-
Geological Society of America Bulletin
doi: 10.1130/B30446.1 2011;123;387-411Geological Society of
America Bulletin
Yildirim Dilek and Harald Furnes
fingerprinting of ancient oceanic lithosphereOphiolite genesis
and global tectonics: Geochemical and tectonic
Email alerting servicesarticles cite this article
to receive free e-mail alerts when
newwww.gsapubs.org/cgi/alertsclick
SubscribeAmerica Bulletin
to subscribe to Geological Society
ofwww.gsapubs.org/subscriptions/click
Permission request to contact
GSAhttp://www.geosociety.org/pubs/copyrt.htm#gsaclick
official positions of the Society.citizenship, gender, religion,
or political viewpoint. Opinions presented in this publication do
not reflectpresentation of diverse opinions and positions by
scientists worldwide, regardless of their race, includes a
reference to the article's full citation. GSA provides this and
other forums for thethe abstracts only of their articles on their
own or their organization's Web site providing the posting to
further education and science. This file may not be posted to any
Web site, but authors may postworks and to make unlimited copies of
items in GSA's journals for noncommercial use in classrooms
requests to GSA, to use a single figure, a single table, and/or a
brief paragraph of text in subsequenttheir employment. Individual
scientists are hereby granted permission, without fees or further
Copyright not claimed on content prepared wholly by U.S. government
employees within scope of
Notes
2011 Geological Society of America
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
INVITED REVIEW ARTICLEINVITED REVIEW ARTICLEINVITED REVIEW
ARTICLEINVITEDNVITED R REVIEWEVIEW A ARTICLERTICLE
ABSTRACT
Ophiolites, and discussions on their origin and signifi cance in
Earths history, have been instrumental in the formulation, testing,
and establishment of hypotheses and theories in earth sciences. The
defi nition, tectonic ori-gin, and emplacement mechanisms of
ophio-lites have been the subject of a dynamic and continually
evolving concept since the nine-teenth century. Here, we present a
review of these ideas as well as a new classifi cation of
ophiolites, incorporating the diversity in their structural
architecture and geochemi-cal signatures that results from
variations in petrological, geochemical, and tectonic processes
during formation in different geo-dynamic settings. We defi ne
ophiolites as suites of temporally and spatially associated
ultramafi c to felsic rocks related to separate melting episodes
and processes of magmatic differentiation in particular tectonic
envi-ronments. Their geochemical characteris-tics, internal
structure, and thickness vary with spreading rate, proximity to
plumes or trenches, mantle temperature, mantle fertility, and the
availability of fl uids. Subduction-related ophiolites include
suprasubduction-zone and volcanic-arc types, the evolution of which
is governed by slab dehydration and accompanying metasomatism of
the mantle, melting of the subducting sediments, and repeated
episodes of partial melting of meta-somatized peridotites.
Subduction-unrelated ophiolites include continental-margin,
mid-ocean-ridge (plume-proximal, plume-distal, and trench-distal),
and plume-type (plume-proximal ridge and oceanic plateau)
ophio-
lites that generally have mid-ocean-ridge basalt (MORB)
compositions. Subduction-related lithosphere and ophiolites develop
during the closure of ocean basins, whereas subduction-unrelated
types evolve during rift drift and seafl oor spreading. The peak
times of ophiolite genesis and emplacement in Earth history
coincided with collisional events leading to the construction of
super-continents, continental breakup, and plume-related
supermagmatic events. Geochemical and tectonic fi ngerprinting of
Phanerozoic ophiolites within the framework of this new ophiolite
classifi cation is an effective tool for identifi cation of the
geodynamic settings of oceanic crust formation in Earth history,
and it can be extended into Precambrian green-stone belts in order
to investigate the ways in which oceanic crust formed in the
Archean.
INTRODUCTION
Ophiolites represent fragments of upper mantle and oceanic crust
(Dewey and Bird, 1971; Coleman, 1977; Nicolas, 1989) that were
incorporated into continental margins during continent-continent
and arc-continent collisions (Dilek and Flower, 2003), ridge-trench
inter-actions (Cloos, 1993; Lagabrielle et al., 2000), and/or
subduction-accretion events (Cawood et al., 2009). They are
generally found along suture zones in both collisional-type (i.e.,
Alpine, Himalayan, Appalachian) and accretionary-type (i.e., North
American Cordilleran) orogenic belts (Fig. 1) that mark major
boundaries be-tween amalgamated plates or accreted terranes (Lister
and Forster, 2009). The geological rec-ord of the evolution of
ocean basins from the rift-drift and seafl oor spreading stages to
the ini-tiation of subduction and fi nal closure (the Wil-
son cycle) is well preserved in most orogenic belts. Magmatism
during each of these phases produces spatially and temporally
associated, mafi c-ultramafi c to highly evolved rock assem-blages.
These rock units, which have varying in-ternal structures,
geochemical affi nities, and age ranges, and originally formed in
different geo-dynamic settings, constitute discrete ophiolite
complexes and can become tectonically juxta-posed in collision
zones (Dilek, 2003).
In the Penrose defi nition (Anonymous, 1972, p. 24), an
ophiolite is described as a distinctive assemblage of mafi c to
ultramafi c rocks that includes, from bottom to top, tectonized
perido-tites, cumulate peridotites, and pyroxenites over-lain by
layered gabbros, sheeted basaltic dikes, a volcanic sequence, and a
sedimentary cover; an ophiolite may be incomplete, tectonically
dismembered, or metamorphosed. This original Penrose defi nition of
ophiolites (Anonymous, 1972) is highly restrictive and does not
refl ect the actual heterogeneity in ophiolite composition and
occurrence, and therefore a more determin-istic approach to defi
ning ophiolites and their ig-neous evolution is needed. In this
paper, we fi rst review the evolution of the ophiolite concept
before and after the formal Penrose defi nition, and we redefi ne
an ophiolite in light of recent observations and diverse data sets
from ophio-lites worldwide. We outline the signifi cance of
ophiolite pulses in Earth history within a global tectonic
framework and introduce a new and more comprehensive classifi
cation of ophiolites based on their distinctive internal
structures, geo-chemical signatures, and regional tectonics. We
then present petrogenetic models for the forma-tion of different
types of ophiolites and discuss the implications of this new
ophiolite classifi ca-tion for the origin of Precambrian oceanic
crust, particularly for some Archean greenstone belts.
For permission to copy, contact [email protected] 2011
Geological Society of America
Ophiolite genesis and global tectonics: Geochemical and tectonic
fi ngerprinting of ancient oceanic lithosphere
Yildirim Dilek1, and Harald Furnes21Department of Geology,
Shideler Hall, Miami University, Oxford, Ohio 45056, USA, and
Faculty of Earth Sciences, China University of Geosciences at
Wuhan, Wuhan 430074, Hubei Province, China2Department of Earth
Science & Centre for Geobiology, University of Bergen, Bergen
5007, Norway
387
GSA Bulletin; March/April 2011; v. 123; no. 3/4; p. 387411; doi:
10.1130/B30446.1; 12 fi gures; 2 tables, Data Repository item
2011131.
E-mail: [email protected].
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
388 Geological Society of America Bulletin, March/April 2011
HISTORICAL BACKGROUND AND NEW DEFINITION OF OPHIOLITES
Early Ideas and Evolving Ophiolite Concept
The term ophiolite was fi rst used in 1813 by a French
mineralogist, Alexandre Brongniart (17701847), in reference to
serpentinites in mlanges; he subsequently redefi ned his defi
ni-
tion of an ophiolite (Brongniart,1821) to include a suite of
magmatic rocks (ultramafi c rocks, gabbro, diabase, and volcanic
rocks) occurring in the Apennines. Gustav Steinmann (18561929)
elevated the ophiolite term to a new concept by defi ning
ophiolites as spatially as-sociated kindred rocks that originally
formed as in situ intrusions in axial parts of geosynclines
(Steinmann, 1927). Steinmann emphasized the common occurrence of
peridotite (serpenti-
nite), gabbro, and diabase-spilite, in association with deep-sea
sedimentary rocks in the Medi-terranean mountain chains and
interpreted the origin of these rocks as differentiated magmatic
units evolved on the ocean fl oor. He considered these rock
assemblages to have developed from a consanguineous igneous process
during the evolution of eugeosynclines. This interpretation
subsequently led to the widely known notion of the Steinmann
trinity.
mid-ocean-ridge
165E15W
7575E
Indonesian belt (Cenozoic)Western Pacific and Cordilleran belts
(Paleozoic-Tertiary)Alpine - Himalayan belt (Jurassic -
Cretaceous)Appalachian - Caledonian - Hercynian - Uralian &
Central Asian belts (early Paleozoic)Tasmanides (Paleozoic)
Sunda Trench
Figure 1. Global distribution of major Phanerozoic orogenic
belts and ophiolite age clusters on a north polar projection.
Signifi cant examples of different ophiolite types with
characteristic geochemistries are marked with symbols used in
Figure 2. Modern mid-ocean ridges and subduction zones (marked by
trenches) where contemporary oceanic lithosphere has been produced
are also depicted. The two major arc-trench rollback systems,
Izu-Bonin-Mariana and Tonga-Kermadec, are the sites of ophiolite
and volcanic-arc generation, which undergo tectonic extension and
trenchward-migrating magmatic construction. The collision zone
between the NW Australian passive margin and the Sunda arc-trench
system where the island of Timor has been emerging during the last
~5 m.y. represents the best modern analogue for ophiolite
emplacement.
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 389
Although Steinmann considered peridotite, gabbro, diabase, and
volcanic rocks in ophio-lites as comagmatic in origin, his
observation that gabbroic and diabasic rocks were intru-sive bodies
in the serpentinized peridotites is an extremely important one
because it differs from the contemporary interpretation of the
Penrose-type ophiolite. It implies that, at least in the Apennine
ophiolites, the gabbros and volcanic rocks are younger than the
peridotites. Steinmann also correctly interpreted the ophio-lites
in the Northern Apennines as thrust sheets tectonically overlying
the Tertiary sedimentary rocks in Tuscany (Steinmann, 1913). This
inter-pretation led to the discovery of allochthonous nappe
sequences in the Alpine-Apennine oro-genic system.
Thayer (1967) discussed the signifi cance of the consanguineous
relationship between ultramafic and associated mafic rocks in
alpine-type peridotites, which were defi ned by Benson (1926)
earlier, and explained how the gabbro, diabase, and other
leucocratic rocks in alpine-type peridotites could have originated
from a single primary peridotitic magma. Jackson and Thayer (1972)
subsequently dis-tinguished harzburgite-type versus lherzolite-type
alpine peridotites. In this subgrouping, the harzburgite-type
alpine peridotites represent the uppermost oceanic mantle, whereas
the less-depleted lherzolite-type alpine peridotites correspond to
the subcontinental mantle and/or to the deeper oceanic mantle,
where partial melting is much less intense. Recent studies of
ophiolites have shown that both harzburgite- and lherzolite-type
peridotites may occur in ophiolites, and that they can be used to
clas-sify ophiolite types and their inferred spread-ing rates of
formation in an oceanic setting (Ishiwatari, 1985; Boudier and
Nicolas, 1985; Nicolas and Boudier, 2003).
