No large earthquakes in fully exposed subducted seamount

Abstract

Bathymetric highs on the ocean floor ultimately sink into highly seismic subduction zones, raising vigorous debates on their potential to trigger or arrest large earthquakes (Mw > 7.5). Many geophysical and seismological studies addressing this problem meet penetration and/or resolution issues and deal with only the most recent earthquakes. We herein present the missing piece of the puzzle with the time-integrated field and petrographic record of a unique, almost intact subducted seamount cropping out along a fossil subduction interface. We document seamount buildup and subduction down to ~30 km, and we show that this seamount did not behave as a large earthquake asperity and may have acted as a barrier. INTRODUCTION The bathymetric roughness of seamounts and seamount chains entering subduction zones (Hillier and Watts, 2007; Bassett and Watts, 2015) has long been suspected to impact the geometry and seismic coupling of the subduction interface (Cloos and Shreve, 1996; Scholz and Small, 1997; Agard et al., 2018). Constraining the size and/or location of megathrust ruptures is critical for assessing earthquake hazard, and it is therefore of crucial importance to understand whether seamounts can limit large earthquake rupture propagation (acting as barriers) or may generate large earthquakes (acting as asperities; Cloos, 1992; Mochizuki et al., 2008; Wang and Bilek, 2011; Geersen et al., 2015; Saffer and Wallace, 2015). Despite spectacular geophysical imaging at the trench (Ranero and von Huene, 2000), seamounts are poorly imaged once subducted beyond ~15 km depth (Kodaira et al., 2000; Singh et al., 2011; Saffer and Wallace, 2015), and their internal deformation is thus beyond reach (Park et al., 1999). Fossil exhumed examples are scarce (MacPherson, 1983), yet they are of utmost interest because they possibly preserve millions of years of seamount evolution on the ocean floor and within the subduction zone. AN EXCEPTIONALLY PRESERVED FOSSIL SEAMOUNT Here, we report the discovery of a unique, fully exposed subducted seamount, namely, the Siah Kuh (SK) massif, and we present its structure and evolution, which we detailed through extensive field and petrological data. This massif crops out within ophiolite fragments of the Neotethys Ocean subducted beneath Eurasia (Agard et al., 2011) in the easternmost portion of the Zagros Mountains, next to and below oceanic blueschists metamorphosed during the Late Cretaceous (Angiboust et al., 2016). This massif rises from Quaternary sediment infill as an 18 × 12-km-wide and ≥1.5-km-high feature (Figs. 1A and 1B), and it is composed of two subunits separated by tectonic contacts (Figs. 1C and 1D): (1) An ~15 × 12 km oval-shaped unit (unit A) to the southwest is composed of up to 3 km of basaltic lava flows and pillow lavas intercalated with pillow breccia. Rhyodacitic subvolcanic rocks intruding this basaltic core are associated with lavas erupted on top of the basalts. Basalts and felsic lavas are overlain by a ≤500-m-thick Late Cretaceous sedimentary sequence, fully exposed on the southern flank of the unit. The base of this sequence consists of a massive limestone cap (10–50 m thick), of reef to lagoon affinity, with recrystallized fossil fragments like urchin spines, foraminifera, and gastropods. The top of the sequence is a variably thick (~100–500 m), mostly detrital, deepening-up sedimentary sequence of tuffaceous sandstone, red clay, pelagic limestone, and olistostromic debris flows. Pillow lavas (up to 1 km thick) emplaced conformably on top of these sediments indicate resumption of volcanic activity after sedimentation (Fig. 1). (2) The smaller crescent-shaped unit to the northeast of SK (unit B) consists of, from bottom to top, serpentinites with meterto decameter-large gabbroic pods and plagiogranites, a layer of massive gabbro overlain by rhyodacitic lavas, and finally kilometer-thick basaltic lavas without significant sedimentary cover. While Unit B resembles classic ocean floor lithostratigraphy, the size and circular shape, amount of volcanism, shallow reef limestone cap—and overlying high-energy deposits (olistostromic sediments, debris flow)—indicate that unit A is a bathymetric anomaly on the seafloor, i.e., a seamount. The deepening-up sedimentary sequence hints at isostatic reequilibration of the oceanic lithosphere after the first magmatic event and/or seafloor subsidence. Unit B was thrust southwestward onto unit A via a high-angle fault rooted in the lowermost, basal serpentinite horizon (Fig. 1D), suggesting initial strain localization beneath the oceanic Moho. Associated fault striations in gabbros strike N50° on average, parallel to the convergence direction during subduction beneath Eurasia (Agard et al., 2011). Smaller thrust faults of similar orientation rooted in sheared sediments (mainly tuffaceous sandstone and pelagic limestone) cut across the northeastern part of unit A and delineate tectonic subunits (A1′–4; Fig. 1C). Small serpentinite bodies are pinched inside these faults, and basalts located in their vicinity lack significant deformation. The Oligocene Zagros collision (Agard et al., 2011) was responsible for the final arching of the whole SK massif into a south-to-southeastvergent anticline (with locally tight refolded