Robert Dunn and Fernando Martinez (SOEST, University of Hawaii, Honolulu, HI 96822)

The opening of back arc basins behind subduction zones progresses from initial rifting near the volcanic arc to seafloor spreading [Karig, 1970]. During this process, the spreading ridge and the volcanic arc separate and lavas erupted at the ridge evolve from heavily subduction influenced (with high volatile contents derived from the subducting plate) to much less so [e.g., Hawkins, 1995; Pearce et al., 1995; Martinez and Taylor, 2002; Escrig et al., 2009].  The present axis of the Eastern Lau Spreading Center (ELSC) is a key region over which the mantle source composition changes from arc-like (Valu Fa ridge to the south) to MORB-like (Central Lau Spreading Center to the north). This and other systematic variations in axial crustal properties have been attributed to the changing proximity of the ridge to the arc volcanic front [Pearce et al., 1995; Martinez and Taylor, 2002]. The observations are consistent with a general decreasing “subduction-influence” in the mantle as the ridge shifts away from the arc. Virtually unstudied until recently however, is the across-axis evolution of subduction influence, which reveals the continuous history of changing mantle influence as each section of ridge has moved further from the arc.

Recent seismic tomography and geophysical studies of the ELSC show that changes in upper crustal seismic velocity closely mirror changes in seafloor morphology and depth and gravity. Together with sparse off axis geochemical data, these observations delineate distinct crustal domains with step-like transitions between them. We infer that the abrupt changes in crustal properties reflect rapid evolution of the mantle entrained by the ridge, such that stable, broad triangular upwelling regions, as inferred for open-ocean mid-ocean ridges [Forsyth et al., 1998; Conder et al., 2002], cannot form near the mantle wedge corner.  Instead, the observations imply a dynamic process in which low-viscosity subduction-influenced mantle buoyantly entrained in the ridge upwelling zone near the arc is suddenly ‘released’ as the ridge system moves away from the arc and switches to passive upwelling, with consequent rapid changes in crustal properties.  This hypothesis also explains rapid geochemical changes measured along the ridge axis [Escrig et al., 2009].  As individual spreading segments move further from the volcanic arc, they ‘let go’ of the low-viscosity arc-influenced mantle at different points in time.

Missing from this story are densely spaced off-axis geochemical samples that are needed to document the melt supply during the history of spreading.  Existing axial samples and the very few off-axis samples cannot be used to test the hypothesis of rapid changes in the mantle source composition with spreading. We suggest a dedicated campaign to collect off-axis samples to further document changes in the chemical nature of the melt supply.