Myron G. Best. Igneous and metamorphic 2003 Blackwell Science

Подождите немного. Документ загружается.

Additional factors that can be of local signiﬁcance in

the thermal history of orogens include magmatic and

strain heating, advective transfer of heat via ﬂuids, and

the heat budget of mineral reactions.

Several means are available to construct P–T paths

in a rock or suite of cogenetic rocks for the terrane

in which they were metamorphosed. Broad paths are

deﬁned by facies series and metamorphic zones.

Thermobarometry on mineral assemblages and zoned

minerals can yield quantitative data but caution must

be exercised because of disequilibrium effects resulting

from sluggish kinetics and overprinting. Further con-

straints are provided by mineral inclusions in poikilo-

blasts and other relic mineral assemblages preserved

in reaction textures; these assemblages can be placed

in petrogenetic grids derived from experimental and

thermodynamic data. High temperatures and complete

mineral equilibration at peak metamorphic conditions

can all but eliminate the prograde path record; for this

reason, retrograde paths are generally more readily

determined.

Chronologic data at speciﬁc P–T points in the meta-

morphic path permit construction of a complete P–T–t

path; such information provides quantitative estimates

of rates of mineral growth and of burial and exhuma-

tion during metamorphism.

Orogens and the metamorphic terranes in them are

inﬁnitely variable in detail. Some Mesozoic–Cenozoic

orogens of the circum-Paciﬁc have paired belts with

high P/T metamorphism developed in the accretionary

wedge and intermediate- to low-P metamorphism

farther inland coincident with the magmatic arc. Most

orogens, however, are not so simple because of multi-

ple episodes of docking and suturing of island arcs,

oceanic plateaus, ophiolite, and continental fragments

onto overriding continental plates. Some orogens ex-

perienced culminating continent–continent collisions

and suturing, only to be later affected by crustal exten-

sion and rifting.

The classic Barrovian metamorphic zones devel-

oped in pelitic rocks at intermediate pressures have

been a standard reference for regional metamorphism

in orogens since the work of George Barrow in the

Scottish Highlands a century ago. Successive prograde

appearance via continuous and discontinuous reactions

of chlorite, biotite, garnet, staurolite, kyanite, sillima-

nite, and sillimanite  K-feldspar ( migmatite) deﬁne

these Barrovian zones. Higher grade minerals and

assemblages of the granulite facies found in other

terranes around the world include cordierite, spinel,

sapphirine, and orthopyroxene  K-feldspar in lieu of

biotite.

Widespread metabasites in Barrovian terranes have

fewer phases formed by mostly continuous reactions,

hindering deﬁnition of P–T conditions from mineral

assemblages. Greenschist-facies metabasites in chlorite

and biotite pelite zones contain epidote  chlorite 

actinolite  albite  quartz  calcite  titanite 

Fe–Ti oxides. In the garnet zone, aluminous horn-

blende replaces actinolite, oligoclase replaces albite,

and almandine-rich garnet may also appear. Chlorite

and epidote disappear in the amphibolite facies (stau-

rolite, kyanite, and sillimanite zones) and pyroxenes

replace hornblende as the dominant maﬁc phase in the

granulite facies.

Intermediate-P metamorphism deﬁnes open-loop,

clockwise P–T paths in orogens where the continental

crust is thickened and thermal relaxation occurs over at

least tens of millions of years. In pelitic rocks, this path

is recorded in prograde kyanite, perhaps preserved as

inclusions in cores of porphyroblasts, followed by later

higher T sillimanite.

Lower P regional metamorphism in some orogens

resembles that developed in contact aureoles sur-

rounding magmatic intrusions except for the concur-

rent deformation that creates tectonite fabric instead

of hornfelsic. Indeed, this Buchan style is sometimes

referred to as regional contact metamorphism because

of widespread more or less synchronous granitoid

intrusions that serve as an important heat source, per-

haps with an additional heat ﬂux from underplating

mantle-derived basaltic magma. The typical ﬂat-loop,

counter-clockwise P–T paths recorded for Buchan

metamorphism reﬂect this external input of heat. The

lack of much change in P during the prograde and

retrograde paths reﬂects the lack of concurrent crustal

thickening and consequent isostatic unrooﬁng adjust-

ments. In pelitic rocks, the prograde Buchan path is

recorded by early andalusite succeeded by sillimanite;

cordierite is widespread, whereas staurolite and espe-

cially garnet are less common, or absent.