In his classic paper published in Crust of the Earth (Geological
Society of America Spe-cial Paper 62), Hess (1955, p. 393) stated
that Steinmanns ophiolite concept was confusing because it obscured
critical relationships of its [ophiolite] various members to the
tectonic cycle. Recognizing the importance of serpenti-nites and
alpine-type peridotites in orogeny and mountain-building episodes,
he argued that ser-pentinites and rocks of Steinmanns trinity are
common in island arcs and that island arcs rep-resent an early
stage in the development of an alpine-type of mountain system (p.
395). Hess was, therefore, advocating an island-arc origin of mafi
c-ultramafi c rock assemblages and ser-pentinized peridotites found
in orogenic belts. This was nearly 20 yr before Miyashiro (1973)
made the fi rst formal and rather controversial call on the
island-arc origin of the Troodos
ophiolite (Cyprus), connecting ophiolite genesis to
subduction-zone processes.
Hess discussed in his 1962 paper that the main oceanic crustal
layer (his layer 3) along the Mid-Atlantic Ridge was made largely
of serpentinite (his Fig. 2, p. 603; Hess 1962), and that the
seismic velocity of this layer would be highly variable, depending
on the magnitude of serpentinization of the peridotite. He
pro-posed that the interface between the oceanic crust (composed
mainly of serpentinite) and the under lying perido tite with
seismic veloci-ties of 7.4 km/s represented the Moho
discon-tinuity. Since he had interpreted serpentinites as hydrated
perido tites, Hess described the Moho beneath the Mid-Atlantic
Ridge as an altera-tion front (phase transition) rather than a
sharp boundary separating the igneous crust from the underlying
mantle (his Fig. 7, p. 612). Although we now know that oceanic
crust is not made of 70% serpentinite, marine geological and
geo-physical studies have documented that the slow-spreading
oceanic crust along the Mid-Atlantic Ridge has a highly
heterogeneous lithological composition and thickness (Dick, 1989).
For ex-ample, thin-crust domains along the ridge axis (i.e.,
magma-poor segment ends) consist of tec-tonically uplifted
ultramafi c rocks with gabbroic intrusions and a thin basaltic
cover (Cannat et al., 1995). This nonuniform thickness and the
heterogeneous lithostratigraphy of the Mid-Atlantic Ridge crust are
remarkably similar to Steinmanns description of the Ligurian
ophiol-ites in the Apennines. It also largely corresponds to Hess
characterization of oceanic crust devel-oped at the Mid-Atlantic
Ridge. This Hess-type crust differs signifi cantly from
Penrose-type oceanic crust in terms of its internal architec-ture,
as discussed in the following.
The Dutch geologist de Roever (1957) re-interpreted the
Steinmann trinity to result of mantle melting, producing the
basaltic rocks on top and the residual ultramafi c rocks at the
bottom. Subsequently, the Swiss petrolo-gist Vuagnat argued that
the peridotite massifs in ophiolites were partial melting residues
in the upper mantle (Vuagnat, 1964), because he thought that the
overwhelming abundance of ultramafi c rocks in ophiolites compared
to the small volumetric occurrence of gabbroic rocks could not
simply be explained by differentiation of submarine outpourings of
basaltic magma. It is important to note that these two papers by de
Roever (1957) and (Vuagnat, 1964) mark in the literature the
beginning of a signifi cant shift in Steinmanns cogenetic ophiolite
concept and of a new paradigm in oceanic crustal evolution.
Recognition of extensional sheeted dike complexes, the existence
of a refractory mantle unit represented by harzburgitic
peridotites
with high-temperature deformation fabrics, fossil magma chambers
in plutonic sequences, and the allochthonous nature of ophiolites
by the mid-1960s was instrumental in the formula-tion of the
ophiolite model and the ophioliteocean crust analogy within the
framework of the new plate-tectonic theory. The ophiolite suite
became an ideal analogue to explain the seismic velocity structure
of modern oceanic lithosphere, as more seismic data became
avail-able from modern ocean basins, particularly from the Pacifi c
Ocean. Combined with obser-vations from the Troodos (Cyprus) and
Semail (Oman) ophiolites in particular, the seismic velocity
structure of modern oceanic crust and its inferred layer-cake
pseudostratigraphy came to be known as the ophiolite model. This
analogy was confi rmed at the fi rst Penrose Conference on
ophiolites in 1972 (Anonymous, 1972), whereby an ideal ophiolite
sequence was defi ned to have a layer-cake pseudostratig-raphy
complete with a sheeted dike complex as a result of seafl oor
spreading. Ophiolites were interpreted to have developed mainly at
ancient mid-ocean ridges through this model. In a uniformitarian
approach, ophiolite geolo-gists then started reconstructing the
evolution of fossil oceanic lithosphere exposed on land as a
product of paleomid-ocean ridges using the ophioliteocean crust
analogy (Gass, 1968; Coleman, 1971; Moores and Vine, 1971; Cann,
2003, and references therein).
Geochemical studies challenged this view of a mid-ocean-ridge
origin of ophiolites as early as the beginning of the 1970s, and
suggested the association of magma evolution with sub-duction
zones. Miyashiro (1973, p. 218) argued that about one-third of the
analyzed rocks of the lower pillow lavas and sheeted dike rocks in
the Troodos ophiolite follows a calc-alkalic trend, suggesting that
the massif was cre-ated as a basaltic volcano in an island arc with
a relatively thin ocean-type crust rather than in a mid-oceanic
ridge. This was the fi rst formal proposal of a subduction-zone
origin of the Troodos oceanic crust that questioned the ruling
hypothesis of a mid-ocean-ridge setting of ophiolite genesis.
Miyashiros geochemical argument on the island-arc origin of the
Troodos ophiolite would start a major paradigm shift in the
ophiolite concept in the wake of the plate-tectonic revolution. The
subsequent scientifi c exchange in the form of discussions and
replies to Miyashiros 1973 paper initiated a long-last-ing debate
about the tectonic setting of ophio-lite genesis. Pearce (1975)
proposed a marginal basin origin for the Troodos massif during the
evolution of an incipient submarine island arc.
Findings from modern subduction-zone en-vironments in the
western Pacifi c prompted
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
390 Geological Society of America Bulletin, March/April 2011
researchers to consider more rigorously the evolution of
ophiolites in spreading environ-ments within the upper plate of
subduction zones (Hawkins, 1977, 2003; Pearce, 2003). This
development, which came about as a col-lective result of ophiolite
studies on land and marine geological and geophysical
investiga-tions in modern convergent margin settings in the oceans,
led to the defi nition of supra-subduction-zone ophiolites in the
early 1980s (Pearce et al., 1984). The forearc environment of the
Izu-Bonin-Mariana arc-trench system is today one of the best
studied (through deep-ocean drilling and submersible diving
surveys) and best understood modern suprasubduction zones that we
consider to be a contemporary suprasubduction-zone ophiolite
factory (Fig. 1; Stern et al., 1989; Stern and Bloomer, 1992;
Reagan et al., 2010; Dilek and Furnes, 2010). Sys tematic
petrological and geochemical in-vestigations of world ophiolites
throughout the 1980s and 1990s demonstrated the signifi cance of
subduction-zonederived fl uids and melting history in development
of ophiolitic magmas (Saunders and Tarney, 1984; Rautenschlein et
al., 1985; Hbert and Laurent, 1990; Thy and Xenophontos, 1991;
Beccaluva et al., 1994; Bdard et al., 1998; Dilek et al., 1999;
Shervais, 2000; Dilek and Flower, 2003). Forearc, embry-onic arc,
and backarc settings in suprasubduc-tion zones became the most
widely accepted tectonic environments of origin.
New Defi nition of Ophiolites
The basic tenet of the 1972 Penrose defi ni-tion is that an
ideal ophiolite has a layer-cake pseudostratigraphy with laterally
persistent and horizontal contacts. The Mohorovicic disconti-nuity
(Moho) is considered to be a petrologi-cal transition zone
separating the crustal and upper-mantle rocks that have a
melt-residua genetic relationship. Studies since 1972 have
demonstrated, however, that most ophiolites have a dynamic
evolution and display a later-ally discontinuous and vertically
heterogeneous crustal architecture and varying geochemical
characteristics due to multiple magmatic epi-sodes and different
mantle sources during their igneous evolution. The fossil Moho also
differs in character in ophiolites; in some, it represents a major
tectonic discontinuity (i.e., detachment fault), whereas in some
others, it is an altera-tion front. However, in some ophiolites it
is a nearly 1-km-thick transition zone reminiscent of the Moho in
slow-spreading young oceanic lithosphere (Dick et al., 2006). The
diversity in the architecture and geochemical fi ngerprints
observed in ophiolites refl ects differences in igneous and
tectonic processes involved in the
formation of oceanic crust in different geo-dynamic
settings.
We defi ne an ophiolite as an allochthonous fragment of
upper-mantle and oceanic crustal rocks that is tectonically
displaced from its pri-mary igneous origin of formation as a result
of plate convergence. Such a slice should include a suite of, from
bottom to top, peridotites and ultramafi c to felsic crustal
intrusive and volcanic rocks (with or without sheeted dikes) that
can be geochronologically and petrogenetically re-lated; some of
these units may be missing in in-complete ophiolites. Ophiolite
emplacement is a process that starts with displacement of oceanic
lithosphere from its primary geodynamic en-vironment and ends with
its incorporation into mountain belts during orogenesis (Coleman,
1971; Dewey, 1976; Searle and Cox, 1999; Gray et al., 2000;
Wakabayashi and Dilek, 2003). Ophiolites are commonly emplaced on a
passive continental margin (buoyant crust) and island arc or in an
accretionary complex. The mag-matic and structural architecture of
an ophio lite may refl ect a product and complex interplay of
successive melting episodes and processes of magmatic
differentiation, spreading rate and geometry, intra-oceanic
faulting, and deforma-tion associated with tectonic extension,
prox-imity to plumes or trenches, mantle temperature and fertility,
and the availability of fl uids during its primary igneous
evolution. Some ophiolites are stratigraphically overlain by
pelagic (chert or limestone) and/or Fe-Mnrich hydrother-mal
sedimentary rocks and are underlain by amphibolite-greenschist
rocks related to their tectonic displacement and emplacement.
