Content uploaded by Oguz H. Göğüş
Author content
All content in this area was uploaded by Oguz H. Göğüş on May 16, 2021
Content may be subject to copyright.
Articles
https://doi.org/10.1038/s41561-021-00746-9
1Department of Earth Sciences, University of Toronto, Toronto, Canada. 2Eurasia Institute of Earth Sciences, Istanbul Technical University, Istanbul, Turkey.
✉e-mail: erkan.gun@mail.utoronto.ca
The continental lithosphere can be extended by a variety
of processes of horizontal (far field/passive rift) and ver-
tical (near field/active rift) tectonics1–3. For instance, the
Alpine–Mediterranean region has an array of locales of rift basins
and metamorphic exhumation in which the continental litho-
sphere undergoes extension in conjunction with plate subduction
and collision4,5. In such settings, continental lithosphere of the
back-arc (overriding) environment extends, and/or the microcon-
tinent (terrane and/or blocks) that accompanies the subducting
oceanic lithosphere undergoes an extension after the collision as
high-pressure exhumation occurs6 (Fig. 1a,b). Both mechanisms
infer that slab-pull forcing transmits stresses through the crust to
yield a trenchward retreat of the subduction (that is, the subduc-
tion rate is faster than the plate convergence)4. A key aspect of these
tectonic deformational events on the overriding continent or in the
microcontinent terrane is their temporal occurrence: the exten-
sion is contemporaneous (syn) or proceeding (post) the collision/
subduction time of the continental block. We propose that there is
another fundamental class of extensional tectonics which occurs
during the pre-collisional stages of the tectonic cycle as microconti-
nents (terranes) drift towards a subduction zone (Fig. 1c), and that
this pre-collisional episode of plate tectonic deformation has been
unrecognized.
Petrological studies in the Western Alps suggest that the burial
and exhumation of polymetamorphosed (high pressure–low tem-
perature) Austro-alpine continental crust (including eclogites) of
the Sesia–Lanzo zone (microcontinent) occurs through the devel-
opment of extensional and/or detachment faulting7 (Figs. 1 and 4a).
Notably, these events occur prior to the Eocene continent–conti-
nent collision in the Alps8. It is probable that there were overprint-
ing effects of deformation, and metamorphism in conjunction with
syn- and post-collision, but the pre-collisional timing of the main
phase of exhumation is unequivocal. This suggests an episode of
extension on the Sesia–Lanzo microcontinent prior to collision
and precludes geodynamic explanations for lithospheric stretching,
such as post-orogenic gravitational collapse9,10.
If microcontinent terranes experience an appreciable extension
prior to their continental docking at collision, this has important
implications. Terrane assembly is a fundamental step of continental
tectonics11 and accreted terranes may include intra-oceanic islands,
rifted microcontinental blocks, seamounts or oceanic aseismic ridges
and oceanic plateaus, which vary in size from a few to several thousand
kilometres12. Hence, the pre-collisional extension in such microconti-
nents could be a widespread phenomenon. For example, the entire
Alpine–Himalayas chain and North American Cordillera are intrinsi-
cally the result of an amalgamation of such a series of terranes13,14.
In this work, we explore the dynamics of how the pre-collisional
extension of microcontinents may evolve and how such events may
manifest in the metamorphic record. Computational forward mod-
els quantify the thermomechanical evolution of microcontinents
during drift and collision stages in an ocean lithosphere subduc-
tion system. The models monitor the metamorphic evolution of dis-
crete packages of crust during the convergence and collision. These
data are reconciled with petrological data of Tethyan microconti-
nent terranes in the eastern Mediterranean and Western Alps that
yielded the geological interpretations for pre-extensional collision.
The pre-collisional extension presents an unexplored and enigmatic
phase of tectonic activity in the plate subduction–collision system
that yields important new clues on the plate tectonic assembly of
the lithosphere.
Models for pre-collisional extension
To test quantitatively the deformational behaviour of microconti-
nents prior to a collision–accretion event, we used the geodynamic
numerical modelling tool SOPALE15. SOPALE is a geodynamic
code that solves plane strain deformation of viscoplastic materials
using arbitrary Lagrangian–Eulerian techniques. Extended Data
Fig. 1 and Extended Data Table 1 describe the configuration of the
Pre-collisional extension of microcontinental
terranes by a subduction pulley
Erkan Gün 1 ✉ , Russell N. Pysklywec 1, Oğuz H. Göğüş 2 and Gültekin Topuz 2
Terrane accretion is a ubiquitous process of plate tectonics that delivers fragments of subduction-resistant lithosphere into a
subduction zone, resulting in events such as ocean plateau docking or continental assembly and orogenesis. The post-collisional
extension of continental terranes is a well-documented tectonic process linked with gravitational collapse and/or trench retreat.
Here we propose that microcontinental terranes can also undergo a substantial extension before their collision with the upper
plate, owing to pull from the trenchward part of the subducting plate. Forward geodynamic numerical experiments demonstrate
that this pre-collisional extension can occur over a protracted phase on microcontinents that are drifting towards a subduction
zone, which distinguishes the deformation from post-collisional extension on the overriding plate, as is traditionally postulated.
The results show that the magnitude of pre-collisional extension is inversely correlated with the size of the microcontinental
terrane and imposed convergence velocity. We find that locations along the Tethyan belts, namely, the Sesia zone and Eastern
Anatolia, are evidence for this style of pre-collisional extension, as this mechanism reconciles with geothermobarometric data
and kinematic analyses. The operation of this subduction pulley reveals that drifting lithospheric plates may undergo substan-
tial tectonic events before the arrival and involvement with regular plate boundary processes.
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles NATurE GEoSciENcE
models. A full explanation of the governing equations and formula-
tion for rheology is given in Methods and in numerous previous
applications of the SOPALE code16–20.
The models test varying microcontinent widths from 120 to
600 km and three cases of convergence velocity: 0 cm yr–1 (litho-
spheric motion is driven only by the subducting ocean plate), and 2
and 4 cm yr–1 convergence velocity (as a proxy for large-scale plate
tectonics, such as ridge push). Supplementary Figs. 1 and 2 show
sets of additional test models that compare a range of different rhe-
ology parameters, a continental overriding plate and an additional
convergence rate; the Supplementary Table lists relevant modelling
parameters.
