Myron G. Best. Igneous and metamorphic 2003 Blackwell Science
Подождите немного. Документ загружается.
rock together with the peak equilibrium assemblage.
The decision as to what constitutes an equilibrium
mineral assemblage must therefore be made cautiously
(see Section 16.1).
The pelitic rocks in the Onawa contact aureole (Fig-
ure 14.25) again serve to illustrate how mineral assem-
blages provide information on the intensive variables
of metamorphism. The highest grade hornfels in the
aureole contain the assemblage quartz alkali feld-
spar sillimanite biotite cordierite Fe–Ti oxide.
(Irregularly shaped andalusite grains may be metast-
able relics from lower grade conditions.) Whereas alkali
feldspar alone is stable over a wide range of P and T,
examination of Figure 14.31 shows that, in the pres-
ence of quartz, alkali feldspar plus sillimanite is stable
only at P greater than 2.25–3.1 kbar and at T greater
than 540–610°C, depending on the exact composition
of the feldspar and the decomposing mica. The upper
P–T limit is about 11 kbar and 850°C (not shown in
Figure 14.31). A possible further constraint is provided
by the occurrence in the hornfels of small segregations
of quartz and alkali feldspar that have a bulk composi-
tion equivalent to the minimum-T valley in the liquidus
of the granite system (Figures 5.24 and 5.25). These
segregations are unlike the granodiorite in the nearby
pluton that drove the contact metamorphism and likely
formed by local partial melting rather than represent-
ing offshoots from the pluton. If they did originate as
local partial melts, the permissible range of P and T for
the highest grade hornfels in the innermost part of the
aureole must lie just above the water-saturated solidus
of granite on the high-T side of the breakdown curve
for muscovite quartz, possibly near 700°C and 3.5
kbar, in Figure 14.31. Pressures in the aureole cannot
have been too high; otherwise kyanite would have been
stabilized. Further limitations on P–T conditions of
the high-grade hornfels could be provided by stability
limits of the particular cordierite composition (at in-
creasing P) and biotite (at increasing T) that occur in
the rock.
Thus, one can envisage a network, or petrogenetic
grid, of intersecting reaction lines bounding the stabil-
ity field of a particular mineral assemblage in a P–T
diagram appropriate for a particular rock composition.
Since first conceived by N. L. Bowen in 1940, many
grids constructed from experimental and thermo-
dynamic data are now available to constrain mineral
equilibria (Spear, 1993).
Mineral thermometers and barometers are yet another
way of obtaining information on P–T conditions of meta-
morphism. These were very briefly introduced in Sec-
tion 5.8 and are discussed more fully in Section 16.11.3.
14.2.7 Metamorphic Facies
Discerning changes in metamorphic conditions is
facilitated where many mineral assemblages in diverse
chemical rock groups—pelitic, calc-silicate, mafic, and
so on—occur within restricted parts of metamorphic
terranes. All of the mineral assemblages in all of the
chemical groups, which constitute a spatially localized
suite of mineral assemblages, may then be compared
with suites in a different part of the same terrane or
in other terranes to determine how P–T conditions
differed.
Such comparison of suites of mineral assemblages
was first made nearly a century ago. In 1911, V. M.
Goldschmidt found a mixed sequence of pelitic, cal-
careous, and quartzo-feldspathic hornfelses surround-
ing magmatic intrusions in the area near Oslo, Norway.
Mineral assemblages constituting the hornfelses are rel-
atively simple; no more than four or five minerals make
up each assemblage in each chemical type. Moreover,
a particular mineral assemblage was found to consis-
tently make up one chemical type of hornfels to the
exclusion of a chemically equivalent assemblage; thus,
anorthite hypersthene was consistently present instead
of the chemically equivalent andalusite diopside.
Goldschmidt concluded that these regular corres-
pondences between mineral assemblage and bulk-rock
chemical composition represented states of mineralo-
gical equilibrium attained under the particular meta-
morphic conditions.
Shortly later, in 1914–15, P. Eskola described re-
gionally metamorphosed rocks in the Orijärvi region of
Finland whose suite of mineral assemblages again
showed consistent and simple correspondences to
bulk chemical compositions. Some assemblages, such
as diopside grossular in calc–silicate rocks and
andalusite cordierite plagioclase in pelites, occur
also in the Oslo aureoles. However, other Orijärvi
mineral assemblages differ from the Oslo assemblages,
but occur in rocks of equivalent bulk chemical com-
position. For example, the Oslo assemblage of alkali
feldspar andalusite occurs in pelites of the same bulk
chemical composition (volatiles excepted) as Orijärvi
pelites made of muscovite quartz. These two pairs of
minerals are related to one another in reaction 14.1.
The following list compares three additional mineral
assemblages in rocks of equivalent volatile-free chem-
ical composition (on the same line).
