a plate tectonic model for the paleozoic and mesozoic...

A plate tectonic model for the Paleozoic and Mesozoicconstrained by dynamic plate boundaries and restored

synthetic oceanic isochrons

G.M. Stamp£i, G.D. Borel *Institute of Geology and Paleontology, University of Lausanne, CH-1015 Lausanne, Switzerland

Received 6 August 2001; received in revised form 15 November 2001; accepted 15 November 2001

Abstract

We developed a plate tectonic model for the Paleozoic and Mesozoic (Ordovician to Cretaceous) integrating dynamicplate boundaries, plate buoyancy, ocean spreading rates and major tectonic and magmatic events. Plates wereconstructed through time by adding/removing oceanic material, symbolized by synthetic isochrons, to major continentsand terranes. Driving forces like slab pull and slab buoyancy were used to constrain the evolution of paleo-oceanicdomains. This approach offers good control of sea-floor spreading and plate kinematics. This new method represents adistinct departure from classical continental drift reconstructions, which are not constrained, due to the lack of plateboundaries. This model allows a more comprehensive analysis of the development of the Tethyan realm in space andtime. In particular, the relationship between the Variscan and the Cimmerian cycles in the Mediterranean^Alpine realmis clearly illustrated by numerous maps. For the Alpine cycle, the relationship between the Alpides senso stricto and theTethysides is also explicable in terms of plate tectonic development of the Alpine Tethys^Atlantic domain versus theNeoTethys domain. ß 2002 Published by Elsevier Science B.V.

Keywords: isochrons; plate boundaries; plate tectonics; models; Paleozoic; Mesozoic; paleo-oceanography

1. Introduction

Recent syntheses by IGCP and Tethys Groupprojects, in which we actively participated, pro-vide an up-to-date framework for reassessingplate tectonic models for the Tethyan realm [1].New plate tectonic concepts were also developed

to integrate new data regarding the geodynamicevolution of pre-Variscan [2] and Variscan Eu-rope [3^5] and of the Cimmerian orogenic cyclein the western Tethys [6,7]. These works followedplate tectonic models developed for the Alpineregion [8,9] and for the Mediterranean region[10]. Our goal during the development of thismodel was to integrate geodynamic constraints(such as ocean spreading rates, plate buoyancy,dynamic plate boundaries and basin evolution),lithologic data (e.g. pelagic series and ophioliteoccurrences) as well as the age and style of majortectonic and magmatic events.

0012-821X / 02 / $ ^ see front matter ß 2002 Published by Elsevier Science B.V.PII: S 0 0 1 2 - 8 2 1 X ( 0 1 ) 0 0 5 8 8 - X

* Corresponding author.Tel. : +41-21-692-4359; Fax: +41-21-692-4305.

E-mail addresses: gerard.stamp£[email protected](G.M. Stamp£i), [email protected] (G.D. Borel).

EPSL 6084 7-3-02 Cyaan Magenta Geel Zwart

Earth and Planetary Science Letters 196 (2002) 17^33

www.elsevier.com/locate/epsl

2. Reconstructions parameters

Numerous parameters were integrated in themodel in an iterative approach. It appeared thatthe most e¤cient procedure was to develop the¢nal model forward in order to check its consis-tency step by step. The weight of the di¡erentparameters varies extremely from one reconstruc-tion to another but also in the same reconstruc-tion from one region to another (for details, seeSections 3 and 4).

2.1. Geological objects and tools

To de¢ne the size of continents, we digitizedtheir present shape or geological boundarieswhere applicable. Where possible, we used a sim-ilar procedure for the terranes keeping theirpresent shape for geographical identi¢cation. Wesystematically used tight ¢ts in order not tounderestimate widespread extension a¡ecting con-tinents and their margins. To use plates and notonly continents, we developed the concept of adynamic plate boundary (e.g. active spreadingridge, subduction zone, and transform fault) byadding to each continent its oceanic surfacethrough time. Thus, paleo- synthetic oceanic iso-chrons have been constructed through time in or-der to de¢ne the location of the spreading ridgesand to restore subducted ocean basins. To simpli-fy the process we worked with symmetrical sea-£oor spreading for the main oceans (Paleo- andNeoTethys).

We combined stratigraphic, sedimentary andpaleontological data to produce subsidence curvesin some key areas. Subsidence analysis character-izes and constrains the timing of major geody-namic events, such as the onset of rifting, subduc-tion of an active ridge or collision of continents,and consequently constrains spreading rates. Pa-leomagnetic (e.g.[11]) and paleobiogeographic in-formation have also been incorporated, to thelimit of our knowledge, to constrain the pre-Me-sozoic location of older plates, whilst existing iso-chrons [12] were used to constrain younger recon-structions. Geological data, mainly regarding theage of accretion/collision, were compiled for mostkey areas (e.g. [13] and [14] for south-east Asia,

[15] for the Tibetan back-arcs, etc.). Where possi-ble, this multidisciplinary approach was appliedto the entire globe in order to generate self-con-strained reconstructions. However, little consider-ation was given to the PaleoPaci¢c due to a lackof any information before the Jurassic. Despitethis, most of the Paleozoic information could begathered through investigation of cordillera con-struction/collapse around the Paci¢c (e.g. [16]).