OPHIOLITE PULSES AND GLOBAL TECTONICS
The distribution of ophiolites in orogenic belts shows spatial
and temporal patterns (Fig. 1), and the clusters of ophiolites with
particular age ranges in different orogenic belts mark clear
pulses, refl ecting peak times of ophiolite genesis and emplacement
in Earth history (Fig. 2). Some of the main ophiolite pulses
overlap in time with major orogenic events that led to the
construction of supercontinents. Examples include the Fama-tinian
(Fmt) and Caledonian (Cld; Baltica- Lau-rentia collision) orogens
in the early Paleozoic, which collectively formed the Gondwana and
Laurasia supercontinents, and the Appalachian-Hercynian (Ap-Hy) and
Altaid-Uralian (Al-Ur) orogens later in the Paleo zoic, which built
the Pangean supercontinent (Fig. 2; Moores et al., 2000). The
sequential collisions of India (In-Eu) and Arabia (Ar-Eu) with
Eurasia during the Neogene, after the emplacement of Neotethyan
ophio lites and elimination of the Neotethyan sea-
ways by subduction, are part of the current as-sembly of a new
supercontinent that has been taking place since the Paleogene.
Paleozoic ophiolites in the Appalachian-Caledonian orogenic
belts (Fig. 1) developed in the Iapetus Ocean and its seaways
between North America and Baltica-Avalonia (van Staal et al., 2009,
and references therein). Ophiolites in Iberia, central Europe, and
northwestern Africa evolved in the Rheic Ocean between
Baltica-Avalonia and Gondwana continental masses (Nance et al.,
2010; Murphy et al., 2010, and references therein). The Paleozoic
ophiolites in the Uralides and the Altaids in central Asia are the
remnants of the Pleionic Ocean, which evolved between the
BalticaEastern Europe and Kazakhstan-Siberian continental masses
(Brown et al., 2006; Windley et al., 2002; Xiao et al., 2004). The
JurassicCretaceous ophiolites of the Tethyan Ocean systems extend
from the Betic-Rif and Pyrenees in the west through the
Alpine-Himalayan orogenic belts in the center to the Indonesian
region in the east (Fig. 1; Hall, 1997; Pubellier et al., 2004;
Bortolotti and Principi, 2005). The Phanerozoic ophio lites in
these col-lisional orogenic belts (i.e., Appalachian, Caledo-nides,
Uralides, and Altaids in central Asia, Betic-Rif and Pyrenees,
Alpine-Himalayan) com-monly show mid-ocean-ridge basalt (MORB) to
island-arc tholeiite (IAT) and boninitic geochem-ical affi nities
(Varfalvy et al., 1997; Bdard et al., 1998; Spadea and DAntonio,
2006; Pag et al., 2009). The ophio lites in the accretionary-type
Western Pacifi c and Cordilleran orogenic belts are slivers of
abyssal peridotites and volcanic ocean islands, seamounts, and
mid-ocean-ridge crust scraped off from downgoing plates, and they
are commonly associated with accretionary mlanges and high-pressure
metamorphic rocks (Cloos, 1982; Waka bayashi, 1999; Ernst, 2005;
Ring, 2008; Hall, 2009; Cawood et al., 2009; Xiao et al.,
2010).
The principal ophiolite pulses during the last 250 m.y. coincide
with the emplacement of plume-related large igneous provinces
(LIPs) and giant dike swarms (Ernst et al., 1995; Yale and
Carpenter, 1998; Coffi n and Eldholm, 2001) and collectively mark
supermagmatic events in Earth history (Fig. 2). The enhanced large
igneous province formation and ophiolite gen-eration in the Late
Jurassic and Cretaceous are particularly noteworthy (Vaughan and
Scarrow , 2003). The evolution of the Tethyan and Carib-bean
ophiolites overlapped with the Cretaceous super plume event (12080
Ma), which was responsible for the formation of oceanic plateaus in
the Pacifi c and Indian Oceans, high global sea levels, and
increased rates of seafl oor spreading (Larson, 1991). The
JurassicCretaceous peri-Caribbean ophiolites (Fig. 1) include
remnants
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 391
Ng PgTertiary
Cretaceous Triassic Permian Carb. Devonian Sil. Ord.
Camb.Mesozoic Paleozoic
setil
oi hpo
roj
am
fo
reb
mu
N hci h
w
ot st
ne
ve
roj
aM
det
aler
era
setil
oi hpo
Jurassic
Central Asian ophiolites
Age (Ma)
Age (Ma)
Figure 2. Ophiolite pulses and the distribution of major
orogenic belts with ophiolite occurrences during the Phanerozoic.
A. Ophiolite pulses and the geographic distribution of Phanerozoic
ophiolites through time. B. Distribution of representative examples
of major ophio-lite types through time. C. Approximate time
intervals for the lifespan of major supercontinents and their
breakup, signifi cant orogenic events, and supermagmatic events
represented by the emplacement of giant dike swarms and large
igneous provinces (LIPs). The main pulses of ophiolite generation
coincide with plate movements leading to the closure of ocean
basins and continental collisions, large mag-matic events (with the
production of large igneous provinces and giant dike swarms), and
the breakup of supercontinents. Major orogenic events are (from
youngest to oldest): Ar-EuArabia-Eurasia collision,
In-EuIndia-Eurasia collision, Al-UrAltaid-Uralian orogenies of
Central Asia, Ap-HyAppalachian-Hercynian orogenies, CldCaledonian
orogeny, FmtFamatinian orogeny, P-Af-BrPan-AfricanBrasiliano
orogenies. NgNeogene; PgPaleogene. For a list of different
ophiolite types, see Table 1.
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
392 Geological Society of America Bulletin, March/April 2011
of proto-Caribbean oceanic crust and the Carib-beanColombian
oceanic plateau (Kerr et al., 1998) and display a complex record of
igneous activity associated with continental rifting, sea-fl oor
spreading, the construction of an oceanic plateau, and the
development of island arcs (Giunta and Oliveri, 2009; Kerr et al.,
2009). The most prominent ophiolite pulse during the Mesozoic
coincided with the breakup of Pangea through discrete episodes of
continental rifting during the Late Triassic and Jurassic (Fig. 2;
Dalziel et al., 2000).
A NEW CLASSIFICATION OF OPHIOLITES
The main ophiolite pulses appear to be tem-porally and spatially
linked to some fi rst-order global tectonic and magmatic events.
These global events and related mantle processes con-
trolled the development of different ophiolite types in
different tectonic environments (Dilek, 2003). We list
representative examples of the main ophiolite types, their ages,
geographic lo-cation, and related references in Table 1. These
ophiolite types are marked in Figures 1 and 2 with different
symbols, indicating formation in different tectonic environments,
as explained in the following section. In Table 2, we also list and
explain a series of abbreviations in refer-ence to different
ophiolite types and all the rel-evant geochemical terminology used
in the next two sections and on the fi gures.
Tectonic Settings of Ophiolite Types
Continental margin (CM) ophiolites form during the early stages
of ocean basin evolution, following initial continental breakup.
These ophiolites are fragments of magma-poor, ocean-
continent transitions (OCT). Modern, in situ ocean-continent
transitions include the Iberia and Red SeaWestern Arabia rifted
margins (Fig. 1). Some classic examples of continental margin
ophiolites include the Jurassic ophio-lites in the Northern
Apennines (Ligurian) and the western Alps (Caby, 1995; Rampone et
al., 2005; Manatschal and Mntener, 2009). These ophiolites consist
of exhumed, subcontinental lithospheric mantle lherzolite directly
overlain by basaltic lavas and intruded by small gab-broic plutons
and rare mafi c dikes. The crustal rocks display normal (N) MORB
geochemical signatures. Continental margin ophiolites cor-respond
to the lherzolite-type (LOT) ophiolites of Ishiwa tari (1985) and
Boudier and Nicolas (1985) and are the products of low degrees of
melting of less-depleted subcontinental litho-spheric mantle and
upwelling asthenosphere (Rampone et al., 2005).
TABLE 1. REPRESENTATIVE EXAMPLES OF MAIN OPHIOLITE TYPES, THEIR
GEOGRAPHIC LOCATIONS, APPROXIMATE AGES, AND RELATED REFERENCES
Ophiolite Location Age (Ma) ReferencesContinental margin type1
Tihama Red Sea, Saudi Arabia 20 Coleman et al. (1972, 1977), Dilek
et al. (2009)2 Ligurian Italy 200 Rampone and Piccardo (2000),
Muntener and Piccardo (2003)
Manatschale and Muntener (2009)3 Ust-Belaya 1 NE Russia 310
Ishiwatari et at. (2003), Sokolov et al. (2003)4 Ust-Belaya 2 NE
Russia 320 Ishiwatari et at. (2003), Sokolov et al. (2003)5 Nurali
S Urals, Russia 410 Spadea et al. (2003)Mid-ocean-ridge type1A
Macquarie Isl. SW Pacifi c 10 Kamentsky et al. (2000), Varne et al.
(2000), Rivizzigno and Karson (2004)1B Taitao S Chile 10 Le Moigne
et al. (1996), Guivel et al. (1999), Lagabrielle et al. (2000),
Shibuya et al. (2007)2 Khoy Iran 98-103 Ghazi and Hassanipak
(2000), Hassanipak and Ghazi (2000)
Khalatbari-Jafari et al. (2004)4 Masirah W Indian Ocean 150
Peters and Mercolli (1998), Peters (2000)5 Horo Kanai Central
Hokkaido, Japan 165180 Ishiwatari et al. (2003)6 Kuyul 1 NE Russia
190 Sokolov et al. (2003)7 Kuyul 2 NE Russia 200 Sokolov et al.
(2003)8 Kuyul 3 NE Russia 210 Sokolov et al. (2003)9 Nurali S Urals
405 Pertsev et al. (1997), Spadea et al. (2003)Plume type1A Loma de
Hiero Venezuela 80 Giunta et al. (2002)1B Bolivar SW Colombia 80
Nivia (1996)2 Nicoya Costa Rica 8995 Kerr et al. (1997a, 1997b),
Sinton et al. (1997), Hauff et al. (2000)3 Peri-Caribbean 1 Cuba,
Puerto Rica, Hispaniola 105 Kerr et al. (1997a, 1997b), Giunta et
al. (2006)4 Peri-Caribbean 2 Cuba, Puerto Rica, Hispaniola 125 Kerr
et al. (1997a, 1997b), Giunta et al. (2006)5 Duarte Hispaniola 140
Lapierre et al. (1997, 1999), Giunta et al. (2006), Escuder Viruete
et al. (2009)6 Loma La Monja Hispaniola 155 Escuder Viruete et al.
(2009)7 Mino-Tamba 1 SW Japan 185 Ichiyama et al. (2008)8
Mino-Tamba 2 SW Japan 200 Ichiyama et al.
(2008)Suprasubduction-zone type1 Zambales Philippines 4044 Yumul et
al. (2000), Encarnacion (2004)2 Antique Panay, Philippines 7580
Dimalanta et al. (2006)3A Troodos Cyprus 9294 Batanova and Sobolev
(2000), Dilek and Furnes (2009)3B Semail Oman 9295 Lippard et al.
(1986), Hacker et al. (1996), Warren et al. (2005)
Dilek and Furnes (2009), Alabaster et al. (1982)3C Kizildag
Turkey 9294 Tinkler et al. (1981), Erendil (1984), Bagci et al.