Ocean-ridge metamorphism (or alteration, as some

would prefer) is a globally signiﬁcant interaction be-

tween seawater and hot basaltic rock at spreading

oceanic rifts. Metasomatism results in mineral assem-

blages resembling those of the zeolite, prehnite–

pumpellyite, greenschist, and locally amphibolite

facies. But the reconstitution is patchy, more or less

monomineralic veins are widespread, and the mineral

zonation is telescoped within the upper few kilometers

of the seaﬂoor in pillow lavas and underlying sheeted

dike complex. Metasomatism creates metal sulﬁdes

and releases silica for seaﬂoor chert deposits.

The most intact oceanic lithosphere exposed on

land is found in a few large slabs of ophiolite, such as

the Semail in Oman. These ophiolites appear to have

been created in back-arc basins underthrust by nearby

subducting continental margins that subsequently

buoyed upward, lifting the overriding oceanic rocks

above sealevel.

Terranes of arc-trench sedimentary deposits

(graywacke), mélange, high P/T facies series rocks,

Metamorphism at Convergent Plate Margins: P–T–t Paths, Facies, and Zones

607

serpentinite, and variably dismembered ophiolite

typify accretionary wedges over subducting oceanic

lithosphere alongside oceanic trenches. Zeolite-facies

rocks characterized by any one of several zeolites coex-

isting stably with quartz are typically developed in

thick prisms of volcanogenic detritus shed off adjacent

volcanic highlands. Locally merging into zeolitic rocks

are those of the higher P and T prehnite–pumpellyite

facies that are characterized by these two phases coex-

isting stably with certain minerals of the greenschist

and blueschist facies. Blueschist-facies rocks are char-

acterized by stable glaucophane  lawsonite and

additionally at higher P by aragonite and a jadeitic

clinopyroxene; stilpnomelane occurs instead of bio-

tite, kyanite instead of andalusite or sillimanite, and

phengite-rich white mica, titanite, and garnet are com-

mon. The diagnostic assemblage in metabasites of mag-

nesian garnet  omphacite deﬁnes the eclogite facies.

High P/T metamorphism of the blueschist and

adjacent facies follows relatively tightly looped clock-

wise P–T paths; these refrigerated rock paths imply

nearly isothermal burial and subsequent immediate

exhumation. Little heating takes place because of the

relatively rapid burial and exhumation in the near-

trench, accretionary-wedge setting over cool subduct-

ing oceanic lithosphere. Consequently, no magmatism

and migmatism take place. If exhumation is not so

rapid, as commonly happens in more mature and

collisional orogens, thermal relaxation via conduc-

tion of heat can transform blueschist-facies mineral

assemblages into greenschist- or amphibolite-facies

assemblages.

Very rare, low- to moderate-T, ultrahigh-P (UHP)

rocks occur in several continent–continent collisional

zones. Although the deﬁning coesite and diamond

inclusions in garnets and other “pressure cell” minerals

are very rarely preserved, other relict UHP mineral as-

semblages that escaped lower-P or higher-T overprint-

ing metamorphism are consistent with their ultradeep

ancestry. These UHP phases and assemblages include

pyrope garnet, Si-rich phengite, ellenbergite, K–Al-rich

clinopyroxene, talc  kyanite, and grossular-rich gar-

net  rutile. Preservation of UHP rocks appears to

stem from their development in continental protoliths

that were rapidly exhumed once buoyant forces took

over in the refrigerated subducting plate.

CRITICAL THINKING QUESTIONS

18.1 Why are pelitic rocks better indicators of

metamorphic grade in Barrovian terranes than

feldspathic sandstones and granitoids?

18.2 Discuss mineralogical changes during progres-

sive Barrovian-style regional metamorphism of

pelitic rocks. Cite ﬁrst appearances of index

minerals or mineral compatibilities and changes

in composition of solid solutions.