2.2. Geodynamic forces and implications

Based on the fact that a continent is nearly al-ways attached to a larger plate, continental drifthas implications for its plate boundaries. There-fore, we consider that only models integratingplate boundaries can be compared to each otherand that reconstructions based on continentaldrift as practised by everybody since Wegenershould be abandoned. Continental drift modelsdo not allow assessment of important geodynamicevents such as the onset of subduction [17], mid-ocean ridge subduction (e.g. [18]) presence/ab-sence of slab rollback and related plate motion(e.g. [19]) nor plate buoyancy criteria (e.g.[20]).Fig. 1 schematizes all these parameters and theirin£uence on plate kinematics. Among other pa-rameters, we assume that cordillera build-up(Fig. 1E) corresponds to subduction of youngbuoyant lithosphere, whereas cordillera collapse(Fig. 1F) and opening of back-arcs correspondto the subduction of heavier lithosphere produc-ing slab rollback. In doing so we could relate, forexample, the evolution of the Australian activemargin (e.g. [21,22]) with the opening of the Neo-Tethys. Furthermore, two types of terranes canexist. The ¢rst one detached from the active mar-gin by slab rollback and opening of back-arc, andhas an active margin and a passive one (Fig. 1B),e.g. the European Hunic terranes. The other type,detached from a passive margin by slab pull, hastwo passive margins (Fig. 1G), e.g. the Cimmerianblocks. Slab pull becomes strong enough to de-tach a terrane only after the subduction of theactive ridge. The sedimentary record of the ter-rane gives the timing of rifting for both oceansand consequently a good control on the spreadingrate.


G.M. Stamp£i, G.D. Borel / Earth and Planetary Science Letters 196 (2002) 17^3318

2.3. Technical tools and constraints

We develop our model using GMAP softwarepackage [23]. More than 100 continents and ter-ranes were used to elaborate the reconstructions.In general, we always tried the simplest geomet-rical solution to move a plate.

The use of Euler poles to move plates has nu-merous consequences; one of them is the geome-try of transform boundaries (small circle) ; wherethey are known (e.g. Levant transform) the chos-en Euler pole has to conform to this geometry.The consequence is that the maximum spreadingof a chosen pole lies along its equator. As a result,an acceptable spreading rate close to the Eulerpole may be unacceptable a few thousands ofkilometers away. We tried to keep sea-£oorspreading at the Euler equator below 20 cm/yr.Spreading rates were calculated at the Euler equa-tor of PaleoTethys and NeoTethys sensu strictowithout taking in account the opening of back-arc basins along the Eurasian margin. To controlthe spreading rates we used the paleo-isochronsand the geological expression of the main geody-namic features, e.g. subduction of an active ridge,uplift of the active margin, implacement of bath-olites and increasing of slab pull (Fig. 1). Themodel implies spreading rates between 8 and 17cm/yr for PaleoTethys. NeoTethyan values aregenerally lower, but the spreading rate increasesfrom 7 cm/yr at 260 Ma to 20 cm/yr between 230and 220 Ma. Although this rate is plausible due tothe length of the subducted slab, such a spreadingrate is very high. The spreading rate decreases to6 cm/yr at 220 Ma (Carnian/Norian). This strongdeceleration can be related to the Cimmerian oro-genic collage [24,25], preceding the onset of sub-duction of NeoTethys under the recently accretedCimmerian terranes.

We used the orthographic (orthogonal) projec-tion to create the reconstructions because thisprojection deforms little the angles and allows agood visual control to choose between alterna-tives. The disadvantage of this projection is thatonly one hemisphere can be displayed. To workwith a plate ¢xed as reference appeared as thebest solution to easily check the geometrical con-sistency of a hypothesis. Thus, we worked with

Fig. 1. Sketches showing the e¡ect of the main driving forces(slab pull, slab buoyancy) on the evolution of oceanic mar-gins. (A) Continental crust; (B) lithospheric mantle; (C) as-thenosphere; (D) oceanic crust; (E) intrusives; (F) volcan-ism; (G) accretionary prism.


G.M. Stamp£i, G.D. Borel / Earth and Planetary Science Letters 196 (2002) 17^33 19

Europe ¢xed in its present-day position and movethe other plates relative to it.

We used Baltica paleopoles as global referencefor paleolatitudes [23]. Baltica poles appeared tobe the best constrained for the Paleozoic. Torsviket al. [26] provide a new dataset for the Mesozoicadjusting the European poles with the NorthAmerican ones. These datasets were checkedagainst paleomagnetic data available in the liter-ature for all relevant plates (e.g. [11,27]). Poles ofcontinents situated at high latitudes generallycome out in a cluster of 15³ around the Balticavalues.

In order to validate the choice of Baltica polesas reference, we constructed an absolute wanderpath of Australia, because we had non-consistentpaleomagnetic poles locations for this plate in re-gard to Gondwana, due to its low-latitude situa-tion. The result shows an acceptable absolutepath also applicable to Gondwana (Fig. 2).

