GEODYNAMIC REGIMES AND TECTONIC SETTINGS FOR …€¦ · METAMORPHISM: RELATIONSHIP TO THE SUPERCONTINENT CYCLE MICHAEL BROWN Laboratory for Crustal Petrology, Department of Geology,

Indian Journal of GeologyVol.80, Nos. 1-4, (2008) p.3-21 (Published January 2010)

GEODYNAMIC REGIMES AND TECTONIC SETTINGS FORMETAMORPHISM: RELATIONSHIP TO THE SUPERCONTINENT CYCLE

MICHAEL BROWN

Laboratory for Crustal Petrology, Department of Geology,University of Maryland, College Park, MD 20742-4211, USA

e-mail: [email protected]

Abstract

Metamorphism associated with orogenesis provides a mineral record that may be inverted toyield ambient apparent thermal gradients. On modern Earth, tectonic settings with lower thermalgradients are characteristic of subduction zones whereas those with higher thermal gradients arecharacteristic of backarcs and orogenic hinterlands. The duality of thermal environments reflects theasymmetry of one-sided subduction, which is the hallmark of modern plate tectonics; a duality ofmetamorphic belts is the characteristic imprint of one-sided subduction in the geological record.Apparent thermal gradients derived from inversion of age-constrained metamorphic P-T data maybe used to identify tectonic settings of ancient metamorphism and to evaluate Precambrian geodynamicregimes. The Neoarchean records the first occurrences of granulite facies ultra-high temperaturemetamorphism (G-UHTM) and medium-temperature eclogite-high-pressure granulite metamorphism(E-HPGM), signifying a change in geodynamics that generated sites of higher and lower heat flowthan those implied by apparent thermal gradients recovered from the older geological record. G-UHTM is dominantly a Neoarchean-Cambrian phenomenon inferred to have developed in settingsanalogous to backarcs and orogenic hinterlands. In addition to Proterozoic occurrences, E-HPGM iscommon in the Paleozoic Caledonides and Variscides, and is inferred to record subduction-to-collisionorogenesis. The occurrence of G-UHTM and E-HPGM belts since the Neoarchean signifies a changeto one-sided subduction of oceanic lithosphere, possibly beginning as early as the Paleoarchean, andwidespread transfer of water into the upper mantle. This change registers the beginning of a 'Proterozoicplate tectonics regime' that evolved during a Neoproterozoic transition to the 'modern plate tectonicsregime' characterized by colder apparent thermal gradients and deep subduction of continental crust.The age distribution of metamorphic belts is not uniform, recording amalgamation of continentallithosphere into supercratons or supercontinents.

Keywords : Geothermal gradients, metamorphism, plate tectonics, supercontinents

1. Introduction

The plate tectonics revolution provideda paradigm to understand the Cenozoic-Mesozoic tectonics of the lithosphere – thestrong outer layer of Earth above the softerasthenosphere – which led to debate about

when Earth first adopted plate tectonics.Consensus is emerging that Earth had (partiallyor completely) adopted plate tectonics bysometime in the Archean (Brown, 2006,2007a, b, 2008), although there are those whoargue for (some form of) plate tectonics asearly as the Hadean (Davies, 2006)—

Metamorphism and Geodynamic Regimes4

consistent with the null hypothesis that platetectonics was the mode of convectionthroughout Earth history – and those whoargue against world-wide modern – stylesubduction before the Neoproterozoic (Stern,2007) – requiring an alternative hypothesis toplate tectonics for the Hadean toMesoproterozoic interval. From a geologicalperspective, we may break into severalcomponents the question of when platetectonics began on Earth. We may ask: whendid the lithosphere first behave as a mosaic ofplates – torsionally-rigid lithosphere elementsbounded by zones of plate generation, platedestruction or transform displacement – andhow far back in time may we identifyindependent horizontal motions from differentcratons; and, when in the rock record do wefirst identify zones of convergence andsubduction of oceanic lithosphere (e.g. Davies,1992)? However, care is required – whatimprint in the geological record do we take toanswer each of these questions (Cawood et al.,2006)?

2. The Hallmark of Plate Tectonics

In the early days of plate tectonics, therelationship between plate tectonics andmetamorphism was addressed by Miyashiro(1972). The first record of blueschist faciesmetamorphism occurs in the Neoproterozoic,and this is commonly taken as one indicatorof the beginning of the modern style ofsubduction (Stern, 2005). However, due tosecular cooling, Brown (2006, 2007a) askedwhether plate tectonics on a warmer Earthmight have left a different imprint in theancient rock record. To answer this question,it is necessary to establish the distinctive

characteristics of plate tectonics that might bepreserved in the rock record. Brown (2006,2007a) argued that the definitive hallmark ofplate tectonics is a duality of thermalenvironments at convergent plate margins (thesubduction zone and the arc-backarc ororogenic hinterland), the asymmetry of whichis a direct consequence of one-sidedsubduction (Gerya et al., 2008). The imprintof this hallmark in the rock record is registeredas the contemporary occurrence of spatiallydistinct contrasting types of metamorphism.

Metamorphic rocks record evidence ofchange in pressure and temperature with time–commonly expressed as a P-T-t path – or burialand exhumation in the thermal environment(s)in which the mineral assemblage(s)equilibrated (e.g. Brown, 1993, 2001). Ineffect, the P-T-t path records the changingspectrum of metamorphic (transient)geotherms characteristic of a particulartectonic setting. Different settings arecharacterized by different ranges of dT/dP andregister different metamorphic imprints in therock record.

Early plate-like behavior may haveallowed crust to float, forming thick stacksabove zones of boundary-layer downwelling,or, for a slightly cooler Earth, forming thickstacks above zones of 'sub-lithospheric'subduction (Davies, 1992). However, thesebehaviors are unlikely to have created dualthermal environments at the site of tectonicstacking, although thickening (with or withoutdelamination of eclogite sinkers) could haveinduced melting deep in the stack generatingtonalite-trondhjemite-granite-type magmas(Foley et al., 2002, 2003). Therefore, inprinciple, the metamorphic imprint imposedby either of these behaviors during the early

M. Brown 5

part of Earth history should be distinguishablefrom that imposed by one-sided subductionand plate tectonics sensu stricto. Anotherchallenge is to assess whether once platetectonics was established it was maintainedto the present day or whether geodynamics onEarth alternated between plate tectonics andsome other mode (Sleep, 2000) or was episodic(O'Neill et al., 2007) or intermittent (Silver andBehn, 2008), particularly in the Mesoarcheanto Mesoproterozoic interval.

In this chapter, I review metamorphismin relation to different types of orogenic systemas recorded in the Phanerozoic record beforeextending the analysis back through most ofthe Precambrian using a published datasetcompiled in 2005 (Brown, 2007a). I discussthe implications of these data in relation to dualthermal environments as the hallmark of platetectonics and identify when the imprint of thisdistinctive feature first appears in the rockrecord (cf. Brown, 2006). Finally, I evaluatethe age distribution of metamorphism since theNeoarchean and consider how this relates tothe supercontinent cycle (cf. Brown, 2007b,2008).