Model EXP-1 has a 120-km-wide microcontinent with a 0 cm yr–1
imposed convergent rate (Fig. 2). The left column in Fig. 2 shows
the model at 5.5 Myr when the subducting oceanic lithosphere is
consumed and the microcontinent begins to collide with the over-
riding oceanic lithosphere. By 7.1 Myr (centre column) there has
been slab break-off of the subducting lithosphere and the micro-
continent shortens. The microcontinental block largely resists being
consumed, owing to the buoyancy of the overlying continental
crust. We tracked a number of discrete crustal rock locations in the
microcontinent to reveal the pressure–temperature (P–T) evolu-
tion: three from the lower crust (indicated by purple, magenta and
teal colours), and one (black) from the upper crust (Fig. 2). The
tracked rock packages were chosen primarily from the lower crust
to consider the full exhumation of the entire crust. These locations
are also horizontally distributed across the microcontinent to detect
any lateral variations of the extension that may manifest. One P–T–
time (P–T–t) path of an upper crustal rock (black) near the colli-
sion zone is also tracked. The P–T–t paths of the tracked rocks are
indicated in the far-right column. Solid lines show the P–T–t paths
of the tracked rocks for the first 5.5 Myr (Fig. 2, right column), and
dashed lines are paths between 5.5 and 7.1 Myr. The lower crustal
rocks at the centre (magenta) and the far end (teal) of the micro-
continent undergo continuous exhumation in the first 5.5 Myr of
the experiment, which corresponds to a 2.5 kbar (0.25 GPa) drop in
pressure (or ~8.7 km exhumation). The lower crustal rock package
close to the subduction zone (purple) first exhumes ~7.4 km, but
then it is rapidly buried as it becomes caught in the collision and
continental subduction. The upper crustal material at the near end
of the microcontinent (black) is buried for the first 5.5 Myr of the
model evolution, but undergoes later exhumation on a fault hanging
wall amidst the collision.
The notable behaviour of the model is the exhumation of crustal
rocks during the initial phase of plate consumption prior to colli-
sion. That is, the microcontinent undergoes pre-collisional exten-
sion as it drifts towards the plate boundary. The later events relate
to collision and depend on the proximity of the rock packages to
the collision front. For instance, near the collision front, the oro-
genically buried upper crustal material (black) starts to unroof. At
the centre of the microcontinent, the lower crustal rock package
(magenta) shows a post-collisional sequence of burial and exhuma-
tion. At the distal end of the microcontinent, the lower crustal mate-
rial (teal) is not appreciably affected by the collision and break-off
events. The lower crustal package in the subducted part of the
microcontinent (purple) continues its burial further into depths
below the lithosphere. This is a consequence of the location of the
slab break-off. The break-off happens in the subducted extent of
the microcontinent, as its lithospheric strength is lower than that of
the oceanic lithosphere21.
For EXP-2, the plate convergence velocity remains at 0 cm yr–1,
but the width of the microcontinent increases to 300 km (Fig. 2).
The experiment shows a similar evolution to that of EXP-1,
with consumption of the downgoing ocean lithosphere and drift
of the microcontinental block to the subduction zone, followed by
microcontinent collision and eventual slab break-off. The P–T–t
N
0 500 1,000 km
Continental block
a
Sesia BriançonnaisAdria Europe
85 Ma
d
c
Slab pull
This work
Extension
Accretion/
exhumation
b
Slab retreat
Extension
Black Sea
Aegean
Sea
Tyrrhenian
Sea
Alboran Sea
Arabian Platform
Anatolia
Pannonian Basin
Atlas Mountains
I
z
m
i
r
–
A
n
k
a
r
a
Adriatic Sea
Iberia
Pyrenees
Alps
Greater Caucaus
EAF
EURASIA
AFRICA
Sesia
Akdağ
Metamorphism
Exhumation
Collision
Sesia Akdağ
100 Ma
80–60 Ma
50 Ma
83 Ma
71 Ma
50 Ma
Pre-collisional extension locations
50° N
48° N
46° N
44° N
42° N
42° N
38° N
36° N
34° N
32° N
30° N
46° E44° E42° E40° E38° E36° E34° E32° E30° E28° E26° E24° E20° E18° E16° E14° E12° E10° E8° E6° E4° E2° E0°2° W4° W6° W8° W10° W 22° E
V
a
r
d
a
r
s
u
t
u
r
e
C
a
r
p
a
t
h
i
a
n
s
N
o
r
t
h
A
n
a
t
o
l
i
a
n
F
a
u
l
t
-
E
r
z
i
n
c
a
n
s
u
t
u
r
e
Z
a
g
r
o
s
s
u
t
u
r
e
D
e
a
d
S
e
a
F
a
u
l
t
H
e
l
l
e
n
i
c
T
r
e
n
c
h
Fig. 1 | Map showing main orogenic belts in the Alpine–Mediterranean region. Red stars indicate locations of the microcontinents that underwent
pre-collisional extensional deformation (inset). a, Tectonic setting of a microcontinent approaching a subduction trench. b, Back-arc extension on the
overriding plate due to slab retreat (modified from Brun and Faccenna6). c, Interpreted extensional tectonics of a microcontinent (continental block) in the
pro-plate, discussed in this article. d, Palaeotectonic interpretation for the Western Alps made by Rosenbaum and Lister35 with pre-accretion locations
of several microcontinents including the Sesia block. The base map in the figure was made with GeoMapApp (www.geomapapp.org) under a Creative
Commons licence CC BY 4.056. EAF, East Anatolian Fault.
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles
NATurE GEoSciENcE
tracks for EXP-2 for the first 5.5 Myr follow similar paths com-
pared to those in EXP-1, but the magnitude of exhumation for the
lower crustal rocks is less than that for EXP-1. For example, for the
lower-crust rocks at the centre of the microcontinent (magenta),
this value is 1.7 kbar (0.17 GPa; ~5.9 km uplift) and it is 2.1 kbar
(0.21 GPa; ~7.4 km uplift) for the rock package at the far end of the
microcontinent (teal). Again, the important feature here is a sub-
stantial pre-collisional extension of the microcontinental block.
Break-off of the subducted slab (which occurs in the microcontinent
rather than ocean lithosphere) occurs at ~7.1 Myr, slightly delayed
with the increased microcontinent width in comparison with that of
EXP-1. The timing of this event also manifests in the P–T–t graphs,
in which the magenta- and teal-labelled rocks continue exhumation
to 7.1 Myr until the slab breaks off.
EXP-3 has a microcontinent increased to 450 km (Fig. 2). The
P–T–t paths of the tracked rock locations reveal a similar petro-
logical evolution to EXP-2, with an initial exhumation and inferred
extension of the microcontinent as it traverses towards the subduc-
tion. However, the amount of exhumation is less than that of both
EXP-1 and EXP-2. For instance, over 5.5 Myr the exhumation for
the magenta crustal package is 1.6 kbar (0.16 GPa; ~5.5 km uplift)
and the teal crustal package is 2.0 kbar (0.20 GPa; ~6.9 km uplift).