3. Tethyan realm: some of the main chosenalternatives

We focussed on the Tethyan realm for whichnumerous continental drifting reconstructionshave been presented in the last 30 yr, in orderto see if plate tectonic models, including dynamic

plate boundaries, could solve some of the maingeological issues. Alternative hypotheses weretested and choices were made. In the process, wediscovered that dynamic plate boundaries,brought from one reconstruction to the next, pro-vide greater constraints than the continental driftmodels proposed so far.

One of the main alternatives touches the evolu-tion of PaleoTethys and the concept of the Hunsuperterrane [2]. This concept regards all Variscanand Avalonian terranes as derived from Gondwa-na in Paleozoic times, in a context of an activeGondwanan margin. Terranes were detachedfrom Gondwana through the opening of back-arc oceans, following the subduction of the mid-ocean ridge of a former peri-Gondwana ocean ^the ProtoTethys. Several stages of successful andaborted opening of back-arc basins can be recog-nized between late Neoproterozoic and Siluriantimes, giving birth ¢rst to the Rheic ocean, thento the PaleoTethys ocean (Figs. 3 and 4). Thecomposite nature of the Hun superterrane impliesthe presence and amalgamation with the Cado-mian and Serindia terranes (Figs. 3 and 4), andallowed us to ¢nd a satisfactory location for theChinese blocks around Gondwana. A positionwhich, in turn, allows a satisfactory location forthese blocks to be determined later on, with aminimum of displacement (Figs. 4 and 5), but stillwithin the limits of present paleomagnetic infor-mation on these blocks (e.g. [11]).

Our favored model is to group the Variscanterranes in two main terranes, the EuropeanHunic terranes and the Asiatic Hunic terranes,until their collision with Laurussia and their sub-sequent lateral transfer and dismemberment alongthe Eurasian margin [5]. The terrane concept usedfor the Variscan domain is based on the presentclose and repeated juxtaposition of metamorphicand non-metamorphic units in the Variscan belt.A general mid-Devonian HP metamorphic eventis recognized in most metamorphic units, whereasthe non-metamorphic units do not record thisevent in term of £ysch deposits or deformation.On the contrary, their geological evolution is gen-erally towards a more mature passive marginstage characterized by pelagic sedimentation.Flysch deposits started only during the Early or

Fig. 2. Absolute polar wander path for the Australian platerelated to Europe ¢xed in its present-day position using Bal-tica paleopoles.



Fig. 3. Explanations in text. Orthographic projection with Europe ¢xed in its present-day position. Paleopoles of Baltica are usedas reference for the paleolatitudes. A complete legend for selected key reconstructions (490 Ma, 360 Ma, 240 Ma, 155 Ma-M25and 84 Ma-34) are available in the EPSL Online Background Dataset1.



even Late Carboniferous on some blocks. To rec-oncile these present-day contradictory juxtaposi-tions, we regard the metamorphic units as belong-ing to the leading edge (i.e. active margin) of asingle, ribbon-like terrane in which HP metamor-phism can be expected, whereas the other unitsrepresent the passive margin side of this terrane.This passive margin, in turn, became an activemargin after the docking of the terrane againstLaurussia. Recent paleomagnetic data con¢rmthe drifting history of the European Hunic terrane[28^30], as well as the absence of collision betweenGondwana and Laurasia before the Carbonifer-ous [31], supporting the concept of a large Paleo-Tethys ocean in Devonian times, leading to aPangea A collisional assembly of continent andterranes in Late Carboniferous times (Figs. 4and 5).

The other main alternative concerns the Neo-Tethys and its prolongation into the east Medi-terranean, Ionian Sea domain. Although the geo-dynamic evolution of NeoTethys is now relativelywell constrained, mainly through recent studies ofits Oman and Himalayan segment (e.g. [25]), itsrelationship with the east Mediterranean basin ismore debatable. A new interpretation considersthe east Mediterranean domain as part of theNeoTethyan oceanic system since the Late Paleo-zoic [24,25,32].

This is supported by:

1. The geophysical characteristics of the IonianSea and east Mediterranean Basin (isostaticequilibrium, seismic velocities, elastic thick-ness) which exclude an age of the sea £ooryounger than Early Jurassic.

2. The subsidence pattern of areas such as theSinai margin, the Tunisian Je¡ara rift, Sicilyand Apulia sensu stricto, con¢rming a LatePermian onset of thermal subsidence for theeast Mediterranean and Ionian Sea basinsand the absence of younger thermal events.

3. The Triassic MORB found in Cyprus (Mamo-

nia complex [33]), that is certainly derived fromthe east Mediterranean sea £oor.

4. Late Permian Hallstatt-type pelagic limestone,similar to that found in Oman, where it com-monly rests directly on MORB [34,35] has alsobeen reported from the Sosio complex [36] inSicily. This Late Permian pelagic macrofaunahas a¤nities with both Oman and Timor andimplies a Late Permian direct deep-water con-nection of the east Mediterranean Basin withthe NeoTethys, also con¢rmed by microfaunal¢ndings in Sicily.

