Myron G. Best. Igneous and metamorphic 2003 Blackwell Science

Подождите немного. Документ загружается.

volcaniclastic rocks, most older associations are sub-

volcanic, shallow plutonic complexes that contain ijo-

lite, the phaneritic nepheline-clinopyroxene plutonic

counterpart of nephelinite. These complexes (Figure

13.39) are typically circular or elliptical in map view,

and the constituent rock units form plugs, arcuate ring

dikes, and cone sheets. Brecciation is widespread.

Rock types include, in addition to carbonatite and

ijolite, alkaline pyroxenite, phonolite, nepheline syen-

ite, and, most characteristic of all, fenite. This is a

metasomatic rock (Figure 13.40) produced by solid-

state transformation of older wall rocks by inﬁltra-

tion of alkaline hydrothermal solutions derived from

the nephelinitic or carbonatitic magmas. The combi-

nation of brecciation, multiple intrusion, assimilation,

and fenitization produces complicated overprinted

fabrics.

Nd and Sr isotope ratios (Bell, 1989) unequivocally

demonstrate that the ultimate source of carbonatite

magmas is the mantle, probably long-lived, metasomat-

ically enriched continental lithosphere. Tangible sam-

ples are represented by rare garnet-phlogopite-calcite

lherzolite inclusions in carbonatite tuffs. Experiments

conﬁrm that carbonatite partial melts can exist under

special mantle conditions (Figure 11.5).

In June 1993 Oldoinyo Lengai erupted carbonatite

lava and ash that contain nephelinitic spheroids that

themselves contain alkali carbonatite segregations

(Table 12.4). These are obviously immiscible silicate-

carbonatite melts that separated from a carbonated sil-

icate magma at low P (Figure 12.8).

13.12 ALKALINE ORPHANS, MOSTLY IN

STABLE CRATONS

After alkaline basalts, nephelinites and phonolites

(Barker, 1996b) are the most widespread and volumi-

nous alkaline rocks. However, all these have received

only a fraction of the attention given to lamprophyres,

lamproites, orangeites, and kimberlites. These four are,

collectively, generally primitive silica-undersaturated

and potassic, maﬁc to ultramaﬁc volcanic and shallow

intrusive volatile-rich rocks. Subsolidus alteration to

clay and carbonate minerals is common because of the

Magmatic Petrotectonic Associations

397

13.39 Idealized spatial relations of a subvolcanic carbonatite-

nephelinite association in east Africa. Intrusion of nephelinite

(ijolite) magma is accompanied by early fenitization of wall

rocks. Eruption of magma to the surface builds a composite

volcano that subsequently experiences caldera collapse. Car-

bonatite magma (C

, dark shaded) intruded into the ijolite and

older country rocks and doming the roof rocks is accompanied

or preceded by a wave of fenitization and brecciation (light

shaded). Resurgence of the magma locally breaches its fenite

envelope. A swarm of later carbonatite cone sheets (C

) is then

emplaced, followed by still later carbonatite dikes (C

). (Re-

drawn from Le Bas, 1987.)

Caldera

Nephelinitic lavas and pyroclastic deposits

Cone sheets

Late fenitization

and brecciation

Ijolite

fenitization

Early

13.40 Hand sample showing fenite reaction zone adjacent to thin

dike of carbonatite in granite host rock. Note acicular alkali

amphiboles in carbonatite and their concentration as aphanitic

grains in the fenite.

Granitic

rock

Fenite

Carbonatite

2cm

high H

O and CO

fugacities of the magmas. Conse-

quently, rocks weather readily, are notoriously poorly

exposed, and tend to underlie soil- or water-ﬁlled topo-

graphic depressions. The four are petrologic “or-

phans” in that they do not have well-constrained chem-

ical, mineralogical, nor in some instances textural

deﬁnitions, as do common magmatic rock types such as

basalt or granodiorite. Each of the four is, in fact, not a

rock type at all, but a diverse clan of rocks loosely re-

lated in some way. Their tectonic settings are also

rather diverse. Most are found in continental regions of

crustal extension or crustal stability (cratons), of which

the African plate is the supreme example; but there are

also occurrences of lamprophyre lava ﬂows in conti-

nental arcs, such as the Western Mexico volcanic belt

just described.

Were it not for their intriguing attributes and the

potential insights they provide on mantle sources,

magma generating processes, and tectonics, these rocks

could be ignored. An added factor in their interest is

economic—three are the sole known primary source of

diamonds.

