The formation of Archean lithosphere

The distinctive composition and physical properties of the stiff, buoyant mantle roots beneath the cratons (Section 11.3.1) result from the chemical depletion and extraction of melts from the primitive mantle. These two processes lowered the density and increased the viscosity of the residue left over from partial melting and resulted in a keel that consists mostly of high-Mg olivine and high-Mg orthopyroxene (O'Reilly et al., 2001; Arndt et al., 2002). Both of these components are absent in fertile (undepleted) mantle peridotite and are rare in the residues of normal mantle melting, such as that which produces modern oceanic crust and oceanic islands. Consequently, most workers have concluded that the distinctive composition is related to unusually high degrees (30-40%) of mantle melting over a range (4-10 GPa) of mantle pressures (Pearson et al., 2002; Arndt et al., 2002). High-degree partial melting of mantle peridotite produces magma of komatiitic (Section 11.3.2) composition and a solid residue that is very similar to the composition of Archean lithospheric mantle (Herzberg, 1999; Arndt et al., 2002).

One radiometric system that has been of considerable use in determining when melt extraction and Archean root formation occurred involves the decay of 187Re to 188Os (Walker et al., 1989; Carlson et al., 2005). The key feature of this isotopic system is that Os is compatible during mantle melting and Re is moderately incompatible. Consequently, any residue left behind after melt extraction will have a lower Re and a higher Os concentration than in either the mantle melts or the fertile mantle. This characteristic allows Re-Os isotope analyses of mantle xenoliths to yield information on the age of melt extraction. The data from mantle xenoliths show that the oldest melting events are Early-Middle Archean in age. Significant amounts of lithospheric mantle also formed in Late Archean times and are associated with voluminous mafic magmatism (Pearson et al., 2002).

Although high-degree partial melting undoubtedly occurred, this process alone cannot explain the origin of the Archean mantle lithosphere. The main reason for this conclusion is that the abundance of komatiite found in the Archean crustal record is much too low to balance the amount of highly depleted peridotite found in the cratonic mantle (Carlson et al., 2005). This imbalance suggests that either a large proportion of komati-

itic magma never reached the surface or other processes must have contributed to root formation. One of these processes is an efficient sorting mechanism that concentrated the unusual components of the Archean mantle lithosphere at the expense of all others (Arndt et al., 2002). The most likely driving force of the sorting is the buoyancy and high viscosity of high-Mg olivine and orthopyroxene, although exactly how this happened is uncertain. The density and viscosity of these minerals depends upon their Mg-Fe ratios and water content, respectively; which are lower in Archean mantle lithosphere compared to normal asthenospheric mantle. Arndt et al. (2002) considered three processes that could have resulted in the mechanical segregation and accumulation of a layer of buoyant, viscous mantle near the Earth's surface during Archean times. First, upwelling buoyant residue in the core of a mantle plume could have separated from the cooler, denser exterior and accumulated during ascent (Fig. 11.2a). In this model, melting begins at high pressure (~200 km depth) and continues to shallow depths, by which point melt volumes are high and the dense residue of early, low-degree melting is swept away by mantle flow. Second, buoyant residue could have segregated slowly as material was transported down subduction zones and recycled through the mantle in convection cells (Fig. 11.2b). Third, some subcontinental lithosphere could be the remnants of an initial crust that crystallized in an Archean magma ocean that formed during the final stages of Earth accretion (Fig. 11.2c). In all three cases, buoyant, viscous material rises and is separated from higher density residue during mantle flow. Whether some combination of these or other processes helped to form the cratonic keels is highly speculative. Nevertheless, they illustrate how several different mechanisms could have concentrated part of the residue of mantle melting into a near-surface layer.

In addition to high-degree partial melting and efficient sorting, most authors also have concluded that the formation and evolution of mantle lithosphere involved a multi-stage history involving many tectonic and mag-matic events (James & Fouch, 2002). However, opinions are divided over whether root construction preferentially involved the underthrusting and stacking of subducted oceanic slabs (Carlson et al., 2005), the accretion and thickening of arc material (Lee, 2006), or the extraction of melt from hot mantle plumes (Wyman & Kerrich, 2002). By applying a range of criteria some geologic studies have made compelling cases that ancient mantle plumes played a key role in the

(a) Model 1: Segregation of residue from an upwelling mantle plume

Komatiite lavas

Primitive^ crust prjithosphere

(b) Model 2: Segregation of recycled refractory residue

Oceanic plateau

Primitive^ crust prjithosphere

Residue of high-degree melting

Residue of low-degree melting

Oceanic plateau

Residue of high-degree melting

Residue of low-degree melting

(c) Model 3: Preservation of remnants of the crust of a magma ocean

Crust

Liquid interior of magma ocean

Fig. 11.2 (a-c) Three possible mechanisms that could allow the segregation and accumulation of high-Mg olivine and orthopyroxene near the surface of the Earth (after Arndt et al., 2002, with permission from the Geological Society of London).