(2005), Dilek et al. (1999)
Dilek and Thy (1998, 2009)4 Xigaze Tibet, China 120126 Aitchison
et al. (2003), Malpas et al. (2003), Zhang et al. (2003)5 Sabah
Northern Borneo 135140 Rangin et al. (1990), Mller (1991)6A Mirdita
Albania 160 Beccaluva et al. (1994), Bortolotti et al. (2002),
Saccani and Photiades (2005)
Dilek et al. (2007, 2008)6B Pindos Greece 160 Capedri et al.
(1980), Saccani and Photiades (2005), Dilek and Furnes (2009)7 Cape
Povorotny Far East Asia 230250 Sokolov et al. (2003)8 Yakuno SW
Japan 270280 Ishiwatari (1985), Ichiyama and Ishiwatari (2004)
(continued)
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 393
Mid-ocean-ridge (MOR) ophiolites may form at plume-proximal
(e.g., Iceland) and plume-distal mid-ocean ridges, trench-proximal
mid-ocean ridges, or trench-distal backarc spreading ridges (Table
2). They generally have a Penrose-type structural architecture
(particularly at the centers of ridge segments) and show N-MORB
(e.g., Argolis-Pindos in Greece), enriched (E) MORB (e.g.,
Macquarie Island), and/or contaminated (C) MORB geochemical affi
ni-ties. N-MORB and E-MORB ophiolites have compositions that are
more depleted and more enriched, respectively, than primitive
mantlederived magmas (Pearce, 2008). C-MORB ophiolites are
crustally contaminated. The Taitao ophiolite in Chile (Fig. 1),
which formed at a trench-proximal Chile Rise (Karsten et al.,
1996), is a type example of C-MORB ophio-lite. It was emplaced into
the South American continental margin as a result of a
ridge-trench
collision (Anma et al., 2009). Mid-ocean-ridge ophiolites, in
general, correspond to class II and III types in Miyashiros (1975)
classifi cation of ophiolites based on the presence of tholeiitic
and alkaline volcanic rocks.
Plume-type (P) ophiolites may form close to plume-proximal
spreading ridges and as part of oceanic plateaus (e.g., Caribbean
Plateau; Kerr et al., 2009). They have thick plutonic and volcanic
sequences (Coffi n and Eldholm, 2001; Kerr et al., 2009), and show
depleted (D-MORB) to enriched (E-MORB) trace-element patterns
(Pearce, 2008).
Suprasubduction-zone (SSZ) ophiolites (e.g., Mirdita, Albania;
Samail, Oman; Troodos, Cyprus; Fig. 1) form in the extending upper
plates of subduction zones, as in the modern Izu- Bonin-Mariana and
Tonga-Kermadec arc-trench rollback systems (Fig. 1; Hawkins, 2003;
Reagan et al., 2010). They may evolve in ex-
tending, embryonic backarc to forearc environ-ments (BA-FA),
forearc settings (FA), and both oceanic and continental backarc
basins (OBA and CBA, respectively; Table 2). The Rocas Verdes
ophiolites in southern Chile are the best examples of
suprasubduction-zone continental backarc basin ophiolites (Saunders
et al., 1979; Stern and de Wit, 2003). Suprasubduction-zone
ophiolites commonly have a Penrose-type struc-tural architecture
and may show a MORBIATboninitic geochemical sequence of igneous
activity. Suprasubduction-zone forearc ophio-lites result from
oceanic crust generation dur-ing the closure of ocean basins and
mark major subduction initiation events (Casey and Dewey, 1984;
Dilek and Furnes, 2010; Pearce and Rob-inson, 2010). The age range
among their vari-ous ophiolitic subunits is commonly less than 10
m.y. (Dilek and Furnes, 2009). They cor-respond to the class I
ophiolites of Miyashiro
TABLE 1. REPRESENTATIVE EXAMPLES OF MAIN OPHIOLITE TYPES, THEIR
GEOGRAPHIC LOCATIONS, APPROXIMATE AGES, AND RELATED REFERENCES
(continued)
Ophiolite Location Age (Ma) ReferencesSuprasubduction-zone type
(continued)9 Magnitogorsk 1 S Urals, Russia 385400 Spadea and
Scarrow (2000), Spadea et al. (2003)
Spadea and DAntonio (2006)10 Baimak-Buribai SW Urals, Russia 420
Spadea and Scarrow (2000)11A Trinity 1 California, USA 440 Brouxel
et al. (1989), Metcalf et al. (2000)11B Solund-Stavfjord SW Norway
440 Furnes et al. (1982); Pedersen (1986), Dunning and Pedersen
(1988)
Pedersen and Furnes (1991), Furnes et al. (1990, 2003, 2006)12
Kudi-Kunlun NW China 460470 Wang et al. (2001, 2002), Yang et al.
(1996)13A Thetford Mines Canada 479 Hebert and Laurent (1989), Page
et al. (2009), Schroetter et al. (2003)13B Bay of Islands Canada
484 Casey et al. (1985), Suhr (1992), Bedard and Hebert (1996)
Varfalvy et al. (1997), Kurth-Velz et al. (2004)13C Betts Cove
Canada 489 Coish et al. (1982), Bedard et al. (1998), Bedard
(1999)14A Karmy SW Norway 474493 Furnes et al. (1980), Pedersen
(1986), Dunning and Pedersen (1988)
Pedersen and Hertogen (1990), Pedersen and Furnes (1991)14B
Gulfjellet SW Norway 489 Furnes et al. (1982), Dunning and Pedersen
(1988), Heskestad et al. (1994)14C Leka NW Norway 497 Prestvik
(1974), Pedersen (1986), Dunning and Pedersen (1988)
Pedersen and Furnes (1991), Furnes et al. (1988, 1992)15 Lachlan
SE Australia, Tasmania 495510 Spaggiari et al. (2003,
2004)Volcanic-arc type1 Itogon Philippines 30 Encarnacion (2004)2A
Coast Range and
Great Valley 1California, USA 140 Shervais et al. (2004)
2B Zedong 1 Tibet, China 127140 Malpas et al. (2003)3A Coast
Range and
Great Valley 2California, USA 155 Shervais et al. (2004), Hopson
et al. (2008)
3B Zedong 2 Tibet, China 155162 Malpas et al. (2003)4A
Smartville California, USA 155165 Saleeby et al. (1989), Dilek et
al. (1990, 1991)4B Josephine Oregon and California, USA 162164
Saleeby et al. (1982), Harper and Wright (1984), Harper et al.
(1994)
Harper (2003a, 2003b)5 DAguilar 1 E Australia 360 Spaggiari et
al. (2003, 2004)6A DAguilar 2 E Australia 380 Spaggiari et al.
(2003, 2004)6B Magnitgorsk 2 S Urals, Russia 370 Spadea et al.
(2003)7A Magnitgorsk 3 S Urals, Russia 385 Spadea et al. (2003)7B
Trinity 2 California, USA 385 Brouxel et al. (1989), Metcalf et al.
(2000)Accretionary type1 Mineoka Central Japan 25 Hirano et al.
(2003), Takahashi et al. (2003), Ogawa and Takahashi (2004)2 Tokoro
Japan 60 Taira et al. (1988), Isozaki (1996)3 Peri-Caribbean 3
Hispaniola, Guatemala,
Aruba-Curacao, Central Cuba 8890 Donnelly (1989), Kerr et al.
(1997b), Sinton et al. (1998)4 Tamba Japan 135 Nakae (2000),
Koizumi and Ishiwatari (2006)5 Solonker 1 Central Asia 240 Xiao et
al. (2003), Chen et al. (2009)6 Solonker 2 Central Asia 250 Xiao et
al. (2003), Chen et al. (2009)7 Ganychalan 1 NE Russia 420 Sokolov
et al. (2003)8 Ganychalan 2 NE Russia 440 Sokolov et al. (2003)9
Ganychalan 3 NE Russia 460 Sokolov et al. (2003)10 Ganychalan 4 NE
Russia 480 Sokolov et al. (2003)11 Ganychalan 5 NE Russia 500
Sokolov et al. (2003v
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
394 Geological Society of America Bulletin, March/April 2011
TABL
E 2.
OPH
IOLI
TE/O
CEAN
IC C
RUST
TY
PES,
THE
IR L
OCA
TIO
NS, A
ND R
EFER
ENCE
S TO
DAT
A SO
URCE
S, A
ND A
BBRE
VIAT
IONS
USE
D IN
THE
TEX
T AN
D TH
E FI
GUR
ES
Oph
iolite
/oce
anic
crus
t typ
eAb
brev
iatio
nsLo
catio
n
No.
anal
.Bo
wen
No.
anal
.M
ulti
No.
anal
.V/
Ti
No.
anal
.Th
-Yb-
NbRe
fere
nce
to d
ata
sour
ces
Cont
inen
tal m
argi
nCM
Inte
rnal
Lig
urid
es,It
aly
272
Otto
nello
et a
l. (19
84), R
ampo
ne et
al. (1
998)
Exte
rnal
Lig
urid
es, I
taly
2611
2619
Vann
ucci
et a
l. (19
93), M
ontan
ini et
al. (2
008)
North
Ape
nnin
e, It
aly
3939
Ferra
ra e
t al.
(1976
)Co
rsica
1313
Becc
aluv
a et
al.
(1977
)M
id-o
cean
ridg
eM
OR
Plum
e-pr
oxim
al m
id-o
cean
ridg
eM
OR
PPIc
elan
d11
937
6739
Sigv
alda
son
(1974
), Hem
ond e
t al. (
1993
)Pl
ume-
dist
al m
id-o
cean
ridg
eM
OR
PDM
acqu
arie
Isla
nd12
1212
12Ka
men
tsky
et a
l. (20
00)
Tren
ch-p
roxim
al m
id-o
cean
ridg
eM
OR
TPTa
itao
Peni
nsul
a, S
. Chi
le31
3131
9Le
Moi
gne
et a
l. (19
96), G
uivel
et al.
(199
9)No
rmal
mid
-oce
an ri
dge
basa
ltNM
ORB
Depl
eted
(in th
e inc
ompa
tible
eleme
nts)
mid
-oce
an-ri
dge
basa
ltDM
ORB
Enric
hed
(in th
e inc
ompa
tible
eleme
nts)
mid
-oce
an-ri
dge
basa
ltEM
ORB
Crus
tally
con
tam
inat
ed m
id-o
cean
-ridg
e ba
salt
CMO
RBTr
ansit
iona
l mid
-oce
an-ri
dge
basa
ltTM
ORB
Plum
eP
Gor
gona
Isla
nd, C
olom
bia
1010
Kerr
et a
l. (19
96a)
Wes
tern
Col
ombi
a85
1984
23Ke
rr et
al.
(1997
a)Ja
mai
ca17
1717
17Ha
stie
et a
l. (20
08)
Cura
cao,
Car
ibbe
an S
ea84
1119
19Kl
aver
(198
7), Ke
rr et a
l. (19
96b)
Oce
an-is
land
bas
alt
OIB
Supr
asub
duct
ion
zone
SSZ
Back
arc
to fo
rear
cSS
Z BA
-FA
Alba
nia
113
4610
245
Dile
k et
al.
