Decompressional melting of asthenosphere under spreading centers has been accepted to produce oceanic lithospheric mantle with vertical compositional variations, but these gradients are much smaller than those observed from ophiolites, which clearly require additional causes. Here we conduct high-density sampling and whole-rock and mineral analyses of peridotites across a Tibetan ophiolitic mantle section (~2 km thick), which shows a primary upward depletion (~12% difference) and local more-depleted anomalies. Thermodynamic modeling demonstrates that these features cannot be produced by decompressional melting or proportional compression of residual mantle, but can be explained by melt-peridotite reaction with lateral melt/rock ratio variations in an upwelling asthenospheric column, producing stronger depletion in the melt-focusing center and local zones. This column splits symmetrically and flows to become the horizontal uppermost lithospheric mantle, characterized by upward depletion and local anomalies. This model provides insights into melt extraction and uppermost-mantle origin beneath spreading centers with high melt fluxes.
© 2022. The Author(s).