Liquid interior of magma ocean

Flotation and/or in-situ crystallization of ol + opx

Flotation and/or in-situ crystallization of ol + opx evolution ofArchean lithosphere (Tomlinson & Condie, 2001; Ernst et al., 2005). Data from seismic profiles, geochronologic studies, and isotopic analyses indicate that many roots were affected by large pulses of mafic magmatism during the Late Archean (Wyman & Kerrich, 2002; James & Fouch, 2002). Other studies, however, have emphasized a subduction zone setting to explain the evolution of Archean mantle lithosphere. Most of the cratons display evidence for the significant modification of cratonic roots by terrane collisions and thickening during at least some stage in their history (James & Fouch, 2002; Schmitz et al., 2004). In support of a subduction zone mechanism, a Late Archean (2.82.6 Ga) fossil subduction zone (Fig. 11.3) has been found within the Abitibi craton in northern Canada using seismic data (Calvert & Ludden, 1999; van der Velden et al., 2006). Nevertheless, it is important to recognize that Archean mantle roots probably resulted from more than one tectonic environment and that no single setting or event is applicable to all cases.

The distinctive rock associations that comprise granite-greenstone belts (Section 11.3.2) provide another important means of evaluating the mechanisms that contributed to the formation and evolution of Archean lithosphere. One of the key questions to answer is whether the komatiitic and tholeiitic lavas that form the majority of the greenstones formed in environments that were broadly similar to modern tectonic environments. For example, if these lavas loosely represent the Archean equivalent of modern mid-ocean ridge basalts, as is commonly believed, then they might be used to infer that much of the volcanism in Archean times involved the creation and destruction of ocean crust (Arndt et al., 1997). However, one of the problems

'tjo"

^ ^Abitibi greenstone domain ^ ^

100 km W-E offset

Opatica gneiss-plutonic domain

^ ^Abitibi greenstone domain ^ ^

100 km W-E offset

Opatica gneiss-plutonic domain

'tjo"

Fig. 11.3 Seismic reflection profile of the Opatica-Abitibi belt in the Superior Province of northern Canada (modified from van der Velden et al., 2006, by permission of the American Geophysical Union. Copyright © 2006 American Geophysical Union). Interpretation is modified from Calvert et al. (1995), Lacroix & Sawyer (1995), and Calvert & Ludden (1999). S, fossil subduction zone; Sh, shingle reflections suggesting imbricated material in the middle crust.

Fig. 11.3 Seismic reflection profile of the Opatica-Abitibi belt in the Superior Province of northern Canada (modified from van der Velden et al., 2006, by permission of the American Geophysical Union. Copyright © 2006 American Geophysical Union). Interpretation is modified from Calvert et al. (1995), Lacroix & Sawyer (1995), and Calvert & Ludden (1999). S, fossil subduction zone; Sh, shingle reflections suggesting imbricated material in the middle crust.

with these comparisons is that no chemically unaltered, complete example of Archean ocean crust is preserved. In addition, the Archean mantle was hotter by some amount than the modern mantle (Section 11.2), which undoubtedly influenced the compositions, source depths, and patterns of the volcanism (Nisbet et al., 1993). These problems have complicated interpretations of the processes that produced and recycled Archean crust and how they may differ from those in modern environments.

Most authors have concluded that the high magnesium contents and high degrees of melting associated with the formation of komatiites reflect melting temperatures (1400-1600°C) that are higher than those of modern basaltic magmas (Nisbet et al., 1993). Exactly how much hotter, however, is problematic. Parman et al. (2004) proposed a subduction-related origin for these rocks similar to that which produced boninites in the Izu-Bonin-Mariana island arc (Fig. 11.4). Boninites are high-Mg andesites that are thought to result from the melting of hydrous mantle in anomalously hot forearc regions above young subduction zones (Crawford et al., 1989; Falloon & Danyushevsky, 2000). If the komatiites were produced by the melting of hydrous mantle, then the depth of melting could have been relatively shallow, as in subduction zones, and the Archean mantle need only be slightly hotter (~100°C) than at present (Grove & Parman, 2004). In this interpretation, shallow melting and subduction result in the formation and thickening of highly depleted mantle lithosphere that some time later is incorporated into the cratonic mantle below a continent.

Alternatively, if the source rocks of komatiites were dry then high ambient temperatures in the Archean

(a) Subduction initiation

Komatiitic crust

(a) Subduction initiation

Komatiitic crust

Fig. 11.4 Conceptual model for the generation of komatiites and cratonic mantle by partial melting in a subduction zone (after Parman et al., 2004. Copyright © 2004 Geological Society of South Africa). (a) Partial melting produces komatiitic magma in a forearc setting. (b) Mature subduction cools and hydrates residual mantle. (c) Obduction of komatiitic crust occurs during collision.

Fig. 11.4 Conceptual model for the generation of komatiites and cratonic mantle by partial melting in a subduction zone (after Parman et al., 2004. Copyright © 2004 Geological Society of South Africa). (a) Partial melting produces komatiitic magma in a forearc setting. (b) Mature subduction cools and hydrates residual mantle. (c) Obduction of komatiitic crust occurs during collision.