13.12.1 Lamprophyres

As implied by their name, lamprophyre rocks are

porphyritic (phyric), and the most common variety—

minette—contains abundant Fe-Mg mica (Greek lam-

pros, means “glistening”). Traditionally, lamprophyres

have been thought of as exclusively dike rocks, and

that characteristic is commonly cited as one deﬁning

attribute, but local occurrences of lava ﬂows (Luhr,

1997) and possibly moderate-size plutons of phaner-

itic rock have also been recognized. Petrologists only

began to study lamprophyres seriously in the last

decade or so of the twentieth century. Thus, much has

yet to be learned of their true nature, the best way to

classify them, the origin of magmas that form lampro-

phyres, and even whether they should continue to be

recognized as a distinct clan of kindred rocks. A work-

ing deﬁnition of lamprophyres can be proposed as fol-

lows (see also Le Maitre, 1989, p. 11; Mitchell, 1994,

p. 142): a diverse group of polygenetic rocks crystal-

lized under volatile-rich conditions and characterized

by abundant maﬁc phenocrysts of biotite-phlogopite

and/or amphibole together with lesser amouts of

clinopyroxene and/or melilite; feldspars are conﬁned

to the matrix along with, either singly or in various

combinations, feldspathoids, carbonate, monticellite,

melilite, mica, amphibole, pyroxene, olivine, per-

ovskite, Fe-Ti oxides, and glass. Strongly zoned and

corroded minerals indicate a lack of equilibrium dur-

ing ﬁnal crystallization.

Lamprophyres in continental arcs commonly occur

as dikes in tonalite-granodiorite plutons. This associa-

tion is so widespread that some petrologists recognize

it as a trilogy, the third component of which are maﬁc

inclusions and larger bodies of texturally and modally

diverse but hornblende-rich rock called appinite

(Pitcher, 1997, Chap. 10).

Mica-rich minettes (Table 13.11) and some kindred

lamprophyres have arc afﬁnities in their relatively de-

pleted Nb and Ta contents, high Ba/Ti, spiky normal-

ized trace element diagrams, and bulk chemical simi-

larities to the absarokite-shoshonite-banakite series

rocks of subduction zones. Shoshonitic rocks, com-

posed mostly of olivine, two pyroxenes, and plagioclase

and formed under low water fugacity, can be consid-

ered as heteromorphs of minettes that form under high

water fugacity, destabilizing plagioclase and anhydrous

maﬁc minerals. This heteromorphism accords with the

proposal of Mitchell (1994) that lamprophyres should

be considered as a facies—a diverse group of rocks

crystallized under similar conditions, in this case, high

water fugacity.

The high incompatible and compatible element con-

centrations of primitive lamprophyres, together with

their somewhat radiogenic Sr and Nd isotopic ratios,

suggest derivation from low degree partial melts of pre-

viously metasomatized mantle. In this vein-plus-wall-

rock melting mechanism (Section 11.5.1), veins of

phlogopite-rich assemblages could have been intro-

duced by aqueous solutions rising into the mantle

wedge above the subducting slab, in some cases as long

ago as the Proterozoic (Tingey et al., 1991).

13.12.2 Lamproite, Orangeite, and Kimberlite Clans

From the late 1800s it was believed that the only pri-

mary source of diamonds was kimberlite; a secondary

source was alluvium derived from kimberlite. How-

ever, in the late 1970s in the West Kimberly area of

Western Australia, rich diamond concentrations were

discovered in poorly exposed volcaniclastic deposits of

lamproite. Better exposures of lamproites in Western

Australia were made famous in studies in the mid-

twentieth century by R. T. Wade and R. T. Prider.

Unlike lamprophyres, the lamproite-orangeite-

kimberlite clans never contain plagioclase, nepheline,

or melilite. All are markedly low in Al

relative to al-

kalies so most are also peralkaline. In contrast to other

rather primitive maﬁc and ultramaﬁc rocks, such as

picrite and komatiite, which have high concentrations

of compatible Cr and Ni, rocks of these clans are also

unusually enriched in incompatible elements (Table

13.11; Figure 11.20). Xenocrystic olivine may in part

account for the high compatible element concentra-

tions. Diamond is not an essential deﬁning mineral. It

is only present in some occurrences and in extremely

minute concentrations; for example, in the famous

Kimberly mine in South Africa where kimberlite was

christened, 24 million tons of rock yielded only 3 tons

of diamond, or 0.125 ppm!

398 Igneous and Metamorphic Petrology

Each clan ranges widely in composition and fabric,

far more than common rock types. Additionally, there

are overlaps in most any rock property. Instead of tight

deﬁnitions, lengthy characterizations of each clan and

contrasts between them must sufﬁce.

Lamproite Clan. Of the 24 known worldwide locales of

clustered lamproite bodies, altogether 100 km

, the

main locales are in southeastern Africa ( Jurassic),

Western Australia (Proterozoic and Miocene), and the

Leucite Hills of Wyoming (Quaternary). All lie in tec-

tonically stable plate interiors, although prior sub-

duction or rifting appears to have metasomatized the

thick continental lithosphere through which magmas

ascended. In Western Australia, lamproite occurs in

narrow pipes that ﬂare at the surface into subcircular

craters. These are ﬁlled with a variety of volcaniclastic

deposits and are intruded by dikes and sills feeding

surface ﬂows. The 2-Ma Leucite Hills consist of lava

ﬂows and small cinder cones.

