Although plate tectonics and mantle plumes were introduced to geology at the same time in the 1960s and early 1970s by J. Tuzo Wilson and Jason Morgan, unlike plate tectonics, which rapidly collected supporters from the Earth Science community, mantle plumes took a back seat. Yes, Hawaii was an example of a mantle plume and as oceanic plates moved over plumes they leave hotspot tracks. The prevailing attitude was one of “this is fine, but let's now move on to plate tectonics where the real excitement is.” For twenty years geoscientists focussed most of their efforts on trying to understand plate tectonics and document examples of it in the geologic record. It was not until the late 1980s that scientists turned some of their attention to mantle plumes, and indeed during the 1990s, when mantle plumes really “became of age”, publications dealing with mantle plumes increased exponentially. Why the long period of dormancy for mantle plumes? I believe it was simply because geoscientists were overwhelmed by plate tectonics-a band wagon effect that influenced all of the Earth Sciences.

I think three things brought mantle plumes to the forefront in the nineties. First is high speed computers, which allowed scientists to numerically model mantle processes in reasonable amounts of time with increased accuracy. Models appeared for the production and ascent of mantle plumes, the effects of mantle phase transitions on plumes, and the interaction of plumes with both the continental and oceanic lithosphere.

Plumes and Supercontinents

Introduction

Supercontinents have aggregated and dispersed several times during geologic history, although our geologic record of supercontinent cycles is only well documented for the last two cycles: Gondwana–Pangea and Rodinia (Hoffman 1989; Rogers 1996). It is generally agreed that the supercontinent cycle is closely tied to mantle processes, including both convection and mantle plumes. Pangea 200 Ma was centered approximately over the African geoid high (Fig. 2.28(a)), and the other continents moved away from this high during breakup of Pangea. Because this geoid high contains many of the Earth's hotspots and is characterized by low seismic-wave velocities in the deep mantle, it is probably hotter than average, as discussed in Chapter 2. Except for Africa, which still sits over the geoid high, continents seem to be moving toward geoid lows, which are also regions with relatively few hotspots and high lower mantle velocities, all of which point to cooler mantle (Anderson 1982). These relationships suggest that supercontinents may affect the thermal state of the mantle as the mantle beneath continents becomes hotter than normal, expands, and produces the geoid highs (Anderson 1982; Gurnis 1988). This is followed by increased mantle plume activity, which may fragment supercontinents or at least contribute to dispersal of cratons.

Introduction

Large igneous provinces, commonly referred to as LIPs, are voluminous occurrences of dominantly mafic igneous rock not directly related to plate tectonic processes (Carlson 1991; Coffin and Eldholm 1994). Included as LIPs are oceanic plateaus, continental flood basalts, passive margin volcanics, ocean-basin flood basalts, submarine ridges, giant dyke (and sill) swarms, and some large layered intrusions. Major LIPs less than about 250 Ma are shown in Figure 3.1. Next to mafic magmas emplaced at ocean ridges, LIPs are the most significant accumulations of mafic igneous rock on and near the Earth's surface. As with hotspot volcanics, LIPs are generally thought to have a mantle plume origin and to account for 5–10% of the heat and magma extracted from the mantle (Davies 1988; Sleep 1990). However, unlike ocean-ridge basalts, which appear to have been continuously erupted throughout geologic time, LIPs may be episodic, which is an intriguing feature of these rocks discussed in Chapter 8.

Volcanic LIPs are dominated by thick, laterally extensive basalt flows, some which have areal distributions of more than 105 km2 and volumes greater than or equal to 106 km3 (Table 3.1). At depth, layered mafic intrusions and sills and dykes may be important. In flood basalt provinces, sills are common and may contribute a significant fraction of the total igneous volume. In some LIPs, felsic and intermediate igneous rocks produced by fractional crystallization or partial melting of crustal rocks are associated with initial and late stages of LIP development (Campbell and Griffiths 1990).

General Features of Mantle Plumes

A mantle plume is generally considered to be a blob of relatively hot, low-density mantle that rises because of its buoyancy. The existence of mantle plumes in the Earth was first suggested by J. Tuzo Wilson (1963) as an explanation of oceanic island chains, such as the Hawaiian–Emperor chain, that change progressively in age along the chain. Wilson proposed that as a lithospheric plate moves across a fixed hotspot (the mantle plume), volcanism is recorded as a linear array of volcanic seamounts and islands parallel to the direction in which the plate is moving. Morgan (1971) championed the idea of mantle plumes, suggesting that flood basalts formed by melting of plume heads, whereas hotspot volcanic chains were derived from partial melting of plume tails. He also showed that closely spaced hotspots on the same plate had not moved significantly relative to each other and suggested this was evidence that the plumes had come from the core–mantle boundary (Morgan 1972). Morgan noted that some hotspot tracks, like the Mascarene–Chagos–Laccadive track in the Indian Ocean, are traceable to flood basalts and can be used to reconstruct paths of opening ocean basins. Richards, Duncan, and Courtillot (1989) recognized at least 10 flood basalt–hotspot track pairs that formed from mantle plumes in the last 250 Myr.

The first laboratory experiments aimed at understanding mantle plumes better were those of Whitehead and Luther (1975), who showed that plume viscosity has an important effect on the shape of a plume.

Introduction

Magmas derived from mantle plumes provide a powerful method to recognize and map compositional domains in the mantle, which in turn constrain mantle processes. The most definitive tracers are isotopic ratios of daughter elements of radioactive nuclides or concentration ratios of incompatible elements. Incompatible elements are transferred almost entirely to magmas upon melting in the mantle. For moderate or large degrees of melting, both isotopic and incompatible element ratios transfer from a mantle source to a derivative magma, providing a geochemical “signature” of the source.

