HAL Id: hal-02733950
https://hal-brgm.archives-ouvertes.fr/hal-02733950
Submitted on 2 Jun 2020
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
History of Subduction Polarity Reversal During
Arc-Continent Collision: Constraints From the
Andaman Ophiolite and its Metamorphic Sole
Alexis Plunder, Debaditya Bandyopadhyay, Morgan Ganerod, Eldert L
Advokaat, Biswajit Ghosh, Pinaki C Bandopadhyay, Douwe J van Hinsbergen
To cite this version:
Alexis Plunder, Debaditya Bandyopadhyay, Morgan Ganerod, Eldert L Advokaat, Biswajit Ghosh, et
al.. History of Subduction Polarity Reversal During Arc-Continent Collision: Constraints From the
Andaman Ophiolite and its Metamorphic Sole. Tectonics, American Geophysical Union (AGU), 2020,
�10.1029/2019TC005762�. �hal-02733950�
History of Subduction Polarity Reversal During
Arc‐Continent Collision: Constraints From the
Andaman Ophiolite and its Metamorphic Sole
Alexis Plunder
1,2
, Debaditya Bandyopadhyay
3,4
, Morgan Ganerød
5
, Eldert L. Advokaat
1,6
,
Biswajit Ghosh
3
, Pinaki Bandopadhyay
3
, and Douwe J. J. van Hinsbergen
1
1
Department of Earth Sciences, Utrecht University, Utrecht, the Netherlands,
2
BRGM, Orléans, France,
3
Department of
Geology, University of Calcutta, Kolkata, India,
4
Department of Geology, University of North Bengal, Darjeeling, India,
5
Geological Survey of Norway NGU, Trondheim, Norway,
6
Department of Physical Geography, Utrecht University,
Utrecht, the Netherlands
Abstract Subduction polarity reversal during arc‐continent collision has been proposed as a key
mechanism to initiate new subduction zones. Despite often interpreted, well‐exposed geological record
that document the reversal is sparse. The ophiolitic lithounits of the Andaman and Nicobar Islands have
been proposed to have formed during the initiation of a new subduction zone following the collision of the
Woyla Arc of Sumatra with Sundaland (Eurasia). We here present new field, petrological and
geochronological data to evaluate the timing of the initiation of Andaman subduction. We targeted the
previously inferred but unstudied metamorphic sole of the Andaman ophiolites that witnessed juvenile
subduction. Thermodynamic modeling reveals that the exposed amphibolites of the sole formed at around
0.9 GPa and 675 °C. We dated two samples of the metamorphic sole using the Ar/Ar method on amphibole,
giving cooling ages of 106.4 ± 2.1 and 105.3 ± 1.6 Ma. This is similar to published ages from plagioclase
xenocrysts in recent Barren Island volcanics and in zircons from a gabbro sample from the Andaman
ophiolite, which we interpret as the age of the original ophiolite formation. The Ar/Ar ages are considerably
older than arc magmatic gabbros and plagiogranites of the overlying ophiolite previously dated at
99–93 Ma and thought to reflect the ophiolite age but recently reinterpreted as a volcanic arc built on the
ophiolite. Combined with the ages of Woyla‐Sundaland collision, we argue that subduction polarity reversal
occurred in a transient period of perhaps some 10 Myr, similar to recent settings.