Komatiite Mantle

Fig. 11.5 Model of komatiitic and tholeiitic basalt formation involving mantle plumes (after Arndt et al., 1997, by permission of Oxford University Press). Model shows the influence of lithospheric thickness on depth of melting where CFB is continental flood basalt, OIB oceanic island basalt, and MORB mid-ocean ridge basalt.

Fig. 11.5 Model of komatiitic and tholeiitic basalt formation involving mantle plumes (after Arndt et al., 1997, by permission of Oxford University Press). Model shows the influence of lithospheric thickness on depth of melting where CFB is continental flood basalt, OIB oceanic island basalt, and MORB mid-ocean ridge basalt.

mantle would have caused melting to begin at depths that were much greater than occurs in subduction zones, possibly in upwelling mantle plumes or at unusually deep levels (~200 km) beneath mid-ocean ridges. The greater depths of melting would produce large volumes of basalt and oceanic crust that was much thicker (20-40 km) than it is today (Bickle et al., 1994). Evidence of large volumes of mafic magma and high eruption rates have suggested that oceanic plateaux and continental flood basalts are the best modern analogues for such thick mafic crust and invites comparisons with Phanerozoic LIPs (Section 7.4.1) (Arndt et al., 1997, 2001). In this latter context, the differences between the modern and ancient rocks are explained by variations in the depth of melting and in the effects of a thick overriding lithosphere (Fig. 11.5). These and a variety of other models (Fig. 11.6) illustrate how information on the depth and source of the melting that produced komatiites has important consequences for both the tectonic setting and the thermal evolution of the early Earth.

2000

1800

1600

1400

1200

2000

1800

1600

1400

1200

Mid-ocean Subduction Modern Archean ridge zone plume komatiite

Fig. 11.6 The range of mantle melt generation temperatures estimated for various modern tectonic settings compared to temperatures inferred for komatiite melt generation by a plume model (black filled oval) and a subduction model (gray filled oval) (after Grove & Parman, 2004, with permission from Elsevier).

Mid-ocean Subduction Modern Archean ridge zone plume komatiite

Fig. 11.6 The range of mantle melt generation temperatures estimated for various modern tectonic settings compared to temperatures inferred for komatiite melt generation by a plume model (black filled oval) and a subduction model (gray filled oval) (after Grove & Parman, 2004, with permission from Elsevier).

A variety of tectonic models also have been postulated for the origin of Archean continental crust. Windley (1981) noted the geologic and geochemical similarities between Archean tonalite-trondhjemite-granodiorite (TTG) suites and exhumed granitoids associated with Andean-type subduction zones (Section 9.8). He considered this to be an environment in which voluminous quantities of tonalite can be produced, and concluded that this represents a reasonable analogue for the formation of these rocks in Archean times. Subsequent work has led to a general consensus that these subduction models are applicable to the Late Archean. However, their applicability to Early and Middle Archean times when thick oceanic crust may have inhibited subduction is more controversial. As an alternative to subduction, Zegers & van Keken (2001) postulated that TTG suites formed by the removal and sinking of the dense, lower part of thick oceanic plateaux. The peeling away, or delamination, of a dense eclogite root results in uplift, extension, and partial melting to produce TTG suite magmas. This process could have returned some oceanic material into the mantle and may have accompanied collisions among oceanic ter-ranes in Early-Middle Archean times. However, the possible absence of subduction creates a problem in that, assuming a nonexpanding Earth (Section 12.3), a high rate of formation of oceanic lithosphere during these times must have been accompanied

by a mechanism by which oceanic lithosphere also was destroyed at high rates.

An important aspect of the origin of TTG suites is the type of source material that melted to produce the magma. Early petrologic studies suggested that these magmas could result from the partial melting of subducted oceanic crust in the presence of water (Martin, 1986). However, more recent work has emphasized other sources, including the lower crust of arcs and the base of thick oceanic plateaux (Smithies, 2000; Condie, 2005a). The importance of the source material is illustrated by a two-stage model proposed by Foley et al. (2003). This model envisages that during the Early Archean, oceanic crust was too thick to be subducted as a unit, and so its lowermost parts delaminated and melted (Fig. 11.7a). These lower roots are inferred to have been pyroxenites that were produced by the meta-morphism of ultramafic cumulate layers. The melting of the pyroxenite did not favor the generation of TTG

melts but produced basaltic melts instead. As the oceanic crust cooled and became thinner, a point was reached in the Late Archean when the entire crust could subduct (Fig. 11.7b). At this time hydrothermally altered crust, such as amphibolite, was introduced into subduction zones and led to the widespread formation of TTG suites. This model supports the view that the formation of the earliest continental crust requires subduction and the melting of a hydrous mafic source.

Was this article helpful?

0 0
Boating Secrets Uncovered

Boating Secrets Uncovered

If you're wanting to learn about boating. Then this may be the most important letter you'll ever read! You Are Going To Get An In-Depth Look At One Of The Most Remarkable Boating Guides There Is Available On The Market Today. It doesn't matter if you are just for the first time looking into going boating, this boating guide will get you on the right track to a fun filled experience.

Get My Free Ebook


Post a comment