For comparison only (not as deﬁning attributes),

chemical analyses in Table 13.11 indicate that, relative

to the orangeite and kimberlite clans, lamproites are rel-

atively rich in SiO

, TiO

, Ba (commonly 5000 ppm),

La (200), and Zr (500) and poor in CO

. When

compared to orangeites-kimberlites that contain abun-

dant carbonate minerals, fresh lamproites generally

have 0.5 wt. % CO

. They are ultrapotassic (molecu-

lar K

O/Na

O  3) and their peralkalinity is mani-

fested in Fe

3

-Ba-rich, Al-poor leucite and sanidine;

Ti-rich, Al-poor phlogopite; and Ti-K-rich richterite

amphibole. Also present are Mg-rich olivine (both phe-

nocrysts and groundmass in some), diopside, glass, and

accessory amounts of apatite, ilmenite, and exotic min-

erals, which include wadeite (K

ZrSi

), priderite (K,

Ba, Fe

3

, Ti oxide), perovskite (Ca, Na, Fe

2

, REEs,

Ti, Nb oxide), and several other minerals (Scott-Smith,

1996). Not all of these phases necessarily occur in a

single sample and some samples can be mineralogically

unique, unlike others in the clan. This attribute, to-

gether with dissimilarites between lamproites in differ-

ent locales and the fact that some lamproities contain

mantle-derived xenocrysts, imposes difﬁculties in accu-

rate characterization of the whole clan.

Orangeite and Kimberlite Clans. The orangeite and

kimberlite clans are hybrid rocks emplaced as explo-

sive breccia in carrot- and funnel-shaped diatremes

fed from intrusive, nonfragmental dike-sill complexes

(Figures 9.25 and 9.26). Rocks are exceptionally rich

(12 wt.%) in CO

 H



, which are chieﬂy se-

questered in primary calcite and in secondary serpen-

tine that replaces olivine and orthopyroxene. The true

composition of the primitive magmas remains uncer-

tain because of differentiation and alteration overprints

and especially because of contamination by abundant

xenoliths and xenocrysts derived from both crust and

mantle. Xenoliths of crustal rock (shale, granite, and

others) and of mantle rock (eclogite and a wide vari-

ety of peridotitic rocks, commonly metasomatically al-

tered; see Section 11.2) are readily identiﬁed. On the

other hand, mantle-derived xenocrysts are difﬁcult,

and commonly impossible, to distinguish from phe-

nocrysts that precipitated from the kimberlite melt at

different places in its excursion from depths of perhaps

200 km or more, where diamond is stable (Figure

11.5). The degree of euhedralism as a guide to discrim-

inating foreign versus cognate crystals may be mislead-

ing because deep euhedral precipitates might be re-

sorbed at shallower depths or physically abraded in

the explosive diatreme. Rocks are characteristically

inequigranular because the polygenetic megacrysts

(5 mm in diameter) are hosted in an aphanitic, never

glassy groundmass of calcite, secondary serpentine,

phlogopite, Mg-spinel, and apatite. Most megacrysts,

commonly about 25 modal %, are Mg-olivine, but in

archetypal kimberlite, lesser megacrysts include Mg-

rich ilmenite, phlogopite, enstatite, purple-red pyrope

(Mg-rich) garnet, and green Cr-rich clinopyroxene; the

latter two serve as colorful “indicator” minerals in the

regolith overlying weathered kimberlites and in down-

stream stream and glacial deposits.

Early work just after the turn of the 1900s revealed

two types of diamondiferous rocks in South Africa. In

subsequent decades, this distinction became confused,

but it was again clariﬁed by Mitchell (1995). The two

clans are:

(1) Archetypal kimberlites (also called basaltic or

group I kimberlites)

(2) Orangeites (also called micaceous or group II kim-

berlites), which are mineralogically, isotopically,

and, with regard to origin, more similar to the lam-

proite clan than to archetypal kimberlites, although

orangeites contain more CO

and compatible Ni

and Cr but less Rb, Ba, Zr, and Ti than lamproites

The only systematic chemical contrast between the

kimberlites and orangeites lies in higher K

O/TiO

the latter. The most striking contrast is signiﬁcantly

higher

Sr/

Sr and lower

143

Nd/

144

Nd in orangeites

than in archetypal kimberlites (Figure 13.41). Miner-

alogical contrasts among the lamproite, kimberlite, and

orangeite clans are set out in Table 13.12.

All known (200+) orangeite bodies lie in the

Archean Kaapvaal craton in South Africa (Figure

13.42), where they occur as early Cretaceous (125–

110 Ma) and early Jurassic (165–145 Ma) swarms of

dikes and diatremes. Worldwide, there are 5000

known bodies of archetypal kimberlites having an esti-

mated volume 5000 km

; most of these are in Pre-

cambrian cratons. They typically occur in clusters

40 km across of a few dozen intrusions (Nixon,

1995). In southern Africa, kimberlite emplacement has

occurred episodically several times since 1600 Ma.