The slab break-off is further delayed compared with the previous
experiments.
800 1,000 1,200 1,400 1,600 1,800
Distance (km)
120 km300 km450 km600 km
Block
width
Slab break-off
Slab break-off
Slab break-off
Slab break-off
Lithostatic Pressure ( kbar )
Temperature (ºC)
200 300 400
4
6
8
5.5 Myr
7.1 Myr
0 100 200 300 400 50
Temperature (
0
2
4
6
8
10
Lithostatic Pressure ( kbar )
Lithostatic Pressure ( kbar )
Temperature (ºC)
200 300 400
4
6
8
5.5 Myr
7.1 Myr
Lithostatic Pressure ( kbar )
Temperature (ºC)
200 300 400
4
6
8
5.5 Myr
7.1 Myr
EXP-1
0 100 200 300 400 500
Temperature (
0
2
4
6
8
10
Lithostatic Pressure ( kbar )
EXP-2
EXP-3
0 MyrTracked crustal rocks
0 Myr
0 Myr
Initial state
Extended state
Initial state
Extended state
Initial state
Extended state
Oceanic
lithosphere
Oceanic
lithosphere
5.5 Myr
7.1 Myr
Lithostatic pressure (kbar)
4
6
8
5.5 Myr
7.1 Myr
0 100 200 300 400 50
Temperature (
0
2
4
6
8
10
Lithostatic Pressure ( kbar )
Lithostatic pressure (kbar)
4
6
8
5.5 Myr
7.1 Myr
2
4
6
8
10
Lithostatic Pressure ( kbar )
Lithostatic pressure (kbar)
Temperature (°C)
200 300 400
Temperature (°C)
200 300 400
Temperature (°C)
200 300 400
Temperature (°C)
200 300 400
4
6
8
5.5 Myr
7.1 Myr
800 1,000 1,200 1,400 1,600 1,800 2,000
Distance (km)
200
300
400
500
600
Depth (km)
0 Myr
Initial state
Extended state
EXP-4
1,400 1,500 1,600 1,700 1,800 1,900 2,000
1,400 1,500 1,600 1,800 1,900 2,000
1,700
200
300
400
500
600
Depth (km)
200
300
400
500
600
Depth (km)
200
300
400
500
600
Depth (km)
800 1,000 1,200 1,400 1,600 1,800 2,000
Distance (km)
800 1,000 1,200 1,400 1,600 1,800
Distance (km)
800 1,000 1,200 1,400 1,600 1,800
Distance (km)
800 1,000 1,200 1,400 1,600 1,800 2,000
Distance (km)
800 1,000 1,200 1,400 1,600 1,800 2,000
Distance (km)
800 1,000 1,200 1,400 1,600 1,800
Distance (km)
1,400 1,500 1,600 1,700 1,800 1,9001,300 2,000
1,400 1,500 1,600 1,700 1,800 1,900
1,300
5.5 Myr 7.1 Myr
Lithostatic pressure (kbar)
4
6
8
Fig. 2 | Model results showing the effect of microcontinent width with 0cmyr–1 imposed convergence rate. Figures from top to bottom show 120, 300,
450 and 600 km microcontinent widths (EXP-1, EXP-2, EXP-3 and EXP-4, respectively). The colour legend for the rheologies in the frames are the same
as those of the model set-up in Extended Data Fig. 1. Frames on the left show the moment at 5.5 Myr when microcontinents reach the subduction zone.
The inset figures show the initial states of the models and chosen points for P–T–t tracking. The initial depths of the tracked rocks are 7 km for the black,
and 28 km for the purple, magenta and teal symbols. The frames in the middle of the figure show the state of the models at 7.1 Myr when slab break-off
has happened or is about to happen. Supplementary Fig. 3 gives the corresponding second invariant of the deviatoric strain-rate tensors. The graphs on
the right show P–T–t paths of the tracked rocks between 0 and 7.1 Myr with 0.16 Myr subsampling intervals. The colours of the lines are the same colours
those of the tracked rocks in the figures of the left and middle columns. The solid lines represent 0–5.5 Myr and the dashed lines are the phase from 5.5 to
7.1 Myr. 1 kbar = 0.1 GPa.
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles NATurE GEoSciENcE
EXP-4 has a microcontinent of 600 km width and a 0 cm yr–1
imposed convergence rate (Fig. 2). Although the P–T–t paths here
are generally similar to those of the previous experiments, again
there is a further decrease in the magnitude of microcontinent
exhumation and a delay in the timing of the slab break-off.
Figure 3 illustrates trends in the models: Fig. 3a shows microcon-
tinent width versus the maximum lithostatic pressure drop (essen-
tially the magnitude of exhumation) of the tracked rock packages
and Fig. 3b shows width versus the β-factor (extension factor) of
the microcontinent. In addition to the data from the models above
(solid lines), data from experiments with imposed convergence
rates of 2 cm yr–1 (long dashed lines) and 4 cm yr–1 (short dashed
lines) are included. These additional imposed convergence rates
compare when conceptual ridge-push (imposed convergence) and
slab-pull forces act together, with when slab pull is the only driving
force in the system. The graph colours correspond with the colours
of the tracked rocks in Fig. 2. For the upper crustal rocks of the
microcontinent near the collision front (black), both microconti-
nent width and convergence rate do not have any substantial effect
on the extension rate (Fig. 3a). However, as noted with the results
above, a trend shows that with increasing continent width there is
a decrease in the magnitude of the pre-collisional exhumation (and
inferred extension) for most packages of microcontinental crustal
rock. This is consistent for all the models with 0, 2 and 4 cm yr–1
convergence rates, although the extension rate is always higher in
the case of no imposed convergence rate over the 2 and 4 cm yr–1
rates. Similarly, the β-factor decreases with increasing microconti-
nent width. Also note that the change in the exhumation rate due
to the convergence rate is reduced with an increase of microconti-
nent width. In other words, the effect of convergence velocity on the
extension rates is more prominent when the microcontinent size is
smaller. This is also evident in Fig. 3b, in which, for instance, the β
extension factors are 1.40, 1.27 and 1.17, respectively, for 0, 2 and
4 cm yr–1 convergence rates at the centre of a 120-km-wide micro-
continent, but they are correspondingly 1.31, 1.21 and 1.16 for a
600-km-wide microcontinent. The extension style in our experi-
ments develops rather as a pure shear type1 than a simple shear22
because the tracked rock packages—horizontally distributed across
the tested microcontinents—do not show differences in exhuma-
tion (Fig. 3a).