1. Introduction
The formation of new subduction zones is critical in the plate tectonic cycle. A fundamental way in which
new subduction zones form is through subduction polarity reversal following arc‐continent collision (e.g.,
Chemenda et al., 2001; Dewey, 1976; Faccenda et al., 2008; Pysklywec, 2001; Stern, 2004; Stern &
Gerya, 2017) where upon ongoing convergence, the arrest of one subduction zone is followed by the
initiation of another, typically within or behind the older arc (Figure 1). The arrest of the older subduc-
tion zone is then often related to arrival of buoyant continental or thickened oceanic lithosphere on the
original downgoing plate in the trench. For instance, the arrival of the Ontong‐Java Oceanic Plateau in
the Vitiaz Trench led to the formation of the New Hebrides subduction zone in the Vitiaz/Melanesian
Arc (Hall, 2002; Knesel et al., 2008). Subduction polarity reversal following arc‐continent collision was
also proposed in the Aleutian Arc (Vaes et al., 2019) or to form the modern Kamchatka subduction zone
(Domeier et al., 2017; Konstantinovskaia, 2001; Shapiro & Solov'ev, 2009; Vaes et al., 2019). Modern
examples where polarity reversal may be ongoing are the Banda Arc in Timor (e.g., Breen et al., 1989;
Harris, 2006; Tate et al., 2015), Taiwan (Chemenda et al., 2001), or the Solomon Arc (Cooper &
Taylor, 1985; Cooper & Taylor, 1987). Despite the widespread recognition of arc polarity reversal in the
geological record, the dynamics and longevity of the reversal process are poorly known. The duration
of polarity reversal, that is, the time between subduction stops upon arc‐continent collision and the time
at which the new subduction is fully developed (see Figure 1) is however not known from direct evi-
dences. Estimating the kinematic history of the transition requires a geological record that allows dating
the arrest of the old subduction (ceasing of the arc) and the beginning of the new one, which is challen-
ging given the destructive nature of active margins.
©2020. American Geophysical Union.
All Rights Reserved.
RESEARCH ARTICLE
10.1029/2019TC005762
Special Section:
Tethyan dynamics: from rift-
ing to collision
Key Points:
• We document the condition of
formation of the metamorphic sole
of the Andaman Island ophiolite
• We provide Ar/Ar ages for the
subduction initiation leading to the
formation of the Andaman Island
ophiolite
• We estimate the duration of
subduction polarity reversal
following an arc‐continent collision
to 10 Myr at least
Supporting Information:
• Supporting Information S1
Correspondence to:
A. Plunder,
a.plunder@brgm.fr
Citation:
Plunder, A., Bandyo padhyay, D.,
Ganerød, M., Advokaat, E. L., Ghosh,
B., Bandopadhyay, P., & van
Hinsbergen, D. J. J. (2020). History of
subduction polarity reversal during
arc‐continent collision: Constraints
from the Andaman ophiolite and its
metamorphic sole. Tectonics, 39,
e2019TC005762. https://doi.org/
10.1029/2019TC005762
Received 12 JUL 2019
Accepted 6 MAR 2020
Accepted article online 10 MAR 2020
PLUNDER ET AL. 1of24
The ophiolitic sequence of the Andaman Islands in the eastern Indian
Ocean, however, may preserve such a record. The Andaman and
Nicobar Archipelago is located in the forearc of the Sunda‐Sumatra sub-
duction zone (Figures 2a and 2b). It exposes a thrusted sequence of vari-
ably depleted and chromitite‐bearing peridotites. The peridotites are
overlain by mafic magmatic rocks and underlain by a
serpentinite‐hosted mélange that contains sheared greenschist‐ and
amphibolite‐facies blocks, as well as radiolarian chert blocks with strati-
graphic ages up to the middle Eocene (Bandopadhyay & Carter, 2017a;
Bandyopadhyay et al., 2020; Ghosh et al., 2009, 2017; Ling et al., 1996;
Sengupta et al., 1990). The ophiolitic rocks are unconformably overlain
by Paleocene to Eocene shallow‐marine sandstones derived from a volca-
nic arc, and after the middle Eocene, the sequence thrusted and tectoni-
cally repeated (Bandopadhyay & Carter, 2017b).
The variably depleted peridotites and the magmatic sequence of the
ophiolites have long been considered to represent one coherent but tec-
tonically dismembered suprasubduction zone (SSZ) ophiolite sequence,
(i.e., an ophiolite with an arc signature but with a structure of oceanic
lithosphere; Pearce et al., 1984; Stern, 2004; for the Andaman ophiolite,
see Ghosh et al., 2009; Pal, 2011; Pedersen et al., 2010). Magmatic rocks
in the sequence include in places plagiogranites (trondhjemites) and are
overlain by Upper Cretaceous (Campanian) radiolarian cherts (Ling
et al., 1996). Whole rock Sm/Nd ages on different component of the
ophiolite (basalt, plagiogranite and peridotite) yielded a 98±8.2 Ma iso-
chron (Bhattacharya et al., 2020). Two trondhjemites samples gave zir-
con U/Pb ages of of 93.6±1.3 Ma (Sarma et al., 2010) and 95±2Ma
(Pedersen et al., 2010) that were interpreted to represent the formation
of Andaman ophiolite crust above a subduction zone. Whereas
Pedersen et al. (2010) interpret them to represent subduction initiation,
other authors argue a geochemical signature typical of a mature arc
rather than an SSZ fore‐arc setting (Jafri et al., 1995; Sarma et al., 2010).