This has also some bearings on the Apulian andTauric plates dynamics and the opening of theAlpine Tethys:

1. Apulia sensu stricto (southern Italy) belongedto the Cimmerian blocks, and then to Gond-wana after the cessation of spreading in theeast Mediterranean domain in Late Triassictimes. This implies that a Permian ¢t wherethe Adriatic plate (northern Italy) is left in itspresent-day position relative to Apulia is im-possible, as it would bring Adria over the Ibe-ric plate [10]. An alternative model [37], givingan even tighter ¢t than the one we propose forIberia, placing Adria south of the French Pyr-enees, is not supported by geological data fromsurrounding areas. We prefer a model whereItaly is split in two, the Adriatic plate reachingits present-day position, with respect to Apulia,during the Late Cretaceous^Paleogene, in awest-directed extrusion process related to theformation of the western Alps.

2. The Tauric plate (southern Turkey) also be-came part of Gondwana in Late Triassic timesuntil the east Mediterranean Ocean started tosubduct under it in the Miocene. In the mean-time, the Tauric plate was located 2000 kmwest of its present-day location with regardto the Pontides area (northern Turkey). Thisimplies that elements found on both sides ofthe Izmir^Ankara suture zone had a quite dif-ferent geological history before their Paleogenejuxtaposition.

3. The opening of the Alpine Tethys is necessarilythe eastward continuation of the central Atlan-

1 http://www.elsevier.nl/locate/epsl, mirror site http://www.elsevier.com/locate/epsl



Fig. 4. See legend of Fig. 3.



tic, as no other oceanic domains opened be-tween Africa and Europe during the Jurassic.This allows its width and average spreadingrate to be determined with precision usingmagnetic anomalies in the Atlantic.

4. Dynamics of the Tethyan realm

Rheic ocean opening, separating Avalonia fromGondwana (Fig. 3), is constrained by subsidenceanalysis [38] as well as paleomagnetism (e.g. [39]),and its docking to Laurentia by deformation [40]and post-collisional deposits [41]. The extensioneastward of Avalonia to Moesia and the Istan-bul^Zonguldak terranes is supported by faunala¤nities [42]. The Rheic ocean should have had,at its early stages, a lateral continuation in theformer eastern prolongation of peri-Gondwananmicrocontinents (e.g. Cadomia, Intra-Alpine Ter-rane). Subduction of oceanic ridge (ProtoTethys)triggered the break-o¡ of Avalonia, whereas inthe eastern prolongation, the presence of the ridgetriggered the amalgamation of volcanic arcs andcontinental ribbons with Gondwana (Ordovicianorogenic event [2]).

Renewed Gondwana-directed subduction led tothe opening of the PaleoTethys (Fig. 3) associatedwith the detachment of the ribbon-like Hunsuperterrane along the northern margin of Gond-wana. Diachronous subsidence patterns ofTethyan margins since the early Paleozoic [6] pro-vide constraints for PaleoTethys opening duringthe Late Ordovician and the Silurian and thesplitting of the Hun terranes in two parts, theEuropean and Asiatic. The latter was a¡ected byterranes amalgamation in Ordovician^Siluriantimes (Serindia terranes comprising parts of northChina and Tarim, e.g. [43,44]).

The Variscan orogeny in Europe was not a con-tinent^continent collision but a major accretion ofperi-Gondwanan terranes, which took placemainly in Devonian times (Fig. 4). The main Va-riscan metamorphic events are then regarded asresulting from this collision, as well as from thesubduction of the PaleoTethys mid-oceanic ridgein the Early Carboniferous. This led to the for-mation of the Variscan cordillera before the ¢nal

collision of Gondwana with Laurasia in Late Car-boniferous times.

The evolution of the Kipchak arc is taken from[45] and of the Asiatic ocean and Asiatic Hunicterranes from [46,47].

Gondwana collided with Laurasia in Late Car-boniferous time, the event being well recorded bydeformation of surrounding foreland basins, butthe ¢nal closure of PaleoTethys in the Tethyandomain took place during the Cimmerian (Trias-sic to Jurassic) orogenic cycle (Fig. 5). The north-ward subduction of PaleoTethys is responsible forthe widespread Late Carboniferous calc-alkalineintrusions and volcanism found in the VariscanAlpine^Mediterranean domain. Slab rollbackalso produced a general collapse of the pre-exist-ing Variscan cordillera and large-scale lateral dis-placement of terranes took place, con¢rmed bypaleomagnetic data [48]. A rift opened betweenAfrica and East Gondwana in the Late Carbon-iferous, joining southward to the Karroo system.This rift was a future aborted branch of Neo-Tethys whereas another rift appeared in theeast Mediterranean area (e.g. Je¡ara basin of Tu-nisia).

NeoTethys opened from the Late Carbonifer-ous to late Early Permian starting east of Austra-lia and progressing to the east Mediterranean area(Fig. 6), as recorded from subsidence patterns ofits southern margin [25]. This opening was asso-ciated with the drifting of the Cimmerian super-terrane and the ¢nal closure of PaleoTethys inMiddle Triassic times. The opening of NeoTethysand detachment of the Cimmerian blocks in Per-mian times was realized through increasing slab-pull forces in the PaleoTethys domain followingthe subduction of its mid-oceanic ridge below theEurasian margin (e.g. accretion of PermianMORB in Iran [49]). At the same time, slab roll-back of the PaleoTethys opened numerous back-arc basins along the Eurasian margin and resultedin the ¢nal collapse of the Variscan cordillera [7].