3. Earth in the Phanerozoic Eon

Currently, there are two main zones ofsubduction into the mantle (Collins, 2003) —the circum-Pacific and the Alpine-Himalayan-Indonesian subduction systems — and twomajor zones of upwelling or superswells(Montelli et al., 2006; Tan and Gurnis,2007) — under southern Africa and the SouthPacific — which define a simple pattern oflong-wavelength mantle convection. ThePhanerozoic has been distinguished byrearrangement of the continental lithosphereduring the final steps in supercontinent

assembly and the early stages ofsupercontinent breakup. The formation ofEurasia and its linking to Gondwana to formPangea involved successive subduction-to-collision orogenic systems — theAppalachian/Caledonian-Variscide-Altaid andthe Cimmerian-Himalayan-Alpine orogenicsystems — within a single (Pangean)convection cell, whereas a complementary(Panthalassan) convection cell, centered on theSouth Pacific superswell, is composed ofocean lithosphere and defined at its outer edgeby circum-Pacific accretionary orogenicsystems (Collins, 2003). The Panthalassan cellwas established by the end of theNeoproterozoic, centered on the formerRodinia and reflecting enhanced heat fluxassociated with the slab graveyard due toRodinia assembly. It reached a maximumaround the Devonian-Carboniferous boundary,and has been in decline since the Pangean cellbegan expanding in the Jurassic, centered onthe former Gondwana and reflecting enhancedheat flux associated with the slab graveyarddue to Gondwana assembly.

3.1. Orogenic systemsThe Circum-Pacific and Cimmerian-

Himalayan-Alpine orogenic systems definetwo orthogonal great circle distributions of thecontinents, which may reflect a simple patternof mantle convection. This inference issupported by a variety of geophysical data(Richards and Engebretson, 1992). Along eachgreat circle the convergent plate margins arecharacterized by a different type of orogenicsystem, which has been the situation during atleast the Cenozoic and Mesozoic Eras (Zwart,1967; Maruyama, 1997; Ernst, 2001; Liou etal., 2004). These types are accretionaryorogenic systems [variants have been called


'Pacific-type' (e.g. Matsuda and Uyeda, 1971)or 'Cordilleran-type' (e.g. Coney et al., 1980)],which form during ongoing subduction, asexemplified by the Phanerozoic evolution ofthe Pacific Ocean rim, and collisional orogenicsystems [sometimes called 'Himalayan-type'(Liou et al., 2004) or 'Turkic-type' (Sengör andNatal'in, 1996)], in which an ocean is closedand arcs and/or allochthonous terranes and/orcontinents collide, as exemplified by theTethysides.

Accretionary orogenic systems varyaccording to whether the over-riding plate isretreating, neutral or advancing (e.g. theTasmanides versus the South AmericanCordillera), and whether suspect terranes,some of which may be far-traveled (cf. Coneyet al., 1980; Johnston and Borel, 2007),become accreted to the over-riding plate (e.g.the North American Cordillera). For Turkic-type orogenic systems, Sengör and Natal'in(1996) argued that the pre-collision history ofone or both of the colliding margins mightinvolve the growth of large accretionaryorogenic systems, with significant juvenile arcmagmatism. Nonetheless, a distinctionbetween the accretionary and the collisionalphases of an orogenic system remains usefulbecause some accretionary orogenic systemsexist for hundreds of millions of years withoutdisruption by collision. In addition, otheraccretionary orogenic systems have thecontinuity of subduction interrupted – but notnecessarily terminated if a new trench isformed outboard of the collider – bysubduction of ocean floor debris or an oceanicplateau or non-terminal collision and suturingof allochthonous terranes or arcs (e.g. Cloos,1993; Collins, 2002).

Major mountain belts of the Circum-Pacific orogenic systems are located in

subduction zone backarcs, which arecharacterized by high heat flow (greater than70 mWm-2 for continental crust with averageradiogenic heat production) and uniformly thinand weak lithosphere over considerable widths(Hyndman et al., 2005a). Subduction zonebackarcs are hot due to shallow convection inthe mantle wedge consequent upon a reductionin viscosity induced by water from theunderlying subducting plate. Mohotemperatures in subduction zone backarcs are800-900ºC and lithosphere thicknesses are 50-60 km, compared to 400-500ºC and 200-300km for cratons (Hyndman et al., 2005a); thedifference results in backarc lithosphere beingat least an order of magnitude weaker thancratons. Parameters such as thecontemporaneous surface heat flow cannot bemeasured directly in ancient orogens, so wemust look to the metamorphic record forinformation. Also, thermal expansion due tothe high lithosphere temperatures is argued toaccount for about 2500 m of elevation insubduction zone backarcs without significantcrustal thickening (Hyndman et al., 2005b),which may be recorded in ancient orogens asvariation in the pressure field. Most Circum-Pacific mountain belts are broad zones of long-lived tectonic activity because they aresufficiently weak to be deformed by the forcesdeveloped at plate boundaries; complexdeformation histories in these orogenicsystems likely relate to changing dynamicsalong the convergent boundaries (Dewey,1975). In accretionary orogenic systemscharacterized by accumulation of suspectterranes, former backarcs continue as a locusof deformation during terrane suturing becausethey remain weaker than the hinterland.

Shortening and crustal thickening inlarge hot orogenic systems, such as the

M. Brown 7

Himalayan orogenic system and the SouthAmerican Cordillera, appears to beaccommodated mostly in the weak lower crust(Beaumont et al., 2006; Sobolev and Babeyko,2005). The upper crust may be uplifted as aplateau, which may be underthrust by theadjacent stable craton, and the upper crust mayremain largely undeformed internally withdeformation localized into high-strain belts.

3.2. The metamorphic realmThe metamorphic realm traditionally is

divided into facies, each represented by agroup of mineral assemblages associated inspace and time that are inferred to registerequilibration within a limited range of P-Tconditions (Fig. 1a). Some rocks characteristicof the granulite facies record temperaturesgreater than 1,000°C at pressures of 0.5-1.2GPa [ultrahigh-temperature metamorphism(UHTM); e.g. Harley, 1998; Brown, 2007a],whereas some eclogite facies rocks recordpressures greater than 5 GPa at temperaturesof 600-1,000°C, and in one case mineralogicalevidence suggests pressures of at least 10 GPa[ultrahigh-pressure metamorphism (UHPM);e.g. Chopin, 2003; Liu et al., 2007; Brown,2007a]. The transition between these two isreferred to as medium-temperature eclogite -high-pressure granulite metamorphism (E-HPGM; e.g. O'Brien and Rötzler, 2003;Brown, 2007a). On modern Earth, the differentmetamorphic facies series leading to granulite- ultrahigh temperature metamorphism (G-UHTM), medium temperature eclogite - highpressure granulite metamorphism (E-HPGM)and high pressure metamorphism - ultrahighpressure metamorphism (HPM-UHPM) aregenerated in different tectonic settings withcontrasting thermal regimes at convergentplate boundary zones (Brown, 2006, 2007a).