Near Chiriyatapu, trondhjemites are found as blocks in an andesitic
agglomerate. The agglomerates display arc characteristics (explosive
intermediate volcanism) and are generally found overlying the ophiolite.
In the field, the absence or really minor presence of any other ophiolitic
clasts (i.e. basalt, peridotite), and their disposition suggest that they are
cogenetic with the andesites. Combined, these data show that there has
been at least 5 Myr of magmatism with a diverse geochemical signature,
suggesting an arc setting for parts of the Andaman ophiolites. Finally,
thin ash layers within Upper Cretaceous radiolarian cherts that interca-
late with pillow lavas of South Andaman also suggest that there was
explosive volcanism during or not long after the magmatic sequence
of the ophiolite formed (Jafri et al., 2006).
Both the 5‐Myr age range as well as the broad range of geochemical signa-
ture and composition of the volcanics of the Andaman ophiolite are sur-
prising (Ghosh et al., 2017). SSZ ophiolites are typically found in fore‐arc settings where they form during
incipient upper plate extension following subduction initiation (e.g., Guilmette et al., 2018; Stern et al., 2012).
Mature volcanic arcs occur at a distance of typically ~100–300 km away from the trench (Dickinson, 1973)
with an average around 166 ± 66 km (Gill, 1981) or between 180 and 275 km (Syracuse & Abers, 2006). The
juxtaposition of the SSZ forearc and arc may thus require tectonic motion from a forearc to a back‐arc
domain or that SSZ spreading occurred within a mature arc setting. In the relatively restricted area of the
Andaman Islands (350 × 40 km), it is surprising to find a 5‐Myr duration at least for magmatism since
already formed crust moves at half‐spreading rate from the spreading centers. Typical magmatic
Figure 1. Conceptual sketch of subduction polarity reversal. The first stage
depicts oceanic subduction below a mature arc. The red lines show the
subduction interface where one lithosphere slides against the other one.
The second sketch depicts the continent entering the subduction zone, and
the red star emphasizes the collision between the continent and the arc.
The dashed red line shows the mechanically weak place where the new
subduction will start. The third cartoons show the effective subduction
polarity reversal. The red lines again show the subduction interface where
one lithosphere slides against the other one. The cartoon is inspired by
Chemenda et al. (2001).
10.1029/2019TC005762
Tectonics
PLUNDER ET AL. 2of24
Figure 2. (a) Location of the Andaman Island in the Indian Ocean realm (modified from Ghosh et al., 2017). (b) Schematic geological map of the Andaman
Islands modified from Ghosh et al. (2017). (c) Sketch of the investigated area redrawn from field observation. Color of the symbols denotes the nature of the
rock. The shape of the symbol denotes the character of the observation/sample: in situ of flying blocks. (d) Field view of the in situ clinopyroxene‐bearing
amphibolite (AN1709a sample). The red star refers to the position of sample AN1709 on the localization map. Both insets show more detailed picture on the in situ
location and of a boulder (sample AN1704). The right panel shows the in situ location of the greenschist‐facies metamorphic sole characterized by the alternation
of mafic and quartz‐rich layers. The pink star refers to the position on the map. Mineral abbreviations are after Whitney and Evans (2010).
10.1029/2019TC005762
Tectonics
PLUNDER ET AL. 3of24
spreading rates are of several tens of kilometer per million years. It means that the ages of magmatic rocks in
narrow ophiolite bodies like on Andaman are typically within 1–2 Myr (e.g., in Oman, Rioux et al., 2016, 2013;
or Turkey, van Hinsbergen et al., 2016; and reference therein; Parlak et al., 2019). Instead, the minimal
5‐Myr transition between SSZ and arc magmatism that are found within a few meters in Andaman suggests
a more or less stationary magma source after the SSZ ophiolite is formed. After it formed around 93 Ma, the
arc is believed to be active at least until the Paleocene to Eocene as attested by arc‐derived sandstone
(Bandopadhyay, 2012).