Oslo suite Orijärvi suite
alkali feldspar andalusite muscovite quartz
alkali feldspar cordierite muscovite biotite
alkali feldspar hypersthene biotite hornblende
anorthite
hypersthene anthophyllite
Comparison of the two suites of assemblages in the two
regions led Eskola (1915) to formulate the concept
of metamorphic facies. A facies is a suite of mineral
Metamorphic Rocks and Metamorphism: An Overview
437
assemblages that is repeatedly found in terranes of all
ages around the world and possesses a regular relation
between mineral composition and bulk chemical com-
position. Thus, the complete Oslo suite of assemblages
constitutes one metamorphic facies and the complete
Orijärvi suite another facies. Eskola’s intent, which
is still valid today, was to use facies as a means of
contrasting differences in P–T conditions of meta-
morphism. Qualitative thermodynamic reasoning led
Eskola to believe that the Oslo facies equilibrated at
relatively higher T and P than the Orijärvi facies, since
the Oslo assemblages consist of less hydrous, denser
minerals.
Turner (1981, p. 54) warns that: “Any [one] facies is
recognized on the basis of a set of mutually associated
rocks collectively covering a wide range of composi-
tion. It cannot be defined in terms of a single rock
type.” Moreover, “contrary to a view repeatedly ex-
pressed in the literature [of petrology]..., the facies
concept is unsuited for use in developing a compre-
hensive classification of individual rock types.”
Since the seminal work of Goldschmidt and Eskola,
there has been countless documentation in terranes of
all ages all over the world of simple mineral assem-
blages corresponding to, and predictable from, bulk
chemical rock compositions and formed under a re-
stricted range of metamorphic conditions. Under the
same restricted conditions in which one facies forms,
rocks of the same chemical composition are composed
of the same mineral assemblage. Any difference in a
mineral assemblage corresponds to a different bulk
chemical composition. This consistency and repetition
of mineral assemblages in space and time is believed to
reflect attainment of equilibrium during recrystallization
—supporting the facies concept.
Although the concept of metamorphic facies has
proven useful, there has been an ongoing debate in the
decades since 1915 as to how facies should be defined
and especially what constitutes a legitimate facies and
under what conditions the suite of mineral assemblages
constituting a particular facies equilibrated (Miyashiro,
1994, Chapter 7). Some workers have proposed nar-
rowly defined transitional facies between Eskola’s that
have been rejected by other petrologists. Eskola and
most following petrologists focused on P–T conditions
as the only intensive variables controlling the stability
of minerals in a particular facies. But chemical activities
of volatiles can also be significant, as in stabilization of
zeolites under low CO
2
activities at low P and T. With
these caveats in mind, Table 14.2 presents a standard
list of facies and their characterizing mineral assem-
blages as used by most petrologists. Each facies has a
name referring to a distinctive attribute. For example,
rocks of the zeolite facies are characterized by zeolites,
and the greenschist facies, probably the most wide-
spread of all, is characterized by schists in which the
green minerals chlorite, actinolite, epidote, and serpen-
tine are predominant.
The approximate P–T ranges over which the stand-
ard facies are interpreted to have formed is shown in
Figure 14.33. Definite P–T boundaries between the
facies cannot be drawn because most mineral assem-
blages form by continuous reactions “smeared” over
some range of P and T as solid solutions break down
into new minerals and interact with C–O–H fluid solu-
tions of differing activities. In addition, despite decades
of effort, experimental and thermodynamic data on
P–T stabilities of mineral assemblages—defining petro-
genetic grids—are still incomplete and not universally
agreed upon.
Two facies recognized in this textbook are found
in shallow crustal contact aureoles and constitute
relatively low-P hornfels facies. The pyroxene–hornfels
facies forms under lower temperatures than the sani-
dinite facies.
14.2.8 Metamorphic Facies Series
In 1961, A. Miyashiro (see 1994) recognized that dif-
ferent regional metamorphic terranes in what were
later identified as different plate tectonic settings
possess different associated facies, or facies series. In
regional metamorphic terranes associated with con-
tinental magmatic arcs overlying subducting oceanic
lithosphere (Figure 14.24), rocks of the greenschist and
amphibolite facies constitute an intermediate-P facies
series (Figure 14.34); these two facies make up most
of the classic Barrovian zones in the Scottish High-
lands (Figure 14.30). Locally, at higher temperatures,
migmatites and/or rocks of the granulite facies may
also be associated and, at lower temperatures, rocks
of the prehnite–pumpellyite and perhaps even the
zeolite facies. A low-P facies series corresponds to the
cordierite–andalusite Buchan zones in pelitic rocks in
east-central Scotland (Figure 14.30). Closer to the
trench in subduction zones, rocks of the high P/T facies
series are metamorphosed under conditions appropri-
ate to the blueschist facies and locally all or part of the
eclogite and the zeolite and prehnite–pumpellyite
facies. The latter two, low-grade, or “subgreenschist,”
facies are typically developed during burial metamor-
phism in forearc–accretionary complexes (Figure 14.24)
where they can also grade into greenschist-facies rocks,
as in the South Island of New Zealand (see Figure
18.30). Paralleling, juxtaposed paired belts of inter-
mediate-P and high-P/T metamorphism have been
found in a number of orogens around the margin of
the Pacific Ocean.