18.3 The ﬁrst and second sillimanite isograds may

not be sharply deﬁned in pelites in Barrovian

terranes, but for two different reasons. Discuss

these.

18.4 What would be the advantages and disadvant-

ages of erecting a new UHP facies distinct

from the eclogite facies? Would this be a valid

metamorphic facies?

18.5 What is a possible pitfall in using the SHRIMP

U–Pb age of a zircon inclusion in a zoned

garnet from a garnet-zone pelitic rock to

constrain a point on its P–T–t path? How valid

is the ﬁssion-track age of an apatite inclusion

in the garnet in constraining the peak

metamorphic T ? Discuss fully.

18.6 How would you account for the high closure

T for the U–Pb isotope system in zircon?

18.7 How might the rate of plate convergence gov-

ern the development of paired metamorphic

belts?

18.8 Explore the geologic consequences of rapid

uplift and erosion on conﬁguration of the

geotherm, age of exposed metamorphic rocks,

and occurrence of hot springs. Suggest and

discuss other implications.

18.9 Heat conduction from a static magmatic body

raises the T of adjacent rocks to only about

one-half that of the body (Section 8.4.1).

How do you reconcile the 860°C recorded

in the metamorphic sole immediately below

the White Hills peridotite whose mineral

assemblage had equilibrated at 950°C?

18.10 Because of lithospheric plate reorganization,

ancient subduction zones are not marked by

active seismicity and arc volcanism. Without

these indicators of active subduction, what

rock types and associations verify the former

existence of a subduction zone and its polarity

(direction of slab dip)?

18.11 Contrast the two ways that ophiolite is

emplaced and occurs in subduction zones.

18.12 Suppose that an amphibole grain grew

progressively during metamorphism on a

clockwise P–T path under blueschist- and

amphibolite- and ﬁnally greenschist-facies

conditions. Describe its compositional zoning,

assuming it never fully equilibrated under any

facies conditions.

18.13 Compare P–T paths of Barrovian, Buchan,

blueschist-facies, and UHP metamorphism.

Explain and account for their differences.

608 Igneous and Metamorphic Petrology

PROBLEMS

18.1 Compare typical rates of heat conduction

and tectonic transport of rock mass. Using

the Jaeger approximation in Section 8.4.1

for a thermally perturbed overthrust rock

slab 20 km thick in a convergent plate

regime, calculate the time required to allow

“substantial” and “practically complete” ther-

mal effects at the center of the slab. Answer:

About 3 and 30 million years. In 3 million

years, what is the distance of transport using

a reasonable plate tectonic rate of movement?

Discuss your results.

18.2 Evaluate the contribution of strain heating by

the mechanical work of shearing during orogen-

esis using equation 11.1. Assume a 100-km-

wide orogen shortened at a rate of 1 cm/y at a

shear stress of 100 bars. Compare this rate of

heating with that due to radioactivity in granite

and, if the model crust is 30 km thick, with that

of the average surface heat ﬂow. Discuss your

results.

18.3 Construct an AKF diagram for pelitic mineral

assemblages of the Barrovian staurolite zone.

How does it compare with the AFM diagram

in terms of representing mineral reactions in

the zone?

Metamorphism at Convergent Plate Margins: P–T–t Paths, Facies, and Zones

609

UNDAMENTAL

UESTIONS

ONSIDERED IN

HIS

HAPTER

1. What was the nature of magmatic and

metamorphic rock-forming processes during the

infancy of planet Earth and how do they compare

with those of more recent geologic time? In other

words, how valid is the concept of

Uniformitarianism?

2. How far back in the geologic past were plate

tectonics and mantle plumes expressions of the

convecting-mantle heat engine driving rock-

forming processes?

3. What is the nature of the oldest preserved

remnants of sialic and maﬁc crust? How closely

do these correspond to younger (Mesozoic–

Cenozoic) continental and oceanic crust?

4. How have continents evolved throughout Earth

history?