Recently, Stüwe (2007) has introducedan alternative division of P-T space based onwhether thermal conditions implied by thepeak metamorphic mineral assemblages inorogenic crust were warmer or cooler than anormal (conductive) continental geotherm(Fig. 1b). On this diagram we may distinguishP-T fields that are reached as a function ofdifferent tectonic processes. For conditionswarmer than a normal continental geotherm,a thermal gradient of approximately1,000ºC/GPa is the practical limit for a conductiveresponse (based on the extreme condition ofthe Moho resting directly on theasthenosphere). Metamorphic belts that recordapparent thermal conditions hotter than thislimit (I use the term 'ultra-high temperature'for this P-T field) require a component ofintracrustal advection-driven heating, perhapsin an arc or due to thinning the sub-crustallithospheric mantle (e.g. Sandiford and Powell,1991), consistent with the subduction zonebackarc model discussed above (Hyndman etal., 2005a, b). For thermal conditions coolerthan a normal continental geotherm, Stüwe(2007) suggests a two-fold division into a'cooler than normal' (note, Stüwe used"colder") and an 'ultra-low temperature' P-Tfield. The cooler than normal P-T field is athermal regime that may be reached by wholelithospheric thickening to the potential energylimit for equilibrium with normal plate tectonicdriving forces (about double normal thicknesscrust or half the normal thermal gradient), P-T conditions that do not necessarily reflectthermal regimes attendant with subduction orthat require unusually rapid exhumation (cf.Sandiford and Dymoke, 1991). In contrast, theultra-low temperature P-T field may only bereached by processes other than normal crustalthickening; subduction is one possible process


Fig. 1. On the left is a P-T diagram to show the principal metamorphic facies and the P-T ranges of different types of metamorphism.HP-UHP metamorphism includes the following - BS = blueschist, AEE = amphibole-epidote eclogite facies, ALE =amphibole lawsonite eclogite facies, LE = lawsonite eclogite facies, AE = amphibole eclogite facies, UHPM = ultrahigh-pressure metamorphism; GS = greenschist facies and A = amphibolite facies; E-HPG = medium temperature eclogite -high pressure granulite metamorphism; and, G = granulite facies, whereas UHTM = the ultrahigh-temperature metamorphicpart of the granulite facies. On the right is an alternative division of P-T space based on whether thermal conditionsimplied by peak metamorphic mineral assemblages were warmer or cooler than a normal (conductive) continental geotherm(cf. Stüwe, 2007), which distinguishes P-T fields that are reached as a function of different tectonic processes.

by which rocks may enter this thermal regime(Stüwe, 2007).

4. Earth in the Archean and ProterozoicEons4.1. Metamorphic regimes through Earthhistory

Metamorphic belts are classified intothree types according to the characteristicmetamorphic facies series registered by thebelt, as follows (Fig. 2a; data from Brown,2007a): HPM-UHPM, characterized bylawsonite blueschist to lawsonite eclogitefacies series rocks and blueschist to eclogiteto ultrahigh-pressure facies series rocks, whereT plotted in Fig. 2a is that registered atmaximum P; E-HPGM, characterized by faciesseries that reach peak P-T in the high-pressuregranulite facies (O'Brien and Rötzler, 2003),where maximum P and T generally are

achieved approximately contemporaneously(Fig. 2a); and, G-UHTM, characterized bygranulite facies series rocks that may reachultrahigh-temperature metamorphicconditions, where P plotted in Fig. 2a is thatregistered at maximum T.

The P-T value for each terrane in Fig.2 records a point on a metamorphic (transient)geotherm, and different apparent thermalgradients are implied by each type ofmetamorphism. These apparent thermalgradients are inferred to reflect differenttectonic settings. HPM-UHPM ischaracterized by apparent thermal gradients of150-350°C/GPa (approximately equivalent to4-10°C/km), and plots across the boundarybetween the "cooler than normal" and "ultra-low temperature" fields. About half of theseterranes require a process other than simplethickening to achieve such cold gradients. We

M. Brown 9

Fig. 2. On the left representative 'peak' metamorphic P-T conditions of metamorphic belts are shown in relation to the metamorphicfacies. Common granulite and granulite facies ultrahigh temperature metamorphism – G-UHTM belts (P at maximum T;light circles are common granulite facies belts whereas dark circles are granulite facies ultrahigh temperature metamorphicbelts - data from Tables 1 and 2 of Brown, 2007a); medium temperature eclogites - high-pressure granulites - E-HPGMbelts (peak P-T; diamonds - data from Table 3 of Brown, 2007a); lawsonite blueschists - lawsonite eclogites and ultrahighpressure metamorphic rocks - HPM-UHPM belts (T at maximum P; light squares are lawsonite-bearing rocks whereasdark squares are ultrahigh pressure belts - data from Tables 4 and 5 of Brown, 2007a). On the right peak P-T values areshown for each terrane in relation to a normal (conductive) continental geotherm. Each of these data records a point ona metamorphic (transient) geotherm, and the different apparent thermal gradients implied by each type of metamorphismare inferred to reflect different tectonic settings and thermal regimes.

know from the global context that all of theseterranes were associated with subduction, soit is likely that subduction was the process thatcreated the ultra-low temperatureenvironment. E-HPGM is characterized byapparent thermal gradients of 350-750°C/GPa(10-20°C/km), and plots across the normalcontinental geotherm but mostly in the fieldwhere heating is a conductive response tothickening. G-UHTM is characterized byapparent thermal gradients >>750°C/GPa(>>20°C/km), and mostly plot across theboundary into the field which requires anadvective component of heating. Figure 3illustrates apparent thermal gradient, which isinferred to relate to tectonic setting, plottedagainst age of peak metamorphism for each

belt. Each type of metamorphism has a distinctrange of apparent thermal gradient, as expectedfrom Fig. 2, and UHPM is restricted to thelate Neoproterozoic and Phanerozoic.However, what is now clear is the dual natureof the thermal regimes represented in themetamorphic record since the Neoarchean.The Neoarchean to Neoproterozoic interval ischaracterized by G-UHTM and E-HPGM,whereas the late Neoproterozoic and thePhanerozoic are characterized by HPM-UHPM and E-HPGM, with only sporadicexamples of G-UHTM post-Cambrian.

Although patterns based on firstoccurrences may be challenged by newdiscoveries, analysis of the data in Fig. 3provides a set of compelling first-order


observations from which to argue that themodern era of ultra-low temperaturesubduction began in the Neoproterozoic, asregistered by the occurrence of HPM-UHPM,but that ultra-low temperature subductionalone is not the hallmark of plate tectonics. Incontrast, G-UHTM and E-HPGM are presentin the exposed rock record back to at least theNeoarchean, registering a duality of thermal

regimes, which has been argued to representthe hallmark of one-sided subuction and platetectonics (Brown, 2006, 2008). Based on thisobservation, plate tectonics processes likelywere operating in the Neoarchean as recordedby the imprints of dual types of metamorphismin the rock record, and this may manifest thefirst record of global one-sided subduction onEarth.

Fig. 3. Apparent thermal gradient in ºC/GPa plotted against age of peak metamorphism in Ma (data from Tables 1-5 of Brown,2007a) for the three main types of metamorphic belt - G-UHTM (circles), E-HPGM (diamonds) and HPM-UHPM(squares) for two time intervals, a. Phanerozoic and Neoproterozoic, and b. Mesoproterozoic to Neoarchean.