Advokaat et al. (2018) recently suggested that when corrected for post‐Cretaceous tectonic motion along
the Sunda forearc, the Andaman ophiolites restore adjacent to and west of the extinct intraoceanic
Woyla Arc that is now located on Sumatra and West Java. This arc collided with the Sundaland margin
in the late Cretaceous. Advokaat et al. (2018) postulated that the Andaman SSZ ophiolites may have
formed during a subduction polarity reversal within or adjacent to the Woyla Arc upon its collision with
Sundaland. In this paper, we study sheared amphibolites below the Andaman ophiolite that are inter-
preted as their metamorphic sole (Pal & Bhattacharya, 2010). Metamorphic soles are slivers of oceanic
crust stripped of the downgoing slab that are formed and underplated beneath the mantle wedge during
subduction infancy. They form during subduction initiation (Wakabayashi & Dilek, 2000) over a short
time span (typically of 1–2 Myr) and cool coevally to the formation of SSZ ophiolite (Hacker, 1994;
Rioux et al., 2016). Worldwide, they share very similar characteristic: they are formed in a warm ther-
mal regime, up to melting condition (850 °C and 1.2 Gpa), in a much higher thermal regime than
expected for a mature subduction zone (Agard et al., 2016; Dubacq et al., 2019; Jamieson, 1986; Soret
et al., 2017; Wakabayashi & Dilek, 2003; Woodcock & Robertson, 1977). They comonly have a 10‐ to
500‐m thickness in total, and parts with different peak metamorphic grades represent different tectonics
slices. The higher temperature part of the sole is made of mafic oceanic crust. The lower parts tend to
show an increase of a sedimentary component (Casey & Dewey, 1984). Cooling ages of metamorphic
soles (i.e., obtained by the Ar/Ar method on amphibole) are globally strongly correlated to the age of
the crust of the overlying SSZ ophiolites, suggesting that decompression and exhumation of soles are
the result of upper plate spreading (Dewey & Casey, 2013; van Hinsbergen et al., 2015). Such cooling
ages may thus serve as a proxy for the age of the SSZ ophiolite spreading (or hyperextension).
However, they provide only a minimum age for the initiation of subduction, which may predate SSZ
spreading by ~10 Myr or more as recently shown by Lu‐Hf garnet ages in soles (Guilmette et al., 2018;
Pourteau et al., 2019).
The metamorphic sole are found below the ophiolites on Middle and North Andaman Islands
(Figures 2a–2c). We establish the pressure‐temperature (PT) condition at which they formed using empiri-
cal thermobarometry and pseudosection modeling. We provide Ar/Ar ages to establish their cooling as
proxy for the timing of SSZ ophiolite formation. We finally discuss our results in terms of temporal
and spatial relationships between SSZ ophiolite evolution, magmatic evolution, and subduction polarity
reversal dynamics.
2. Geological Setting
2.1. Regional Setting of the Andaman‐Nicobar Islands
The Andaman and Nicobar Islands are located at the northernmost extension of the present‐day
Sunda‐Sumatra subduction system where the Indo‐Australian Plate is subducting below Eurasia
(Figure 2a; Curray, 2005, 1989; McCaffrey, 1992, 2009). To the east of the Andaman and Nicobar Islands
is the Andaman Sea, an approximately N‐S spreading, Neogene pull‐apart basin that formed in the
back‐arc region of the Andaman subduction zone as a result of formation of the Burma‐Andaman
fore‐arc sliver, partitioning highly oblique India‐Sundaland convergence over a trench in the west and a
transform system in the east (Curray, 2005). Restoring the opening of this basin brings the
Cretaceous‐Paleogene rock record of the Andaman Islands toward the south, juxtaposed in Oligocene time
against West Sumatra (Curray, 2005).
On West Sumatra lies the accreted, formerly intraoceanic Woyla volcanic arc. This arc formed in earliest
Cretaceous time within the Neotethys ocean (Hall, 2012). Paleomagnetic data from the Woyla Arc give
paleolatitudes that are consistent with the arc having formed on the Australian Plate that until the
10.1029/2019TC005762
Tectonics
PLUNDER ET AL. 4of24