INTRODUCTION

In this concluding chapter, we continue to pursue

a fundamental concept introduced in Chapter 1 and

examined throughout this textbook: rocks provide a

record of how the Earth works, how gravitational and

thermal energy drive movement of matter in rock-

forming processes. However, in this chapter, we focus

not on the distinct and separate records preserved in

magmatic and metamorphic rocks but rather on an in-

tegrated view of the rock-forming systems in a speciﬁc

interval of time—the Precambrian. During this ﬁrst

3.9 Gy, or 87%, of Earth history deﬁned by the lack of

conspicuous fossils, unique and unusually widespread

associations of rocks were created; these include ultra-

maﬁc komatiite lava ﬂows, megacrystic and massif-

type anorthosite, rapakivi granites, and vast terranes of

tonalite, trondhjemite, and granodiorite orthogneiss.

Do these unusual associations merely reﬂect a much

greater amount of time for their creation? Or, do they

imply something fundamentally different about rock-

forming processes in the youthful Earth? In the more

likely latter case, signiﬁcant contrasts between these

earliest rocks and “modern” rocks, formed since the

creation of the oldest preserved oceanic crust by

seaﬂoor spreading in the Jurassic, raise important

questions regarding global geologic systems in the

premodern planet. How did a presumably hotter

convecting-mantle heat engine drive tectonic pro-

cesses during the Precambrian? Were there moving

lithospheric plates as today? Were there continents

and ocean basins as during modern times?

Despite the difﬁculties in interpreting Precambrian

rocks and their formative geologic systems, trying to

understand them is a tantalizing challenge that cannot

be ignored. New discoveries, new techniques of study,

and new ways of thinking about Precambrian rocks in

the coming decades will be needed to resolve specu-

lative and conﬂicting interpretations.

The exposed volume of the Precambrian record

preserved in continents on Earth decreases expon-

entially in older rocks and its clarity is increasingly

clouded by overprinting, commonly multiple, episodes

of recrystallization, deformation, metasomatism, and

migmatization, or partial melting (see, for example,

Figure 14.39). Original ﬁeld relations are difﬁcult, if

not impossible, to interpret unambiguously in such

metamorphic rocks. It is not always possible to view

the progressive effects of a single overprint from initial

Precambrian

Rock

Associations

CHAPTER

through intermediate to the ﬁnal state of the changing

rock system, as in the examples shown in Figures

14.16, 14.17, and 14.38.

Important keys to unlocking the complexities of

Precambrian geology are precise isotope geochrono-

logy (Special Interest Box 19.1) and trace element

and isotope geochemistry. Precise dating can establish

time(s) of recrystallization, reveal the cooling history

of a rock by determining closure temperatures in dif-

ferent datable minerals (Figure 18.3), and tell when an

episode of deformation occurred by dating the retro-

grade minerals in a shear zone. Stratigraphic and struc-

tural dilemmas can be resolved, such as whether a thick

sequence of metamorphic rock layers originated by im-

bricate stacking of thrust sheets of rocks of similar age

or by sequential deposition and emplacement. Isotopic

systematics permit a distinction between older contin-

ental rock that was recycled or reworked via melting

or metamorphism versus juvenile additions to contin-

ental crust that were created by magmas extracted

from the mantle, and possibly when that extraction

occurred (“model age”). Using these laboratory tech-

niques, virtually the entire history of some rocks can

potentially be unraveled, from their time of crystalliza-

tion from magma through burial and multiple episodes

of recrystallization and deformation to ﬁnal uplift and

exhumation. But this laboratory work can be meaning-

less without careful ﬁeld observations and mapping to

establish an accurate spatial and temporal context for

rock masses. Many hours of patient ﬁeld work to ﬁnd a

critical relationship between rock units can be just as

important as the time spent in the determination of an

isotope ratio.