Magmatic Petrotectonic Associations

399

Implications of Diamonds and Inclusions in Them.

Diamond-bearing kimberlites and orangeites occur

only on Archean cratons (2.5 Ga), whereas barren

kimberlites lie in peripheral terranes that were sub-

jected to post-Archean tectonism. On the other hand,

diamond-bearing lamproites occur on Archean and

Proterozoic cratons. Diamonds commonly contain in-

clusions of many different minerals (visible with a 10

hand lens) that provide signiﬁcant information not only

on their origin but also on the origin of the host magma

and perhaps even the tectonic history of the planet

(Haggerty, 1999).

Sm-Nd and Rb-Sr ages of garnet and pyroxene in-

clusions, and the assumed syngenetic host diamonds,

are 900 to as much as 3300 Ma. Hence, most diamonds

are much older than their host rocks (1600–90 Ma in

southern Africa). The chemical composition of inclu-

sions indicates that diamonds crystallized and were

stored and preserved in stable craton roots at depths of

mostly 120–200 km and temperatures 1200–1500°C

(Helmstaedt and Gurney, 1995; Gurney and Zweistra,

1995). This implies an unusually low geothermal gradi-

ent (8–10°C/km), especially for what is generally be-

lieved to have been a hotter planet during the Precam-

400 Igneous and Metamorphic Petrology

13.41 Radiogenic isotope ratios in potassic maﬁc-ultramaﬁc volcanic rocks compared with oceanic basalts (Figure 13.8). Note that orangeites

(dark shade) are far more enriched than worldwide archetypal kimberlites (light shade) and are isotopically more like lamproites (diag-

onally ruled). Note also provinciality of lamproites; those in Western Australia are more enriched in

Sr than those in Leucite Hills of

Wyoming. BE, bulk Earth. (Redrawn from Mitchell and Bergman, 1991.)

0.702 0.706 0.710 0.714

0.718

Sr/

0.5116

0.5120

0.5124

0.5128

0.5132

143

144

LEUCITE

HILLS

Phl-rich

Lamproites

WESTERN

AUSTRALIA

Lct-rich

Ol-rich

Orangeites

Archetypal kimberlites

HAWAII

MORB

Lamprophyres

−4

−8

−12

−16

−18

Table 13.12. Mineralogical Comparison of Lam-

proites, Archetypal Kimberlites, and Orangeites

RCHETYPAL

AMPROITES

IMBERLITES

RANGEITES

Olivine mm, ppp mmmm, ppp mm, pp

Phlogopite mmm, ppp, mm, pp, ggg mmm, ppp,

ggg ggg

Leucite ppp, ggg g

Ti-K richterite ppp, ggg gg

Sanidine ppp, ggg ggg

Diopside ggg ggg

Monticellite ggg

Spinel gg gggg gg

Perovskite gg ggg gg

Apatite ggg ggg gggg

K-Ba-titanites ggg g ggg

Zr-silicates ggg g ggg

Mn-ilmenite g gg ggg

Calcite gggg ggg

Pyrope mmm

Blank, absent; x, very rare; xx, rare; xxx, common; xxxx, abun-

dant; g, groundmass; p, phenocryst; m, megacryst.

(Data from Mitchell, 1996.)

brian. Very rare inclusions even appear to be from the

mantle transition zone at a depth of 410–670 km, and

some may have crystallized in the deeper, lower mantle

(e.g., Harris et al., 1997). After storage for long periods

in relatively cool, presumably nonconvecting mantle,

diamonds were caught up in ascending, younger, ge-

netically unrelated magmas. Pieces of the rock in

which the diamonds were originally lodged now occur

as diamond-bearing xenoliths of eclogite and peridotite

(chieﬂy depleted harzburgite) in the host lamproites,

orangeites, and kimberlites. Disaggregation of these

xenoliths into the much larger volumes of entraining

magmas produced abundant megacrysts of indicator

minerals—olivine, pyroxene, garnet, and others—as

well as the orders-of-magnitude-more-diluted xenocrys-

tic diamond population.

Note the hierarchy of petrologically signiﬁcant in-

clusions: Minute mineral grains that serve as isotopic

clocks and as geothermobarometers are hosted in dia-

monds that occur as grains in mantle eclogite and

harzburgite which are included as xenoliths in lam-

proites, and kimberlites, which are hosted in Precam-

brian craton country rock.

Origin of Magmas and Relation to the Diamond Source.

Genesis of lamproite, orangeite, and kimberlite mag-

mas must take into account their chemical and iso-

topic compositions, episodic production since the late

Archean, and presence of entrained diamond-bearing

eclogite and peridotite derived from deep, cool, non-

convecting lithospheric mantle, mostly of Archean age.