(2008
)Cy
prus
56Ra
uten
schl
ein
et a
l. (19
85), A
ucla
ir an
d Lu
dden
(198
7),
Taylo
r (199
0)Tu
rkey
6140
6125
Dile
k an
d Th
y (19
98, 2
009)
Om
an13
415
113
57Li
ppar
d et
al.
(1986
), Eina
udi e
t al. (
2003
), God
ard et
al. (2
003)
Fore
arc
SSZ
FANe
wfou
ndla
nd47
2247
23Be
dard
(199
9)O
cean
ic ba
ckar
cSS
Z O
BAW
este
rn N
orwa
y80
280
2Fu
rnes
et a
l. (20
06, a
nd re
feren
ces t
herei
n)Co
ntin
enta
l bac
karc
SSZ
CBA
Sout
hern
Chi
le67
Elth
on (1
979),
Saun
ders
et al.
(197
9), St
ern an
d Elth
on (1
979),
Ster
n (19
80)
Volca
nic
arc
VALu
zon,
Phi
lippi
nes
5339
Evan
s et
al.
(1991
), Yum
ul e
t al.
(2000
)No
rth C
asca
des,
Was
hing
ton,
USA
66
6M
etzg
er e
t al. (
2002
)No
rthwe
ster
n Ca
liforn
ia,
USA
9322
9340
Harp
er (1
984),
Harp
er et
al. (1
988,
2003
a, 20
03b)
Sier
ra N
evad
a, C
alifo
rnia
, US
A4
Dile
k et
al.
(1991
)
Tota
l num
ber o
f ana
lyses
1902
283
1581
336
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 395
(1975) and harzburgite-type (HOT) ophiolites of Ishiwatari
(1985) and Boudier and Nicolas (1985), which are the products of
high degrees of melting of depleted, harzburgitic mantle. Both
suprasubduction-zone oceanic backarc basin and continental backarc
basin ophiolites form as a result of seafl oor spreading in
ensi-matic and ensialic settings (respectively).
Volcanic-arc (VA) ophiolites form in ensi-matic arc settings
(e.g., the Philippines, SE Asia; Sierra Nevada, California). They
have a poly-genetic crustal architecture with a deformed, older
oceanic basement, mafi c lower crust composed of gabbroic plutons
and hypabys-sal intrusions, moderately to well-developed
dioritic-tonalitic middle crust, andesitic to rhyo-litic extrusive
rocks and dikes (locally sheeted) forming the upper crust, and
volcaniclastic cover (locally subaerial). These crustal units
display tholeiitic to calc-alkaline geochemi-cal signatures.
Volcanic-arc ophiolites differ from suprasubduction-zone ophiolites
based on their thicker and more fully developed arc crust with
calc-alkaline compositions. The age range among various ophiolitic
subunits in volcanic-arc ophiolites can be longer than 2030 m.y.
(Dilek et al., 1991).
Accretionary-type ophiolites, occurring in subduction-accretion
complexes of active mar-gins, contain fragments of any of the
previ-ously outlined ophiolite types and are locally associated
with pelagic-hemipelagic sedimen-tary rocks and trench-fi ll
sediments that may have been deposited on them prior to and after
their incorporation into the accretionary prism. These ophiolites
may have diverse lithologi-cal assemblages, metamorphic grades,
styles of deformation, and chemical affi nities with no genetic
links between them, since they consist of tectonic slices of
oceanic rocks scraped off from downgoing plates (e.g., Mineoka
ophiolite in central Japan; Ogawa and Takahashi, 2004). They become
progressively younger in age structurally downward within
subduction-accre-tion complexes. We do not treat these ophiolites
separately in our discussion here because they do not show a
distinctive lithological construc-tion, and hence they lack a
unique geochemical fi ngerprint.
Geochemical Fingerprinting of Ophiolite Types
We use a selection of diagrams to character-ize the geochemical
signatures of some well-preserved examples of the types of
ophiolites distinguished here. These diagrams are based on an
extensive database (compiled from our own analytical work and the
extant literature) that is summarized in Table 2. The literature we
used
in our ophiolite classifi cation and geochemical-tectonic fi
ngerprinting is presented in the GSA Data Repository.1
Since lavas and dikes in ophiolites are, in gen-eral, subject to
various degrees of hydrothermal alteration and greenschist- to
amphibo lites-facies metamorphism in intra-oceanic conditions, it
is important to use elements that are relatively stable during such
processes in order for us to determine their primary geochemical
composi-tions. Several studies have been carried out on the element
behavior of magmatic rocks that were variably altered and
metamorphosed. In general, the mobility of an element relates to
the water-rock interactions during reaction (e.g., Bickle and
Teagle, 1992). Low-temperature ex-perimental studies of reaction
between basalt and seawater have demonstrated minor leaching of Fe
and Si and enrichment of Na and Mg; on the other hand, Al, Ti, and
P are the least mobile elements, and Ca is variably depleted (Scott
and Hajash, 1976; Seyfried et al., 1978). The trace elements Y, Zr,
Nb, V, Cr, Co, Ni, rare earth ele-ments (REEs), Th, and Ta are
generally rela-tively immobile (Coish, 1977; Hellman et al., 1979;
Shervais, 1982; Seyfried and Mottl, 1982; Dickin and Jones, 1983;
Dungan et al., 1983; Mottl, 1983; Staudigel and Hart, 1983;
Seyfried et al., 1988; Gillis and Thompson, 1993). A study on the
behavior of transition metals (Ti, V, Ni, Cr, Co, Cu, Zn, Fe, Mn)
and Mg in meta basic rocks suggests relatively little mobility
dur-ing medium to high degrees of metamorphism (Nicollet and
Andriambololona, 1980). During hydrothermal alteration of basaltic
pillow lavas, Ba shows variable alteration trends (Humphris and
Thompson, 1978), and Pb becomes mod-erately to strongly depleted
(Teagle and Alt, 2004). Alteration (palagonitization) of the glass
rind of pillow lavas results in enrichment of K, Rb, and Cs,
particularly the latter two (Hart, 1969; Staudigel and Hart, 1983).
Therefore, we paid particular attention in constructing the
geo-chemical diagrams presented here to use those elements that are
relatively stable during hydro-thermal alteration.
In Bowen diagrams (Fig. 3) demonstrating the compositional
variability in upper-crustal units (lavas and dikes), the
subduction-related suprasubduction-zone and volcanic-arc
ophio-lites show larger variability in SiO2 and TiO2 at given MgO
contents than the subduction-unre-lated continental margin,
mid-ocean-ridge, and plume ophiolites. The highest variability with
respect to these two elements is represented by
the suprasubduction-zone backarc- to forearc-type ophiolites,
whereas the suprasubduction-zone forearc-type ophiolites show
invariably low TiO2 (Fig. 3B). The largest spread in MgO is
exhibited by the subduction-unrelated plume-type ophiolites (Fig.
3A). In MORB-normalized multi-element diagrams, the con-tinental
margin, mid-ocean-ridge, and plume ophiolites display fl at
patterns between V and Zr, and an increase toward the most
incompati-ble elements (i.e., Ba, Rb, Cs; Fig. 4A). In the same
multi-element diagrams, the patterns of the suprasubduction-zone
and volcanic-arc ophiolites display much larger variability; they
are generally enriched in the most incompati-ble, nonconservative
elements (Cs, Rb, Th) and show generally negative Ta and Nb and
positive Pb and Sr anomalies (Fig. 4B).
In a Ti-V discrimination diagram (Shervais, 1982), the
continental margin, mid-ocean-ridge, and plume ophiolites straddle
the fi eld defi ned by the ratios between 20 and 50, typical of
mid-ocean-ridge basalts (Fig. 5A), whereas the suprasubduction-zone
and volcanic-arc ophio-lites show a wider scatter of Ti/V ratios
between 50 (Fig. 5B). However, the subtypes of both the
subduction-related and subduction-unrelated ophiolites demonstrate
pronounced differences in their Ti-V distributions. For the
subduction-unrelated types, the Ti-V data of the lavas and dikes
for the plume subtype hardly overlap with those of the continental
margin and mid-ocean-ridge trench-proximal subtypes (Fig. 5A).
Similarly, for the subduction-related ophiolite types, the mafi c
lavas and dikes of the suprasubduction-zone forearc subtype
ex-clusively plot in the boninite fi eld and do not overlap with
those of the suprasubduction-zone oceanic backarc basin subtype
(Fig. 5B). By far, the suprasubduction-zone backarc to forearc
subtype shows the largest range in the Ti-V dia-gram (Fig. 5B).
This dispersion of Ti/V ratios is a result of a large geochemical
range from boninite and island-arc tholeiite to MORB mag-mas that
occur in subduction-infl uenced igne-ous systems (Shervais, 1982;
Dilek et al., 2007; Dilek and Furnes, 2009).
In the Nb/Yb versus Th/Yb diagram (Pearce, 2008), the lavas and
dikes of the continental margin, mid-ocean-ridge, and plume
ophiolites plot within the mantle array (Fig. 6A), whereas those of
the suprasubduction-zone and volcanic-arc ophiolites show a signifi
cant shift away from this mantle array, toward the
subduction-related Mariana arc fi eld (Fig. 6B). These fi ve
elements (Ti, V, Th, Yb, Nb), which we have used in dis-criminating
possible tectonic settings of ophio-litic magma generation, are
most immobile during metamorphism and alteration; therefore, they
are most reliable as proxies to differentiate
1GSA Data Repository item 2011131, Data source for geochemistry
and tectonics of different ophio-lite types used in Tables 1 and 2,
and for Figures 36, is available at
http://www.geosociety.org/pubs/ft2011.htm or by request to
[email protected].
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
396 Geological Society of America Bulletin, March/April 2011
between subduction-related and other magmas (Shervais, 1982;
Pearce, 2008), particularly when utilized together with other
informative geochemical techniques and fi eld-oriented re-gional
tectonic constraints.