Because of the almost universal metamorphism that

has affected Precambrian rocks, standard geochemical

and isotopic techniques employed for fresh younger

rocks may not be applicable or valid because con-

centrations and ratios of many elements and isotopes

can be perturbed. Before using any geochemical

data, whole suites of samples should be carefully and

Precambrian Rock Associations

611

Special Interest Box 19.1 Age determinations

in old rocks

Standard methods for dating of Phanerozoic rocks

that utilize Rb–Sr (Section 2.6.2) and K–Ar isotopic

systems are generally not suitable for older Pre-

cambrian rocks because of metamorphic overprints

that alter parent-daughter ratios. The Sm–Nd sys-

tem is less susceptible to open-system alteration

and has yielded some useful ages. The U–Pb system

has proven to be the most useful on ancient rocks,

primarily using zircon (ZrSiO

) grains because they

not only contain measurable concentrations of the

parent U that substitutes for Zr but also are

relatively immune to effects of metamorphism and

melting and, in fact, can develop datable over-

growths during these processes. Other U-bearing

minerals such as titanite, baddeleyite, garnet, rutile,

and monazite have also yielded useful data. The

refractory nature of zircon is manifest in its occur-

rence as Precambrian restite xenocrysts in Ceno-

zoic silicic rocks whose magmas were generated

by partial melting of continental crust. Another

desirable attribute of zircon is that it provides two

independent chronometers (

238

U and

235

U) which

serve as an internal check on closed system beha-

vior. A drawback of zircon for dating is its absence

in maﬁc and ultramaﬁc rocks.

Many zircon grains have experienced multiple

episodes of growth during the history of their host

rock; these overgrowth zones are visible under

high-power magniﬁcation. Dating of individual

zones is made possible with a sensitive high resolu-

tion ion micro probe (SHRIMP) such as the one

employed at the Australian National University

in Canberra by Compston and colleagues (e.g.

Compston and Pidgeon, 1986) and at the Geolo-

gical Survey of Canada in Ottawa by Stern and

colleagues (e.g. Stern and Bleeker, 1998). These

workers have determined ages of the oldest known

rocks and minerals on Earth. A high energy beam

of oxygen ions is used to ionize a small hemispher-

ical volume of zircon within a growth zone on a

polished section through the interior of the grain

(Figure 19.1). The released U, Th, and Pb isotopes

are separated and ratios measured in the high resolu-

tion mass spectrometer for the age determination

(Faure, 1986).

Much less expensive but equally precise dating

as SHRIMP, as well as trace element analyses, can

be performed by laser ablation inductively coupled

plasma mass spectrometry (The Lattice, Mineralo-

gical Society of America, February 2001).

A less precise method of dating monazite grains

has been described by Williams et al. (1999). Mon-

azite (Table 2.3) is a widespread accessory min-

eral in magmatic and metamorphic rocks, contains

signiﬁcant amounts of Th and U and little non-

radiogenic Pb, and has very slow diffusion rates of

relevant atoms. Its value as a chronometer is greatly

enhanced because the analyses can be made in

polished sections with an electron microprobe—a

standard analytical tool in most geological labora-

tories. Ages can be directly correlated with fabric

and mineralogical features in the same polished

thin section to elucidate important episodes in the

tectonic, magmatic, and metamorphic history of the

rock (e.g. Section 17.3.3).

systematically scrutinized petrographically and chem-

ically. Constancy of element ratios, correspondence

between chemical variations and those predicted by

experimental data and petrographic observation,

preserved magmatic textures, and lack of second-

ary minerals are useful indications, but no proof, of

pristine condition. Generally less mobile elements

(Figure 2.22), including HFSE (Ti, Zr, Ta, Nb, Th),

REE (La to Yb), Al, and Mg, are more reliable for

petrogenetic interpretations.

Most Precambrian rocks are exposed in cratons;

these are large segments of continental lithosphere that

have remained coherent and geologically more or less

stable since their formative processes. Other more or

less synonymous terms for a craton are shield, block,

and platform. Cratons are made of one or more ter-

ranes in which the rocks have some degree of internal

petrologic and tectonic coherence. Obviously, any

relationship to then-prevailing plate boundaries and

processes, if they indeed were operative, has been

erased by younger events.

For our purposes here, the oldest rocks on Earth

are Archean. (According to the most recent geologic

time scale of Harland et al. (1990), pre-Archean, or

Priscoan, time pre-dates 4.0 Ga. Subsequent to this

publication, terrestrial rocks 4.0 Ga were dis-

covered.) Proterozoic time is 2500 to 570 Ma and

Phanerozoic time (Paleozoic, Mesozoic, and Cenozoic

eras) is 570 Ma.

The organization of this chapter follows a quasi-

chronologic order, beginning with Archean rocks

and following with Proterozoic rocks. Lastly, some

current and controversial ideas regarding the evolution

through time of the continental crust are considered.

