EGU25-13559, updated on 15 Mar 2025
https://doi.org/10.5194/egusphere-egu25-13559
EGU General Assembly 2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
Mechanical and metasomatic evolution of a developing mantle wedge from subduction initiation to obduction
Alissa Kotowski1, Andrew Keats1, Hester Smit1, Jippe van Broekhoven1, Matthew Tarling2, Marguerite Godard3, Oliver Plümper1, Martyn Drury1, and Eric Hellebrand1
Alissa Kotowski et al.
  • 1Utrecht University, Earth Sciences, Utrecht, Netherlands (a.j.kotowski@uu.nl)
  • 2Department of Earth, Ocean and Atmospheric Sciences, University of British Columbia, Vancouver, Canada
  • 3French National Centre for Scientific Research (CNRS), Géosciences Montpellier, France

Geodynamic models suggest that plate boundary shear zones require mechanically weak materials to form. However, peridotites in proto-plate boundary hanging walls are inherently strong and experience cooling from >1000°C to <500°C over ~10 Myr during subduction initiation. Without a micro-physical or metasomatic mechanism to weaken the olivine-rich mantle, it will resist strain localization with cooling. Serpentinites are often credited with facilitating lithosphere-scale strain localization, but proto-interface temperatures exceed ~550°C at 20–30 km depth, and therefore are too hot for serpentine to be stable. What, therefore, are the roles of both olivine and serpentine in plate boundary formation?

To address this, we present structural and geochemical data from a fossilized subduction interface at Mont Albert (Québec, Canada). This Ordovician ophiolite records subduction initiation and subsequent obduction during the Taconian Orogeny (~450–500 Ma). Field and microstructural observations show that spinel peridotites in distributed shear zones evolved from mylonitic to ultramylonitic fabrics under increasingly hydrous conditions toward the paleo-plate contact. Olivine Crystallographic Preferred Orientation (CPO) transitions from A- and D-type fabrics in mylonites to weaker AG- and B-type fabrics in ultramylonites, accompanied by grain size reduction from ~60–80 μm to ≤20 μm, and phase mixing of olivine-orthopyroxene metasomatic layers. These transitions are consistent with a mechanical switch from dislocation creep to diffusion-accommodated creep, with sustained grain size reduction through phase mixing and growth of hydrous phases such as chlorite and amphibole.

At the paleo-plate contact, a ~10–20 m thick zone of ultramylonites is heavily serpentinized (75–90%). This zone contains finely layered, well-aligned lizardite (confirmed with Raman spectroscopy), Fe-oxides (hematite and magnetite), and relict olivine ± orthopyroxene, amphibole layers. No antigorite was identified. We interpret serpentinization as largely static and post-kinematic with respect to the incredibly strong fabrics in contact ultramylonites, supported by observations of undeformed lizardite mesh textures and hematite-decorated grain boundaries in coarser lizardite aggregates.

Bulk rock geochemical analyses along a 40 m transect in the hanging wall of the paleo-plate boundary reveal mantle Al2O3 (wt%), chondrite-normalized [Yb], and HREE concentrations all decrease systematically with distance from the contact, highlighting pimary compositional layering. Ce, Sr, and Pb show subtle enrichment at the contact where rocks are most heavily serpentinized. However, LREE and other fluid-mobile element distributions are highly variable, suggesting limited chemical overprinting associated with the serpentinizing fluid.

Our findings suggest that high-temperature ductile deformation initially localized due to hydrous phase introduction, facilitating deformation near the paleo-plate contact despite cooling conditions through shifts in deformation mechanisms. Based on the chemical data and the micro-textural observations of static lizardite, we infer that plate boundary serpentinization was late-stage and occurred under very low-temperatures (<300°C) from a highly oxidizing fluid. Serpentinites therefore did not aid strain localization or obduction but instead formed post-kinematically, locking the shear zone and forcing obduction-related strain to migrate elsewhere.

How to cite: Kotowski, A., Keats, A., Smit, H., van Broekhoven, J., Tarling, M., Godard, M., Plümper, O., Drury, M., and Hellebrand, E.: Mechanical and metasomatic evolution of a developing mantle wedge from subduction initiation to obduction, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13559, https://doi.org/10.5194/egusphere-egu25-13559, 2025.