The normalized trace element diagram (Figure 11.20)

for lamproite and kimberlite is similar to the humped

pattern for oceanic island basalts (OIBs) but the more

incompatible elements are more enriched by about an

order of magnitude than in OIB. This suggests deriva-

tion of magmas by smaller degrees of melting of an

OIB-like, depleted mantle source or by melting of a

metasomatically enriched source. The high ratio of

light REEs/heavy REEs suggests garnet in the source

Magmatic Petrotectonic Associations

401

13.42 Global occurrence of kimberlite, orangeite, and lamproite. Generalized extents of Archean cratons shaded. Major diamond deposits

shown by larger circles (lamproite) and squares (kimberlite). Lower left is enlarged southern Africa Kaapvaal craton, showing clusters of

kimberlite and orangeite diatremes with ages of emplacement ranging from 1600 to 60 Ma. (Redrawn from Mitchell, 1995.) Diamond-

bearing orangeite and kimberlite occur only on Archean cratons, but diamondiferous lamproites are found in both Proterozoic and

Archean cratons. (Redrawn from Nixon, 1995; Haggerty, 1999.)

0 miles 250

4000km

Namibia

PROTEROZOIC

100

110

120

South Africa

137

CRATON

Botswana

500

165

145

125

115

120

1600

1200

Lamproite

Kimberlite

120

Orangeite

Kimberlite

cluster

190

KAAPVAAL

Table 13.13. Principal Factors in the Origin of Two Petrotectonic Associations

ECTONIC

ETTING

Oceanic rift

Continental arc

AUSE OF

AGMA

ENERATION

Decompression of

upwelling mantle

Reduction of solidus T in

subarc mantle wedge by

increase in water

concentration; increase

in T and perhaps in

volatile concentration of

lower crust due to

intruded mantle-derived

maﬁc magmas

AGMA

OURCE

(

)

AND

OMPONENTS

Mantle peridotite at

depths of 40–80 km

previously depleted in

incompatible elements

Aqueous solutions and

partial melts from

subducting oceanic

basalt crust; depleted

upper mantle

peridotite hydrated

and metasomatized by

aqueous solutions and

melts; continental crust

IFFERENTIATION

ROCESS

(

)

Polybaric olivine

fractionation of

magmas ascending

from source; low-P

fractionation of Pl 

Cpx  Ol in crustal

magma chambers but

moderated by mixing

with draughts of more

primitive mantle-

derived magma

Concurrent fractional

crystallization of

primitive mantle-

derived magmas and

assimilation of crustal

rocks (AFC); magma

mixing

OCK TYPE

(

)

Relatively uniform

tholeiitic basalt

(MORB) and gabbro

containing Q and Hy;

rare ferrobasalt,

ferrodacite, and

plagiogranite;

peridotite

Andesite, dacite, rhyolite,

dioritic rocks, tonalite,

granodiorite, granite;

uncommon adakite,

basalt, and gabbro

HARACTERISTIC

HEMICAL AND

SOTOPIC

OMPOSITION

Relatively low concentrations

of incompatible elements;

low

Sr/

Sr and high

143

Nd/

144

Evolved (nonprimitive)

chemical compositions;

high incompatible, low

compatible and HFS

elements (Ta, Nb), spiky

trace element arc

signature; variable and

commonly relatively high

Sr/

Sr and 

O and

low

143

Nd/

144

ABRICS AND

IELD

ELATIONS

Submarine basalt lava

ﬂows, commonly

pillowed, and associated

clastic deposits; sheeted

dikes; rare dikes of

plagiogranite; isotropic

gabbros; layered

cumulate gabbros and

peridotites; altered

metamorphic textured

(deformed) peridotites

Composite volcanoes; lava

and debris ﬂows;

widespread pyroclastic

deposits; calderas;

typically huge batholiths

made of hundreds of

individual plutons

emplaced over tens of

millions of years

Magmatic Petrotectonic Associations

403

residue, agreeing with the deep source implied by the

xenocrystic diamonds. Figure 13.41 shows that magmas

creating the archetypal kimberlite clan have a source

that is near bulk Earth and like that of asthenospheric,

plume-related OIB, exempliﬁed by Hawaiian magmas.

In striking contrast, orangeite and lamproite clans have

higher

Sr/

Sr and lower

143

Nd/

144

Nd, so their mag-

mas must have a major component from mantle litho-

sphere metasomatically enriched during the Precam-

brian.

It is possible that nonconvecting mantle underly-

ing old cratons is a remnant of ancient mantle plumes

(Herzberg and O’Hara, 1998; Haggerty, 1999) and has

experienced an unusual long-term history of metaso-

matism and a younger episode of magma generation, as

suggested by markedly different Sr isotope ratios of

lamproites in Figure 13.41. Whether the diamondifer-

ous eclogites in the subcratonic mantle represent

frozen high-P basaltic melts or remnants of ancient

subducted oceanic crust (Helmstaedt and Gurney,

1995) is controversial. Likewise, the source of the car-

bon for diamonds and the CO

dissolved in the kim-

berlitic magmas—whether from ancient subducted

crust or primeval mantle—is also controversial. These

questions bear on the problematic origin of Archean

continents and lithosphere (see Section 19.8).