Although magmatic rock-type names are used through-

out this chapter, it must be kept in mind that basalts,

granites, and so on are almost invariably metamorphosed

to some degree; accordingly the preﬁx “meta” is usually

implied.

19.1 THE YOUNG EARTH—A BRIEF

OVERVIEW

Accretion of the proto-Earth and other planetary

bodies in the solar system, as represented in datable

meteorites, is believed to have occurred at about

4.5 Ga (Allègre et al., 1995). Production of heat during

transformations of kinetic and gravitational potential

energy during accretion and differentiation of a denser

metallic core from a silicate mantle, supplemented by

decay of short-lived isotopes of

Fe and especially

Al, was sufﬁcient to melt at least the upper mantle

(Bickle, 1990). The apparent similarity between the

composition of the mantle of the Earth and that of the

whole Moon and its relative depletion of volatiles, as

well as other arguments, have led to the hypothesis that

the Moon was ejected from the proto-Earth after core

segregation during the impact of a Mars-sized object. If

such an impact did indeed occur, the energy input into

the proto-Earth would have been sufﬁcient to create

extensive melting and formation of a “magma ocean”

(see, for example, Tonks and Melosh, 1993). It is

further thought that the primeval crust on this magma

ocean was likely remelted and recycled by vigorous

convection and intense meteorite bombardment until

about 3.8 Ga. (Meteorite bombardment of Earth is

assumed because of the presence of widespread craters

of this age on the adjacent inner planets of the solar

system and the Moon.) This primeval heat was retained

through much of the infancy of the Earth when mantle

temperatures were estimated to be as much as 200–

300°C greater than at present (Abbott and Hoffman,

1984; Bickle, 1990). However, these estimates are

model-dependent, subject to future revision, and not

universally accepted. Campbell and Grifﬁths (1992),

for example, conclude the early mantle was no hotter

than at present.

Minute fragments of this earliest primeval crust may

be represented in detrital zircons as old as 4270 Ma

found in 3000 Ma quartzites and metaconglomerates

in Western Australia (Maas et al., 1992). Trace element

compositions of the zircons are consistent with a

612 Igneous and Metamorphic Petrology

19.1 Back-scatter electron microscope image of a complexly zoned

zircon from the tonalitic Acasta Gneiss Complex, Slave

Province (Figure 19.27), Northwest Territories, Canada. The

polished surface of the grain was etched in HF acid vapor to

enhance internal textural details such as the barely visible

oscillatory zoning in the magmatic core (high Th/U ratio)

overgrown by an unzoned metamorphic rim (low Th/U ratio)

and areas of late ﬂuid alteration. Oval “shadows” are pits

sputtered off (excavated into) the polished surface by the ion

beam of the SHRIMP. Isotopic analysis of sputtered mater-

ial from three spots within the zircon core yielded an age of

4028  21 Ma. The metamorphic rim of the grain yielded an

age of about 3600 Ma. The three deeper pits in the core were

revisited with a higher intensity beam for in situ trace element

analysis. Image courtesy of Richard Stern and Wouter Bleeker,

Geological Survey of Canada, Ottawa.

granitoid source, while 3900 Ma overgrowths imply

further crystallization deep in a differentiated contin-

ental source of substantial thickness.

The oldest known rocks on Earth are the Acasta

orthogneisses (metatonalite and metagranodiorite)

exposed over an area of only about 40 km

in the Slave

Province of northwestern Canada (see Figure 19.27)

that contain zircons as old as 4030 Ma (Special

Interest Box 19.1; see also Stern and Bleeker, 1998;

Bowring and Williams, 1999). Other rocks 3.8 Ga are

preserved as small scraps totaling only a few 100 km

in West Greenland (Figure 19.2) and nearby north-

ern Labrador as well as China and Antarctica. Rocks

3.6 Ga are exposed over an aggregate area of about

6000 km

in Ukraine, Western Australia, and South

Africa, in addition to the locales of 3.8 Ga rocks.

Virtually all rocks 3600 Ma are high-grade meta-

morphic rocks of the amphibolite and granulite facies,

many of which have been substantially modiﬁed during

younger events. Accordingly, their original nature can

be highly speculative.

