The global mid-ocean-ridge system is characterized by several regional-scale depth and geochemical anomalies. A prominent depth discontinuity between the East Pacific Rise and the Pacific-Antarctic Rise also coincides with a geochemical discontinuity that has been suggested by previous workers to indicate a boundary between distinct mantle- upwelling domains with separate convective histories. We propose a plate-kinematic origin for this discontinuity in which different rates of asthenospheric sequestration and spreading-center migration result in different equilibrium depths for different spreading centers. Absolute plate motions determine both the rate at which asthenosphere is converted to lithosphere (i.e., the sequestration rate) and the rate at which the spreading center moves relative to hotspots (i.e., the migration rate). If limits on the consumption (i.e., the sequestration/migration ratio) of asthenosphere by spreading centers are determined by the thickness and flux of asthenosphere, then the fast-spreading, slowly migrating East Pacific Rise should have a deeper equilibrium depth than the slower-spreading, rapidly migrating Pacific-Antarctic Rise. Sustained, localized asthenospheric consumption by the East Pacific Rise contrasts with the lower consumption and abundance of asthenospheric flux from hotspots near the Pacific-Antarctic Rise. A similar mechanism could explain the discontinuity between the localized depth anomaly on the Southwest Indian Ridge near the Bouvet hotspot and the much broader, but deeper, anomaly on the adjacent Mid-Atlantic Ridge, where asthenosphere is being transformed to lithosphere at more than three times the rate of the Southwest Indian or American-Antarctic Ridge. Geochemical evidence is consistent with the notion of deeper, more extensive melting under both of the spreading centers with anomalously high sequestration/migration ratios.