On spreading modes and magma supply at slow and ultraslow mid-ocean ridges

Abstract

The ultraslow eastern Southwest Indian Ridge (SWIR) offers an opportunity to study the effect of magma supply on an ultraslow mid-ocean ridge starting from quasi-melt-free detachment-dominated spreading, and transitioning to volcanic spreading as one nears prominent axial volcanos. Detachments in the quasi-melt-free mode extend along-axis 60 to 95 km and have a lifetime of 0.6 to 1.5 myrs. They cut into their predecessor's footwall with an opposite polarity, causing part of the footwall lithosphere to experience further deformation, hydrothermal alteration, sparse magmatism and possibly thermal rejuvenation, in a hanging wall position. The accretion of the oceanic lithosphere in this context therefore occurs in two distinct stages over the lifetime of two successive detachment faults. We examine the transition from this nearly amagmatic detachment-dominated mode to the more common volcanic mode of spreading, showing that it occurs along-axis over distances ≤ 30 km. It involves a significant thinning of the axial lithosphere and a gradual decrease of the amount of tectonic displacement on faults, as the magmatic contribution to the divergence of the two plates increases. We develop a conceptual model of this transition, in which magma plays a double role: it fills the space between the diverging plates, thus reducing the need for displacement along faults, and it modifies the thermal state and the rheology of the plate boundary, affecting its thickness and its tectonic response to plate divergence. Based on a comparison of the ultraslow eastern SWIR, with the faster spreading Mid-Atlantic Ridge, we show that the activation of the volcanic, or of the detachmentdominated modes of spreading is connected with the volume of magma supplied per increment of plate separation, over a range of axial lithosphere thickness, and therefore over a range of the M ratio defined by (Buck et al., 2005) as the relative contribution of magma and faults to plate divergence (M is smaller, for a given volume of melt per increment of plate separation, if the plate is thicker). We therefore propose that M does not fully explain the variability in faulting styles observed at slow and 2 ultraslow ridges and propose that rheological changes induced by magma (melt itself is weak, hydrothermally altered gabbro-peridotite mixtures are weak, and melt heat sustains more vigourous hydrothermal circulation) also play a key role, resulting in contrasted potentials for strain localization, footwall flexure on faults and the development of detachment faults.

Internal Structure of the Oceanic Lithosphere at a Melt‐Starved Ultraslow‐Spreading Mid‐Ocean Ridge: Insights From 2‐D Seismic Data

“…While the fault damage zones at mid‐ocean ridge systems documented in the geological literature are narrow (<400 m), for example, (Hayman & Karson, ), our results confirm findings of Momoh et al () and indicate that a thicker damage zone is likely even for short‐offset faults (present offset at young D1 axial detachment fault is estimated as <4 km; Cannat et al, ). Following on the discussion in Momoh et al (), we propose that this young yet thick damage zone may have formed in three ways: (1) due to distributed simultaneous deformation over a thick domain, (2) as a finite damage zone due to strain localizing on a succession of subparallel individual fault segments, or (3) by a combination of both styles.…”

“…The black lines in the model indicate the inferred traces of the active and inactive (dashed) detachment faults. If correct, our interpretation suggests that the shallow parts of detachment fault systems (<5 km) are steep (45-55°) at an early stage of the fault's activity (represented by the presently active axial detachment), and rotate to a shallower-dipping angle (25°) after about 16 km of horizontal displacement (estimated for detachment fault D3; Cannat et al, 2019).…”

“…We propose that these lower angles may result from the more significant footwall rotation incurred in the detachment fault (D3) system as it terminated after an estimated 16 km of horizontal displacement (Cannat et al, ). Compared to the present‐day detachment fault damage zone, which has so far accommodated ~4 km of horizontal displacement only (Cannat et al, ), and assuming a linear relationship between horizontal offsets and emergence angles, then the flexural rate in the shallow footwall of these detachment faults is ~2.5°/km of fault displacement. This is significantly less than the flexural rate that could be estimated (assuming a similar initial dip of the fault at shallow depth) for the shallow footwall of MAR corrugated OCCs: the emergence angle of the 13°N active detachment fault is less than 10° (Smith et al, ) for a total horizontal displacement of ~12 km (MacLeod et al, ).…”

“…Geochemistry, Geophysics, Geosystems fault system is also discussed and compared with existing geological observations, using the two-stage conceptual framework proposed for nearly amagmatic seafloor spreading in Cannat et al (2019): (1) tectonic exhumation in the footwall of a detachment fault, and (2) further faulting, alteration, and magmatic infiltration in the hanging wall of the next detachment fault. On this basis, we propose a simplified conceptual model to explain the seismic reflection characteristics of nearly amagmatic oceanic lithosphere.…”

“…Bathymetric map of our study area with the traces of the emergence and breakaway of successive detachment faults (Cannat et al, ). Plate separation is approximately north‐south (indicated by the white arrows).…”