M. Brown 11

The distribution of these types ofmetamorphism throughout Earth history isdisplayed in Fig. 4, together with intervals ofsupercontinent amalgamation. Changes in themetamorphic record broadly coincide with thetransitions from the Archean to Proterozoicand Proterozoic to Phanerozoic Eons, andimply a different style of tectonics in theArchean in comparison with the Proterozoicand in the Proterozoic in comparison with thePhanerozoic. Overall, the restricted time spanof different types of metamorphism throughEarth history and the periods of metamorphicquiescence during the Proterozoic Eon suggesta link with the supercontinent cycle and major

events in the mantle.4.1.1. Caveats

There are several caveats about possiblebias in this record. It is commonly argued thatgoing back through time increases loss ofinformation by erosion of the older record.However, the data in Figs. 3 and 4 plot inparticular periods and there is a cleardistinction between the pre-Neoproterozoic,where UHPM does not occur, andNeoproterozoic and younger belts, whereUHPM is common. These observations areinconsistent with a progressively degradedrecord with increasing age. Extrapolation backin time also raises questions about partial-to-

Fig 4. Plot of apparent thermal gradient in ºC/GPa versus age of peak metamorphism in Ma for the three main types ofmetamorphic belt – G-UHTM (circles), E-HPGM (diamonds) and HPM-UHPM (squares); an approximate conversion toºC/km is also shown (data from Tables 1-5 of Brown, 2007a). Also shown are the timing of supercraton amalgamation(Vaalbara, Superia and Sclavia) and supercontinent amalgamation (Nuna (Columbia), Rodinia, Gondwana as a step toPangea), and the start of the Proterozoic plate tectonics and modern plate tectonics regimes in relation to thermal gradientsfor the three main types of metamorphic belt.


complete overprinting by younger events,which is a concern in any metamorphic study.However, our ability to recognize the effectsof overprinting and to 'see through' them hasimproved significantly, and overprinting hasbeen avoided in compiling the dataset used forthis analysis (Brown, 2007a). Finally, it islikely that some Mesozoic-Cenozoic G-UHTM rocks have not yet been exposed atEarth's surface, leading to bias in the youngerpart of the record.

4.1.2. Paired metamorphic belts revisitedOrogens may be composed of belts with

contrasting types of metamorphism that recorddifferent apparent thermal gradients. Classicpaired metamorphic belts are composed of aninboard HT-LP belt, commonly withpenecontemporaneous granites, juxtaposedalong a tectonic contact against an outboardHP-LT belt; they are found in accretionaryorogens of the circum-Pacific (Miyashiro,1961).

In accretionary orogens, HP-LTblueschists and low-temperature eclogites aregenerated in the subduction zone, and amajority of lawsonite eclogites are associatedwith the HP-LT terranes of the circum-Pacificaccretionary orogens (Tsujimori et al., 2006).Complementary HT-LP metamorphism may begenerated in an arc-backarc or orogenichinterland. Events that may occur at the trenchinclude ridge subduction, entry into thesubduction channel of an ocean floortopographic high and evolution of platekinematics leading to change from a retreatinghinge to an advancing arc. These events mayaffect the thermal structure of the over-ridingplate, may lead to a change in the displacementvector across the axis of plate divergence oracross an associated transform as a ridge is

subducted and may change the tectonics fromextension to shortening (e.g. Brown, 1998b,2002; Collins, 2002). HP-LT blueschists andlow-temperature eclogites represented eductedand accreted materials that have beentranslated along a convergent margin as aforearc terrane – due to changes in platekinematics – to juxtapose metamorphic beltsof contrasting type during a single orogeniccycle along a common convergent margin (e.g.Brown, 1998a, b, 2002).

In Miyashiro's original classification oftypes of metamorphism (Miyashiro, 1961), anintermediate P/T type of metamorphism wasincluded for unpaired belts such as those inthe Scottish Highlands and the NorthernAppalachians, although in both cases themedium P/T metamorphic belt (Barroviantype) is juxtaposed against an intermediatelower P/T metamorphic belt (Buchan type),which is also the case in the eastern Himalayain Nepal. Miyashiro (1973) subsequentlysuggested that ". . . paired and unpaired (single)metamorphic belts form by the samemechanism, and an unpaired belt representspaired belts in which the contrast between thetwo belts is obscure, or in which one of thetwo belts is undeveloped or lost."

I suggest we may extend the concept of'paired metamorphic belts' more widely thanaccretionary orogens, outside the originalusage by Miyashiro (1961), to subduction-to-collision orogenic systems. The modern platetectonics regime is characterized by a dualityof thermal environments in which twoprincipal types of regional-scale metamorphicbelts are being formed contemporaneously.Brown (2006, 2008) considers this duality tobe the hallmark of one-sided subduction andthe characteristic imprint of plate tectonics inthe ancient rock record to be the broadly

M. Brown 13

contemporaneous occurrence of twocontrasting types of metamorphism reflectingthe duality of thermal environments. On thisbasis, I propose the following:Penecontemporaneous belts of contrastingtype of metamorphism that record differentapparent thermal gradients, one warmer andthe other colder, commonly juxtaposed byplate tectonics processes, may be called 'pairedmetamorphic belts'.

Thus, the combination ofpenecontemporaneous G-UHTM with E-HPGM in adjacent terranes may be describedas a paired metamorphic belts. This extendsthe original concept of Miyashiro (1961)beyond the simple pairing of HT-LP and HP-LT metamorphic belts in circum-Pacificaccretionary orogenic systems, and makes itmore useful in the context of the relationshipbetween thermal regimes and tectonic setting.This is useful in subduction-to-collisionorogenic systems, where an accretionary phaseis overprinted by a collision phase that will beregistered in the rock record by the imprint ofpenecontemporaneous higher-P/lower-Tmetamorphism at the suture and higher-T/lower-P metamorphism in the arc-backarc ororogenic hinterland (Brown, 2006).

4.2. The Archean-to-Proterozoic transitionGranulite facies ultrahigh temperature

metamorphism (G-UHTM) is dominantly aNeoarchean-Cambrian phenomenondocumented during four periods —Neoarchean, Orosirian, Ectasian-Stenian andEdiacaran-Lower Cambrian — synchronouswith formation of supercratons/supercontinents; it may also be inferred atdepth in Mesozoic–Cenozoic orogenicsystems (Brown, 2007a). The first occurrenceof G-UHTM signifies a change in

geodynamics that generated sites of very highheat flow, perhaps analogous to modern arcsand subduction zone backarcs or orogenichinterlands. In the Neoarchean, G-UHTMoccured in the Kaapvaal Craton and theSouthern Marginal Zone of the Limpopo Belt,southern Africa, and in the Lewision Complexof the Assynt terrane, Scotland (the Badcallianevent), at ca. 2.72-2.69 Ga, and within theNapier Complex, East Antarctica and theAndriamena Unit, Madagascar, at ca. 2.56-2.46 Ga (references in Brown, 2007a, 2008).

Although rare in the Neoarchean-to-Paleoproterozoic transition, medium-temperature eclogite - high-pressure granulitemetamorphism (E-HPGM) also is firstrecognized in the Neoarchean and occurs atintervals throughout the Proterozoic andPaleozoic rock record (Brown, 2007a). E-HPGM belts complement G-UHTM belts, andare generally inferred to record subduction-to-collision orogenesis. The oldest occurrenceis represented by eclogite blocks withinmélange in the Gridino Zone of the EasternDomain of the Belomorian Province, wherethe eclogite facies metamorphism appears tohave been reliably dated at ca. 2.72 Ga andthe P-T data of 1.40-1.75 GPa and 740-865ºCare well-characterized. This occurrence is oneof the earliest records of E-HPGM within asuture zone and is critical in evaluating thestart of plate tectonics; the fact that theseNeoarchean ages come from blocks in amélange cannot be avoided.

