Available online at www.sciencedirect.com

Tectonophysics 444 (2007) 63 – 79 www.elsevier.com/locate/tecto

Modelling the extension of heterogeneous hot lithosphere Dimitrios Sokoutis a,⁎, Giacomo Corti b , Marco Bonini b , Jean Pierre Brun c , Sierd Cloetingh a , Thomas Mauduit a , Piero Manetti d a

Netherlands Centre for Integrated Solid Earth Sciences, Faculty of Earth and Life Sciences, De Boelelaan 1085, Vrije Universiteit, 1081 HV Amsterdam, The Netherlands b CNR, Istituto di Geoscienze e Georisorse (IGG), Unità Operativa di Firenze, via G. La Pira 4, I-50121 Firenze, Italy c Geosciences Rennes UMR 6118 CNRS, Université de Rennes 1, Campus de Beaulieu, F-35042 Rennes Cedex, France d Università degli Studi di Firenze, Dipartimento di Scienze della Terra, via G. La Pira 4, I-50121 Firenze, Italy Received 21 July 2006; received in revised form 24 July 2007; accepted 2 August 2007 Available online 29 August 2007

Abstract The consequences of weak heterogeneities in the extension of soft and hot lithosphere without significant previous crustal thickening has been analysed in a series of centrifuge models. The experiments examined the effects of i) the location of heterogeneities in the ductile crust and/or in the lithospheric mantle, and ii) their orientation, perpendicular or oblique to the direction of bulk extension. The observed deformation patterns are all relevant to the so-called “wide rifting” mode of extension. Weak zones located in the ductile crust exert a more pronounced influence on localisation of deformation in the brittle layer than those located in the lithospheric mantle: the former localise faulting in the brittle crust whereas the latter tend to distribute faulting over a wider area. This latter behaviour depends in turn upon the decoupling provided by the ductile crust. Localised thinning in the brittle crust is accompanied by ductile doming of both crust and mantle. Domains of maximum thinning in the brittle crust and ductile crust and mantle are in opposition. Lateral differences in brittle crust thinning are accommodated by lateral flow in the ductile crust and mantle. This contrasts with “cold and strong” lithospheres whose high strength sub-Moho mantle triggers a necking instability at the lithosphere-scale. This also differs from the extension of thickened hot and soft lithospheres whose ductile crust is thick enough to give birth to metamorphic core complexes. Thus, for the given lithospheric rheology, the models have relevance to backarc type extensional systems, such as the Aegean and the Tyrrhenian domains. © 2007 Elsevier B.V. All rights reserved. Keywords: Backarc extension; Analogue modelling; Weakness zones; Continental lithosphere rheology

1. Introduction Modes of extension strongly depend on the rheological layering of the continental lithosphere (Buck, 1991) but the location and distribution of rifts are often controlled by pre-existing heterogeneities (e.g. Ziegler and Cloetingh, 2004). Among the parameters that ⁎ Corresponding author. Tel.: +31 20 5986583. E-mail address: [email protected] (D. Sokoutis). 0040-1951/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2007.08.012

control the rheological layering of continental lithosphere, the thermal state is likely to be the most important one. When the Moho temperature reaches 700–800 °C (which roughly corresponds to surface heat flow of 70–75 mW/m2) the mantle peridotite becomes entirely ductile inducing a drop in bulk lithospheric strength (Buck, 1991; Ranalli, 2000; Alfonso and Ranalli, 2004). This range of Moho temperatures makes the difference between “cold and strong” lithosphere, whose highest strength is located in the