Between 240 and 220 Ma northward subduc-tion of the PaleoTethys triggered the opening ofback-arc oceans along the Eurasian margin fromAustria to China (Fig. 7). The fate of these Per-mo-Triassic marginal basins is quite di¡erentfrom area to area. Some closed during the Cim-



merian collisional events (Karakaya, Agh-Dar-band, Ku«re), others (Meliata^Maliac^Pindos) re-mained open and their delayed subduction in-duced the opening of younger back-arc oceans(Vardar, Black Sea).

The evolution of the Tibetan marginal oceans isderived from the work of [15]. Concomitant LatePermian^Triassic opening of the marginal Melia-ta^Maliac^Pindos oceans (within the Eurasianmargin) and NeoTethys (within the northern mar-gin of Gondwana) accelerated the closure ofPaleoTethys in the Dinaro^Hellenide region.Late Permian to Middle Triassic melanges andfore-arc basin remnants found in the Dinarides,Hellenides and Taurides, indicate a ¢nal closureof this Paleozoic ocean at that time (Eocimmerianevent). In northern Turkey and Iran, the collisionof the Cimmerian terranes with the Eurasian mar-gin was more complex due to the presence ofoceanic plateaus and Marianna-type back-arcbasins between the two domains (e.g.[50]).

NeoTethys subduction (Fig. 8) recordedthrough the onset of magmatism along its north-ern Iranian margin, created a strong slab pull,which contributed to the break-up of Pangeaand the opening of the central Atlantic Ocean inEarly Jurassic time. This break-up extended east-ward into the Alpine Tethys trying to link withnew Eurasian back-arc oceans like the IzmirÂn-kara and south Caspian Ocean. The diachronoussubduction of the NeoTethys spreading ridge isheld responsible for major changes in the LateJurassic to Early Cretaceous plate tectonics andthe ¢nal break-up of Gondwana, as well as thedetachment of the Argo-Burma terrane o¡ Aus-tralia, and the detachment of the Indian plate. Ifthe collision of the Argo-Burma terrane with Ma-laysia can be placed before the Santonian, thereare still some doubts about the timing of its sep-aration from Australia. We adopted here a solu-tion which does not ¢t rheological constraints butfollows general views on a Late Jurassic driftingof Argoland. The problem is the necessity to de-velop a far too long transform to link up the newArgo mid-ocean ridge with the still-existing Neo-Tethys ridge north of the Indian plate. An EarlyCretaceous drifting of Argoland together with In-dia would be far less problematic.

Between the Valanginian and the Santonian,the subduction of the NeoTethyan mid-oceanicridge south of Iran together with the subductionof the remnant VardarÎzmirÂnkara oceans trig-gered a strong transtensional stress partly respon-sible for the break-up of Gondwana and theopening of a north^south oceanic realm fromthe Mozambique area up to the NeoTethys (Fig.9). Changes in the subduction style in the Paleo-Paci¢c around Australia and Antarctica is cer-tainly also responsible for the break-up of easternand western Gondwana.

The position of the Indian plate with respect toAfrica is de¢ned by the oceanic isochrons in theSomalia^Mozambique basins from Late Jurassicto Late Cretaceous. The rotation of east Gondwa-na (comprising the future Indian plate) with re-spect to Africa was responsible for intra-oceanicsubduction within the NeoTethys along a paleo-transform and the onset of spreading of the Sem-ail Marianna-type back-arc (Fig. 9). The directionof the Semail intra-oceanic subduction is con-trolled by the age of the oceanic crust on eachside of the transform fault, the older one subduct-ing under the younger one.

The complex obduction/collision of the Vardarplate with the surrounding passive margins, fol-lowed the Late TriassicÊarly Jurassic intra-oce-anic subduction of the Meliata^Maliac oceans,related to the southward subduction of the Ku«reOcean. The north-east-directed subduction of thecombined remnant Vardar and IzmirÂnkaraoceans, generated the collision of an intra-oceanicarc with the Austro^Carpathian and Balkanideareas in late Early Cretaceous times. All theseevents are well recorded through formation of ac-cretionary melanges, fore-deep basins, deforma-tion, magmatism and metamorphism. At aboutthe same time, the Iberic plate became part ofthe African plate and drifted eastward, its easternpromontory, the Brianc°onnais, becoming impli-cated in the Alpine orogenic system.

The Alpides orogenic system sensu stricto(Alps, Carpathian and Balkans) is directly relatedto the evolution of the Maliac^Maliata/Vardarevolution and the fact that the subduction of theMaliac^Meliata slab was directed southward. Theorogenic prism moved from the Austro-Carpathi-



an area to the Alpine Tethys and extended west-ward to the western Alps in Late Cretaceous time.

In contrast, the Tethysides system is related toa north-directed subduction of the NeoTethys andassociated back-arc basins (e.g. Semail andSpongtang), the large NeoTethys slab was entirelysubducted, leaving hardly any ophiolitic traces;all NeoTethyan ophiolites being derived fromCretaceous back-arc basins

5. Conclusions

The plate tectonic model presented here with itsdynamic plate boundaries and synthetic paleo-iso-chrons o¡ers new information on oceans widths,on the timing of major tectonic events and on theprocesses responsible for their existence and dis-appearance. Geodynamic principles provide stableconstraints when geological information is sparse.