14.2.9 Metamorphic Field Gradients and P T t Paths
A useful concept related to facies series is that of a
metamorphic field gradient. This is a line (or a radial
sector) on a P–T diagram showing how interpreted
438 Igneous and Metamorphic Petrology
Metamorphic Rocks and Metamorphism: An Overview
439
conditions of peak metamorphism vary across a ter-
rane, such as in the intermediate-P and the high-P/T
facies series (Figure 14.34). Other gradients are illus-
trated in Turner (1981, Figure 11-5). A field gradient
does not represent a single geotherm (geothermal gra-
dient) that prevailed in the crust during metamorphism
because geotherms actually vary from place to place in
a regional terrane that develops in an orogen. Geotherms
are greater near heat sources such as underplatings of
mantle-derived magma and less near heat sinks such
as cool subducting old lithosphere. In addition, the
metamorphic field gradient reflects how and when
the terrane was uplifted, eroded, and exposed after
metamorphism. Deep parts of orogens tend to be
isostatically uplifted most, so that once more or less
horizontal isograd surfaces (Figure 14.24) are tilted;
after erosion they obliquely intersect the ground sur-
face (Figure 14.30). Finally, field gradients are a record
of deepening P–T conditions in the crust but not in a
true vertical direction, as for an actual geotherm. To
further understand what a metamorphic field gradient
represents it is necessary to introduce P–T–t paths.
The complete set of P–T conditions that a rock
now exposed in an outcrop at the surface experienced
during its time-dependent (t-dependent) metamorphic
history—from initial burial and heating to eventual up-
lift, cooling, and exposure—is embodied in its P–T–t
path. This path has an initial prograde path segment
typically involving devolatilization reactions followed
by a return retrograde segment, where equilibrating
mineral reactions are typically limited or nonexistent.
Determination of the specific P–T–t path for a
rock or set of rocks characterizing an entire terrane is
obtained in two ways. These are considered in more
detail in Section 18.1 but a brief summary suffices
here. In the “reverse” method, the petrologist uses
P (kbar)
20
16
12
8
4
0
Depth (km)
70
60
50
40
20
10
0 200 400 600
T (°C)
800 1000 1200
30
0
Base of continental crust
ECLOGITE
Pl out
Jd+Qtz
BLUESCHIST Ab
GREENSCHIST
3
2
1
4
5
6
PRH-PMP
ZEOLITE
Ky Sil
And
AMPHIBOLITE
s
o
l
i
d
u
s
G
r
a
n
i
t
e
Basalt
solidi
GRANULITE
PX HRFLS SANID
14.33 Metamorphic facies (capital letters) and a few of the very
numerous defining mineral reactions, dashed where approxim-
ately located. Data from Spear (1993) and Bucher and Frey
(1994). Compare with Figures 14.1 and 14.24. P–T boundaries
between facies are not precisely defined as shown by the fact
that some reaction lines pass through a facies field and because
reactions involving solid solutions are smeared over a range
of P and T, rather than being a univariant reaction line as for
end-member compositions. “Pl out” line indicates maximum
P stability of plagioclase in mafic rocks. Solidi of basalt and
granite from Figure 14.1. See Table 14.1 for mineral abbrevia-
tions. Numbers on curves refer to following reactions, where
reactant phases on left are stable at lower T:
1. Analcite quartz albite H
2
O
2. 2 lawsonite 5 glaucophane tremolite 10 albite
2 chlorite
3. 6 tremolite 50 albite 9 chlorite 25 glaucophane
6 zoisite 7 quartz 14 H
2
O. Also 13 albite 3 chlo-
rite quartz 5 glaucophane 3 paragonite 4 H
2
O
4. 25 pumpellyite 2 chlorite 29 quartz 7 tremolite
43 zoisite 67 H
2
O
5. 4 chlorite 18 zoisite 21 quartz 5 Al-amphibole
26 anorthite 20 H
2
O. Also 7 chlorite 13 tremolite
12 zoisite 14 quartz 25 Al-amphibole 22 H
2
O.
Also albite tremolite Al-amphibole quartz (ap-
proximate reaction)
6. Hornblende clinopyroxene orthopyroxene Ca-
plagioclase H
2
O (approximate reaction)
P (kbar)
Depth (km)
20
18
16
14
12
10
8
6
4
2
0
70
50
40
30
High-P/T
Intermediate-P
Low-P
Base of continental crust
60
20
10
0
0 200 400 600 800 1000 1200
T (°C)
14.34 Representative metamorphic field gradients in regional
metamorphic terranes. See also Turner (1981, Figure 11.5).
Mineral reaction lines from Figure 14.33 are shown for refer-
ence so that corresponding facies series can be read. High-P/T
field gradient and facies series (typically blueschist and
prehnite–pumpellyite with or without zeolite and eclogite fa-
cies) form in near-trench accretionary wedges (Figure 14.24).