Geochemical characterization of different types of ophiolites
allows us to distinguish two
major groups, one related to or least infl uenced by
subduction-zone processes and the other un-related to subduction
zones. The suprasubduc-tion-zone ophiolites that formed in backarc
and incipient arcforearc tectonic environments (e.g., Mirdita,
AlbaniaDilek et al., 2007, 2008; Troodos, CyprusRobinson et al.,
2003;
Pearce and Robinson, 2010), in a forearc set-ting (e.g., Betts
Cove, CanadaBdard, 1999), and as a volcanic arc (e.g., Smartville,
Califor-niaDilek et al., 1991) display the most pro-nounced
variations in geochemical patterns. On the other hand,
trench-distal backarc ophiolites that formed in oceanic or
continental settings,
Figure 3. Bowen diagrams show-ing the relationships between
MgO-SiO2 and MgO-TiO2 for subduction-unrelated ophiolites (i.e.,
continental margin, plume, and mid-ocean-ridge types) (A1 and A2),
and subduction-related ophiolites (i.e., volcanic-arc and supra
subduction-zone [SSZ] types) (B1 and B2). The mid-ocean-ridge type
(MOR) is subdivided into three sub-types, i.e., plume-proximal
(PP), plume-distal (PD), and trench-proximal (TP). The
supra-subduction-zone type (SSZ) is subdivided into four subtypes,
i.e., backarc to forearc (BA-FA), forearc (FA), oceanic backarc
(OBA), and continental backarc (CBA). Data sources (listed in the
GSA Data Repository [see text footnote 1]): Continental margin
typeFerrara et al. (1976), Becca luva et al. (1977), Ottonello et
al. (1984), Vannucci et al. (1993), Rampone et al. (1998),
Montanini et al. (2008). Plume typeKerr et al. (1996a, 1996b,
1997), Hastie et al. (2008). Mid-ocean-ridge types, includ-ing PP
subtypeSigvaldason (1974), Hemond et al. (1993); PD
subtypeKamenetsky et al. (2000); TP subtypeLe Moigne et al. (1996),
Guivel et al. (1999). Volcanic-arc typeYumul et al. (2000), Evans
et al. (1991), Metzger et al. (2002), Harper (1984), Harper (2003a,
2003b), Harper et al. (1988), Dilek et al. (1991).
Supra-subduction-zone types, includ-ing BA-FA subtypeDilek et al.
(2008), Lippard et al. (1986), Einaudi et al. (2003), Godard et al.
(2003), Auclair and Lud-den (1987), Rautenschlein et al. (1985),
Taylor (1990), Dilek and Thy (1998, 2009), Y. Dilek (personal
observation, 1998). FA subtypeBdard (1999); oceanic backarc
basin-subtypeFurnes et al. (2006, and references therein); and
continental backarc basin-subtypeSaunders et al. (1979), Stern and
Elthon (1979), Stern (1979, 1980), Elthon (1979).
A1. Subduction-unrelated
400 5 10 15 20 25 30 0 5 10 15 20 25 30
0 5 10 15 20 25 30 0 5 10 15 20 25 30
50
60
70
MgO (wt. %)
SiO
2 (w
t. %)
Cont. margin Plume MOR (PP)MOR (PD)MOR (TP)
A2. Subduction-unrelated
0
1
2
3
TiO
2
Cont. margin Plume MOR (PP)MOR (PD)MOR (TP)
B1. Subduction-related
40
50
60
70
MgO (wt. %)
SiO
2 (w
t. %)
Volc. arcSSZ (BA-FA)SSZ (FA)SSZ (OBA)SSZ (CBA)
B2. Subduction-related
0
1
2
3
MgO (wt. %)
TiO
2Volc. arcSSZ (BA-FA)SSZ (FA)SSZ (OBA)SSZ (CBA)
(wt. %
)(w
t. %)
MgO (wt. %)
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 397
e.g., the Solund-Stavfjord ophiolite in West Norway (Furnes et
al., 2006) and the Rocas Verdes ophiolites in the southernmost
Andes, Chile (Saunders et al., 1979; Stern and De Wit, 2003), show
weaker geochemical evidence of subduction. The groups of ophiolites
that are entirely unrelated to subduction processes are the
continental margin, mid-ocean-ridge, and plume ophiolites.
PETROGENESIS OF OPHIOLITE TYPES IN DIFFERENT TECTONIC
SETTINGS
Figure 7 depicts the petrogenesis of subduc-tion-related and
subduction-unrelated types of ophiolites in different tectonic
settings. The petro-genesis of a subduction-unrelated continental
margin ophiolite involves slow exhumation and
limited partial melting of subcontinental mantle lherzolite
(Fig. 8A) and upwelling astheno-sphere in response to lithospheric
extension and continental rifting (Fig. 7A1; Rampone et al., 2005;
Piccardo et al., 2009). Multiple intrusions of MORB-type magma form
small olivine gab-bro pods and dikes (Fig. 8A) and cause basaltic
eruptions on the seafl oor (Figs. 7A2 and 8B). Extensional
tectonics and associated faulting
A. Subduction-unrelated
0.1
1
10
100
Cs Rb Ba Th U Ta Nb K La Ce Pb Pr Sr P Nd Zr Hf Sm Eu Gd Ti Tb
Dy Y Ho Er Tm Yb Lu V Sc Co Cr Ni
Roc
k/M
ORB
Cont. marginPlumeMOR (PP)MOR (PD)MOR (TP)
B. Subduction-related
0.1
1
10
100
Cs Rb Ba Th U Ta Nb K La Ce Pb Pr Sr P Nd Zr Hf Sm Eu Gd Ti Tb
Dy Y Ho Er Tm Yb Lu V Sc Co Cr Ni
Roc
k/M
OR
B
SSZ (BA-FA)SSZ (FA)SSZ (OBA)Volc. arc (MORB-like)Volc. arc
(IAT-bon)
Figure 4. Mid-ocean-ridge-basalt (MORB)normalized multi-element
diagrams, showing average values for subduction-unrelated (A) and
subduction-related (B) ophiolites. IATisland-arc tholeiite;
bonboninite. Different types and subtypes of ophiolites are
explained in Figure 3. Normalizing values (in ppm) are: Cs (0.007),
Rb (0.56), Ba (6.3), Th (0.12), U (0.047), Ta (0.13), Nb (2.33), K
(1079), La (2.5), Ce (7.5), Pb (0.3), Pr (1.32), Sr (90), P (314),
Nd (7.3), Zr (74), Hf (2.05), Sm (2.63), Eu (1.02), Gd (3.68), Ti
(7614), Tb (0.67), Dy (4.55), Y (28), Ho (1.01), Er (2.97), Tm
(0.456), Yb (3.05), Lu (0.455), V (300), Sc (40), Co (40), Cr
(275), and Ni (100). The elements have been placed in order of
their relative incompatibility with spinel-lherzolite mantle (after
Pearce and Parkinson, 1993). Data sources (listed in the GSA Data
Repository [see text footnote 1]): Continental margin typeMontanini
et al. (2008); plume typeKerr et al. (1996b, 1997), Hastie et al.
(2008); mid-ocean-ridge types, including plume-proximal
subtypeHemond et al. (1993); plume-distal subtypeKamenetsky et al.
(2000); trench-proximal subtypeLe Moigne et al. (1996), Guivel et
al. (1999); volcanic-arc typeHarper (2003b); suprasubduction-zone
types, including BA-FA subtypeDilek et al. (2008), Dilek and Thy
(1998), Y. Dilek (personal observation, 1998); FA subtypeBdard
(1999); and oceanic backarc basin subtypeH. Furnes (personal
observation, 1997).
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
398 Geological Society of America Bulletin, March/April 2011
and shearing may cause tectonic brecciation of the lavas (Fig.
8C).
Oceanic crust formation at oceanic spreading axes involves
decompression melting of uprising asthenosphere and focused upward
ascent of the melt into a melt lens and associated crystal mush
zone (Fig. 7A1). Magma injection into a narrow, ~250-m-wide
region (Rubin and Sinton, 2007) above the melt lens causes crustal
accretion via diking and eruption on the seafl oor along the ridge
axis. Lavas and dikes have compositions more depleted in
incompatible trace elements
than magmas generated from primitive mantle. Locally, melts
derived from incompatible- elementenriched mantle sources may
segre-gate and rise to form off-axis intrusions and to feed
near-ridge, E-MORB lavas. Studies of core samples from modern ocean
ridges have shown that variations in rates of magma supply and the
thermal structure beneath the spreading axis con-trol the mode of
magmatic accretion and the ar-chitecture of oceanic crust produced.
A low and episodic supply of magma to a slow-spreading ridge
creates a cold environment in which ex-tensional faulting and
crustal attenuation accom-pany seafl oor spreading. Amagmatic
extension can result in exhumation of serpentinized upper-mantle
peridotite on the seafl oor, and highly thinned lower crust and
sheeted dikes (Fig. 7A2; Cannat et al., 1995; Dick et al., 2006).
On the other hand, a voluminous supply of magma and the existence
of a crustal melt lens at shallow depth (Fig. 7A1) beneath
fast-spreading ridges create a hot environment, in which continuous
magma emplacement keeps pace with seafl oor spreading.
Contemporaneous extension and dik-ing produce Penrose-type oceanic
crust under-lain by a
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 399
by rapid slab rollback leading to extension and seafl oor
spreading in the upper plate (Fig. 7B1). In the subduction
initiation stage, magma is fi rst produced by decompressional
melting of deep and fertile lherzolitic mantle and produces the
earliest crustal units with MORB-like compo-
sitions (Figs. 8D8F). Fluids derived from the subducted slab
have little infl uence on melt evolution at this early stage. The
subsequent phases of melting are strongly infl uenced by slab de
hydra tion and related mantle metasoma-tism, melting of subducting
sediments, repeated
episodes of partial melting of metasomatized perido tites, and
mixing of highly enriched liquids from the lower fertile source
with re-fractory melts in the melt column beneath the extending
protoarc-forearc region (Fig. 7B1). Melt aggregation, mixing, and
differentiation
A. Subduction-unrelated
0.01
0.1
1
10
Nb/Yb
Th/Y
b
Cont. marg.PlumeMOR (PP)MOR (PD)MOR (TP)
Marianaarc-basin
Cont. crust
B. Subduction-related
0.01
0.1
1
10
1001010.10.01
1001010.10.01
Nb/Yb
Th/Y
b
SSZ (BA-FA)SSZ (FA)SSZ (OBA)Volc. arc
Marianaarc-basin
Cont. crustOIB
E-MORB
N-MORB
Figure 6. Geochemical data from the subduction-unrelated (A) and
subduction-related (B) ophiolite types and their subtypes (see Fig.
3 for explanation) plotted in Nb/Yb-Th/Yb discrimination diagram
(after Pearce, 2008). OIBocean-island basalt; E- and N-MORBenriched
and normal mid-ocean-ridge basalt. Data sources (listed in the GSA
Data Repository [see text footnote 1]): Continental margin
typeVannucci et al. (1993), Rampone et al. (1998), Montanini et al.
(2008); plume typeKlaver (1987), Kerr et al. (1997), Hastie et al.
(2008); mid-ocean-ridge types, including plume-proximal
subtypeHemond et al. (1993); plume-distal subtypeKamenetsky et al.
(2000); trench-proximal subtypeLe Moigne et al. (1996), Guivel et
al. (1999); volcanic-arc type: Metzger et al. (2002), Harper
(2003a, 2003b); suprasubduction-zone types, including BA-FA
subtypeDilek et al. (2008), Einaudi et al. (2003), Godard et al.
(2003), Dilek and Thy (1998), Y. Dilek (personal observation,
1998); FA sub-typeBdard (1999); oceanic backarc basin subtypeH.