64

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

sub-Moho mantle, from “hot and weak” lithosphere, whose only high strength layer is the brittle crust. When extended strong and weak lithospheres give contrasting modes of rifting (Buck, 1991; Brun, 1999). The presence of a high strength mantle favours the development of lithospheric-scale necking that can lead to extreme crustal thinning, allowing the mantle to reach the surface (Brun and Beslier, 1996). In the absence of a high strength mantle, the low viscosity of both lower crust and lithospheric mantle lead to a wider distribution of the deformation and to the maintenance of a flat Moho despite lateral variations in upper crustal stretching (Block and Royden, 1990; Buck, 1991). Therefore, because rock rheology is strongly dependent on temperature, the initial thermal state of the lithosphere plays a decisive role on the localisation of extension — i.e. narrow rifting versus wide rifting. Inherited heterogeneities can occur throughout the whole lithosphere. For instance, pre-existing faults and shear zones may represent a mechanical weakness at upper and lower crustal levels (Daly et al., 1989; Sibson, 1995; Ranalli, 2000). In Europe, several deep seismic profiles have illustrated the reactivation of pre-existing fabrics of Caledonian or Hercynian age during Mesozoic or Cenozoic extensional events (Cheadle et al., 1987; Klemperer and Mathews, 1987; Brun et al., 1991; Brun and Tron, 1993; Ziegler and Cloetingh, 2004). It has also been suggested that the source of structural inheritance could be related to the preservation, within the uppermost mantle, of a lattice preferred orientation of olivine crystals formed during previous tectonic phases (Vauchez et al., 1998; Tommasi and Vauchez, 2001). Geophysical data suggesting that the structural inheritance controlling rift development may be hidden in the upper lithospheric mantle also supports this view (e.g. Babuška et al., 2002). Thus, zones of weakness can either extend through the whole lithosphere, or be localised in the lithospheric mantle, or in the crust. Both analogue and numerical modelling have been used to argue and illustrate the strong control that can be exerted by the pre-existing weakness zones on rift evolution and symmetry/asymmetry of the resulting structures (Braun and Beaumont, 1987; Dunbar and Sawyer, 1988, 1989; Braun and Beaumont, 1989; Harry and Sawyer, 1992; Hopper and Buck, 1996; Corti et al., 2003a,b, 2004; Corti, 2004; Van Wijk, 2005). However, most of these studies have considered the extension of strong lithosphere, whereas extension often develops in regions of high surface heat flow like in back-arc-type tectonic environments (Collins, 2002; Hyndman et al., 2005; Currie et al., 2004). One of the major differences should be the mechanical behaviour of the sub-Moho mantle

(where the heterogeneities are located) that can either be of Mohr–Coulomb-type in the former category of lithospheres or dominantly non-linear viscous in the latter. In the present study, we investigate the effects of preexisting zones of weakness in soft lithosphere through centrifuge models that have undergone extension. Previous models (e.g., Corti, 2004) have shown that pre-existing weakness zones in the lower crust are able to control the architecture of continental rifts and related transfer zones. Here we analyse the influence of inherited fabrics on rifting in a more detail by considering weakness zones that are i) located in the lithospheric mantle and/or in the lower ductile crust and ii) trend perpendicular and/or oblique to the direction of extension. The position and symmetry of fault patterns that develop in the upper crust are studied as functions of depth, location and orientation of the weakness zones. The tectonic response of these models of soft and hot lithospheres is compared to those of strong and cold lithospheres. The applicability of the experimental results is discussed in particular with reference to extensional environments of backarc-type. 2. Modelling set-up 2.1. General definition of the models Our models intend to represent a non-thickened lithosphere in a region of high surface heat flow, as occurs in backarc basins. Consequently, they are made of an upper brittle crust, a ductile lower crust and a fully ductile lithospheric mantle. Three types of models are used in the present paper: a) experiments investigating lithospheric extension with laterally uniform rheological stratification; b) experiments that simulate extension with a weakness zone in the upper lithospheric mantle trending at variable angles to the direction of extension and c) models investigating extension with weakness zones in both the upper lithospheric mantle and ductile crust trending at variable angles. Similar models that were carried out using the same technique but without crust and mantle weak zones (Corti, 2005) provide a context for the results of the present paper. In the following, we refer to “α” as the angle between the weakness zone in the ductile crust and normal to the direction of extension, whereas “β” represents the angle between this latter direction and the weak mantle trend (Fig. 1c; Table 1). All models have been submitted to symmetrical boundary displacement (see Fig. 1b). The experiments were performed in an artificial gravity field of gravitational acceleration 100 g at the

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

HRTL of Uppsala University (Sweden) and at the Tectonic Modelling Laboratory of the CNR-IGG at the Earth Sciences Department of Firenze (Italy) (Fig. 1).