Intermediate-P field gradient and facies series (typically green-
schist and amphibolite with or without zeolite, prehnite–
pumpellyite, and granulite facies) in Barrovian zones form
farther inland more or less coincident with the magmatic arc
along the central axis of orogens. Low-P field gradient and
facies series (typically greenschist and amphibolite facies) in
Buchan zones occur in a few orogens where pressures are
unusually low for the prevailing temperatures.
Table 14.2. Mineral assemblages in the conventionally recognized metamorphic facies listed according to major chemical groups of metamorphic rocks
F
ACIES
Zeolite
Prehnite–
pumpellyite
Greenschist
Amphibolite
Granulite
Blueschist
Eclogite
Pyroxene
hornfels
Sanidinite
From Turner (1981), Yardley (1989), and Blatt and Tracy (1996). Stable diagnostic mineral assemblages are denoted by “” in listing, possible additional stable phases in assemblage by ; otherwise the
list only indicates likely minerals. Quartzo-feldspathic rocks include silicic magmatic rock protoliths as well as feldspathic/lithic sandstones and arc-related volcaniclastic deposits.
*Plagioclase is usually more calcic than about An
24
but a stably coexisting albite (An
5
) has also been reported because of the existence of a narrow solvus in sodic plagioclases.
M
AFIC ROCKS
(
ALL ASSEMBLAGES
F
E
–T
I OXIDES
)
Laumontite; heulandite
Prehnite pumpellyite chlorite
albite epidote; actinolite takes place
of prehnite at higher T; lawsonite
albite chlorite occurs at higher P
Albite chlorite actinolite
epidote titanite quartz white
mica calcite; stilpnomelane is
widespread at lower T and biotite at
higher T where hornblende also occurs
Hornblende oligoclase
*
epidote
almandine-rich garnet titanite
biotite chlorite quartz
Plagioclase clinopyroxene
orthopyroxene hornblende olivine
(low P); Plagioclase clinopyroxene
orthopyroxene garnet
hornblende (medium P); Plagioclase
clinopyroxene garnet quartz
hornblende (high P)
Glaucophane lawsonite aragonite
jadeitic clinopyroxene chlorite
albite titanite pumpellyite
actinolite or hornblende
stilpnomelane epidote garnet
Omphacite pyrope-rich garnet
kyanite rutile quartz or coesite
Essentially as granulite facies
Near subsolidus basalt assemblage
P
ELITIC ROCKS
(
ALL ASSEMBLAGES
QUARTZ
F
E
–T
I OXIDES
)
Mixed-layer clays
White mica/illite chlorite albite
stilpnomelane
Muscovite chlorite albite paragonite
graphite rutile magnetite hematite
carbonate epidote K-feldspar Fe-Ti
oxides stilpnomelane (low Al) pyrophyllite
chloritoid (the latter two in high-Al rocks);
biotite in the biotite zone; almandine-rich
garnet in the garnet zone
Biotite muscovite plagioclase almandine-
rich garnet cordierite aluminosilicate
chlorite alkali feldspar magnetite graphite;
staurolite in the staurolite zone; kyanite in the
kyanite zone; sillimanite in the sillimanite zone
Alkali feldspar plagioclase scapolite
cordierite garnet rutile ilmenite
magnetite graphite olivine corundum
spinel kyanite (high P) sillimanite (moderate
P); Orthopyroxene sapphirine (high T )
Glaucophane lawsonite albite phengite
paragonite garnet chlorite epidote
kyanite chloritoid titanite
Pyrope-rich garnet carpholite phengite
chloritoid chlorite kyanite coesite
Essentially as granulite facies but andalusite is
typical aluminosilicate; cordierite biotite tends
to be more stable than almandine garnet, except
where Fe
2
/Mg is high; silica deficient rocks
contain spinel, corundum, and alkali feldspar in
lieu of an aluminosilicate
Sanidine, tridymite, cordierite, mullite
(3Al
2
O
3
·2H
2
O), glass, clinopyroxene, spinel,
corundum (in silica-poor rocks)
Q
UARTZO
-
FELDSPATHIC ROCKS
(
ALL ASSEMBLAGES
QUARTZ
F
E
–T
I OXIDES
)
Heulandite analcite and laumontite albite
(wairakite may occur in lieu of heulandite; some
rocks contain chlorite)
Albite chlorite pumpellyite prehnite
stilpnomelane white mica titanite epidote
carbonate; actinolite takes place of prehnite
at higher T; lawsonite is stable at higher P
Albite epidote muscovite chlorite
titanite stilpnomelane actinolite; biotite is
stable at higher T
Plagioclase alkali feldspar biotite
muscovite hornblende
Alkali feldspar plagioclase garnet
kyanite orthopyroxene clinopyroxene
hornblende magnetite ilmenite
Jadeitic clinopyroxene lawsonite
muscovite chlorite titanite glaucophane
Essentially as granulite facies
Sanidine, tridymite, cordierite, glass,
clinopyroxene
C
ALCAREOUS AND
CALC
–
SILICATE ROCKS
Calcite, dolomite, quartz, talc, clays
Calcite, prehnite, albite, quartz,
chlorite
