On the cause of the asymmetric distribution of seamounts about the Juan de Fuca ridge: ridge-crest migration over a heterogeneous asthenosphere

Abstract

The distribution of non-hotspot seamounts in the northeast Pacific is highly asymmetric; small seamount chains and isolated edifices are numerous on the Pacific plate, but nearly absent on the Juan de Fuca plate. We propose a hypothesis for the asymmetric generation of seamounts near a ridge axis in which upwelling and early melting of upper mantle heterogeneities occur in advance of a spreading centre which migrates with respect to the asthenospheric frame of reference. Plate motion solutions indicate that the Juan de Fuca ridge is migrating to the west at a ridge-perpendicular rate of 20 mm a −1, which is large compared to the half-spreading rate of 30 mm a −1. Migration of the ridge axis will result in the initiation of upwelling in the upper mantle in advance of the spreading centre. If slightly enriched, lower melting temperature heterogeneities are present in the upper mantle, they will intersect their solidus at a greater depth and will begin to melt earlier than the host peridotite. Seamount volcanism will occur preferentially on the Pacific plate because of two factors: Firstly, more asthenosphere and hence more early-melt heterogeneities must ascend to supply the Pacific plate which is translating more rapidly than the relatively stationary Juan de Fuca plate. Secondly, the asthenosphere that is required to ascend to supply the thickening of the Juan de Fuca plate will have been flushed of its significant shallow-level heterogeneities by the previous advance and passing of the ridge; few should remain to cause volcanism on the Juan de Fuca plate. Testable physical and petrologic predictions of this model can be identified. For example, within a small seamount chain, the age difference between individual seamounts and the crust on which they lie will decrease toward the ridge axis. Variation in lava chemistry along a small seamount chain, which is interpreted to represent magmatism from a single large heterogeneity, will provide the major discriminant of this hypothesis. Systematic variations should be observed, with earlier formed edifices being constructed of more enriched lavas, and later (younger) ones of more depleted lavas. Isotope ratios should be unaffected by varying degrees of partial melting, but could display increased dilution towards the ridge axis as melting of the host mantle and magma mixing become significant.