Microplate extension by slab pull and a subduction pulley
Our geodynamic models demonstrate that when microcontinents
drift towards a subduction zone, they can undergo substantial
extensional deformation, as evidenced by crustal exhumation and
stretching factors (Fig. 2). This extensional deformation is present
in all of the quantitative experiments here, regardless of the size of
the microcontinent or the microcontinental terrane convergence
rate (Fig. 3). The extension rate in a microcontinent is inversely
correlated with the microcontinent width and the convergence rate.
For instance, the exhumation rate of the lower crustal rocks of a
600-km-wide microcontinent is ~40% lower compared with that
of a 120-km-wide one in the case no imposed convergent velocity.
Further models show that with a sufficient imposed convergence
velocity, the pre-collisional extension can be effectively shut off
(Supplementary Fig. 2).
We propose that a main driving mechanism for the
pre-collisional extension of the microcontinent terrane is the verti-
cal slab pull transferred across a ‘subduction pulley’ to the micro-
continent during its drift towards subduction and ultimate collision.
A mechanical analogue was superimposed on top of the subduc-
tion–microcontinent accretion system to illustrate this mechanism
(Fig. 4). Oceanic lithosphere older than 10 Myr should resist defor-
mation when tensile stress acts on it23, hence, in the mechanical sys-
tem, the ocean lithosphere is represented as semi-rigid blocks that
resist deformation (Fig. 4). There is (as yet) no evidence of exten-
sional deformational regions on oceanic plates except for back-arcs;
that is, no similar pre-collision extension of a subducting ocean
plate. However, continental lithosphere embedded within the oce-
anic lithosphere is generally weaker under the same physical condi-
tions than the oceanic lithosphere and is a preferable location for
extensional deformation and rifting21,24–27. This is due to the rheol-
ogy and thickness differences between the oceanic and continental
crust. In our models, although the 30-km-thick microcontinental
crust is represented by weak wet quartzite28, the 7-km-thick oce-
anic crust is a strong Maryland diabase29. Given such differences
between the mechanical properties of the continental versus oce-
anic lithospheres, our mechanical model posits that the tensile
stress on the subducting plate is focused at the microcontinent,
which yields pre-collisional extension (Fig. 4 and Supplementary
Video 1).
120 300 450 600
0.5
1.0
1.5
2.0
Pressure drop
(P0 – P1) (kbar or 10–1 GPa)
Microcontinent width (km)
a
120 300 450 600
1.1
1.2
1.3
1.4
β-factor
Microcontinent width (km)
b
Fig. 3 | Plots showing the magnitudes of pre-collisional extension depending on tested parameters. a, The effect of microcontinent width and
convergence rate on exhumation rates from 0 to 5.5 Myr. The horizontal axis shows the tested microcontinent widths and the vertical axis indicates the
lithostatic pressure drop (due to unroofing or exhumation) in the selected rock locations. Teal, magenta and purple lines show the exhumation rates
of the lower crustal rocks in a microcontinent; purple represents the rocks in the front end, magenta those in the centre and teal those in the far end
of the microcontinent. The black line shows upper crustal rocks in the front of the microcontinent. The locations of these points are indicated in Fig. 2.
Convergence rates are denoted by solid, long-dashed and short-dashed lines for 0, 2 and 4 cm yr–1, respectively. b, β-factors (extension factor) for the
lower crustal rocks at the centre of the microcontinents, which depends on the convergence rates and microcontinent widths for the pre-collisional stage.
Extension factors are calculated by β = li/lf (ref. 1) (li and lf indicate initial and the final lithospheric thicknesses, respectively). Solid, long-dashed and
short-dashed lines represent β-factors for convergence rates of 0, 2 and 4 cm yr–1 , respectively.
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles
NATurE GEoSciENcE
A notable phenomenon in the experiments is the location of the
slab break-off after the microcontinent accretion. Conceptually,
it is normally considered that the location of a post-collisional
slab break-off is in the subducted oceanic lithosphere at variable
depths30–32; however, in our experiments break-off happens in a
subducted portion of the microcontinent instead. This is due to
the same factors discussed above: the weaker microcontinent ter-
rane is a preferential site to break compared with the oceanic plate
(of course, this is not to say that detachment could not occur in
an oceanic lithosphere in other circumstances). An intriguing
consequence of this is that it can yield a deep subduction of the
lower crust (tracked purple rock package in Fig. 2) from the first
~30–40 km subducted portion of the microcontinent (Fig. 2). This is
in accordance with previous calculations, which suggest that down-
going lithospheric mantle can carry an ~10-km-thick detached
lower crust into the asthenosphere33.
Geological record of pre-collisional extension
Pre-collisional extension is supported by geological evidence from
the Western Alps and eastern Mediterranean–Caucasus region. The
Sesia zone, as a ribbon-like microcontinent in the Western Alps,
drifted and accreted to Adria with the closure of the Ligurian–
Piemontese ocean in the Eocene34,35. The metamorphism in the
Sesia zone started ~100 million years ago (Ma) and exhumation
of the metamorphic rocks to the shallow crustal levels occurred
~60 Ma according to K–Ar cooling ages, prior the Tertiary collision
of the Western Alps8. Although there are studies that suggest the
Sesia zone thinned during Jurassic rifting36 and/or its crustal rocks
exhumed during Late Cretaceous dextral transpression37, neither of
these is able to explain why it later cooled to below 300–350 °C dur-
ing the Palaeocene38, that is, prior to collision. According to Avigad8,
Sesia pre-collisional exhumation occurred along a ductile shear
zone and showed a normal sense of kinematic motion (Fig. 5a),
as evidenced by S–C structures, mica fish and rotated porphyro-
clasts. The comparison of a calculated P–T–t track from one of our
experiments (EXP-3) against a variety of P–T–t calculations from
petrological studies in the Sesia zone indicates that the results are
compatible for the last stage of the exhumation period (Fig. 5b)37–41.
The convergence rate estimations between Europe and the Adria/
middle Mediterranean plate are 1–2 cm yr–1 (refs. 42,43), which is con-
sistent with our experiments where the subduction pull creates a
~2.1 cm yr–1 convergence in EXP-3.