Furnes (personal observation, 1997)
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
400 Geological Society of America Bulletin, March/April 2011
can take place at many levels within this melt column, and
repeated melting of the hydrated mantle leaves behind a highly
depleted, olivine- and orthopyroxene-rich source. This
subarc-forearc melting column produces island-arc tholeiite magma
that is emplaced into and forms lavas overlying crustal units with
MORB-like compositions. Rising temperatures in the mantle wedge,
triggered by increased asthenospheric diapirism and lateral fl ow
of hot mantle (the slab edge effect of Pearce and Robinson, 2010)
and further infl ux of slab-derived fl uids result in shallow
partial melting of the ultrarefractory peridotites (harzburgites),
forming Mg- and silica-rich, hydrous, boninitic melts. Replace-ment
of the primary olivine by orthopyroxene (opx) grains in the
peridotites and the pres-ence of hydrous minerals (i.e.,
amphibole), as observed in most of the suprasubduction-zone
ophiolites, indicate that the orthopyroxenite forms by the reaction
of the preexisting olivine with these boninitic melts (Umino and
Kushiro, 1989; Dilek and Morishita, 2009; Morishita et al., 2010).
The orthopyroxenite thus repre-sents a reaction product between the
migrating melt and the host peridotite in the upper mantle, whereas
the harzburgite is the residual, depleted peridotite of the partial
melt that produced the orthopyroxenite (Fig. 8G). It is likely,
therefore, that geochemical features of boninitic melts are
acquired as a result of interaction of migrating melts with
depleted peridotites in the mantle wedge (Varfalvy et al., 1997).
The harzburgite-dunite-orthopyroxenite suite in the upper-mantle
peridotites of suprasubduction-zone ophiolites are melting residues
and melt migration path-ways in the mantle wedge during the
incipient stage of arc construction. Boninitic dikes and lavas
commonly represent the youngest rock units crosscutting and
overlying the earlier-formed igneous suites in suprasubduction-zone
forearc ophiolites (Figs. 8H and 8K8L). Supra-subduction-zone
ophiolites hence generally display a characteristic, sequential
evolution of MORB to island-arc tholeiite to boninitic igne-ous
activity, which manifests itself in a vertically and laterally
well-developed chemostratigraphy (Fig. 7B2; Dilek and Furnes,
2009), as also observed in the modern Izu-Bonin-Mariana forearc
system (Reagan et al., 2010).
The initial stage of construction of a volcanic-arc ophiolite
involves basic magma. With continued subduction and infi ltration
of arc magmas, the hydrated mafi c crust is partially melted to
form tonalitic magmas, and this to-nalitic crust grows in thickness
as the volcanic arc matures (Fig. 7B1). Residual mafi c crust can
be transformed into peridotitic restite, and consequently the Moho
becomes a fossil melt-ing front (Tatsumi et al., 2008).
Volcanic-arc
ophiolites thus consist of an older oceanic litho-spheric
foundation overlain by a mature arc suite, complete with gabbroic
plutons and mas-sive diabase in the mafi c lower crust, dioritic to
tonalitic middle crust, and andesitic to rhyolitic lavas, dike
intrusions, and pyroclastic and vol-cani clastic rocks in the upper
crust (Fig. 7B2). The construction of a volcanic arc is a result of
prolonged subduction (~2040 m.y.) not termi-nated by colliding
continental blocks, as is the case in the evolutionary history of
suprasubduc-tion-zone ophiolites (Dilek and Flower, 2003).
Sheeted dikes (Figs. 8I8J) are tabular intru-sions of magma fl
owing laterally and vertically along fractures produced by
spreading-related tensile stresses, and they form along a narrow
axial zone beneath central rifts along ocean ridges and above
subduction zones. The exis-tence of sheeted dikes in ophiolites is
conven-tionally interpreted as strong evidence for the origin of
ancient oceanic crust now exposed on land by seafl oor spreading
(Gass, 1990; Moores and Vine, 1971) and is generally regarded as an
essential component of an ophiolite. However, the generation of a
sheeted dike complex re-quires a delicate balance between the rates
of spreading and magma supply for a sustained period such that
suffi cient melt is produced to keep pace with extension in the
rift zone (Rob-inson et al., 2008). In the upper plates of
sub-duction zones, the extension is a consequence of the rate of
slab rollback exceeding the rate
of plate convergence, whereas the magma sup-ply is related to
the temperature profi le and the abundance and nature of fl uids in
the mantle wedge, the age and lithological makeup of the subducting
slab, and the history and extent of melting in the mantle source
(Kincaid and Hall, 2003; Robinson et al., 2008). It is rare for the
balance between spreading and magma supply rates to be maintained
in a suprasubduction-zone setting of oceanic crust formation. In
the absence of this balance, a sheeted dike com-plex will not form
fully, or even at all, and may instead be replaced by magmatic infl
ation and the emplacement of plutons, underplating the extrusive
sequence (where the rate of magma supply exceeds the spreading
rate), or by amagmatic tectonic attenuation of the oceanic crust
(where the spreading rate exceeds the rate of magma supply). This
phenomenon may explain the scarcity of sheeted dike complexes in
nearly 90% of the world ophiolites (Robin-son et al., 2008), and
should be considered in interpretations of the architecture of
putative ancient oceanic crust, particularly in Archean greenstone
belts.
Continental margin, mid-ocean-ridge, and plume ophiolites may
show pronounced varia-tions in trace-element abundances,
particularly for the most incompatible elements, which may be
related to both different degrees of melting and mantle fertility,
but which do not defi ne any partic-ular geochemical evolutionary
trend (Fig. 7A3).
Figure 7 (on following page). Tectonic settings and processes of
subduction-unrelated (A1) and subduction-related (B1) ophiolite
types, columnar sections depicting simplifi ed structural
architecture of the ophiolite types (A2B2), and generalized changes
in element concentration during their evolution (A3B3). Note that
the scale varies from the crust to the mantle in B1. Panels A3 and
B3: For subduction-unrelated types (continental margin [CM],
mid-ocean-ridge [MOR], and plume [P]), there is no distinct,
regular change with time. There may be large (for the most
incompatible elements) to moderate (less incom-patible to
compatible elements) changes in the element concentrations, as
indicated by the vertical arrows. For the subduction-related
ophiolites, there is a distinct element change from the youngest to
the oldest components of the ophiolites. The blank horizontal
arrows pointing in opposite directions in B3 indicate that the
compositions of mid-ocean-ridge ba-salt (MORB)like to island-arc
tholeiite (IAT) to boninitic may change to lower or higher contents
of the elements indicated. Abbreviations: A1 (CM-type): U.
Crustupper crust; L. Crustlower crust; Serp. perid.serpentinized
peridotite; A1 (P-type): Cont.con-tinent; B1: MORBmid-ocean-ridge
basalt; IATisland-arc tholeiite; BONboninite. A2 (CM type): Serpt.
perd.serpentinized peridotite; Serp. brecciaserpentinized breccia;
Ppillow lava; Lhzlherzolite; Ol-gabbroolivine gabbro; A2 (MOR
type): Interm.intermediate; Neovolc.neovolcanic; TZtransition zone;
MMoho; DFde-tachment fault. A2 (P type): Gbgabbroic to komatiitic
intrusions; ultr. sillultramafi c sill; picr. bas.picritic basalt;
plw brecciapillow breccia. B2 (suprasubduction-zone type): MORB,
IAT, BON; same as in B1; And.andesitic lava; Trndj. Ntrondhjemite
intrusions. B2 ( volcanic-arc type): Rhy.rhyolite; And.
lavaandesitic lava; Gran./ton.granite/tonalite plutons; Gbgabbro;
Didiorite; DMdepleted mantle; L, M, and HREElight, middle , and
heavy rare earth elements.
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 401
Dep
th (k
m)
010203040
100 50 0 50 100
010203040
km
PostriftSynrift sediments
Dep
th
1000
2000
100
4000
Dep
th (km
)
0
1000
2000
100
4000
0
nonconservative
Backarc
P
dikes
Subcontinentalmantle (Lhz)
Sea level (SL)Serp. breccia/
ophicalcite Chert
CM type
massivelava
SL
UndifferentiatedOcean Crust
Plume source
P typeMOR type
SL
TZ
DF
0.3 km
Fast Interm. Slow
0.5 km 0.5 km Depleted mantle
P
1 km
SLSSZ type
TimeIATMORB-like
daciteBon.
And.lava
Rhy.
Gran./ ton.
DM DMStronglydepleted mantle
SL
VA typevolcaniclastic/
pyroclastic rocks
10 km
Basalt lava
A2
A1
A3
B1
B2
B3
Subduction - unrelated ocean crustContinental margin (CM)
type
Mid-ocean ridge (MOR) type
Plume (P) type
CM - MOR - P types SSZ - VA types
Subduction - related ocean crustSuprasubduction - Zone (SSZ)
type
Volcanic arc (VA) type
Shallowintrusion
0.5km
Serpt. perd.
Depleted mantle
M
15 km
Depleted mantle
GbDi
Ol-gabbro
Neovolc.zone
youngpluton
plw breccia
GbGb
600C300C
1200C900C
Partialmelt.zone
fluid flow
Gb ultr. sill
picr. bas.
Trndj Gb
sheeteddikes
andesite
20 km
10km
120km
120 km
AsthenosphereSubcontinental
10 m
.y.
Figure 7.
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
402 Geological Society of America Bulletin, March/April 2011
A B
C D
E F
Lherzolite
Olv-gabbrodikes
Pillow lava
D1
D2
D1D2
Layeredgabbro
Gabbro
Dikes
Dike
Figure 8 (on this and following page). Field photos from
continental margin and various suprasubduction-zone ophiolites,
depicting their in-ternal structure and the crosscutting
relationships of different ophiolitic subunits. (A) Lherzolitic
peridotites of the Jurassic In-Zecca ophio-lite (continental margin
type) in the Ligurian ophiolites (eastern Corsica) intruded by
irregular olivine gabbro dikes and veins. (B) Pillow lavas with
normal mid-ocean-ridge basalt (N-MORB) geochemical affi nities,
resting directly on serpentinized peridotites of the In-Zecca
ophiolite. (C) Tectonically brecciated pillow lavas (in B), showing
cataclastic shearing in and around the pillow-shaped fl ows. (D)
Layered gabbro rock in the 493 Ma Karmy ophiolite
(suprasubduction-zone backarc to forearc [BA-FA] type) in western
Norway intruded by ba-saltic dikes (D1) with MORB affi nities that
are in turn crosscut by boninitic dikes (D2). (E) Sheeted
dikegabbro transition zone (Karmy ophiolite), where leucocratic
gabbros and basaltic dikes show mutually intrusive relationships in
a Penrose-type crustal pseudostratig-raphy. (F) Pillow lavas with
island-arc tholeiite (IAT) geochemical affi nities in the Karmy
ophiolite crosscut by an island-arc tholeiite dike.