SUMMARY

Relating the wide diversity of magmatic rocks to their

tectonic setting provides rich insights into the way the

Earth works—a central theme of this textbook. Inter-

pretive models of magma origin can constrain global

models proposed by geophysicists and tectonic geolo-

gists, and vice versa. Of fundamental concern is the

thermal budget of the Earth, the way its vast store of

thermal energy is dissipated through motion of litho-

spheric plates and rise of mantle plumes and the mag-

matism associated with these convective processes.

From a narrower perspective, study of petrotec-

tonic associations focuses attention on speciﬁc magma

sources, conditions of generation, and subsequent

processes of differentiation of primitive magmas that

have produced distinct rock types and rock suites in

speciﬁc tectonic settings. Rather than a lengthy sum-

mary of all of the associations described in this chapter,

only two strongly contrasting associations are set out in

Table 13.13.

CRITICAL THINKING QUESTIONS

13.1 Prepare a detailed summary of each of the

petrotectonic associations described in this

chapter using the format in Table 13.13. Two

are provided as examples.

13.2 How does the composition of oceanic rocks

constrain models proposed by geophysicists

concerning the nature of the mantle and con-

vection in it?

13.3 In the complementary Sr and Nd isotope com-

positions of MORB and continental crust (Fig-

ure 2.27) why is MORB closer to the mantle ar-

ray than the continental crust?

13.4 How does the composition of rocks in ophio-

lites constrain their tectonic interpretation?

13.5 What was the state of stress during emplace-

ment of the Karoo dikes and sills shown in Fig-

ure 13.20b?

13.6 What special geologic conditions in Iceland

compared to other oceanic islands might be re-

sponsible for the exceptionally large proportion

of silicic rocks?

13.7 Explain the paradox of the mica-rich lamproite

in Table 13.11, column 7, that contains norma-

tive quartz but modal leucite (see also problem

13.2). How would you classify this rock? Is it

silica-oversaturated or silica-undersaturated?

13.8 What do kimberlitic and lamproitic rocks tell

us about the history of cratons (long-term sta-

ble sectors of continents unaffected by tecton-

ism) and their thermal state?

13.9 Diamonds contain mineral inclusions that crys-

tallized deep in the mantle at high T as long

ago as 3 Ga. How is it possible for these to be

preserved for our study?

13.10 Why does orthopyroxene never crystallize un-

der equilibrium conditions in silica-undersatu-

rated alkaline magmas?

13.11 From phase diagrams in Chapters 5 and 11

how might you account for the contrasting

compositions of olivines in Hawaiian tholeiitic

and alkaline basalts (Section 13.2.1)?

PROBLEMS

13.1 Make primitive-mantle-normalized trace ele-

ment diagrams of granitic and rhyolitic rocks

from Tables 13.7, 13.8, 13.9, and 13.10. Com-

pare these patterns and comment on what they

tell about the source of magmas and overprint-

ing differentiation processes.

13.2 Calculate the normative composition of the pla-

giogranite in Table 13.7 and the phlogopite-rich

lamproite in Table 13.11. How do these norms

compare with the corresponding modes? See

also critical thinking question 13.7.

UNDAMENTAL

UESTIONS

ONSIDERED IN

HIS

HAPTER

1. What is the nature of subsolidus changes in the

fabric and composition of rocks that occur during

metamorphism?

2. What transfers of energy and matter in geologic

systems bring about metamorphism?

3. In what geologic settings does metamorphism

occur, and why?

4. What role do metamorphic rocks play in our

understanding of crustal evolution, global plate

tectonics, and how the Earth works?

INTRODUCTION

In the dynamic Earth, changes in geologic systems res-

ulting from transfers and transformations of energy

and from movement of rock, magma, and ﬂuids are

continually occurring, chieﬂy near margins of lithos-

pheric plates. States of thermodynamic equilibrium are

perturbed, causing rock systems to seek new, lower

energy, more stable states by adjustment of their fabric

and composition. These equilibrating adjustments that

take place in the solid state at elevated temperatures

are called metamorphism. Equilibration processes and

the kinetic factors that control their rates vary widely,

depending on the fabric and composition of the parent

rock, the T and P of the evolving rock system, the com-

position of ﬂuids in it, and the prevailing state of stress.

Changing geologic conditions in metamorphic systems

range between those where buried sediment is ce-

mented and lithiﬁed in the sedimentary process known

as diagenesis at relatively low T and P, up to where

rock partially melts (Figure 14.1). Simply put, during

metamorphism, rocks are hot enough to recrystallize

but not hot enough to melt.