Calcite, dolomite, quartz, muscovite,
albite, K-feldspar, chlorite, zoisite
Calcite, dolomite, quartz, diopside,
tremolite, forsterite, grossular,
anorthite, hornblende, clinozoisite
Calcite, dolomite, quartz, diopside,
scapolite, anorthite, forsterite,
wollastonite, graphite
Aragonite, calcite
Essentially as granulite facies
vesuvianite wollastonite
Anorthite wollastonite diopside
in silica-rich rocks; calcite, wollas-
tonite, melilite (Ca
2
MgSi
2
O
7
), larnite
(Ca
2
SiO
4
), merwinite (Ca
3
MgSi
2
O
8
),
monticellite in silica-poor rocks
preserved, “frozen in” disequilibrium attributes of the
rock and works backwards, estimating the conditions
of its metamorphism. Rarely has a rock perfectly equi-
librated to the last peak metamorphic condition so that
all previous equilibrations along its prograde path
are totally erased; some rocks also record retrograde
adjustments. P–T conditions through time can be
inferred from, for example, partial overprints of one
mineral assemblage on another and arrested reaction
textures in which reactant phase grains are incom-
pletely converted to products; armoured relics and
other mineral inclusions in poikiloblasts (Section
14.1.2) are a rich source of information. Mineral
thermobarometers can evaluate P–T histories of com-
positionally zoned porphyroblasts. Isotopic dating
methods furnish chronologic constaints. Data obtained
by these reverse techniques are commonly compared
with “forward” theoretical heat-flow models in which
radioactive, magmatic, and crustal thickening heat
sources and rates of conductive and advective heat
transfer and uplift and erosion are parameters.
How T and P change along metamorphic paths
through time is infinitely variable depending on the
details of heat and mass transfer. A hypothetical but
common type of clockwise P–T–t path in orogens is
shown in Figure 14.35. Rock is progressively buried
beneath younger deposits or by overthrust sheets, in
both cases thickening the crust and causing both T and
P to increase in the rock. The increase in T occurs
slowly because of slow rates of heat conduction; the
actual rate of increase with time is a function of the rate
of burial, heat production in the crust by decay of ra-
dioactive K, Th, and U (greatest in granite and shales),
and the pre-existing geothermal gradient. The increase
in P, in contrast, is instantaneous and depends on the
rate of burial. After reaching their respective peak
values, P and T subsequently decline as the rock mass
is uplifted (perhaps due to isostasy), eroded, and
cooled before finally being exhumed where it can be
studied in outcrops by the petrologist. Note that the
maximum, or peak, T lags behind that of the peak P
because P adjusts instantaneously during burial,
whereas it takes some considerable time for the rock
buried to some specific depth to heat up to its peak T.
In fact, the rock can still be heating and attain its
“peak” metamorphic T of equilibration during decom-
pression after uplift has started.
A question can now be posed that the astute reader
might have already considered: does a rock experienc-
ing prograde metamorphism recrystallize and equili-
brate incrementally along the same path as represented
by the metamorphic field gradient? Has a rock in the
sillimanite Barrovian zone progressed in time up
through each lower grade zone along the field gradient
for the entire terrane? The negative answer can be seen
by careful study of the interrelations among geothermal
Metamorphic Rocks and Metamorphism: An Overview
441
gradient, metamorphic field gradient, and P–T–t path
in Figure 14.36.
14.3 WHY STUDY METAMORPHIC
ROCKS? METAMORPHIC
PETROLOGY AND CONTINENTAL
EVOLUTION AND TECTONICS
Metamorphic rocks provide valuable, and even unique,
insights into orogenesis, tectonic evolution of the con-
tinental crust, and the evolution of the entire planet
throughout its recorded history in rocks as old as 4
Ga (Chapter 19). Protoliths and field relations tell
about the character of the primary rock—whether it
was deposited on the surface of the Earth, or repres-
ents a segment of the oceanic lithosphere accreted
onto land as an ophiolite, or innumerable other pos-
sibilities. Field relations, fabrics, and mineral com-
positions are records of episodes of deformation and
recrystallization driven by burial, tectonic convergence
and thickening of crust during orogenesis, and sub-
sequent uplift. P–T–t paths can be constructed with
constraining chronologic data from isotopic analyses.