19.2 ARCHEAN GRANITOID–

GREENSTONE TERRANES

Remnants of the oldest crust exposed in Archean

cratons consist of basically three components (Good-

win, 1996). About 10% is a supracrustal sequence of

volcanic and sedimentary rocks deposited in mostly

aqueous environment on the surface of the Earth and

about 60% is a plutonic suite of tonalite, trondhjemite,

and granodiorite (TTG) of roughly the same age as the

supracrustal sequence. Both have been recrystallized

and deformed to varying degrees, locally very strongly.

The supracrustal rocks typically lie in elongate, in-

folded synformal belts between domical masses of

TTG orthogneiss (Figures 9.14, 19.3, 19.6–19.8). Be-

cause the predominantly maﬁc rocks in these belts have

Precambrian Rock Associations

613

48°N

Isua

Godthåb

Fiskenæsset

Archean

gneisses

40°W

66°N

63°N

60°N

40°N

Archean (?) gneisses

intruded by

Proterozoic

plutons

Proterozoic

rocks

48°W

200km0

100miles0

19.2 Precambrian rocks of southern Greenland exposed along glacier-free coastal areas. Foliation in gneisses shown by thin lines, dashed

in Proterozoic rocks. Stippled pattern is granitic rocks. Rectangle encloses area of Figure 19.21a. Redrawn from Myers (1976). These

exposures are in the central part of the North Atlantic Craton that was rifted apart during opening of the Atlantic Ocean (Figure 13.23);

the western part is exposed in Labrador and the eastern in the British Isles.

been metamorphosed into greenstones under typically

greenschist-facies conditions, they are referred to as

greenstone belts. These have been extensively char-

acterized in the 800-page compendium edited by de

Wit and Ashwal (1997). Some TTG plutons were in-

truded as magma into the supracrustal rocks, whereas

others served as a platform onto which the supracru-

stal rocks were deposited. In many places, the con-

tact between orthogneisses and supracrustal rocks is

tectonic (sheared or faulted).

About 30% of Archean terranes is a younger

suite of calc–alkaline, potassic granites that are now

orthogneisses that are generally less metamorphosed

than the TTG orthogneisses.

Ongoing ﬁeld and laboratory investigations of

Archean granitoid–greenstone terranes are attempting

to resolve several basic questions. Are Archean con-

tinental and oceanic crust represented in granitoids of

the TTG suite and greenstones, respectively? Or, does

the typically higher grade of metamorphism of the for-

mer versus the lower grade of the latter simply signify

fundamentally different crustal types, or perhaps sim-

ply different levels of erosion? Do Archean basalts in

greenstone belts resemble modern oceanic ridge basalts

(MORB)? Or, are they more similar to subduction-

related island arc, back-arc, or plume-related oceanic

plateau basalts? Or, are they unlike rocks in any of these

more recent tectonic settings? Can greenstone belts be

compared to Mesozoic accreted terranes of oceanic rocks?

19.2.1 General Character of Greenstone Belts

Virtually all continents are host to small fragments of

Archean greenstone belts 3000 Ma. But younger

greenstone belts are widespread and many geologists

have recognized what they believe to be Phanerozoic

greenstone belts (de Wit and Ashwal, 1997).

614 Igneous and Metamorphic Petrology

25°S

20°S

Ocean

Craton

35°E

Zimbabwe

25°E

Great Dyke

Belt

Mobile

Limpopo

Bushveld

Complex

Craton

Kaapvaal

Indian

35°E

0 km 400

miles 200

19.3 Southeastern Africa showing Archean greenstone belts (black) in the Zimbabwe and Kaapvaal Cratons with their bordering and

transecting Proterozoic mobile belts (long lines). Trend lines in part of the Zimbabwe Craton show form of foliation in largely

tonalite–trondhjemite–granodiorite (TTG) gneiss domes between greenstone belts. Stippled areas are maﬁc–ultramaﬁc intrusive rocks of

the Great Dyke and Bushveld Complex. Box to east of Bushveld Complex encloses Barberton Greenstone Belt shown at larger scale in

Figure 19.4. Redrawn from Hunter and Hamilton (1978) and Phaup (1973).