Many sedimentary and some magmatic rock-forming

systems can be directly observed. However, our study

of metamorphic rock-forming systems can never enjoy

this beneﬁt because of their concealment well beneath

the surface of the Earth, where P and T are elevated.

Consequently, equilibrating metamorphic processes

can generally only be understood from observations

made of their rock products long after the meta-

morphism has terminated, combined with logic and

inferences from necessarily simpliﬁed thermodynamic

models and laboratory experiments.

Like magmatism, metamorphism involves transfer of

heat and mass. Consequently, many concepts developed

in preceding chapters apply equally well to metamor-

phic systems as to magmatic ones. Examples include

concepts of thermodynamics and kinetics (Chapter 3),

phase equilibria (Chapter 5), diffusive transport of atoms

and heat (Chapter 6), and stress-deformation relation-

ships and rheologic behavior (Chapter 8).

Nonetheless, there are signiﬁcant contrasts between

magmatic and metamorphic systems. Magmatic beha-

vior is dominated by a melt, or silicate liquid, that con-

tains dissolved volatiles and interacts with crystalline

phases. In contrast, subsolidus metamorphic systems

lack a melt but usually have a volatile ﬂuid that

interacts chemically, physically, and thermally with the

Metamorphic

Rocks and

Metamorphism:

An Overview

CHAPTER

crystalline aggregate and can have a catalyzing kinetic

effect on metamorphic equilibration, hastening and

facilitating changes in the rock system. Overall, how-

ever, chemical reaction rates tend to be slower than in

magmatic systems because of the lower (subsolidus)

temperatures. Hence, mineralogical disequilibrium

and relics of incomplete overprinted reactions are

widespread, which can be both a curse and a blessing

for the petrologist trying to unravel a metamorphic his-

tory. Metamorphic rocks commonly bear the imprint

of nonhydrostatic states of stress in the form of

anisotropic fabrics developed during ductile ﬂow, such

as foliated and lineated schists and intricately folded

layered rock (e.g. Figure 8.4). In contrast, bodies of

crystallizing magma only locally acquire anisotropic

fabric.

The purpose of this chapter is to introduce meta-

morphic rocks as aggregates of minerals whose com-

position and fabric reﬂect adjustments to new states

of equilibrium brought about by changes in intensive

variables and states of stress. The record of these

changing states frozen in metamorphic rocks provides

crucially signiﬁcant insights into the evolution of the

crust of the Earth. In conjunction with magmatic rocks,

metamorphic rocks tell about the evolution of contin-

ents and global tectonic processes dating from the

earliest history of the Earth (see Chapter 19).

Chemical formula and abbreviated names of meta-

morphic rock-forming minerals are listed in Table 14.1

and select analyses are shown in Table A.2 in Appen-

dix A.

Turner (1981) and Miyashiro (1994) provide a

comprehensive coverage of the mineralogical, ﬁeld,

and tectonic aspects of metamorphic rocks. Useful

introductions include Yardley (1989), Blatt and Tracy

(1996), and Winter (2001). Numerous illustrations of

metamorphic rocks in thin section together with suc-

cinct summaries of metamorphism and metamorphic

rocks can be found in Nockolds et al. (1978), Williams

et al. (1982), and Yardley et al. (1990).

14.1 EXAMPLES OF EQUILIBRATION IN

METAMORPHIC ROCKS

The concept of metamorphism as a change to a new

more stable state of equilibrium in dynamic geologic

systems carries an implication of some sort of path

from the initial state to the ﬁnal state. This path from

the parent rock, or protolith, to the ﬁnal metamorphic

rock product is the process of metamorphism driven

by changes in intensive parameters and states of stress.

Although the driving “force” for metamorphism

commonly involves an increase in T and P, such as

accompanies deepening burial of surﬁcially deposited

sediment or increase in T around a magmatic intrusion,

many other changes in geologic conditions can also

drive re-equilibration of a rock system. Determination

of the nature of this often complex path is the central

challenge facing the petrologist.

This section explores some examples of metamor-

phic paths involved in the changes in composition and

fabric of solid protoliths.

14.1.1 Incipient Metamorphism: Crystallization

of New Minerals and Preservation of Relict

Protolith Fabrics

The upper and lower limits of metamorphism are

easily cited in words or portrayed in diagrams (Fig-

ure 14.1), but not easily discerned in rocks. At the

upper limit, because the beginning of partial melting

Metamorphic Rocks and Metamorphism: An Overview

405

P (kbar)

Not attained on Earth

SUBDUCTION ZONE NEAR TRENCH

CONTINENTAL OROGEN

OCEANIC RIDGES, CONTACT

AUREOLES, ISLAND ARCS

Dry basalt

Granite

COLLISION

ZONE

(Ultra-high-P

 30 kbar)

Base of continetal crust

ran

ite

“Wet”

Depth (km)

0 200 400 600 800 1000 1200

T (°C)

“

”

14.1 P–T conditions of metamorphism. Redrawn from Spear (1993).

Solidi of basalt and granite under dry and water-saturated con-

ditions are indicated. Most metamorphism (darker shading)

takes place below about 850°C, at temperatures below the

water-saturated (“wet”) basalt solidus and at pressures gener-

ally 10 kbar corresponding to depths of 37 km depth,

based on an average continental crustal density of 2.7 g/cm

However, metamorphism can occur very locally at tempera-

tures up to the solidus of dry basalt (lighter shading) and in

subduction zones near trenches and in continental collision

zones to depths of 100 km or more (ultrahigh P 30 kbar).

