incompatible trace element depleted magmas during detachment faulting – is there a genetic link?
Detachment faults accommodate large amounts of slip along slow spreading ridges but the reasons why they form, and their relationship to magmatism, are not well understood. Basalt geochemistry along ridge- perpendicular time lines that intersect oceanic core complexes have been collected from two locations: (i) the Atlantis Bank core complex on the SWIR [1], and (ii) the 15°N core complex on the MAR. In both cases, magmas emplaced into the crust synchronously with core complex formation both extend to much more depleted compositions, and show a wider range of compositions, than those emplaced since detachment faulting stopped. This signature is clearest in the most incompatible trace element ratios such as Th/La and La/Sm. At the Atlantis Bank core complex, where samples of peridotite have also been collected along the same time line (0-14 Myr) they show little systematic variation in composition. The lack of change in peridotite composition indicates that the maximum extent of melting did not change during core complex formation since these come from the top of the melting column. Changes in the abundances of highly incompatible elements in melts emplaced into the crust synchronous with core complex formation in two different locations suggests a genetic link between melt generation and core complex formation. Changing the ratios of highly incompatible elements is most readily achieved by changing the contribution to the aggregated melt from low F melts. This could be achieved by changing from passive to active upwelling leading to a smaller contribution to the aggregated melt from the 'wings' of the melting column. Changing from passive to active upwelling would not change the maximum degree of melting, consistent with peridotite compositions. Active mantle upwelling would impart stresses on the base of the lithosphere that may play a role in allowing long-lived detachment faults to form. In this model, lithospheric deformation style would provide a window into mantle dynamics. Since the observation of changing melt composition coinciding with core complex formation has been made in two locations it looks likely that there is a deep (melting column) control on the formation of core complexes. [1] L. A. Coogan et al., Chemical Geology 207, 13 (2004).