Additional evidence for a pre-collisional extension comes from
Eastern Anatolia in the Tethyan belt (Fig. 5c). Eastern Anatolia,
the eastward elongation of the Anatolide block, is a microconti-
nental fragment bounded by the Izmir–Ankara–Erzincan (IAE)
Subducting
oceanic plate
Slab pull
Overriding
oceanic plate
Microcontinent
Strong oceanic
lithosphere
Weak continental
lithosphere
Asthenosphere
Subducting
oceanic plate
Slab pull
Overriding
oceanic plate
Microcontinent
Strong oceanic
lithosphere
Accretion and
slab break-off
a
c
b
Fig. 4 | The mechanical analogue demonstrates the fundamental dynamics of the tectonic system. This analogue consists of a mass, pulley, dashpot and
string that connects these elements to each other. In our analogy, the elements represent slab weight (slab pull), trench, a relatively weak microcontinent
and a strong oceanic lithosphere. a, The initial state in which the microcontinent drifts towards the subduction zone. b, The microcontinent (dashpot)
extends during its journey to the trench owing to the tensional force applied by the slab pull (the mass) across the subduction zone (pulley). c, The
accretion and slab break-off, which tends to occur in the weaker microcontinent.
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles NATurE GEoSciENcE
suture zone in the north and the Bitlis suture zone in the south.
Calculations based on palaeomagnetic studies again estimate a
slow convergence rate, ~1.3 cm yr–1 in the past 80 Myr, between
Africa and Eurasia for the Anatolia region44,45. On the basis of field
geological constraints and a magmatic lull, continental collision
along the IAE suture zone is constrained to have occurred by the
Palaeocene–Early Eocene46,47. However, most estimates for the con-
tinental collision along the Bitlis–Zagros suture range from the Late
Eocene to Early Miocene48,49. The basement of East Anatolia is made
up of Late Cretaceous high-temperature/low-to-middle-pressure
metamorphic rocks intruded by Late Cretaceous mafic to felsic
intrusions, and locally their non-metamorphic equivalents50,51.
The metamorphic rocks and the intrusions were exhumed to the
surface by the Maastrichtian, as deduced by unconformably over-
lying Maastrichtian reefal limestone. Crustal-scale detachment
faults juxtapose the greenschist-facies and upper amphibolite-facies
domains near Akdağ51 (G.T. et al., manuscript in preparation)
(Fig. 5c). Likewise, a crustal-scale detachment that separates
greenschist- and amphibolite-facies domains is observed in the
Taşlıçay area, which is located ~140 km to the east of the Akdağ area
(Fig. 5c). This sequence of events—exhumation completed by the
end of Late Cretaceous, and the continental collision along the IAE
Issime
Gabby
Valsesia
Val Gressoney
0 5 km
7° 40′ E 7° 50′ E 8° E
45° 40′ N
45° 50′ N
300 400
5
10
Temperature (°C)
Lithostatic pressure (kbar)
Zucali et al.39; Regis et al.40
Regis et al.40; Rubbato et al.41
Babist et al.37
500
Oberhänsli et al.38
200
This work (EXP-3)
Sesia zone
N
Triassic to Upper Cretaceous
carbonate platform
Late Cretaceous
mafic-to-acidic intrusions
Early Jurassic and Late
Cretaceous ophiolite and
melanges
Late Cretaceous
ophiolite and melanges
Post-Cretaceous
cover
Suture
Late Cretaceous
amphibolite-facies
metamorphic rocks
Late Cretaceous
greenschist-facies
metamorphic rocks
Possible detachment fault
Late Cretaceous low-to-high grade
metamorphic rocks containing
local high-pressure metamorphic rocks
Arabian Platform
East Anatolia
Eastern Pontides
Pütürge massif
Lake Van
N
IAE suture
39° E
42° E
39° N
a b
c
EMS
Metaquartzdiorite in EMS
II DK
Gneiss minuti (paraschist–
granitoid complex)
Oligocene volcanics
Ivrea zone
Oligocene intrusives
Detachment fault
Thrust
Black Sea
Aegean
Sea
Arabian Platform
Anatolia
-
Greater Caucaus
EAF
EURASIA
Akdağ
0 500 km
Tyrrhenian
Sea
Adriatic
Sea
Western
Alps
Sesia
0500 km
Detachment fault
0 5 10 km
Tasllçay
a
F
a
u
l
t
Z
a
g
r
o
s
s
u
t
u
r
e
-
E
r
z
i
n
c
a
n
s
u
t
u
r
e
I
z
m
i
r
-
A
n
k
a
r
a
H
e
l
l
e
n
i
c
T
.
N
o
r
t
h
A
n
a
t
o
l
i
a
n
F
a
u
l
t
39° N
B
i
t
l
i
s
–
Z
a
g
r
o
s
s
u
t
u
r
e
I
A
E
s
u
t
u
r
e
Akdag
Akdag
Fig. 5 | The geological evidence for pre-collisional continental terrane extension. a, Geological map of the Sesia zone in the Western Alps8. The
microstructures and normal-sense ductile shear zone between the eclogitic micaschists (EMS) (orange) and seconda zona Diorito–Kinzigitica (II DK)
(light blue) units suggest that Cretaceous EMS unit decompressed and exhumed before the Cenozoic collision of the Sesia microcontinental terrane.
b, Comparison of various petrological data from the Sesia zone to the calculated P–T–t tracks from EXP-3. c, Location and geological map of the East
Anatolia microcontinent, between the IAE and Bitlis–Zagros suture zones in Eastern Anatolia51 (G.T. et al., manuscript in preparation). The Late Cretaceous
metamorphic rocks (orange) and mafic-to-acidic intrusions (red) located near Akdağ and Taşlıçay were exhumed to the surface along the detachment
fault shown in the figure before its collision along the IAE suture zone in the Palaeocene–Early Eocene (see text for references). The greyscale location
maps were made with GeoMapApp (www.geomapapp.org) under a Creative Commons licence CC BY 4.056. Panel a adapted with permission
from ref. 8, Elsevier.
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles
NATurE GEoSciENcE
suture zone occurring later (Palaeocene–Early Eocene)—indicates
a major extension in Eastern Anatolia in a pre-collisional setting.
We believe that this new pre-collisional extension idea explains
some unresolved geodynamic and petrological aspects related to
microcontinent terrane accretion processes. Until now, it has been
proposed that extension/exhumation processes take place after the
accretion of the microcontinents to an overriding plate, namely
syn- and post-collisional extension, due to gravitational collapse52
or back-arc extension4,53–55.
The majority of petrological studies of extension during micro-
continent terrane accretion are interpreted as post-collisional. This
may be due to the overprinting of post-collisional metamorphism
over the pre-collisional one. Our findings suggest an alternative
hypothesis. Ultimately, the hidden or ignored pre-collisional exten-
sion/exhumation process could be important and taken into con-
sideration when investigating the tectono-metamorphic history of
microcontinental terranes.