Three principal geologic environments of metamorphism have

contrasting P/T metamorphic ﬁeld gradients. Subduction

zones near the trench are high P/T (relatively high-P and low

T). Ocean ridge, island arc, and contact aureoles around shal-

low magmatic intrusions are low P/T (low-P and moderate to

high-T). The most widespread environment is in the contin-

ental crust in orogenic belts where intermediate P/T prevails.

There is no sharp distinction between these three gradients,

nor with the distinction between low-grade metamorphism

and diagenesis, where sediment is made into rock at lowest P

and T. The base of the continental crust (Mohorovidic discon-

tinuity; heavy line on right-hand axis) lies at depths ranging

from about 35 km to as much as 70 km beneath the Andes and

Himalaya Mountains.

406 Igneous and Metamorphic Petrology

Table 14.1. Metamorphic Rock-Forming Minerals

Actinolite Act Ca

(Mg,Fe

3

)

(Si

)(OH,F)

Albite* Ab NaAlSi

Almandine* Alm Fe

2

Aluminosilicate Als polymorphs of Al

SiO

Amphibole Am complex hydrous silicates

Analcite (analcime) Anl NaAlSi

·H

Andalusite And Al

SiO

Andradite* Adr Ca

(Fe

3

,Ti)

Ankerite Ank Ca(Mg,Fe

2

,Mn)(CO

)

Annite* Ann K

2

(OH)

Anorthite* An CaAl

Anthophyllite Ath (Mg,Fe

2

)

(OH,F)

Antigorite Atg Mg

(OH)

Aragonite Arg CaCO

Biotite Bt K

(Fe

2

,Mg)

6–4

(Fe

3

,Al,Ti)

0–2

(Si

6–5

2–3

(OH,F)

Brucite Brc Mg(OH)

Calcite Cal CaCO

Carpholite Car (Mn,Mg,Fe

2

)(Al,Fe

3

)

(OH)

Chlorite Chl (Mg,Fe

2

,Fe

3

,Mn,Al)

(Si,Al)

(OH)

Chloritoid Cld (Fe

2

,Mg,Mn)

(Al,Fe

3

)(OH)

(SiO

)

Chrysotile Ctl Mg

(OH)

Clinoptilolite Clp (Na,K)

(Al

)·24H

Clinopyroxene Cpx (Ca,Mg,Mn,Fe

2

,Fe

3

,Ti,Al)

(Si,Al)

Coesite Cs SiO

Cordierite Crd (Mg,Fe)

·H

Corundum Crn Al

Diopside* Di CaMgSi

Dolomite Dol CaMg(CO

)

Epidote Ep Ca

3

O(SiO

)(Si

)(OH)

Fayalite* Fa Fe

SiO

Forsterite* Fo Mg

SiO

Garnet Grt see Alm, Adr, Grs, Prp Sps end-members

Glaucophane Gln Na

(Mg,Fe

2

)

(Al,Fe

3

)

(OH)

Graphite Gr C

Grossular* Grs Ca

Hedenbergite* Hd CaFeSi

Hematite* Hem Fe

Heulandite Hul (Ca,Na

)

(Al

)·24H

Hornblende Hbl (Na,K)

0–1

(Mg,Fe

2

,Fe

3

,Al)

6–7.5

2–0.5

(OH)

Illite Ill K

1.5–1.0

(Si

6.5–7.0

1.5–1.0

)(OH)

Ilmenite* Ilm FeTiO

Jadeite* Jd NaAlSi

Kaolinite Kln Al

(Si

)(OH)

K-feldspar* Kfs KAlSi

Kyanite Ky Al

SiO

Laumontite Lmt Ca

·H

Lawsonite Lws CaAl

(OH)

·H

Magnesite Mgs MgCO

Magnetite* Mag Fe

Monticellite Mtc Ca(Mg,Fe)SiO

Montmorillonite Mnt (

Ca,Na)

0.7

(Al,Mg,Fe)

(Si,Al)

(OH)

·nH

Muscovite* Ms K

(Si,Al)

(OH,F)

Olivine Ol (Mg,Fe)

SiO

Omphacite Omp (Ca,Na)(Mg,Fe

2

,Fe

3

,Al)Si