The occurrence of both G-UHTM andE-HPGM belts since the Neoarchean manifeststhe onset of a 'Proterozoic plate tectonicsregime,' which may have begun locally duringthe Mesoarchean to Neoarchean and may onlyhave become global during the Neoarchean-to-Paleoproterozoic transition (Brown, 2007b,


2008). This premise is consistent withaggregation of continental crust intoprogressively larger units to form supercratons,perhaps indicating a change in the pattern ofmantle convection during the transition to theProterozoic Eon.

The emergence of one-sided subductionrequires lithosphere with sufficient strengththat the over-riding plate can sustain theassociated bending stresses and the presenceof weak hydrated rocks above the subductioninterface (Gerya et al., 2008). Theserequirements were met as basalt became ableto transform to eclogite during subductionreleasing water to the overlying mantle rocks.Secular change in thermal regime to allow thistransformation appears to have been gradual,occurring regionally first during theMesoarchean to Neoarchean — leading to thesuccessive formation of the supercratonsVaalbara, Superia and Sclavia – andworldwide during the Paleoproterozoic –evidenced in belts that suture Nuna (Columbia)– unless this distribution is an artifact of (lackof) preservation or thorough overprinting ofeclogite in the exposed Archean cratons.

The transition to a Proterozoic platetectonics regime resulted in stabilizedlithosphere in which cratons form the cores ofcontinents that grew dominantly by marginalaccretion. It coincided with the first occurrenceof ophiolite-like complexes in Proterozoicsutures and the increase in d18O of magmasthrough the Paleoproterozoic, which mayreflect maturation of the crust, the beginningof recycling of supracrustal rocks and theirincreasing involvement in magma genesis viasubduction (references in Brown, 2007b,2008). Although the style of Proterozoicsubduction remains cryptic, the change intectonic regime whereby interactions between

discrete lithospheric plates generated tectonicsettings with contrasting thermal regimes wasa landmark event in Earth history (Brown,2006, 2008). The absence of HPM-UHPMterranes before the Ediacaran may relate to thesubducting lithosphere. For example, thelithosphere might have been strong enough toallow shallow subduction involvingtransformation of basalt to eclogite andhydration of the overlying mantle but may nothave been strong enough to allow deepsubduction or to provide a mechanism foreduction of continental crust if, indeed, it wasever subductable before the Ediacaran.

5. Ancient Earth

5.1. Pre-Neoarchean metamorphismThe Eoarchean-Mesoarchean rock

record records P-T conditions characteristicof low-to-moderate-P – moderate-to-high-Tmetamorphic facies. In greenstone beltsmetamorphic grade varies from prehnite-pumpellyite facies through greenschist faciesto amphibolite facies and, rarely, into thegranulite facies; ocean floor metamorphism ofthe protoliths is common. In high-gradeterranes, ordinary granulite faciesmetamorphism and multiple episodes ofanatexis are the norm.

In southern West Greenland, in the IsuaSupracrustal Belt (ISB), the metamorphism ispolyphase — Eoarchean and Neoarchean —and an age of ca. 3.7 Ga has been argued for

the early metamorphism. P-T conditions of0.5-0.7 GPa and 500-550°C (or up to 600°C)have been retrieved from the ISB (referencesin Brown, 2008). These data yield warmapparent thermal gradients of 800-1,000°C/GPa (~23-28°C/km), which is just within therange for a purely conductive response to

M. Brown 15

thickening, but may reflect thinner lithospherewith higher heat flow than later in Earthhistory. Also in southern West Greenland, inmafic rocks on the eastern side of InnersuartuutIsland in the Itsaq Gneiss Complex of theFæringehavn terrane, where the gneissesrecord several events in the interval 3.67-3.50Ga, early orthopyroxene + plagioclaseassemblages record poorly defined but ratherordinary granulite facies conditions whereasoverprinting garnet + clinopyroxene + quartzassemblages probably record the widespreadNeoarchean high-pressure granulite faciesmetamorphism associated with final terraneassembly in the interval 2.715-2.650 Ga(Nutman and Friend, 2007). This is consistentwith plate tectonics processes operating onEarth by the Neoarchean (Brown, 2006).

It has been suggested that some ArcheanTTG suites and eclogite xenoliths scavengedby kimberlites from deeper levels in Archeancratons are respectively melt andcomplementary residue from a basaltic source,such as the low MgO eclogites from the Koidukimberlite within the TTG crust of the ManShield in West Africa, where imprecisePaleoarchean formation ages for the eclogitesand TTG crust overlap and are permissive forcrustal growth by partial melting of theprotolith of the eclogite xenoliths (Barth et al.,2002). Values of d18O for the eclogites lieoutside the mantle range, suggesting theprotolith may represent ocean floor, which inturn permits Archean crustal growth in the ManShield to have occurred in a convergent marginsetting involving subduction of oceanlithosphere. High MgO eclogites in the ManShield and elsewhere are inferred tocorrespond to metamorphosed cumulates fromhydrous basalt magmas emplaced beneaththick continental arcs, supporting a role for

arc magmatism in the formation of cratonicroots in the Mesoarchean to Neoarchean.

In South Africa, recent work on the ca.3.23 Ga metamorphism of the Barberton andrelated greenstone belts has yielded P-T datafrom the Onverwacht Group greenstoneremnants of 0.8-1.1 GPa and 650-700°C, fromthe southern Barberton greenstone belt of 0.9-1.2 GPa and 650-700°C, and fromamphibolite-dominated blocks of supracrustalrocks in tectonic mélange from the InyoniShear Zone of 1.2-1.5 GPa and 600-650°C(references in Brown, 2008). Apparent thermalgradients are in the range 450-700°C/GPa(~13-20°C/km), within the limit for aconductive response to thickening.Consequently, a conclusion that these HP-LTconditions represent metamorphic evidence forsubduction-driven tectonic processes duringthe evolution of the early Earth may bepremature. The lowest of these gradients isclose to the highest gradient retrieved fromPhanerozoic HPM-UHPM terranes inferred tohave been associated with subduction (Brown,2007a). Thus, although the Inyoni Shear Zonemay represent a suture recording a site offormer subduction, the apparent thermalgradients derived from the P-T conditionsretrieved from the high pressure rocks do notprovide unambiguous evidence of subduction.

5.2. Pre-Neoarchean tectonicsBlueschists are not documented in the

Archean Eon and there is no metamorphicimprint of subduction of continental crust tomantle conditions and return to crustal depths(although this could be interpreted to mean thatthe subducted continental crust was notreturned). However, the chemistry ofMesoarchean-to-Neoarchean eclogitexenoliths and the chemistry and paragenesis


of Neoarchean diamonds in kimberlites withincratons suggest some process associated withsupercraton formation at convergent marginswas operating to take basalt and othersupracrustal material into the mantle by theNeoarchean Era (references in Brown, 2008).This may have been some form of subduction,but the hotter ocean lithosphere may have beenweaker due to lower viscosity leading to morefrequent slab breakoff and slab breakup, andpreventing exhumation.