65

Table 1 Models parameters Model α angle 1 2 3 4 5

– 90° – 90° 30°

β angle

Extension velocity (mm/s)

Bulk extension (%)

90° 90° 30° 30° 90°

5 10− 1 3 10− 1 5 10− 1 5 10− 1 5 10− 1

60 60 65 60 55

2.2. Experimental deformation

Fig. 1. (a) Top view of centrifuge cabinet showing loading conditions in the centrifuge, where w represents the angular velocity and CF is the centrifuge force field. (b) Schematic representation of the experimental set-up. Removal of rectangular Plexiglas spacers allows the models, under the centrifuge force field, to extend by filling the empty space. For facilitating the description, the experimental results are discussed by reference to the central axis of the models. BC: brittle crust; DC: ductile crust; LM: lithospheric mantle; DCWZ: ductile crust weak zone; MWZ: mantle weak zone. (c) Model top view showing the different configurations of the weakness zones in the experiments, where: “α” is the angle between the DCWZ and the normal to the direction of extension and “β ” is the angle between the MWZ and the normal to the direction of extension.

The rheologically stratified models, with dimensions 8.5 × 7.0 × 2.5 cm, were built inside a transparent Plexiglas box with internal dimensions 19.0 × 7.0 × 10.0 cm. Rectangular plastic blocks (spacers) confined the models that were allocated in a pendulum system, which rotated in response to the centrifugal body force. Progressive removal of spacers was operated during successive runs in the centrifuge (Fig. 1b). This allowed a vertical thinning and lateral expansion of models in response to the centrifugal body force field simulating gravity in nature and resulting in a plane strain deformation (see also Mulugeta, 1988; Bonini et al., 2001; Mulugeta and Ghebreab, 2001). Layer thicknesses were chosen such as sand layers, in particular, that were thick enough to allow a clear visualisation of fault patterns. The Plexiglas wall was lubricated with Vaseline oil in order to minimise boundary shear. Accordingly, model top views (e.g. Fig. 1c) show that neither passive markers nor faults display disturbance against lateral walls. The model surface was free and the basal model boundary was in contact with a low viscosity asthenospheric fluid. Therefore, neither the top nor the base of the model underwent significant shear stress. The extension velocity imposed on the models was of ≈ 3.6 × 10− 1 mm/s as an average (see Table 1 for details). All models have been deformed up to 65% bulk extension. After deformation, the surface of the models has been covered with sand in order to preserve the topography. Models have been then soaked in water, frozen and sectioned. 2.3. Experimental rheology and materials The models represent a three-layer lithosphere with embedded weak zones floating above a fluid-like model asthenosphere (Fig. 2). The experimental set up is based on the assumption that thermal and/or mechanical processes decreased the mantle strength during previous deformation phases. Rheological stratification of models was reproduced by using dry quartz-feldspar sand

66

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

Fig. 2. Experimental strength profiles at the onset of deformation, calculated in the central part of the models. The values of the integrated lithospheric strength are also reported. All initials used here and for the rest of the figures are the same as in Fig. 1.

mixture for simulating the brittle crust and viscous mixtures for the ductile crust and upper lithospheric mantle. The rheological behaviour of the silicone mixtures representing the ductile experimental units was analysed by a coni-cylindrical viscometer. For details see Table 2, where the parameters of the material properties are summarised. 2.4. Scaling of models Scaling of the models to the natural prototype was achieved by maintaining similarity in geometry, dynamics, kinematics and rheology (e.g., Ramberg, 1981). The models were intended to represent a geometrical scale of h⁎ = hm/hn ≈ 3 × 10− 7 (where h⁎is the adimensional model length ratio, and hm and hn are the corresponding lengths in model and nature, respectively). This implies that 1 cm in the experiment

corresponds to ≈ 30 km in nature. Dynamic similarity was achieved by maintaining the resistance of the single layers and the analogue lithosphere scaled to nature. Similarity conditions were checked by calculating dimensionless ratios relating gravitational stresses to differential stresses for brittle and ductile deformation (e.g., Corti et al., 2004). Scaling analysis indicates that the models simulate extension, at rates of ∼ 3 cm/yr, of a ∼ 80-km thick continental lithosphere characterised by a total resistance between 7.5 × 1012 and 1.1 × 1013 Pa m. The experimental strength profiles, reported in Fig. 2, are valid for the very early stages of deformation. 2.5. Limitations of modelling Since lithosphere rheology and thermal boundary conditions are simplified in the models, extrapolation of model results to the nature must be carried out with care.