The PaleoTethys opened as a back-arc basindue to the slab rollback of the Rheic and Asiaticoceans, resulting in the rifting of a ribbon ofsuperterranes from the Gondwana margin. TheVariscan orogeny in Europe is not a continent/continent collision but the accretion of these ter-ranes. The Late Carboniferous^Early Permiancalc-alkaline intrusions and volcanism found inthe Variscan Alpine^Mediterranean domain is re-lated to the northward subduction of the Paleo-Tethys. Slab rollback of the latter produced thecollapse of the pre-existing Variscan cordilleraand led to the opening of NeoTethys.

NeoTethys replaced PaleoTethys even whenPangea was stable during Permian and Triassictimes, pointing out the key role played by slab-pull forces on ocean evolution and consequentlyon plate distribution.

The major changes in the Late Jurassic to EarlyCretaceous plate tectonics can be associated withthe diachronous subduction of the NeoTethys ac-tive ridge along the Eurasian northern margin.The slab pull forced opened the Argo AbyssalPlain and detached the Argo-Burma terranefrom Australia, possibly together with the Indianplate.

The subsequent Valanginian rotation of eastGondwana relative to Africa induced an intra-

oceanic subduction within the NeoTethys southof Iran and the onset of spreading of the SemailMarianna-type back-arc ocean.

The history of the western Tethys is complex;cannibalism of three generations of back-arcs(Maliac, Maliata, Vardar) took place and resultedin the Alpides orogenic system. Such a great com-plexity in a small region tests the limits of thereconstruction process.

Acknowledgements

We thank all the scientists involved in theIGCP 369 project, J. Mosar for his pioneeringwork on the subject and J. von Raumer for shar-ing his knowledge of the Paleozoic. H. Kozur hasactively participated in the choice of alternativemodels. This manuscript bene¢ted from the gen-erous e¡ort and constructive criticism of M. Gur-nis, R.D. Mu«ller and B. Murphy. Project partlysupported by FNRS Grant numbers 20-53P646.98and 20-61P885.00.[RV]

References

[1] G.M. Stamp£i, G.D. Borel, W. Cavazza, J. Mosar, P.A.Ziegler, Palaeotectonic and palaeogeographic evolution ofthe western Tethys and PeriTethyan domain (IGCP Proj-ect 369), Episodes 24 (2001) 222^228.

[2] J. von Raumer, G.M. Stamp£i, G.D. Borel, F. Bussy, Theorganisation of pre-Variscan basement areas at the north-Gondwanan margin, Int. J. Earth Sci. (2001) in press.

[3] I. Vavassis, A. DeBono, G.M. Stamp£i, D. Giorgis, A.Valloton, Y. Amelin, U^Pb and Ar^Ar geochronologicaldata from the Pelagonian basement in Evia (Greece): geo-dynamic implications for the evolution of Paleotethys,Schweiz. Mineral. Petrogr. Mitt. 80 (2000) 21^43.

[4] G.M. Stamp£i, G. Borel, J. von Raumer, Terrane accre-tion in the Variscan domain, in: Basement Tectonics 15,La Corun¬a, Spain, 2000, 167^169.

[5] G.M. Stamp£i, J. von Raumer, G.D. Borel, The Palaeo-zoic evolution of pre-Variscan terranes: From peri-Gond-wana to the Variscan collision, in: J.R. Martinez-Catalan,R.D. Hatcher, R. Arenas, F. Diaz Garcia (Eds.), VariscanAppalachian Dynamics: The Building of the Upper Pa-leozoic Basement, Geological Society of America SpecialPaper, 2002, in press.

[6] G.M. Stamp£i, Tethyan oceans, in: E. Bozkurt, J.A. Win-chester, J.D.A. Piper (Eds.), Tectonics and Magmatism inTurkey and Surrounding Area, Geological Society ofLondon Special Publication 173 (2000) 163^185.



[7] P.A. Ziegler, G.M. Stamp£i, Late Paleozoic^Early Meso-zoic plate boundary reorganisation: collapse of the Varis-can orogen and opening of Neotethys, in: R. Cassinis(Ed. 25), The Continental Permian of the Southern Alpsand Sardinia (Italy) Regional Reports and General Cor-relations, Annali Museo Civico Science Naturali, Brescia,2001, 17^34.

[8] G.M. Stamp£i, R.H. Marchant, Geodynamic evolution ofthe Tethyan margins of the Western Alps, in: O.A. P¢¡-ner, P. Lehner, P.Z. Heitzman, S. Mueller, A. Steck(Eds.), Deep Structure of the Swiss Alps ^ Results fromNRP 20, Birkhau«ser AG., Basel, 1997, 223^239.

[9] G.M. Stamp£i, J. Mosar, R. Marchant, D. Marquer, T.Baudin, G. Borel, Subduction and obduction processes inthe western Alps, in: A. Vauchez, R. Meissner (Eds.),Continents and Their Mantle Roots, Tectonophysics,296 (1998) 159^204.