Thus, attributes of metamorphic rocks and the terranes
T (°C)
P (kbar)
800
600
400
200
0
P (kbar)
12
8
4
0
12
8
4
0
400
T (°C)
600 8002000
t
1
t
PP
T
PP
T
PT
t
PT
t
2
Time
(a)
(b)
Compression
Heating
Peak P
Peak T
Decompression
UPLIFTBURIAL
Cooling
Prograde
Retrograde
t
1
t
2
14.35 Hypothetical P–T–t path. (a) Variation in P and T through
time in a parcel of rock material that was initially deposited at
surface of Earth during progressive burial, followed by uplift
and erosion back to the surface. Note that the peak tempera-
ture, T
PT
, lags well behind the peak pressure, T
PP
, because of
the slow conduction of heat in contrast to the instantaneous
rise in P during burial. (b) P–T–t variations in top diagram are
combined into a closed clockwise P–T–t path loop. Prograde
path occurs at increasing T, mostly during progressive burial.
Subsequent retrograde path follows decreasing T during uplift
and erosion.
442 Igneous and Metamorphic Petrology
in which they are exposed testify of mountain- and
continent-building phenomena even though lofty topo-
graphic edifices have long since been eroded away.
During the last two decades of the twentieth cen-
tury, metamorphic petrologists made astounding dis-
coveries of coesite- and diamond-bearing rocks in
plate-collision belts. After initial deposition on the
surface of the Earth as sediment, the rocks must have
been subsequently buried to and subsequently rapidly
exhumed from depths of at least 100–120 km. These
discoveries of ultrahigh-P metamorphic rocks that
equilibrated at only modest temperatures spawned an
exciting new field of geology that challenged conven-
tional tectonic thinking of what happens where plates
converge and how continents evolve. Ultrahigh-P
metamorphic rocks test uniformitarian tectonic con-
cepts because the thickest known crusts today are only
60–80 km in the high central Andes subduction zone
and the Tibetan Plateau–Himalaya Mountains collision
zone. How could relatively low density continental
rocks be depressed to such depths, displacing denser
mantle rocks? What tectonic processes exhumed the
“refrigerated” ultrahigh-P rocks for petrologic exami-
nation in surface outcrops?
Another significant result of research on metamor-
phic rocks is the extent to which orogenic terranes
have been subjected to successive overprinting events.
In contrast to a simple one-stage P–T–t path for some
geologically recent orogens (Figure 14.35), older ones,
especially in the Precambrian, can be quite tangled;
multiple overprints of metamorphism and deformation
can be recorded (Figure 14.37). In some places, the ef-
fects of an overprint can be walked through in the field
so that the petrologist can obtain a clear understanding
of the metamorphic path (Figure 14.38). However, had
the transition from the initial state not been exposed to
view but only the final pervasive product, elucidating
the nature of the protolith and metamorphic path
would be challenging and likely result in ambiguous
interpretations. Figure 14.39 illustrates another pos-
sible consequence of polymorphism that, although
hypothetical, emphasizes the importance of an open
mind in interpreting metamorphic bodies.
SUMMARY
The metamorphic path of changing intensive variables
and states of stress causing reconstitution of solid rock
during metamorphism is infinitely variable. Insights
into the nature of the path are provided by relict and
newly imposed fabrics, compositional attributes, and
field relations of rocks in metamorphic terranes.
Protoliths of metamorphic rocks include virtually all
magmatic and sedimentary rocks as well as previously
recrystallized and deformed metamorphic rocks that are
widespread in long-lived orogenic belts. The identity of
the protolith is revealed in preserved relict fabrics and
mineral constituents—particularly in fine-grained, low-
grade rocks—and by the bulk chemical and mineral
composition, unless the rock has been significantly
metasomatized.
Recrystallization of existing stable phases and crys-
tallization of newly stabilized minerals during solid-
state metamorphism yield distinctly different fabrics
from those created in magmatic systems by crystal-
lization from melts. Where hydrostatic states of
stress prevail, such as in many thermal contact meta-
Pressure and depth
T
Metamorphic field gradient
Geotherm
Peak T
P-T-t path
14.36 Meaning and derivation of a metamorphic field gradient.
Compare Figure 14.35. The gradient line represents the locus
of peak metamorphic temperatures (circles) at their prevailing
P for several parts of a metamorphic terrane that are inferred
from the mineral assemblages in rocks from each part. A com-
plete P–T–t path (full line) for one part is shown together with
partial paths (long dashed lines) from other parts of the same
metamorphic terrane. Note that a P–T–t path follows the
initial geotherm in its part of the terrane only during the early
phase of prograde metamorphism. In other words, the meta-
morphic field gradient is oblique to a family of geotherm lines
for the terrane.
P
T
D
1
D
2
M
1
D
5
M
4
D
4
M
3
D
3
M
2
14.37 Complex hypothetical P–T–t path for a metamorphic terrane
in an orogen subjected to polymetamorphism with multiple
episodes of deformation (D
1
, D
2
,...) and recrystallization
(M
1
, M
2
,...).
Metamorphic Rocks and Metamorphism: An Overview
443
phic minerals, such as kyanite and staurolite, together
with other aluminosilicates, especially garnet, commonly
nucleate sluggishly and form large porphyroblasts that
may be crowded with smaller mineral inclusions in
poikiloblasts.