Most supracrustal greenstone sequences are 5–10 km

thick (e.g. Figure 19.5) but commonly polyphase

folding (e.g. Figure 19.7) and apparent stacking of

thrust sheets hinder assessment of original stratigraphic

thicknesses. This style of deformation indicates hor-

izontal shortening in the Archean crust that is consist-

ent with, but does not prove, lateral plate motion; that

is, seaﬂoor spreading.

The oldest rocks in greenstone belts are commonly

metamorphosed tholeiitic basalt. Many belts also have

minor proportions of ultramaﬁc komatiite (discussed

later). Both were intruded as sills and extruded as lava

ﬂows, many of which were subaqueous, forming pil-

low lavas and hydromagmatic volcaniclastic deposits

(Sylvester et al., 1997). Distinctive shallow-marine de-

posits of accretionary lapilli and ash are traceable up to

30 km along strike. Andesites are a minor rock type in

Archean greenstone belts except for thick sequences in

the Superior Province of Canada (Figure 19.8) and in

the Yilgarn block of Western Australia (Figure 19.6),

where they form thick lava-debris-ﬂow sequences sim-

ilar to those in modern composite volcanoes. Silicic

lava ﬂows and volcaniclastic deposits are volumetric-

ally minor in all Archean terranes; they tend to occur

in the upper parts of greenstone sequences, overlying

the older basaltic part. Proterozoic greenstones belts

generally contain more andesite and felsic volcanic

rocks than most Archean and the amount of associated

sedimentary rock increases with decreasing age through

the Precambrian (Condie, 1994). This suggests a

gradual trend from submarine basalt–komatiite lava

plains or oceanic plateaus during the Archean to

more evolved composite volcanoes that emerged above

sea level during the Proterozoic.

Study of the metamorphism in greenstone belts

has lagged behind that of other terranes. A paradox

Precambrian Rock Associations

615

Granitoid plutons

(orthogneiss), mostly

TTG; age in Ma

Thrust fault; teeth

on upper plate

Subvertical fault

or shear zone

Moodies Group

Mostly sedimentary rocks

Mostly maﬁc-ultramaﬁc volcanic

rocks, lesser felsic; minor chert

Fig Tree Group

Weltevreden Formation (3290)

Mendon Formation (3300)

Kromberg Formation (3420)

Hoogenoeg Formation

Komati Formation

Theespruit Formation (3540)

Sandspruit Formation

Undivided volcanic rocks

20miles

3437

3481

3180

3227

Barberton

3250

31°E

26°S

Komati River

19.4 Barberton Greenstone Belt and surrounding orthogneiss domes in southern Africa. See Figure 19.3 for location. Domes are metamor-

phosed granitoid plutons, mainly of TTG. Rock units (Moodies is youngest, Sandspruit oldest; ages in Ma) are highly generalized, as are

faults. Redrawn from Lowe and Byerly (1999) and de Wit et al. (1992).

616 Igneous and Metamorphic Petrology

6 km

Theespruit

Formation

Basalt and komatiite intruded

by early Archean TTG

Komatiite

Pillow

basalt

Tholeiitic

basalt

lava ﬂows

Komatiitic basalt

with pillows and

spinifex fabric

Pillow

basalt

Basalt lava

ﬂows

Basaltic

volcaniclastic rocks

Chert (massive)

Chert (thin bedded)

Volcaniclastic rocks

Conglomerate

Sandstone

Moodies GroupFig Tree GroupHooggenoeg FormationKomati Formation

Mendon

Formation

Kromberg

Formation

Komatiite

Sandspruit

Formation

19.5 Stratigraphic section for the southern Barberton Greenstone Belt, South Africa. Column on the right is an expanded view of the Hoogge-

noeg Formation. Redrawn from Lowe and Byerly (1999).

in some belts is an apparent lack of progressive

increase in grade with depth, despite stratigraphic

thicknesses of many kilometers. Major points of the

metamorphism of greenstone belts include (Wilkins,

1997):

1. Facies range from prehnite–pumpellyite through

greenschist to amphibolite and locally to the gran-

ulite (Figure 19.9). Rocks of subgreenschist and

greenschist facies are the most widespread. Retro-

grade effects are common.