Online content
Any methods, additional references, Nature Research report-
ing summaries, source data, extended data, supplementary infor-
mation, acknowledgements, peer review information; details of
author contributions and competing interests; and statements of
data and code availability are available at https://doi.org/10.1038/
s41561-021-00746-9.
Received: 20 April 2019; Accepted: 29 March 2021;
Published: xx xx xxxx
References
1. McKenzie, D. Some remarks on the development of sedimentary basins.
Earth Planet. Sci. Lett. 40, 25–32 (1978).
2. Şengör, A. M. C. & Burke, K. Relative timing of riing and volcanism on
Earth and its tectonic implications. Geophys. Res. Lett. 5, 419–421 (1978).
3. Turcotte, D. L. & Emerman, S. H. Mechanisms of active and passive riing.
Dev. Geotecton. 19, 39–50 (1983).
4. Royden, L. H. Evolution of retreating subduction boundaries formed during
continental collision. Tectonics 12, 629–638 (1993).
5. Jolivet, L. & Faccenna, C. Mediterranean extension and the Africa–Eurasia
collision. Tectonics 19, 1095–1106 (2000).
6. Brun, J.-P. & Faccenna, C. Exhumation of high-pressure rocks driven by slab
rollback. Earth Planet. Sci. Lett. 272, 1–7 (2008).
7. Compagnoni, R. et al. e Sesia–Lanzo zone, a slice of continental crust with
Alpine high pressure-low temperature assemblages in the Western Italian
Alps. Rend. Soc. Ital. Mineral. Petrol. 33, 281–334 (1977).
8. Avigad, D. Pre-collisional ductile extension in the internal western Alps (Sesia
zone, Italy). Earth Planet. Sci. Lett. 137, 175–188 (1996).
9. Dewey, J. F. Extensional collapse of orogens. Tectonics 7, 1123–1139 (1988).
10. Rey, P., Vanderhaeghe, O. & Teyssier, C. Gravitational collapse of the
continental crust: denition, regimes and modes. Tectonophysics 342,
435–449 (2001).
11. Price, R. A. e Cordilleran foreland thrust and fold belt in the
southern Canadian Rocky Mountains. Geol. Soc. Lond. Spec. Publ. 9,
427–448 (1981).
12. Nur, A. & Ben-Avraham, Z. Oceanic plateaus, the fragmentation of
continents, and mountain building. J. Geophys. Res. 8, 3644–3661 (1982).
13. Coney, P. J., Jones, D. L. & Monger, J. W. H. Cordilleran suspect terranes.
Nature 288, 329–333 (1980).
14. Sigloch, K., McQuarrie, N. & Nolet, G. Two-stage subduction history under
North America inferred from multiple-frequency tomography. Nat. Geosci. 1,
458–462 (2008).
15. Fullsack, P. An arbitrary Lagrangian–Eulerian formulation for creeping ows
and its application in tectonic models. Geophys. J. Int. 120, 1–23 (1995).
16. Pysklywec, R. N., Beaumont, C. & Fullsack, P. Lithospheric deformation
during the early stages of continental collision: numerical experiments and
comparison with South Island, New Zealand. J. Geophys. Res. Solid Earth 107,
https://doi.org/10.1029/2001JB000252 (2002).
17. Gray, R. & Pysklywec, R. N. Geodynamic models of mature continental
collision: evolution of an orogen from lithospheric subduction to continental
retreat/delamination. J. Geophys. Res. Solid Earth 117, B03408 (2012).
18. Göğüş, O. H., Pysklywec, R. N., Şengör, A. M. C. C. & Gün, E. Drip tectonics
and the enigmatic upli of the Central Anatolian Plateau. Nat. Commun. 8,
1538 (2017).
19. Bodur, Ö. F., Göğüş, O. H., Pysklywec, R. N. & Okay, A. I. Mantle lithosphere
rheology, vertical tectonics, and the exhumation of (U)HP rocks. J. Geophys.
Res. Solid Earth 123, 1824–1839 (2018).
20. Heron, P. J., Pysklywec, R. N. & Stephenson, R. Lasting mantle scars lead to
perennial plate tectonics. Nat. Commun. 7, 11834 (2016).
21. Vink, G. E., Morgan, W. J. & Zhao, W.-L. Preferential riing of continents: a
source of displaced terranes. J. Geophys. Res. 89, 10072–10076 (1984).
22. Wernicke, B. Uniform-sense normal simple shear of the continental
lithosphere. Can. J. Earth Sci. 22, 108–125 (1985).
23. Lynch, H. D. & Morgan, P. e tensile strength of the lithosphere and the
localization of extension. Geol. Soc. Lond. Spec. Publ. 28, 53–65 (1987).
24. Molnar, P. Continental tectonics in the aermath of plate tectonics. Nature
335, 131–137 (1988).
25. Ranalli, G. & Murphy, D. C. Rheological stratication of the lithosphere.
Tectonophysics 132, 281–295 (1987).
26. Kohlstedt, D. L., Evans, B. & Mackwell, S. J. Strength of the lithosphere:
constraints imposed by laboratory experiments. J. Geophys. Res. 100,
17587 (1995).
27. van den Broek, J. M. & Gaina, C. Microcontinents and continental fragments
associated with subduction systems. Tectonics 39, e2020TC006063 (2020).
28. Gleason, G. C. & Tullis, J. A ow law for dislocation creep of quartz
aggregates determined with the molten salt cell. Tectonophysics 247,
1–23 (1995).
29. Mackwell, S. J., Zimmerman, M. E. & Kohlstedt, D. L. High-temperature
deformation of dry diabase with application to tectonics on Venus. J.
Geophys. Res. Solid Earth 103, 975–984 (1998).
30. Davies, J. H. & von Blanckenburg, F. Slab breako: a model of lithosphere
detachment and its test in the magmatism and deformation of collisional
orogens. Earth Planet. Sci. Lett. 129, 85–102 (1995).
31. van Hunen, J. & Allen, M. B. Continental collision and slab break-o: a
comparison of 3-D numerical models with observations. Earth Planet. Sci.
Lett. 302, 27–37 (2011).
32. Freeburn, R., Bouilhol, P., Maunder, B., Magni, V. & van Hunen, J. Numerical
models of the magmatic processes induced by slab breako. Earth Planet. Sci.
Lett. 478, 203–213 (2017).
33. Molnar, P. & Gray, D. Subduction of continental lithosphere: some constraints
and uncertainties. Geology 7, 58 (1979).
34. Coward, M. & Dietrich, D. Alpine tectonics—an overview. Geol. Soc. Lond.
Spec. Publ. 45, 1–29 (1989).