Tonalite-trondhjemite-granite suites(TTGs) dominate the Archean rock record, buttheir origin has been controversial. One recentview suggests that many older TTGs formedby melting of garnet amphibolite of broadlybasaltic composition hydrated by interactionwith sea water, whereas melting of eclogiteincreased in importance through theMesoarchean to Neoarchean, as shown by anincrease in Nb/Ta in TTGs (Foley et al., 2002,2003). Melting of garnet amphibolite may beachieved in the lower part of thickened basalticcrust or by subduction on a warmer Earth.Conceptually, this is consistent with earlyplate-like behavior that allowed crust to floatto form thick stacks above zones of boundary-layer downwelling, or, for a slightly coolerEarth, thick stacks above zones of 'sub-lithospheric' subduction, and this indeed mayhave been the convective mode of theEoarchean to Paleoarchean. By theMesoarchean to Neoarchean, and possibly asearly as the Paleoarchean, subduction wasoperating in some regions of the Earth; thissubduction was likely characterized by warmergeotherms that enabled melting of subductinggarnet amphibolite and, with secular cooling,a change to melting of subducting eclogite.

Worldwide, the oldest survivingPaleoarchean crustal remnants generally are

composed of juvenile TTGs formed prior to aperiod of polyphase granulite faciesmetamorphism in the interval 3.65-3.60 Ga,and extreme thermal conditions typical of G-UHTM are not generally registered before theNeoarchean. This appears to becounterintuitive, since we might expect thehigher abundance of heat-producing elementsmight have led to higher crustal heatproduction and hotter orogens. However, it ispossible that contemporary heat loss throughoceans and continents was higher andlithosphere rheology was generally weaker onearly Earth. This is permissive for a 'crèmebrûlée' lithosphere rheology structure; also, itis consistent with: (1) modeling suggestingdominance of unstable subduction for platecollision regimes with very hot geotherms inwhich convergence is accommodated by pureshear thickening and development ofgravitational (Rayleigh-Taylor) instabilities;(2) modeling that rules out the commonly-proposed flat subduction model for early Earthtectonics; (3) the proposition that early plate-like behavior may have allowed crust to floatto form thick stacks above zones of boundary-layer downwelling, or, for a slightly coolerEarth, thick stacks above zones of 'sub-lithospheric' subduction; and, (4) the absenceof E-HPGM and HPM-UHPM in the earlyEarth rock record, inferred herein to by moretypical of plate collisions in the Proterozoicand Phanerozoic (references in Brown, 2008).

One alternative tectonic model is thatof continental overflow (Bailey, 1999),whereby continental-style crust would over-ride oceanic-style lithosphere to facilitatepartial melting of the underlying plate andformation of TTGs. Such a process may havecreated an environment favorable forsubduction and may have been the initial step

M. Brown 17

required to initiate subduction and enable platetectonics. Continental overflow might lead toshallow subduction, but this requires that themantle viscosity was large enough to resistrapid sinking of the oceanic-type lithosphereplate, which may not have been the case forthe Archean mantle.

Assuming a hotter early Earth,numerical modeling that takes into accountearly depletion of the upper mantle suggeststhis feature might have limited the ability ofthe ocean lithosphere to subduct and mighthave impacted the viability of early platetectonics (Davis, 2006, 2007); this modelingalso suggests spatial and temporal variabilityin crustal thicknesses that might have allowedsome plates to subduct whereas other platesmight have been blocked to formprotocontinental crust. In an alternativescenario without early depletion of the uppermantle, numerical modeling suggests that thelower viscosity and higher degree of meltingfor a hotter fertile mantle might have led toboth a thicker crust and a thicker depletedharzburgite layer forming the oceanlithosphere (van Hunen and van den Berg,2008). Compositional buoyancy resulting fromthe thicker crust and harzburgite layers mightbe a serious limitation for subductioninitiation, and a different mode of downwellingor 'sub-lithospheric' subduction might havecharacterized early Earth. An additionalproblem for deep subduction is posed by thelower viscosity, which might lead to morefrequent slab breakoff and possible breakupof the crust and harzburgite layers of thesubducting ocean lithosphere. In this case thelower viscosity of the slab might have beenthe principal limitation inhibiting platetectonics on early Earth (van Hunen and vanden Berg, 2008; Gerya et al., 2008).

For thinner Archean ocean lithosphere,modeling also suggests that plate motions inthe Archean might have been faster than onmodern Earth, which may explain the scarcityof accretionary prisms in the Archean rockrecord since high convergence rates will favorsubduction erosion over accretion. Anotherconsequence of a hotter Earth is the reducedsubsidence of lithosphere in extension, whichwill limit the accommodation space forpassive-margin sediments and contribute to thescarcity of passive-margin sequences in theArchean. Formation of diamonds in theMesoarchean to Neoarchean requiresgeotherms similar modern Earth, which in turnprobably reflects the presence of cool mantleroots beneath continents. Stretching continentsunderlain by cool mantle roots may yieldpassive margins similar to those on modernEarth. Thus, the development of recognizablepassive margins may have occurred only afterthe stabilization of cratons (Bradley, 2007;Hynes, 2008).

6. Concluding Discussion

Stabilization of cratonic lithosphere andaggregation into supercratons occurred duringthe Paleoarchean-Neoarchean, probably bymobile-lid convection, although apart fromgeochemical arguments, the evidence is sparseand resulting belts are smaller-scale thanmodern orogenic systems. Metamorphismassociated with orogenesis provides a mineralrecord that may be inverted to yield ambientapparent thermal gradients. On modern Earth,tectonic settings with lower thermal gradientsare characteristic of subduction zones whereasthose with higher thermal gradients arecharacteristic of backarcs and orogenichinterlands. If a duality of thermal regimes is


the hallmark of one-sided subduction and platetectonics, then a duality of metamorphic beltsis the characteristic imprint of one-sidedsubduction and plate tectonics in the record.Ideally, these belts should be 'paired' broadlyin time and space, although the spatialarrangement may be due to late orogenicjuxtaposition. Apparent thermal gradientsderived from inversion of age-constrainedmetamorphic P-T data can be used to identifytectonic settings of ancient metamorphism andevaluate Precambrian geodynamic regimes.