Incompatible element distributions for basalts derived from four different mantle sources are shown in Figure 5.1 normalized to primitive mantle composition. Elements are arranged from most incompatible (during lherzolite melting) on the left (Rb) to least incompatible on the right (Y). Because incompatible elements are largely transferred from source to liquid, the element distributions in each basalt should reflect the element distributions in the respective mantle sources. Ocean-ridge basalts (MORB) record a depleted mantle source in which the most incompatible elements are most depleted. This source, which appears to reside in the upper mantle, is one of the restite reservoirs formed as continental crust has been extracted from the mantle over time (e.g., Hofmann 1997). Notice that oceanic plateau basalts (Ontong Java) record a mantle source somewhat less depleted in highly incompatible elements than MORB and that many oceanic island (OIB) and island arc basalts are derived from sources enriched in these elements.

Introduction

A significant but volumetrically minor amount of dominantly basaltic volcanism occurs within plates as linear chains of volcanoes that grow older in the directions of plate motion (Wilson 1963; Morgan 1971). Examples of this style of volcanism are the Hawaiian–Emperor chain in the Pacific, the Yellowstone–Snake River plain in the western United States, and the Ninetyeast Ridge in the Indian Ocean. These volcanic tracks appear to form over hotspots, which are believed to be the surface manifestations of mantle plumes (Fig. 2.1). As mentioned in Chapter 1, Wilson (1963) suggested that hotspot tracks form as oceanic crust moves over relatively stationary magma sources in the uppermost mantle. Partial melting of plumes, when they intersect the mantle solidus near the base of the lithosphere, leads to large volumes of magma, which are partially erupted or intruded at or near the Earth's surface. In addition to hotspots, two broad mantle upwellings provide the return flow caused by subduction. The mantle upwellings elevate the Earth's surface up to a few hundred meters, and because they elevate the temperature of the uppermost mantle, they also cause minor but widespread melting, giving rise to volcanism and mafic underplating of the crust. Most of the major hotspots on Earth today occur within the mantle upwellings (Fig. 2.1).

It is commonly thought that mantle plumes and mantle upwellings begin life at the D″ thermal boundary layer just above the core–mantle interface.

Introduction

There is considerable interest in the role that mantle plumes may have played in Archean (≥2.5 Ga) magma production, crustal underplating, production of oceanic plateaus, and in cooling of the mantle (Abbott 1996; Tomlinson and Condie 2001). Mantle temperatures must have been higher in the Archean, and hence the sinking of buoyant oceanic lithosphere into the mantle can account for only a small fraction of Archean heat loss (Bickle 1986; Davies 1992). Although plumes today account for no more than about 10% of the Earth's heat loss, the question of whether plumes were more important in cooling the mantle during the Archean is an important unknown in terrestrial thermal history (Davies 1993). If plumes were more important in the early stages of Earth history, as suggested by Fyfe (1978), perhaps the Archean Earth was more like Venus is today.

In this chapter we will review methods that have been used to identify Archean mantle plumes by using igneous rocks derived from plume sources and discuss the results in terms of mantle evolution.

Tracking Plumes into the Archean with Greenstones

Overview

How do we evaluate the role of mantle plumes in the early history of Earth? Three general approaches to this question are well established. First, it is well known that rock associations in oceanic plateaus and flood basalts differ from those in oceanic crust, oceanic islands, and arc systems (Condie 1997a; Kerr et al. 2000).

Introduction

Oceanic plateaus and large aseismic ridges are difficult to subduct (Vogt et al. 1976; Nur and Ben-Avraham 1982; Burke et al. 1978; Burke 1988); therefore, they may be accreted to continents (Abbott and Mooney 1995; Saunders et al. 1996; Kerr et al. 1997a). Some investigators have suggested they represent a major component of continental growth (Abbott 1996; Condie 1997c). Buoyancy models by Cloos (1993) indicate that oceanic plateaus up to about 17-km thick are easily subductable, although they may be partially obducted during collision.

When oceanic plateaus and aseismic ridges (hereafter collectively referred to as oceanic plateaus) reach about 30 km in thickness, they become too buoyant to subduct (Cloos 1993; Abbott 1996). Those formed during the last 200 Myr comprise a significant volume of crust in oceanic basins (3.7 × 106 km3) (Schubert and Sandwell 1989), and if these were all accreted to the continents, the total volume of continental crust would increase by about 5%. Although several remnants of oceanic plateaus accreted to continental crust have been described, there is a curious sparsity of large accreted oceanic plateaus. This is especially puzzling because many relatively large ocean plateaus exist in the oceans today, and these should collide and at least partially accrete to the continents in the future. Of the several reasons for a paucity of oceanic plateaus in the geologic record, Saunders et al. (1996) favor recycling of oceanic plateaus into the mantle at subduction zones.

Introduction

Before describing models of plume behavior, it is important to review the fluid characteristics of mantle plumes. Whereas a fluid is a substance that can undergo an unlimited amount of deformation, a solid will undergo only limited deformation before it breaks. Another distinction is that many solids will deform a certain amount for a given force and then return to their original shape when the force is removed. In contrast, a fluid will keep deforming as long as the force is applied and will stop deforming when the force is removed, but it will not return to its original shape. A Newtonian fluid is a material whose rate of deformation is proportional to the applied force. In the mantle, flowage occurs in response to stresses, which result in strain of the mantle rocks. The proportionality between stress and strain rate is expressed as the viscosity of a fluid. For a viscous fluid undergoing very slow flow, as in the mantle, driving forces are in balance with viscous resisting forces, and the mantle can usually be considered to be an incompressible fluid. Whether all or only some of the mantle behaves as a Newtonian fluid is a subject of considerable uncertainty and debate.