Table 2 Characteristics of experimental materials for the different series Material

Prototype layer

Thickness Density (mm)

Dry Quartz–Feldspar Sand mixture Rhodorsil Gomme 70009 (manufactured by Rhone Poulenc, France)+ Qz Sand 5:4.5(% in weight) Ductile crust mixture + Oleic acid 5:1 (% in weight) Rhodorsil Gomme 70009 (manufactured by Rhone Poulenc, France) + Qz Sand 5:5 (% in weight) Lithospheric mantle mixture + Oleic acid (% in weight) Gypsum Glycerol mixture

Brittle crust (BC) Ductile crust (DC)

10

1360 kg m− 3 μ = 0.6

5

1420 kg/m3

Brittle behaviour Power-law

1430 kg/m3

Power-law

1475 kg/m3

Power-law

10

1480 kg/m3

Power-law

15

1600 kg/m3

Newtonian

Ductile Crust 5 Weak Zone (DCWZ) Lithospheric 10 mantle (LM) Mantle Weak Zone (MWZ) Asthenosphere

(⁎) Measured at room temperature (23 °C ± 0.5 °).

Coefficient of Cohesion Rheological Power-law internal friction characteristics parameters 80 Pa s

n = 1.2 A = 2.5 10−6 (⁎)

n = 1.3 A = 4.5 10−6 (⁎) n = 1.3 A = 5 10−7 (⁎) n = 1.7 A = 1.5 10−7 (⁎) η = 40 Pa s (⁎)

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

Our initial lithospheric layering was intended to reflect thermal conditions of backarc-type environments at the onset of extension – i.e., hot lithosphere – but the models do not take into account any thermal effect that could occur during extension. Specifically, for backarc-type settings such an assumption can be justified by the fact that extension is generally fast and therefore the contribution of possible thermal readjustments could be limited. This prevented rheological modifications in the layering and destabilisation of both the brittle–ductile transition in the crust and the Moho discontinuity. Similarly, syn-extension surface processes as erosion and sedimentation were not considered during the experi-

67

ments. Despite these simplifications, the current analogue models were able to trace the large-scale response to extension of the continental lithosphere and allowed analysis of the effects of pre-existing crustal and/or mantle weaknesses on the rifting process. 3. Experimental results 3.1. Main outcomes from previous works Previous experimental work has shown that extension of brittle–ductile multilayers with only one brittle layer lying above one or several ductile layers resulted in

Fig. 3. Final deformation stage of model-1, with orthogonal MWZ, at 60% bulk extension. (a) Initial set-up; (b) top surface photograph. The line AB indicates the position of the cross-section; (c) cross-section photograph with interpretation of the fault pattern (note that the white sand on top of the model has been added after deformation to preserve the surface topography); (d) plot of horizontal distance versus normalised thickness, which is defined as the ratio of individual layer thickness to total experimental lithospheric thickness. The maximum values of horizontal stretching occur in the MWZ, while the thinning of the brittle crust and mantle are in opposite phase. The maximum thinning of the lithosphere and brittle crust appears at the same place. In this figure and for the following the initials, i and f correspond to initial and final stage, respectively.

68

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

Fig. 4. Model-2 at 60% bulk extension with orthogonal MWZ and DCWZ. (a) Initial set-up; (b) top view photograph. The line AB indicates the position of the cross-section; (c) photograph of the cross-section with fault pattern interpretation. The cross-section depicts classical rift geometry, with major thinning below the central rift depression associated with asthenosphere upwelling; (d) plot of horizontal distance versus normalised thickness. It displays mirrortype symmetry on the response to thinning between the experimental brittle crust and mantle. The maximum thinning of the lithosphere and brittle crust appears at the same place.