The Neoarchean records the firstoccurrence of granulite facies ultra-hightemperature metamorphism (G-UHTM) andmedium-temperature eclogite-high-pressuregranulite metamorphism (E-HPGM),signifying changes in global geodynamics toallow sites of higher and lower heat flow thanthat required by apparent thermal gradientsrecovered from the earlier record. Theoccurrence of G-UHTM and E-HPGM beltssince the Neoarchean signifies change toglobally 'subductable' oceanic lithosphere,possibly beginning as early as thePaleoarchean, and transfer of water into uppermantle. The first appearance of dualmetamorphic belts also registers operation ofa global 'Proterozoic plate tectonics regime',which evolved during a Neoproterozoictransition to the 'modern plate tectonics regime'characterized by subduction of continentalcrust deep into the mantle and its (partial)return from depths of up to 350 km. ThisNeoproterozoic transition has implications fortransfer of water deeper into mantle. The agedistribution of metamorphic belts is notuniform, occurring in periods that correspondto amalgamation of continental lithosphereinto supercratons (Vaalbara/Superia/Sclavia)or supercontinents (Nuna (Columbia),

Rodinia, and Gondwana as a step to Pangea).The rock record indicates a dramatic

change in the thermal environments of crustalmetamorphism through the Neoarchean to theArchean-to-Proterozoic transition, which mayregister the switch to a 'Proterozoic platetectonics regime'. This premise is consistentwith aggregation of the crust into progressivelylarger units to form supercratons (Bleeker,2003), perhaps indicating a change in thepattern of mantle convection during thetransition to the Proterozoic Eon. Theemergence of plate tectonics requires forcessufficient to initiate and drive subduction, andone-sided subduction requires lithosphere withsufficient strength to subduct coherently andhydration of the overlying mantle rocks (Geryaet al., 2008). These requirements likely weremet as basalt became able to transform toeclogite during subduction. Secular change inthermal regime to allow this transformationappears to have been gradual, occurringregionally first during the Mesoarchean toNeoarchean – leading to the successiveformation of the supercratons Vaalbara,Superia and Sclavia – and worldwide duringthe Paleoproterozoic – evidenced in orogenicbelts that suture Nuna (Columbia) – unless thisdistribution is an artifact of preservation orthorough overprinting of eclogite in theexposed Archean cratons.

The transition to a 'Proterozoic platetectonics regime' resulted in stabilizedlithosphere in which the cratonic cores ofcontinents grew by marginal accretion.Furthermore, the transition coincides with thefirst occurrence of ophiolite-like complexesin Proterozoic suture zones (Moores, 2002)and with the increase in d18O of magmasthrough the Paleoproterozoic, which Valley etal. (2005) argue may reflect maturation of the

M. Brown 19

crust, the beginning of recycling ofsupracrustal rocks and their increasinginvolvement in magma genesis via subduction.Although the style of Proterozoic subductionremains cryptic, the change in tectonic regimewhereby interactions between discretelithospheric plates generated tectonic settingswith contrasting thermal regimes was alandmark event in Earth history. The transitionto the 'modern plate tectonics regime' throughthe Neoproterozoic is one in which ocean crustbecame thinner whereas the lithospherebecame thicker (e.g. Moores, 2002), and the

References

Bailey, R.C. 1999. Gravity-driven continental overflowand Archaean tectonics. Nature 398, 413-415.

Barth, M.G., Rudnick, R.L., Carlson, R.L., Horn, I. andMcDonough, W.F. 2002. Re-Os and U-Pbgeochronological constraints on the eclogite-tonaliteconnection in the Archean Man Shield, West Africa.Precambrian Research 118, 267-283.

Beaumont, C., Nguyen, M.H., Jamieson, R.A. and Ellis,S. 2006. Crustal flow modes in large hot orogens, inLaw, R.D., Searle, M.P., and Godin, L., eds., Channelflow, ductile extrusion and exhumation in continentalcollision zones. Geological Society SpecialPublication 268, 91-145.

Bleeker, W. 2003. The late Archean record: a puzzle inca. 35 pieces. Lithos 71, 99-134.

Bradley, D.C. 2007. Age distribution of passive marginsthrough Earth history and tectonic implications. Eos(Transactions, American Geophysical Union) 88Joint Assembly Supplement, abstract U44A-03.

Brown, M. 1993. P-T-t evolution of orogenic belts andthe causes of regional metamorphism. Journal of theGeological Society, London 150, 227-241.

Brown, M. 1998a. Unpairing metamorphic belts: P-Tpaths and a tectonic model for the Ryoke Belt,southwest Japan. Journal of Metamorphic Geology16, 3-22.

Brown, M. 1998b. Ridge-trench interactions and high-

T-low-P metamorphism, with particular reference tothe Cretaceous evolution of the Japanese Islands, inTreloar, P.J., and O'Brien, P.J., eds., What drivesmetamorphism and metamorphic reactions?Geological Society Special Publication 138, 131-163.

Brown, M. 2001. From microscope to mountain belt:150 years of petrology and its contribution tounderstanding geodynamics, particularly thetectonics of orogens. Journal of Geodynamics 32,115-164.

Brown, M. 2002. Plate margin processes and 'paired'metamorphic belts in Japan: Comment. Earth andPlanetary Science Letters 199, 483-492.

Brown, M. 2006. A duality of thermal regimes is thedistinctive characteristic of plate tectonics since theNeoarchean. Geology 34, 961-964.

Brown, M. 2007a. Metamorphic conditions in orogenicbelts: A record of secular change. InternationalGeology Review 49, 193-234.

Brown, M. 2007b. Metamorphism, plate tectonics, andthe supercontinent cycle. Earth Science Frontiers 14,1-18.

Brown, M. 2008. Characteristic thermal regimes of platetectonics and their metamorphic imprint throughoutEarth history: when did Earth first adopt a platetectonics mode of behavior? In: Condie, K., Pease,V. (Eds.), When did Plate Tectonics Begin?Geological Society of America Special Paper 440,

pattern of Brasiliano-Pan-African orogens thatsutured the disparate continental fragments ofGondwana was replaced by semicircum-ferential large-scale plate-margin orogenicsystems (e.g. the Terra Australis, theAppalachian/Caledonian-Variscide-Altaid andthe Cimmerian-Himalayan-Alpine orogenicsystems).

AcknowledgementsI thank Chris Clark and Dave Kelsey

for perceptive review comments that enabledme to improve this contribution.


97-128.Cawood, P.A., Kröner, A., and Pisarevsky, S. 2006.

Precambrian plate tectonics: Criteria and evidence.GSA Today 16, 4-11.

Chopin, C. 2003. Ultrahigh-pressure metamorphism:tracing continental crust into the mantle. Earth andPlanetary Science Letters 212, 1-14.

Cloos, M. 1993. Lithospheric buoyancy and collisionalorogenesis: Subduction of oceanic plateaus,continental margins, island arcs, spreading ridges,and seamounts. Geological Society of AmericaBulletin 105, 715-737.

Collins, W.J. 2002. Nature of extensional accretionaryorogens. Tectonics 21, doi: 10.1029/2000TC001272.

Collins W.J. 2003. Slab pull, mantle convection, andPangaean assembly and dispersal. Earth andPlanetary Science Letters 205, 225-237.

Coney, P.J., Jones, D.L., and Monger, J.W.H. 1980.Cordilleran suspect terranes. Nature 28, 329-333.

Davies, G.F. 1992. On the emergence of plate tectonics.Geology 20, 963-966.

Davies, G.F. 2006. Gravitational depletion of the earlyEarth's upper mantle and the viability of early platetectonics. Earth and Planetary Science Letters 243,376-382.

Davies G. F. 2007. Controls on density stratification inthe early mantle. Geochemistry GeophysicsGeosystems, 8, Q04006, doi:10.1029/2006GC001414.

Dewey, J.F. 1975. Finite plate implications - someimplications for evolution of rock masses at platemargins. American Journal of Science A275, 260-284.

Ernst, W.G. 2001. Subduction, ultrahigh-pressuremetamorphism, and regurgitation of buoyant crustalslices - implications for arcs and continental growth.Physics of the Earth and Planetary Interiors 127, 253-275.