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 403
GH
I
JK
L
Myl
oniti
c ga
bbro
Bon
initi
csi
ll
D2
D1
D3
Bon
initi
cla
va
Cpx-
Harz
burg
ite
Orth
opyr
oxen
itedi
kes
& ve
ins
Figu
re 8
(con
tinue
d). (G
) Clin
opyr
oxen
e po
rphy
rocl
ast-b
eari
ng h
arzb
urgi
te in
the
Mid
dle
Jura
ssic
(165 M
a) Ea
stern
Mird
ita op
hiolit
e of
Alb
ania
(sup
rasu
bduc
tion-
zone
BA-
FA ty
pe),
cross
cut b
y ne
twor
ks o
f ort
hopy
roxen
ite d
ikes
, dik
elet
s, an
d ve
ins.
Thes
e in
trus
ions
re
pre-
sen
t bon
initi
c m
elt c
hann
els t
hat m
igra
ted
upw
ard
into
the
refr
acto
ry h
arzb
urgi
te (s
ee D
ilek a
nd M
orish
ita, 2
009;
Mor
ishita
et al
., 20
10).
(H) P
lastic
ally d
eform
ed la
yere
d ga
bbro
s in
the 9
2 M
a K
izild
ag o
phio
lite i
n so
uthe
rn T
urk
ey (s
upra
subd
uctio
n-zo
ne FA
ty
pe), i
ntru
ded
by
a b
onin
itic
sill a
nd a
dik
elet
. (I
J) Sh
eeted
dike
swar
ms (m
oder
ately
to ve
rtica
lly di
pping
) in t
he K
izilda
g oph
iolite
. (K)
Bas
alti
c an
desit
e di
kes (
D1) w
ith an
islan
d-ar
c th
olei
ite a
ffi n
ity, in
trud
ed b
y pl
agio
gran
ite d
ikes
(D2),
whic
h are
in tu
rn cr
oss
cut b
y a
late
-sta
ge b
onin
itic d
ike
(D3).
(L) B
onini
tic la
vas (
sak
alavit
es) in
the K
izilda
g oph
iolite
. See
Dile
k and
Thy
(200
9) for
de
tails
.
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Dilek and Furnes
404 Geological Society of America Bulletin, March/April 2011
Suprasubduction-zone and volcanic-arc ophio-lites show a
characteristic geochemical evo-lution. In the early stages of their
formation, magmas are MORB-like, but during repeated episodes of
melting, their mantle source be-comes progressively depleted in the
most incompatible elements. The geochemical evolu-tion of
suprasubduction-zone and volcanic-arc ophiolitic magmas is
characterized by low abun-dances of incompatible elements (Cs, Rb,
Ba, U, Ta Nb, and light [L] REEs) in basaltic andesites, andesites,
and dacites, which commonly occur in the upper parts of their
extrusive sequences, and in young crosscutting dikes in sheeted
dike complexes. With repeated melting, the residual mantle source
is progressively enriched in oli-vine and orthopyroxene, the
principal hosts of compatible elements such as Ni, Co, Cr, and
Sc.
At a later stage in the magmatic evolution of
suprasubduction-zone ophiolites, there is a change from depletion
to enrichment in incom-patible element contents in the younger
igneous rocks relative to MORB; the more incompatible an element
is, the more pronounced its enrich-ment becomes in many
suprasubduction-zone ophiolite lavas. This phenomenon suggests that
the mantle source undergoes enrichment of highly mobile elements
during or before the ex-traction of MORB-like magmas from it. It is
the nonconservative, highly incompatible elements, Cs, Rb, Th, and
U, that show the most pro-nounced change from depletion to
enrichment during the late-stage evolution (Fig. 7B3); the other
highly incompatible but conservative ele-ments, such as Ta and Nb,
remain depleted (e.g., Pearce and Parkinson, 1993). Pb and Sr seem
to be enriched at an earlier stage than the other non-conservative
incompatible elements, and these elements, particularly Pb,
increase in concentra-tion from the island-arc tholeiite magmatic
stage to the fi nal boninite activity (Fig. 7B3).
Enrichment of the source mantle in slab-derived, nonconservative
elements is a com-plex process that may involve fl uids released
from altered oceanic crust and its sedimentary cover and felsic
magmas generated by partial melting of subducted sediments (Pearce
and Parkinson, 1993; Hawkesworth et al., 1997; Macdonald et al.,
2000; Elburg et al., 2002). Thus, during the generation of
subduction-related ophio lites, two dominant, contempora-neous
processes oper ate to continuously modify the source region and are
responsible for the typical trace-element patterns of the magmas
produced: (1) Repeated episodes of partial melt-ing progressively
deplete the mantle source in incompatible elements and enrich it in
compat-ible elements. Inhomo geneities in the mantle source and
variable degrees of partial melting could also result in variable
concentrations of
incompatible elements in the magmas produced. (2) The mantle
melt source becomes enriched in highly incompatible,
nonconservative elements (particularly Cs, Rb, Ba, Th, U)
transported in subduction-derived fl uids and/or felsic melts.
Application to Precambrian Greenstone Belts
We have selected three Precambrian green-stone belts ranging in
age from Paleoprotero-zoic (Jormua, Finland) to Neoarchean (Wawa,
Canada) and Paleoarchean (Isua, Greenland), for the purpose of
comparing the published geo-chemical data for the volcanic and
subvolcanic rocks of these sequences with the Phanerozoic ophiolite
types as classifi ed herein.
Isua Supracrustal BeltThe mafi c-ultramafi c units of the ca.
3.8 Ga
Isua supracrustal belt in Greenland occur in two major
tectonostratigraphic units, namely the un-differentiated
amphibolites (UA) and Garben-schiefer amphibolites (GA) (e.g.,
Nutman et al., 1984, 1997; Rosing et al., 1996; Komiya et al.,
1999; Furnes et al., 2007, 2009). The undifferen-tiated
amphibolites unit contains all major litho-logical units of a
typical Penrose-type, complete ophiolite sequence, whereas the
Garbenschiefer amphibolites unit is composed dominantly of
volcaniclastic and volcanic rocks that are com-monly found in
immature island arcs.
Wawa Greenstone BeltsThe 2.7 Ga Wawa greenstone belt of the
Superior Province in Canada consists of Al-undepleted and
Al-depleted komatiites and Mg- and Fe-tholeiites (Polat et al.,
1998, 1999). Compositionally, these mafic volcanic and plutonic
rocks are comparable to Phanerozoic ocean plateau basalts that
subsequently were tectonically imbricated with primitive arc
ba-salts (Polat et al., 1998, 1999).
Jormua ComplexThe 1.95 Ga (Peltonen et al., 1996) mafi c to
ultramafi c rocks of the Jormua Complex (JC) occur in the
central part of an early Protero-zoic (2.31.92 Ga) metasedimentary
sequence that is surrounded by Archean basement rocks in
northeastern Finland (Kontinen, 1987; Peltonen et al., 1996). The
Jormua Complex includes pillow lavas and volcanic breccias, a
sheeted dike complex, mafi c cumulates, and upper-mantle
peridotites, and it is tectonically disrupted into several blocks.
The thickness of the Jormua Complex varies, and in places the lava
sequence rests directly upon the upper-mantle rocks, typical of the
Ligurian ophiolites in the Apennines. The crustal architecture of
the
Jormua Complex is reminiscent of that seen in slow-spreading
oceanic crust and in continental margin ophiolites (Peltonen et
al., 1996, 2003).
SummaryIn the Bowen diagrams (MgO-TiO2), the
younger Garbenschiefer amphibolites of the Isua supracrustal
belt plot exclusively in the fi eld of subduction-related
ophiolites, whereas the un-differentiated amphibolites plot both in
the subduction-related and subduction-unrelated fi elds (Fig. 9A).
The Wawa and Jormua meta-basalts plot predominantly in the fi elds
of plume and continental margin types of the subduction-unrelated
ophiolites, respectively (Figs. 9B and 9C). In the multi-element
diagrams, the undif-ferentiated amphibolites of Isua plot within
the fi eld of subduction-related ophiolites and display their
characteristic features, such as positive Pb anomalies, negative Nb
and Ta anomalies, and strong enrichment of Ba and Th. On the other
hand, the Garbenschiefer amphibolites show strong depletion of the
middle (M) REEs, a typi-cal feature of boninites (Fig. 10A). The
Wawa and Jormua metabasalts plot within the fi eld defi ned by the
subduction-unrelated ophiolites and display the same features of fl
at to moder-ately enriched patterns as the incompatibility of the
ele ments increase (Figs. 10B and 10C). In the Ti-V discrimination
diagram, the Isua data plot in two distinct fi elds, with the
Garben-schiefer amphibolites exclusively in the boninite fi eld
(Ti/V < 10), whereas the undifferentiated amphibolites have Ti/V
ratios of 2030 (Fig. 11A) in the mixed MORB and island-arc fi elds
(Shervais, 1982). The volcanic and dike rocks of the Wawa and
Jormua sequences, on the other hand, plot entirely within the plume
and conti-nental margin types, respectively, of
subduction-unrelated ophiolites (Figs. 11B and 11C). In the
Nb/Yb-Th/Yb discrimination diagram, all the Isua data plot in the
subduction-related fi eld (Fig. 12A), whereas the Wawa and Jormua
data plot in the subduction-unrelated fi eld (Fig2. 12B and 12C).
The Wawa data defi ne a large spread between N-MORB and
oceanic-island basalt (though mostly between N-MORB and E-MORB),
while the Jormua data cluster tightly around E-MORB (Figs. 12B and
12C).
The geochemical character of the metavol-canic and intrusive
rocks of the three selected Precambrian greenstone belts indicates
that they originated in different tectonic environments. Thus,
compared with the geochemical evolu-tion of Phanerozoic ophiolites,
the Paleoarchean Isua rocks most likely represent a
supra-subduction-zone forearc basin subtype ophio-lite, as
suggested by Furnes et al. (2009). The Neo archean Wawa greenstone
belt, on the other hand, is more akin to the structural and
geo-
on February 17, 2011gsabulletin.gsapubs.orgDownloaded from
-
Ophiolite genesis and global tectonics
Geological Society of America Bulletin, March/April 2011 405
chemical character of plume-type ophiolites, in agreement with
the interpretations of Polat et al. (1999). The early Proterozoic
Jormua Complex resembles, both structurally and geochemically,
continental margintype ophiolites, consistent with the
interpretations of Peltonen et al. (2003).
CONCLUSIONS
Ophiolites are diverse in their internal struc-ture, geochemical
makeup, and emplacement mechanisms, and they form in different
tec-
tonic environments during the Wilson cycle evolution of ancient
ocean basins from rift-drift and seafl oor spreading stages to
subduction ini-tiation and closure phases. Mafi c-ultramafi c to
felsic rock assemblages that originally formed in different
tectonic settings may eventually be-come nested in collision zones,
forming distinct ophiolite complexes with signifi cant diversity in
their structural architecture, geochemical fi ngerprints, and
emplacement mechanisms. Differences in the magmatic and structural
archi tecture of ophiolites result from their prox-
imity to plumes or trenches, rates and geometry of spreading,
mantle temperatures and fertility, and the availability of fl uids
in the tectonic set-ting of formation during their primary igneous
evolution. Ophiolites are broadly su