Anisotropic tectonite fabrics created under non-
hydrostatic states of stress are typical of regional
metamorphic terranes in orogenic belts. Ductile flow
commonly accompanied by solid state crystallization
produces fine-grained mylonites in high-strain-rate
(a)
(b) (c)
14.38 Progressive ductile deformation and recrystallization of Archean (3.1–3.4 Ga) Ameralik basalt dikes and their host 3.8 Ga Itsaq (Amît-
soq) gneisses in West Greenland. (a) Undeformed dike in granodioritic augen gneiss. (b) Ductilely deformed dike and host rock. The
feldspar augen (see Figure 15.7) have been flattened so that the foliation of the gneiss is more pronounced; this foliation and the dike
have been folded. (c) More intensely deformed dike and host rock. Slightly less ductile dike rock, here a hornblende plagioclase
amphibolite, has been dismembered into crude, sausage-like boudins (see Figure 15.8); these are encased in a swirled gneissic matrix in
which the original feldspar augen are intensely flattened lenses. Some coarse-grained leucocratic segregations are visible near the black
boudins. Geological Survey of Greenland photographs courtesy of Victor R. McGregor. Reproduced with permission from McGregor VR.
1973. The early Precambrian gneisses of the Godthåb district, west Greenland. Phil. Trans. Royal Soc. London, Series A, 273: 343–58.
morphic aureoles, the imposed isotropic fabric is a
response to mutual competition for common space
during grain growth and equilibration to the most
stable grain boundary configuration of lowest possible
surface energy. Larger grains of a given phase are
more stable than smaller and grains of denser phases
higher in the crystalloblastic series tend to be more
euhedral in contrast to the more common carbonate
and felsic minerals that are typically polygonal with
triple boundary angles near 120°. Uniquely metamor-
444 Igneous and Metamorphic Petrology
Flow of heat and advection of hydrothermal solu-
tions in regional metamorphic terranes and contact
aureoles produce thermal and chemical gradients that
are responsible for zones of contrasting mineral assem-
blages. Successive zones are delineated by isograds
marked by the first appearance upgrade of a critical in-
dex mineral that persists through the zone of the same
name and commonly into still higher grade zones. In
regional orogenic terranes, progressive metamorphism
of pelitic rocks along intermediate-P metamorphic field
gradients creates the Barrovian zonal sequence chlo-
rite, biotite, garnet, staurolite, kyanite, and sillimanite.
In some terranes, peak metamorphism is associated
with partial melting and development of migmatite in
deeper levels of the continental crust. Consideration
of mineral assemblages in all other protoliths in these
terranes shows metamorphism corresponds to the
greenschist–amphibolite facies series. Contact aureoles
and Buchan regional zones that contain andalusite
and cordierite in pelites were metamorphosed along
lower-P field gradients.
Nearer the trench in active subduction zones, forearc
rocks are subjected to high P/T metamorphic field gra-
dients that create regional terranes of the blueschist and
locally prehnite–pumpellyite, zeolite, and eclogite facies.
The tectonic and thermal evolution of deeply
eroded ancient orogens is recorded in the fabric and
mineralogical features of polymetamorphic rocks in
regional terranes.
CRITICAL THINKING QUESTIONS
14.1 What attributes of metamorphic rocks
distinguish them from magmatic?
shear zones and coarser foliated and lineated tectonites
throughout regional terranes. Metamorphic differenti-
ation can create compositional layering where none
existed in the protolith. Pre-existing layers can be
transposed by ductile deformation into new layers.
H
2
O–CO
2
fluids carrying dissolved minerals react
with and replace existing minerals in rocks through
which they advect in the process of metasomatism.
These percolating hydrothermal solutions are common
in contact aureoles surrounding granitic intrusions
where carbonate rocks are reconstituted into calc–
silicate skarns. Colossal volumes of advecting seawater
in the global ocean ridge system variably metasomatize
and hydrate hot basaltic rocks that are subsequently
subducted to generate arc magmas.
Metamorphic paths in regional terranes of orogens
are typically clockwise in P–T space. Equilibrating
prograde reactions at increasing T and P along the
metamorphic path progressively eliminate lower grade
mineral assemblages while liberating volatiles up to about
the peak value of T at the prevailing P and volatile
condition. This peak equilibrium mineral assemblage
remains essentially unchanged in most metamorphic
rocks, with decreasing T and P along the retrograde
segment of the metamorphic path where kinetic rates
rapidly decline so that equilibrating mineral reactions
are incomplete, at best. Only locally where fluids
gain access to the retrograding rock, typically along
fractures, and where localized ductile shear “catalyze”
recrystallization do rocks equilibrate mineralogically.
Where volatiles are absent and no deformation occurs,
high-grade volatile-poor mineral assemblages can
remain unchanged indefinitely in a metastable state,
even though temperatures are elevated.
14.39 Hypothetical but not impossible polymetamorphic overprinting of an apparently simple compositionally layered rock body (upper left)
that may, in fact, have experienced a complex metamorphic history!
