An upper mantle seismic discontinuity beneath the Galápagos Archipelago and its implications for studies of the lithosphere‐asthenosphere boundary

Abstract

An upper mantle seismic discontinuity (the Gutenberg or G discontinuity), at which shear wave velocity decreases with depth, has been mapped from S-to-p conversions in radial receiver functions recorded across the Galapagos Archipelago. The mean depth of the discontinuity is 91 6 8 km beneath the southeastern archipelago and 72 6 5 km beneath surrounding regions. The discontinuity appears deeper beneath the portion of the Nazca plate that we infer passed over the Galapagos mantle plume than elsewhere in the region. We equate the depth of the G discontinuity to the maximum depth extent of anhydrous melting, which forms an overlying layer of dehydrated and depleted mantle. We attribute areas of shallow discontinuity depth to the formation of the dehydrated layer near the Galapagos Spreading Center and areas of greater discontinuity depth to its modification over a mantle plume with an excess temperature of 115 6 30 C. The G discontinuity lies within a high-seismic-velocity anomaly that we conclude forms by partial dehydration and a gradual but steady increase in seismic velocity with decreasing depth after upwelling mantle first encounters the solidus for volatile-bearing mantle material. At the depth of the solidus for anhydrous mantle material, removal of remaining water creates a sharp decrease in velocity with depth; this discontinuity may also mark a site of melt accumulation. Results from seismic imaging, the compositions of Galapagos lavas, and rare-earth-element concentrations across the archipelago require that mantle upwelling and partial melting occur over a broad region within the dehydrated and depleted layer. We conclude that the G discontinuity beneath the archipelago does not mark the boundary between rigid lithosphere and convecting asthenosphere.

Mantle plume capture, anchoring, and outflow during Galápagos plume‐ridge interaction

“…Gravitational ridgeward flow of melts of recycled oceanic crust in deep channels would be enhanced by the slope of the rheological boundary layer. This varies from ∼72 ± 5 km at the ridge to ∼91 ± 8 km above the plume [ Byrnes et al ., ]. For the past ∼5 Ma the GSC has maintained a distance of order ∼150–250 km north of the Galápagos plume via a series of ridge jumps, no one of which appears to be larger than ∼30–50 km [ Mittelstaedt et al ., ].…”

“…This region is exceptional because the islands are widely distributed between the upwelling plume and adjacent intermediate‐spreading ridge (Figure ), and the compositions of recently erupted basalts provide a broad aperture into underlying mantle processes. Also, there are high‐resolution seismic data from which to constrain spatial variations in lithospheric thickness, melt distribution, and temperature [ Villagómez et al ., , 2011, 2014; Gibson and Geist , ; Byrnes et al ., ]. At least five different models of plume‐ridge interaction have been proposed for Galápagos.…”

“…At least five different models of plume‐ridge interaction have been proposed for Galápagos. These include: (i) flow of plume material in a sublithospheric channel [ Morgan , ; Schilling et al ., ; Verma and Schilling , ; Braun and Sohn , ]; (ii) a spreading “puddle” model involving radial flow away from the plume stem [ Schilling et al ., ; Shorttle et al ., ]; (iii) eastward bending of the plume head in the direction of motion of the Nazca Plate [ Richards and Griffiths , ; Geist , ; White et al ., ; Harpp and White , ]; (iv) gravitational spreading along the base of the lithosphere [ Bercovici and Lin , ]; and (v) solid‐state transport of plume material beneath a dehydrated, high‐viscosity mantle lid [ Kokfelt et al ., ; Ito and Bianco , ; Villagómez et al ., ; Byrnes et al ., ].…”

Predictions and Observations for the Oceanic Lithosphere FromS‐to‐PReceiver Functions andSSPrecursors

“…Beneath older seafloor SS precursors, S ‐to‐ P and multiple S wave bounces are more consistent with a discontinuity at constant depth, ~60 km, although with some degree of scatter, at older ages (Gaherty et al, ; Kawakatsu et al, ; Schmerr, ; Tan & Helmberger, ; Tharimena, Rychert, Harmon, & White, ). There are also many reports from ocean islands, although these regions may not be representative of unaltered ocean lithosphere (Byrnes et al, ; Li et al, ; Lodge & Helffrich, ; Rychert et al, , ; Rychert & Shearer, ; Vinnik et al, ). Active source studies have been used to argue for an even sharper discontinuity (8% P wave velocity drops over <1 km) at 72–88 km depth beneath the seafloor, accompanied by a deeper velocity increase of similar magnitude, interpreted as a 10 to 18‐km‐thick melt rich channel representing the base of the plate (Mehouachi & Singh, ; Stern et al, ).…”

A 1.5 Ma record of plume-ridge interaction at the Western Galápagos Spreading Center (91°40′–92°00′W)

“…These models are broadly divided into 1) transport to the GSC along channels within the base of the lithosphere (e.g., Morgan, 1978;Schilling et al, 1982;Verma and Schilling, 1982;Braun and Sohn, 2003); 2) deflection of the Galápagos plume head primarily eastwards due to eastward migration of the Nazca Plate but with some of the material in the north reaching the ridge axis (Richards and Griffiths, 1989;Geist, 1992;White et al, 1993;Harpp and White, 2001); 3) gravity driven plume dispersal along the base of the lithosphere (Bercovici and Lin, 1996;Hoernle et al, 2000); 4) radial outflow of plume material away from its stem along the base of the lithosphere to the spreading center (Schilling et al, 2003;Shorttle et al, 2010); 5) subsolidus transport of plume material beneath a viscous residual plug to the ridge (Kokfelt et al, 2005;Ito and Bianco, 2014;Villagomez et al, 2014;Byrnes et al, 2015); and 6) melt transport via veins and channels below the anhydrous peridotite solidus (Gibson et al, 2015).…”