[10] G.M. Stamp£i, J. Mosar, The making and becoming ofApulia, Mem. Sci. Geol. (Univ. Padova) spec. vol. 3rdworkshop Alpine geology (51/1), 1999, 141^154.

[11] R. Van der Voo, Paleomagnetism of the Atlantic, Tethysand Iapetus Oceans, Cambridge University Press, Cam-bridge, 1993, 411 p.

[12] R.D. Mu«ller, J.Y. Royer, L.A. Lawver, Revised platemotions relative to the hotspots from combined Atlanticand Indian ocean hotspot tracks, Geology 21 (1993) 275^278.

[13] C.S. Hutchinson, Geological Evolution of South-EastAsia, Oxford Monographs on Geology and GeophysicsClarendon Press, Oxford, 1989, 368 pp.

[14] R.H. Findley, The Song Ma anticlinorium, northern Viet-nam: the structure of an allochthonous terrane containingan Early Paleozoic island arc sequence, J. Asian Earth Sci.15 (1998) 453^464.

[15] A. Yin, T.M. Harrison, Geologic evolution of the Hima-layan-Tibetan orogen, Annu. Rev. Earth Planet. Sci. 28(2000) 211^280.

[16] J.J. Veevers, P. Morgan, S.Y. O'Relly, M.R. Walter, E.Scheibner, D.A. Clark, P.W. Schmidt, S. Clark, C. Gaina,R.D. Mu«ller, B.L.N. Kennett, M.A.J. Williams, Billion-Year Earth History of Australia and Neighbours inGondwanaland, Gemoc. Press, Sydney, 2000, 388 pp.

[17] M. Gurnis, Rapid continentla subsidence folowing theinitiation and evolution of subduction, Science 255(1992) 1556^1558.

[18] Z. Gvirtzman, A. Nur, Plate detachment, asthenosphereupwelling, and topography across subduction zones,Geology 27 (1999) 563^566.

[19] C. Lithgow-Bertelloni, M.A. Richards, The dynamics ofCenozoic and Mesozoic plate motions, Rev. Geophys. 36(1998) 27^78.

[20] M. Cloos, Lithospheric buoyancy and collisional orogen-esis : subduction of oceanic plateaus continental margins,island arcs, spreading ridges, and seamounts, Geol. Soc.Am. Bull. 105 (1993) 715^737.

[21] G. Caprarelli, E.C. Leitch, Magmatic changes during thestabilisation of a cordilleran fold belt : the Late Carbon-

iferous-Triassic igneous history of eastern New SouthWales, Australia, Lithos 45 (1998) 413^430.

[22] P.H. Dirks, M. Hand, W.J. Collins, R. O¥er, Structural-metamorphic evolution of the Tia Complex, New Englandfold belt ; thermal overprint of anaccretion-subductioncomplex in a compressional back-arc setting, J. Struct.Geol. 14 (1992) 669^688.

[23] T.H. Torsvik, M.A. Smethurst, Geographic mapping andpaleoreconstruction package (GMAP). Software descrip-tion and examples of paleoreconstructions, 1994.

[24] G. Stamp£i, J. Marcoux, A. Baud, Tethyan margins inspace and time, in: J.E.T. Channell, E.L. Winterer, L.F.Jansa (Eds.), Paleogeography and paleoceanography ofTethys, Palaeogeography, Palaeoclimatology, Palaeoecol-ogy 87 (1991) 373^410.

[25] G.M. Stamp£i, J. Mosar, P. Favre, A. Pillevuit, J.-C.Vannay, Permo-Mesozoic evolution of the westernTethyan realm: the Neotethys/East-Mediterranean con-nection, in: W. Cavazza, A.H.F.R. Robertson, P.A. Zie-gler (Eds.), Peritethyan Rift/Wrench Basins and PassiveMargins (IGCP 369) Peri-Tethys Memoir 6, Bull. Muse-um Nat. Hist. Nat., Paris, 2001, pp. 51^108.

[26] T.H. Torsvik, R. VanderVoo, J.G. Meert, J. Mosar, H.J.Walderhaug, Reconstructions of continents around theNorth Atlantic at about the 60th parallel, Earth Planet.Sci. Lett. 187 (2001) 55^69.

[27] C. Klootwijk, Phanerozoic Con¢gurations of GreaterAustralia: Evolution of the North West Shelf, AustralianGeological Survey Organisation, Canberra, 1996.

[28] M. Scha«tz, V. Bachtadse, J. Tait, H. So¡el, Paleomagneticresults from lower Devonian sediments of the southernAlps, Schr. Alfred-Wegener-Stiftung Terra Nostra 97(1997) 152^155.

[29] H. Torsvik, E.A. Eide, Phanerozoic Paleogeography andGeodynamics with Atlantic Details, NGU Report 98.001,1998, 82 p.

[30] M. Krs, P. Pruner, To the paleomagnetic investigations ofpaleogeography of the Barrandian terrane Bohemian mas-sif, Acta Universitatis Carolinae, Geologica 43 (1999)519^522.

