Skip to main content


Meeting abstract

Foulger, G.R., Do plumes exist?, Fourth National Meeting of Earth Science of Mexico, Juriquilla, Mexico, 1 – 5 November, 2004.

Do Plumes Exist?

G. R. Foulger
Dept. Earth Sciences, University of Durham, U.K.

Mantle plumes were originally proposed in 1971 by W. Jason Morgan to explain intraplate volcanism, and large-volume ridge-centred volcanism, that seemed to be not explicable by plate tectonics. He defined precisely their characteristics and consequences, such that the original, classical plume hypothesis was readily testable. However, subsequent research tended to not confirm the predictions. Large igneous provinces have been found to be not preceded by the predicted uplift, many “hot spots” are not associated with time-progressive volcanic tracks, seismic anomalies extending down into the deep mantle have often not been seen, and at many “hot spots” there is no evidence that they are hot.

Nevertheless, instead of the theory being abandoned as is done, for example, in medical research if a drug is found to not produce the predicted results, the plume model was retained. The problems were dealt with by progressively adapting the theory in an ad hoc fashion to include the unpredicted observations. Plumes have been proposed to come from almost any depth, to rise vertically or tilt, to flow for thousands of kilometres laterally, to have narrow or broad conduits, to have no plume head, one head, or multiple heads, to produce steady or variable flow, to be long- or short-lived, to speed up or slow down, to have a source that is either geochemically depleted, enriched, or both, to have either high or low 3He/4He and to be either hot or cold. Thus, the contemporary plume theory is not falsifiable, but has become so flexible it has become a data-independent, a priori assumption. Many research papers are nothing more than reports that list new data and then suggest how the plume model must be adapted to fit them. Such an approach is unscientific. Worst, it is distracting many scientists from seeking scientific explanations for various volcanic regions.

There has recently been renewed interest in considering alternative models for “hot spots” that may fit the observations with few ad hoc assumptions or appeals to coincidence. One of these proposes that all surface volcanism is essentially a shallow phenomenon, has nothing to do with the deep mantle, and volcanic “anomalies” are simply the by-products of plate tectonics. This theory attributes “hot spots” to permissive volcanism in areas of extension. The volumes of melt produced, which may be large in the case of tholeiitic provinces such as Hawaii, or small in the case of alkaline volcanism in continental rifts, are controlled by the fertility of the underlying source. The presence of eclogite or pyroxenite in the source, or refertilised periditote, will result in larger volumes of magma than if the extending region is underlain by depleted peridotite. Volatile content and temperature will also affect melt volumes, but in a secondary way.

A third of all the world’s “hot spots” lie on or close to spreading plate boundaries. Extending intraplate regions such as the East African Rift and back-arc basins are also commonly associated with “hot spots”. Fertility may be introduced to the mantle by subducting slabs. The crustal portion transforms to eclogite at depth. Other sources of fertility are the metasomatised mantle lithosphere of subducted slabs and continental mantle lithosphere, which may delaminate following thickening as a result of continental collision. Refertilised mantle may have a solidus as much as 200¾C lower than that of depleted mantle peridotite. The melting of such material beneath an extending area may yield several times as much melt as would be extractable at the same temperature from depleted peridotite.

This alternative model for the genesis of “hot spots” raises new questions and challenges. Can the melt volumes observed be quantitatively modeled? How should seismic tomography images be interpreted? How hot are “hot spots”? Are deep mantle plumes physically possible? What is the relationship between large igneous provinces and volcanic chains? Can geochemical observations be reconciled with a fertile source at relatively normal temperatures? What is the origin of high 3He/4He? Right or wrong, the present challenge to the plume hypothesis and the innovative thinking it is encouraging, is unearthing many novel new research problems that have long gone unrecognised.


last update 3rd February, 2006