Folding
Felsic diking and
migmatization
Mafic diking
Recrystallization,
metamorphic
differentiation
Flattening and
boudinage
Ductile shearing
and transposition
14.2 Contrast the factors controlling grain size in
metamorphic and magmatic rocks. Grain
shape.
14.3 Metamorphism of a shale, a graywacke, or a
basalt yields rocks of varied mineral composi-
tion depending on the metamorphic grade. In
contrast, different grades of metamorphism
of a pure limestone or of a pure quartz
sandstone yield only marble and quartzite,
respectively. Why?
14.4 Is the meta-andesite in Figure 14.3b a result
of prograde or retrograde metamorphism?
Discuss.
14.5 Discuss the different ways that compositional
layering can originate in metamorphic bodies.
14.6 How would you distinguish between a relict
plagioclase phenocryst in a metamorphosed
granodiorite, a plagioclase porphyroclast in a
mylonite, and a plagioclase porphyroblast in a
granoblastic rock?
14.7 Why are migmatites not found in metamorphic
terranes formed near the trench in
subduction zones (high-P/T facies series)?
14.8 What evidence might be found in a single hand
sample for prograde metamorphism? For retro-
grade metamorphism? Cite examples.
14.9 Cite mineralogical evidence that the rock in
Figure 14.6d is higher grade than in Figure
14.6c and that it is higher grade than in
Figure 14.6b?
14.10 Why is there no single metamorphic field
gradient for all metamorphic terranes?
14.11 Cite three examples of polymetamorphism
from the illustrations of metamorphic rocks in
this chapter.
14.12 What factors govern the width of a contact
metamorphic aureole at a particular level of
erosion?
14.13 If a metasomatic rock has no relict fabric,
how can the nature of the original rock be
determined?
14.14 The glacier in Figure 14.11b has the appear-
ance of a metamorphic tectonite yet the ice on
a microscopic scale in Figure 14.11a does not.
Why?
14.15 Contrast the concepts of metamorphic facies
and metamorphic zones.
14.16 Where would rocks of the sanidinite facies be
expected to form?
14.17 Propose reasons why protoliths of sedimentary
evaporites, mostly gypsum and halite, and of
highly alkaline rocks, such as kimberlites, are
essentially nonexistent in metamorphic terranes.
14.18 Why are anisotropic fabrics more common in
metamorphic rocks than magmatic?
14.19 Discuss the nature of the metamorphic differ-
entiation (differential movement of ions) in the
growth of the garnet porphyroblasts in the rock
shown in Figure 14.8. Are any assumptions
necessary?
14.20 The lowest grade rocks in terranes of inter-
mediate facies series are typically of the
greenschist facies, rather than the zeolite and
prehnite–pumpellyite facies as in near trench
terranes. Propose two reasons why this might
be the case.
14.21 From Table 14.2, track critical changes in
diagnostic mineral assemblages in intermediate
facies series in the most common magmatic
and sedimentary protoliths. Track changes in
metabasites in the high-P/T facies series.
14.22 Contrast the fabrics of cataclastic, mylonitic,
and granoblastic rocks. Describe grain size and
shape, preferred dimensional orientation of
grains, and presence and character of frag-
ments. What are the contrasting processes and
conditions of production of these fabrics?
PROBLEMS
14.1 How much linear strain (Equation 8.2) is
represented in a deformed cobble from the
Raft River Mountains metaconglomerate
(Figure 14.13) that is 4 cm thick (c, the mini-
mum axis), 12 cm wide (b, the intermediate
axis), and 120 cm long (a, the maximum axis)?
What two assumptions must be made to answer
this question? Hint: the volume of an ellipsoid is
abc/6.
14.2 A skarn is composed of the equilibrium mineral
assemblage pyrite garnet diopside calcite.
Sketch mutual grain shapes in the aggregate,
making sure that each phase contacts every
other phase somewhere.
14.3 Make the simplistic assumption that Barrovian
isograds in northern Scotland correspond to
paleo-isothermal surfaces that intersect the
present erosion surface. Sketch a cross-section
from Glasgow on the south to the northeast
end of the exposed Great Glen Fault showing
how postmetamorphic uplift and erosion of
the Caledonide Orogen could have created the
present configuration of the isograds in Fig-
ure 14.30.
14.4 Write balanced stoichimetric equations showing
the chemical equivalence of Orijärvi and Oslo
mineral assemblages listed in Section 14.2.7.
Metamorphic Rocks and Metamorphism: An Overview
445
14.6 It is usually believed that the sillimanite zone is
of higher grade than the kyanite zone in Bar-
rovian terranes. In a P–T diagram based on the
stability fields of these two Al
2
SiO
5
polymorphs
(Figure 14.31), show how a orogenic clockwise
P–T–t path could make kyanite a retrograde
product.
446 Igneous and Metamorphic Petrology
14.5 On photocopies of Figures 14.1, 14.33, and
14.34 draw lines representing 10, 50, and
100°C/km geotherms. Relate these reference
geotherms to facies series and explain them
in terms of the geologic environments of
metamorphism.