[31] J. Tait, M. Scha«tz, V. Bachtadse, H. So¡el, Palaeomag-netism and Palaeozoic palaeogeography of Gondwanaand European terranes, in: W. Franke, V. Haak, O.Oncken, D. Tanner (Eds.), Orogenic Processes: Quanti¢-cation and Modelling in the Variscan Belt, GeologicalSociety of London Special Publication 179 (2000) 21^34.

[32] G.M. Stamp£i, A. Pillevuit, An alternative Permo-Triassicreconstruction of the kinematics of the Tethyan realm, in:J. Dercourt, L.-E. Ricou, B. Vrielinck (Eds.), Atlas TethysPalaeoenvironmental Maps. Explanatory Notes, Gau-thier-Villars, Paris, 1993, 55^62.

[33] J. Malpas, T. Calon, G. Squires, The development of alate Cretaceous microplate suture zone in SW Cyprus, in:H.M. Prichard, T. Alabaster, N.B.W. Harris and C.R.Neary (Eds.), Magamatic Processes and Plate Tectonics,Geological Society of London Special Publication 76(1993) 177^195.



[34] A. Pillevuit, Les blocs exotiques du Sultanat d'Oman,evolution paleogeographique d'une marge passive £exur-ale, Memoires de Geologie (Lausanne) 17 (1993) 249 pp.

[35] S. Niko, A. Pillevuit, T. Nishida, Early Late Permian(Wordian) non-ammonoid cephalopods from the HamratDuru Group, central Oman Mountains, Trans. Proc. Pa-laeontol. Soc. Japan N.S. 183 (1996) 522^527.

[36] H. Kozur, First evidence of Middle Permian AmmoniticoRosso and further new stratigraphic results in the Perm-ian and Triassic of the Sosio Valley area, Western Sicily,in: Proc. 1st Croatian Geological Congress 1, ZbornikRadova, Opatija, 1995, pp. 307^310.

[37] U.G. Wortmann, H. Weissert, H. Fuink, J. Hauck, Alpineplate kinematics revisited the Adria problem, Tectonics 20(2001) 134^147.

[38] J.K. Prigmore, A.J. Butler, N.H. Woodcock, Rifting dur-ing the separation of Eastern Avalonia from GondwanaEvidence from subsidence analysis, Geology 25 (1997)203^206.

[39] T.H. Torsvik, A. Trench, I. Svensson, H.J. Walderhaug,Paleogeographic signi¢cance of mid-Silurian paleomag-netic results from southern Britain ^ major revision ofapparent polar wander path for eastern Avalonia, Geo-phys. J. Int. 113 (1993) 651^668.

[40] J.D. Keppie, Northern Appalachian terranes and theiraccretionary history, in: D. Dallmeyer (Ed.), Terranes inthe circum-Atlantic Paleozoic orogens, Geological Societyof America Special Paper 230 (1989) 159^192.

[41] P.F. Friend, B.P.J. Williams, M. Ford, E.A. Williams,Kinematics and dynamics of Old Red Sandstone basins,in: P.F. Friend, B.P.J. Williams (Eds.), New perspectiveson the Old Red Sandstone, Geological Society of London,Special Publication 180 (2000) 29^60.

[42] W.T. Dean, O. Monod, R.B. Rickards, O. Demir, P. Bul-tynck, Lower Palaeozoic stratigraphy and palaeontologyKaradere-Zirze area, Pontu Mountains, northern Turkey,Geol. Mag. 137 (2000) 555^582.

[43] A. Yin, S. Nie, A Phanerozoic palinspastic reconstructionof China and its neighboring regions, in: A. Yin, M.Harrison (Eds.), The tectonic Evolution of Asia, Cam-bridge University Press, Cambridge, 1996, 442^485.

[44] Q.-R. Meng, G.-W. Zhang, Timing of collision of thenorth and south China blocks: controversy and reconcia-lation, Geology 27 (1999) 123^126.

[45] A.M.C. Sengo«r and B.A. Natl'in, Paleotectonics of Asia:fragments of a synthesis, in: A. Yin, M. Harrison (Eds.)The Tectonic Evolution of Asia, Cambridge Univ. Press,1996, pp. 443^486.

[46] L.P. Zonenshain, M.I. Kuzmin, M.V. Kononov, Absolutereconstructions of the Paleozoic oceans, Earth Planet. Sci.Lett. 74 (1985) 103^116.

[47] Institute of Tectonics of Lithospheric Plates, Moscow,The Paleogeographic Atlas of Northern Eurasia, 1997,26 maps.

[48] J.B. Edel, The rotations of the Variscides during the Car-boniferous collision paleomagnetic constraints from theVosges and the Massif Central (France), Tectonophysics332 (2001) 69^92.

[49] A.W. Ruttner, Southern borderland of Triassic Laurasiain NE Iran, Geol. Rundsch. 82 (1993) 110^120.

[50] A.I. Okay, Was the Late Triassic orogeny in Turkeycaused by the collision of an oceanic plateau?, in: E.Bozkurt, J.A. Winchester, J.D.A. Piper (Eds.), Tectonicsand Magmatism in Turkey and the Surrounding Area,Geological Society of London Special Publication 173(2000) 25^41.



a plate tectonic model for the paleozoic and mesozoic...

Documents