The evolution of scientific thinking in the mantle plume concept: Criticism and non-plume synthesis

Alan D. Smith, Charles Lewis

Research output: Contribution to journalArticlepeer-review

4 Citations (Scopus)


While directly offering a partial explanation for the origin of intraplate volcanism and a reference frame for plate tectonics, the mantle plume model also exerts an indirect control on models for the driving forces of plate motions, mantle convection and crust-mantle evolution. The apparent flexibility of the plume model has led to an impression that it represents the most viable explanation; however, promotion of the model has been at the expense of self-evaluation and consideration of non-plume models. The latter is illustrated by the fact that a contrasting picture of the Earth can be made by constructing a synthesis of alternatives dismissed at various stages in the evolution of the plume model. In this alternative synthesis, intraplate volcanism is derived from concentrations of volatile-bearing minerals (wetspots) at shallow levels in the asthenosphere which originate from delamination or thermal erosion of continental mantle which has been subjected to metasomatism during subduction or failed rifting events. In opening ocean basins, extrapolation of intraplate tracks into continental sutures/lineaments indicates an origin from continental mantle eroded and cycled toward the ridge axis. Intraplate volcanism in long-lived ocean basins can be traced to enriched domains of asthenosphere created during continental rifting or collision events, displaced to the east of the continent as a result of westward plate lag. This 'differential rotation' is an internal effect within the Earth arising from the transmission of stresses through the asthenosphere. The proximity of the solidus for volatile-bearing peridotite to the asthenosphere adiabat makes the enriched domains susceptible to melting in response to pervasive shearing which results throughout the asthenosphere from flow imposed by eastward movement of the mesosphere and plate motion induced by boundary forces, without need to invoke thermal anomalies (plumes). Stationary melt-collection layers are formed at the intersection of opposing asthenospheric flow regimes when plate motion occurs in the opposite direction to mantle flow. Melt is released when lithospheric stress trajectories intersect the stationary layer, creating an illusion of fixed deep-seated melting anomalies. Formation of volcanic lines is favoured when plate motion occurs in the same direction as mantle flow. Correspondingly, the fate of subducted oceanic crust from the Proterozoic to Recent, as modelled from the Nd-Hf isotopic variation in MORB, has been re-mixing with the depleted mantle (marble-cake model). The existence of a very depleted mantle reservoir in the Early Archean is suggested from the Nd isotopic variation in komatiites, and is compatible with a tectonic scenario where crust was less efficiently re-mixed with the mantle on account of subducting lithosphere melting at shallow (<200km) depth due to high mantle temperatures. Slabs would therefore be less efficient for cooling the mantle, such that the thermal regime is self-stabilising. A large early crust is thus both a consequence, and cause of, hot mantle temperatures, and is compatible with a thermal regime marked by komatiite genesis along ocean ridges. That this alternative model is simpler than, and lacks the internal contradictions and paradoxes associated with the plume model, suggests it was not an absence of viable alternative models nor inadequate data which led to the monopoly on geodynamic interpretations exerted by the plume model.

Original languageEnglish
Pages (from-to)1-40
Number of pages40
JournalJournal of the Geological Society of China
Issue number1
Publication statusPublished - 1999 Feb 1

All Science Journal Classification (ASJC) codes

  • Geology

Fingerprint Dive into the research topics of 'The evolution of scientific thinking in the mantle plume concept: Criticism and non-plume synthesis'. Together they form a unique fingerprint.

Cite this