35. Rosenbaum, G. & Lister, G. S. e Western Alps from the Jurassic to
Oligocene: spatio-temporal constraints and evolutionary reconstructions.
Earth Sci. Rev. 69, 281–306 (2005).
36. Manzotti, P., Ballèvre, M., Zucali, M., Robyr, M. & Engi, M. e
tectonometamorphic evolution of the Sesia–Dent Blanche nappes
(internal Western Alps): review and synthesis. Swiss J. Geosci. 107,
309–336 (2014).
37. Babist, J., Handy, M. R., Konrad-Schmolke, M. & Hammerschmidt, K.
Precollisional, multistage exhumation of subducted continental crust: the
Sesia Zone, western Alps. Tectonics 25, TC6008 (2006).
38. Oberhänsli, R., Hunziker, J. C., Martinotti, G. & Stern, W. B. Geochemistry,
geochronology and petrology of Monte Mucrone: an example of EO-alpine
eclogitization of Permian granitoids in the Sesia–Lanzo Zone, Western Alps,
Italy. Chem. Geol. Isot. Geosci. 52, 165–184 (1985).
39. Zucali, M., Spalla, M. I. & Gosso, G. Strain partitioning and fabric evolution
as a correlation tool: the example of the eclogitic micaschists complex in the
Sesia–Lanzo Zone (Monte Mucrone–Monte Mars, Western Alps, Italy).
Schweiz. Mineral. Petrogr. Mitteil. 82, 429–454 (2002).
40. Regis, D. et al. Multiple metamorphic stages within an eclogite-facies terrane
(Sesia zone, Western Alps) revealed by –U–Pb petrochronology. J. Petrol.
55, 1429–1456 (2014).
41. Rubatto, D. et al. Yo-yo subduction recorded by accessory minerals in the
Italian Western Alps. Nat. Geosci. 4, 338–342 (2011).
42. Hsü, K. J. Time and place in Alpine orogenesis—the Fermor Lecture. Geol.
Soc. Spec. Publ. 45, 421–443 (1989).
43. Hunziker, J. C., Desmons, J. & Martinotti, G. Alpine thermal evolution in the
central and the western Alps. Geol. Soc. Spec. Publ. 45, 353–367 (1989).
44. Torsvik, T. H. et al. Phanerozoic polar wander, palaeogeography and
dynamics. Earth Sci. Rev. 114, 325–368 (2012).
45. van Hinsbergen, D. J. J. et al. Tectonic evolution and paleogeography of the
Kirsehir Block and the Central Anatolian Ophiolites, Turkey. Tectonics 35,
983–1014 (2016).
46. Okay, A. I. & Şahintürk, Ö. in Regional and Petroleum Geology of the Black
Sea and Surrounding Region (ed. Robinson, A. G.) 291–311 (American
Association of Petroleum Geologists, 1997); https://doi.org/10.1306/
M68612C15
47. Schleiarth, W. K., Darin, M. H., Reid, M. R. & Umhoefer, P. J. Dynamics of
episodic Late Cretaceous–Cenozoic magmatism across Central to Eastern
Anatolia: new insights from an extensive geochronology compilation.
Geosphere 14, 1990–2008 (2018).
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles NATurE GEoSciENcE
48. Agard, P., Omrani, J., Jolivet, L. & Mouthereau, F. Convergence history across
Zagros (Iran): constraints from collisional and earlier deformation. Int. J.
Earth Sci. 94, 401–419 (2005).
49. Okay, A. I., Zattin, M. & Cavazza, W. Apatite ssion-track data for the
Miocene Arabia–Eurasia collision. Geology 38, 35–38 (2010).
50. Rolland, Y. et al. Evidence for ~80–75 Ma subduction jump during
Anatolide–Tauride–Armenian block accretion and ~48 Ma Arabia–Eurasia
collision in Lesser Caucasus–East Anatolia. J. Geodyn. 56–57, 76–85 (2012).
51. Topuz, G., Candan, O., Zack, T. & Yılmaz, A. East Anatolian plateau
constructed over a continental basement: no evidence for the East Anatolian
accretionary complex. Geology 45, 791–794 (2017).
52. England, P. & Houseman, G. Extension during continental convergence, with
application to the Tibetan Plateau. J. Geophys. Res. 94, 17561 (1989).
53. Uyeda, S. & Kanamori, H. Back-arc opening and the mode of subduction. J.
Geophys. Res. 84, 1049 (1979).
54. Heuret, A. & Lallemand, S. Plate motions, slab dynamics and back-arc
deformation. Phys. Earth Planet. Inter. 149, 31–51 (2005).
55. Göğüş, O. H. Riing and subsidence following lithospheric removal in
continental back arcs. Geology 43, 3–6 (2015).
56. Ryan, W. B. F. et al. Global multi-resolution topography synthesis. Geochem.
Geophys. Geosyst. 10, Q03014 (2009).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2021
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles
NATurE GEoSciENcE
Methods
Model description. e numerical experiments were performed using the
two-dimensional plane strain viscous–plastic nite element code, SOPALE15,
which has been used to model a range of geodynamic problems16,18,20,57–59. A series
of comprehensive benchmarking experiments veried the accuracy of SOPALE, in
addition to comparisons with formal geodynamic benchmark study results15,60,61.
SOPALE uses an arbitrary Lagrangian–Eulerian mesh system to solve the
governing thermomechanical equations for viscous–plastic deformation of
two-dimensional media. The numerical code assumes the crust and mantle are
incompressible viscous-plastic materials and it solves the governing equations of
the conservation of mass, momentum and internal energy15. Density is defined as a
function of temperature:
ρ
=
ρ0
[
1
−
α
(
T
−
T0
)]
where α, ρ0 and T0 are the coefficient of thermal expansivity, reference material
density and reference temperature, respectively. At each computational node, the
deviatoric stress (
σ′
ij
=
min
(
σ
y
;σ
v
)
) is determined as the lesser value of either a
yield stress σy or a viscous stress σv. In this way, materials in the models deform
dynamically, essentially according to a viscous–plastic strength envelope. This can
be expressed as:
σ
′
ij
=
min
(
σy;σv
)
For the frictional plastic yield stress, a pressure-dependent Drucker–Prager
yield criterion was used, which is equivalent to the Coulomb criterion in plane
strain15:
σy
=
Psin ϕ
+
c0
Here, P, ϕ and c0 are the pressure, internal angle of friction and cohesion,
respectively. The viscous stress is defined as:
σv
=
2ηe
˙
I
′
2
where
˙
I
′
2
is the second invariant of the deviatoric strain rate tensor and ηe is the
effective viscosity:
ηe=
(
3
−(n+1)
2n21−n
n
)
fA
−1
n˙
I
′(
1+n
2n)
2e(Q
nRT )
In this expression, A, f, n, Q, T and R are a viscosity parameter, scaling parameter
and power exponent, and the activation energy, temperature and ideal gas
constant, respectively (the first three of which are determined from uniaxial
laboratory experiments; see the citations in Extended Data Table 1). The term
G
=
(
3
−(n+1)
2n21−n
n
)
is necessary to convert the data from uniaxial laboratory
experiments into a state of stress that is independent of the choice of coordinate
system16.