Foley, S., Tiepolo, M. and Vannucci, R. 2002. Growthof early continental crust controlled by melting ofamphibolite in subduction zones. Nature 417, 837-840.

Foley, S.F., Buhre, S. and Jacob, D.E. 2003. Evolutionof the Archaean crust by delamination and shallow

subduction. Nature 421, 249-252.Gerya, T.V., Connolly, J.A.D. and Yuen, D.A. 2008. Why

is terrestrial subduction one-sided? Geology 36, 43-46.

Harley, S.L. 1998. On the occurrence and characterisationof ultrahigh-temperature (UHT) crustalmetamorphism: in Treloar, P.J., and O'Brien, P.J.,eds., What drives metamorphism and metamorphicreactions? Geological Society Special Publication138, 75-101.

Hyndman, R.D., Currie, C.A. and Mazzotti, S.P. 2005a.Subduction zone backarcs, mobile belts, andorogenic heat. GSA Today 15, 4-10.

Hyndman, R.D., Currie, C. and Mazzotti, S. 2005b. TheOrigin of Global Mountain Belts: Hot SubductionZone Backarcs: Eos, Transactions of the AmericanGeophysical Union 86, abstract T11B-0367.

Hynes, A. 2008. Effects of a warmer mantle on thecharacteristics of Archean passive margins, inCondie, K.C., and Pease, V., eds., When did platetectonics begin on planet Earth? Geological Societyof America Special Paper 440, doi: 10.1130/2008.2440(07).

Johnston, S.T. and Borel, G.D. 2007. The odyssey of theCache Creek terrane, Canadian Cordillera:Implications for accretionary orogens, tectonicsetting of Panthalassa, the Pacific superswell, andbreak-up of Pangea: Earth and Planetary ScienceLetters 253, 415-428.

Liou, J.G., Tsujimori, T., Zhang, R.Y., Katayama, I. andMaruyama, S. 2004. Global UHP metamorphism andcontinental subduction/collision: The Himalayanmodel. International Geology Review 46, 1-27.

Liu, L., Zhang, J., Green, H.W., Jin, Z. and Bozhilov,K.N. 2007. Evidence of former stishovite inmetamorphosed sediments, implying subduction to> 350 km. Earth and Planetary Science Letters 263,180-191.

Maruyama, S. 1997. Pacific-type orogeny revisited:Miyashiro-type orogeny proposed. The Island Arc6, 91-120.

Matsuda, T. and Uyeda, S., 1971, On the Pacific-typeorogeny and its model - extension of the paired beltsconcept and possible origin of marginal seas.

M. Brown 21

Tectonophysics 11, 5-27.Miyashiro, A. 1961. Evolution of metamorphic belts.

Journal of Petrology 2, 277-311.Miyashiro, A. 1972. Metamorphism and related

magmatism in plate tectonics. American Journal ofScience 272, 629-656.

Miyashiro, A. 1973. Paired and unpaired metamorphicbelts. Tectonophysics 17, 241-254.

Montelli, R., Nolet, G., Dahlen, F.A. and Masters 2006.A catalogue of deep mantle plumes. New results fromfinite-frequency tomography. Geochemistry,Geophysics and Geosystems 7, Q11007, doi:10.1029/2006GC001248.

Moores, E.M. 2002. Pre-1 Ga (pre-Rodinian) ophiolites:Their tectonic and environmental implications.Geological Society of America Bulletin 114, 80-95.

Nutman, A. and Friend, C.R.L. 2007. Adjacent terraneswith c. 2715 and 2650 Ma high-pressuremetamorphic assemblages in the Nuuk region of theNorth Atlantic Craton, southern West Greenland:Complexities of Neoarchean collisional orogeny.Precambrian Research 155, 159-203.

O'Brien, P.J. and Rötzler, J. 2003. High-pressuregranulites: formation, recovery of peak conditionsand implications for tectonics. Journal ofMetamorphic Geology 21, 3-20.

O'Neill, C., Lenardic, A., Moresi, L., Torsvik, T.H. andLee, C.-T.A. 2007. Episodic Precambriansubduction. Earth and Planetary Science Letters, 262doi: 10.1016/j.epsl.2007.04.056.

Richards, M.A. and Engebretson, D.E. 1992. Large-scalemantle convection and the history of subduction.Nature 355, 437-440.

Sandiford, M. and Powell, R. 1991. Some remarks onhigh-temperature-low-pressure metamorphism inconvergent orogens. Journal of MetamorphicGeology 9, 333-340.

Sengör, A.M.C. and Natal'in, B.A., 1996, Turkic-typeorogeny and its role in the making of the continentalcrust. Annual Review of Earth and Planetary Sciences

24, 263-337.Silver, P.G. and Behn, M.D., 2008, Intermittent Plate

Tectonics? Science 319, 85-88.Sleep, N.H., 2000, Evolution of the mode of convection

within terrestrial planets. Journal of GeophysicalResearch 105, 17,563-17,578.

Sobolev, S.V. and Babeyko, A.Y. 2005. What drivesorogeny in the Andes? Geology 33, 617-620.

Stern, R.J., 2005. Evidence from ophiolites, blueschists,and ultrahigh-pressure metamorphic terranes that themodern episode of subduction tectonics began inNeoproterozoic time. Geology 33, 557-560.

Stern, R.J. 2007. When and how did plate tectonicsbegin? Theoretical and empirical considerations.Chinese Science Bulletin 52, 578-591.

Stüwe, K. 2007. Geodynamics of the lithosphere.Springer, 504p.

Tan, E. and Gurnis, M. 2007. Compressiblethermochemical convection and application to lowermantle structures. Journal of Geophysical Research112, B06304, doi: 10.1029/ 2006JB004505.

Tsujimori, T., Sisson, V.B., Liou, J.G., Harlow, G.E. andSorensen, S.S. 2006. Very-low-temperature recordof the subduction process. A review of worldwidelawsonite eclogites. Lithos 92, 609-624.

Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C.,Spicuzza, M.J., Basei, M.A.S., Bindeman, I.N.,Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H.,Sinha, A.K. and Wei, C.S. 2005. 4.4 billion years ofcrustal maturation: oxygen isotope ratios ofmagmatic zircon. Contributions to Mineralogy andPetrology 150, 561-580, doi: 10.1007/s00410-005-0025-8.

van Hunen, J., and van den Berg, A.P. 2008. Platetectonics on the early Earth: Limitations imposed bystrength and buoyancy of subducted lithosphere.Lithos, 103 doi: 10.1016/j.lithos.2007.09.016 (inpress).

Zwart, H.J. 1967. The duality of orogenic belts. Geologieen Mijnuow 46, 283-309.

VOL. 80, Nos.1-4, 2008 (Published January 2010) ISSN 0970-1354

KOLKATA

INDIAN JOURNAL OF GEOLOGY

Special Issue

Geodynamic Regimes, Global Tectonics and Evolution of Precambrian Cratonic Basins in India

Documents

GEODYNAMIC REGIMES AND TECTONIC SETTINGS FOR …€¦ · METAMORPHISM: RELATIONSHIP TO THE SUPERCONTINENT CYCLE MICHAEL BROWN Laboratory for Crustal Petrology, Department of Geology,