The experiment set-up has a 100-km-thick microcontinent (30 km continental
crust and 70 km lithospheric mantle) drifting towards an intra-oceanic subduction
zone with a convergence rate of 0, 2 and 4 cm yr–1, a 90-km-thick oceanic
lithosphere (7 km oceanic crust and 83 km lithospheric mantle) and sublithospheric
mantle as deep as 660 km at the closed bottom of the box (Extended Data Fig. 1).
An imposed convergence velocity, if present, is applied at the right lithospheric
boundary of the model by introducing new lithosphere into the box. The
left boundary lithosphere is fixed. An evenly distributed outward flux of the
sublithospheric mantle is prescribed in the left and right side of the model box
to compensate for the inward flux of the lithosphere from the top-right side. The
model box has a free top surface which allows the topography to evolve naturally
during the model evolution. Erosion and deposition processes are not included in
the experiments. A portion of the oceanic lithosphere is already subducted into
the asthenosphere initially, which creates a plate driving force with progressive
sinking of the oceanic lithosphere. This slab pull, in turn, causes the drift of the
microcontinent to the subduction zone. As an initial approach, differently sized
microcontinents (120, 300, 450 and 600 km) were tested to observe the effect of
the width, if any, on the deformation of the drifting microcontinent. Regardless
of the lithosphere type, either continental or oceanic, lithospheric mantle is the
same material for every plate, namely wet Åheim dunite62. Continental crust is
represented by wet quartzite28 and the oceanic crust is a Maryland diabase29. All the
rheological and physical parameters used in the experiments are listed in Extended
Data Table 1. The mantle lithosphere on the left and right boundaries of the model
set-up has relatively high cohesion values (300 MPa) to prevent deformation at the
model box boundaries in case of the presence of imposed convergence velocities.
These cohesion values were initially chosen for practical reasons and they do not
have any significant effect on the pre-collisional extension of microcontinents, as
evidenced by our supplementary test models (Supplementary Figs 1 and 2).
Data availability
The data—the outputs of the numerical experiments—that support the findings of
this study are available at https://doi.org/10.5683/SP2/9RDQYB.
Code availability
The SOPALE modelling code was developed by the Dalhousie University
Geodynamics group; this is not a freely available open-source code.
Documentation for the code and the address of the developer research group may
be found at http://geodynamics.oceanography.dal.ca/sopaledoc.html.
References
57. Göğüş, O. H. & Pysklywec, R. N. Near-surface diagnostics of dripping or
delaminating lithosphere. J. Geophys. Res. 113, B11404 (2008).
58. Göğüş, O. H. Geodynamic experiments suggest that mantle plume caused
Late Permian Emeishan Large Igneous Province in Southern China. Int. Geol.
Rev. https://doi.org/10.1080/00206814.2020.1855602 (2020).
59. Memiş, C. et al. Long wavelength progressive plateau upli in Eastern
Anatolia since 20 Ma: implications for the role of slab peel‐Back and
Break‐o. Geochem. Geophys. Geosyst. 21, e2019GC008726 (2020).
60. Buiter, S. J. H. et al. e numerical sandbox: comparison of model results for
a shortening and an extension experiment. Geol. Soc. Lond. Spec. Publ. 253,
29–64 (2006).
61. Buiter, S. J. H. et al. Benchmarking numerical models of brittle thrust wedges.
J. Struct. Geol. 92, 140–177 (2016).
62. Chopra, P. N. & Paterson, M. S. e role of water in the deformation of
dunite. J. Geophys. Res. Solid Earth 89, 7861–7876 (1984).
63. Hirth, G. & Kohlstedt, D. L. Water in the oceanic upper mantle: implications
for rheology, melt extraction and the evolution of the lithosphere. Earth
Planet. Sci. Lett. 144, 93–108 (1996).
Acknowledgements
This research was enabled in part by support provided by SciNet and Compute Canada
(www.computecanada.ca). A modified version of the SOPALE (2000) software was used
to run numerical models. The SOPALE modelling code was originally developed by P.
Fullsack at Dalhousie University with C. Beaumont and his Geodynamics group. We
used Ö. F. Bodur’s script to plot the P–T–t paths. Funding for this research was provided
by an NSERC Discovery Grant (RGPIN-2019-06803)-RNP and TÜBİTAK grant
(114Y226) to G.T.
Author contributions
E.G. designed and carried out the numerical experiments and interpreted the results
with R.N.P. O.H.G. and G.T. helped with the geological background of the Alpine–
Mediterranean region. E.G. and R.N.P. developed ‘the subduction pulley’ hypothesis and
wrote the manuscript with inputs and comments from all the authors.
Competing interests
The authors declare no competing interests.
Additional information
Extended data is available for this paper at https://doi.org/10.1038/s41561-021-00746-9.
Supplementary information The online version contains supplementary material
available at https://doi.org/10.1038/s41561-021-00746-9.
Correspondence and requests for materials should be addressed to E.G.
Peer review information Nature Geoscience thanks Wim Spakman and the other,
anonymous, reviewer(s) for their contribution to the peer review of this work. Primary
Handling Editor: Stefan Lachowycz.
Reprints and permissions information is available at www.nature.com/reprints.
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
Articles NATurE GEoSciENcE
Extended Data Fig. 1 | Initial conditions of model EXP-2. The figure shows dimensions, geometry, materials, geotherm, Eulerian and Lagrangian grid
dimensions, and convergence velocity direction. The upper left side of the model box is fixed. The imposed velocity, if present, is applied at the location
of the blue arrow by introducing new oceanic lithosphere into the box. Orange arrows show evenly distributed outward flux to compensate inward flux.
The red line on the right shows the change of geotherm with depth. The inset table lists experiment numbers according to microcontinent widths and
convergence rates.
NATURE GEOSCIENCE | www.nature.com/naturegeoscience
A preview of this full-text is provided by Springer Nature.
Content available from Nature Geoscience
This content is subject to copyright. Terms and conditions apply.