distributed faulting — i.e., wide rifting (Faugere and Brun, 1984; Brun et al., 1994; Corti, 2005; Tirel et al., 2006). Three experiments realised in the frame of the present project, but not presented here, confirmed this conclusion. The brittle crust deformation consists of closely spaced tilted fault-bounded blocks and/or conjugate normal faults, distributed throughout the models. Nearly uniform thinning dominates the ductile layers, so that the deformation mechanism may be broadly considered as of pure shear type. In these models, characterised by distributed faulting, high strain rates – i.e., strong coupling between brittle and ductile layers – favour parallel faulting and block tilting (Faugere and Brun, 1984; Brun, 1999). Conversely, low strain rates – i.e., weak brittle–ductile coupling – favour

conjugate normal faulting and the development of horst and graben structures (Allemand, 1990; Brun, 1999). Lateral variations of viscosity or the presence of viscosity anomalies below the brittle–ductile interface strongly modifies the pattern of faulting. A local anomaly with a viscosity of only one order of magnitude lower than the rest of the ductile layer can initiate a strong strain localisation in the overlying brittle layer leading to the exhumation of the ductile layer and formation of a core complex (Brun et al., 1994). Where a series of tilted blocks affect the brittle layer, the direction of fault dip depends on the sense of shear in the underlying ductile layer (Faugere and Brun, 1984) and therefore tilted blocks can be used as shear criteria for the kinematic analysis of more complicated models.

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

3.2. Weakness orthogonal to the direction of extension 3.2.1. Weakness only in lithospheric mantle 3.2.1.1. Model-1 (Fig. 3). In reference to the central axis (Fig. 1b), the deformed model-1 exhibits mirror-type symmetry in terms of both thinning variations and location as well as internal symmetry of fault zones (Fig 3). Faulting in the brittle crust is situated in three domains that are located one above and two aside the weak mantle zone (MWZ) (Fig. 3b, c). Each of them is associated with a smooth upward bending of the brittle– ductile transition (Fig. 3c, d). In the ductile layers,

69

maximum values of horizontal stretching occur in the MWZ, as is also demonstrated by its marked thinning. All three faulting domains display conjugate fault patterns. In the central domain faulting leads to a symmetrical variation in layer thickness. In contrast, the two lateral domains exhibit asymmetrical thickness variations with the maximum layer thinning located toward the central axis. The opposite sense of thinning of the brittle crust and mantle is clearly shown in the plot of horizontal distance versus normalised thickness (Fig. 3d). In this work the term “normalised thickness” is defined as the ratio of individual layer thickness to total experimental

Fig. 5. Final deformation stage of model-3 with oblique MWZ. (a) Initial set-up; (b) top surface photograph of the model at 65% bulk extension. The lines AB and CD specify the position of the cross-sections; (c) cross-section photograph with interpretation of the fault pattern. In this cross-section the MWZ is located to the right of the central axis of the model while in line CD (the photograph of the section is not presented here) the MWZ is located to the left of the central axis. They exhibit mirror-type fault pattern symmetry; (d) plot of horizontal distance versus normalised thickness. The response of the mantle and brittle crust towards thinning is the same but in opposite direction, while the maximum thinning of the experimental lithosphere and brittle crust takes place at the same spot.

70

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

Fig. 6. Final deformation stage of model-4 with oblique MWZ and orthogonal DCWZ. (a) Initial set-up; (b) top surface photograph of the model at 60% bulk extension. The line AB specifies the position of the cross-section; (c) cross-section photograph with interpretation of the fault pattern. This cross-section shows two main domains of faulting, located at the extremities of the ductile crust anomaly. The conjugate fault pattern is slightly asymmetric; (d) plot of horizontal distance versus normalised thickness. Note that the response to thinning of the brittle crust and mantle is in the opposite sense; additionally, the maximum thinning of the lithosphere and brittle crust appears at the same place.

lithospheric thickness. From the same plot we can deduce that the maximum thinning of the lithosphere and brittle crust appears at the same place. 3.2.2. Weaknesses in both ductile crust and lithospheric mantle 3.2.2.1. Model-2 (Fig. 4). Faulting in the brittle crust can be divided in three domains orthogonal to the bulk extension direction. The deformation is localised at the central part (central axis) of the model-2 where the more developed central, wider graben domain is accompanied by two less developed marginal ones (Fig. 4b, c). The conjugate fault pattern is present in all three domains but

with a very pronounced normal fault at the right side of the central axis. This fault is located at the edge of the ductile crust anomaly. The cross-section depicts classical rift geometry, with major thinning below the central rift depression associated with asthenosphere upwelling (Fig. 4c, d). This area corresponds to the main central zone of deformation observed in the brittle layer, while the two others are related to marginal grabens. No significant thinning of the ductile layers is observed below the marginal grabens, whereas strong thinning characterises the central graben. The plot of horizontal distance versus normalised thickness displays mirror-type symmetry of the response to thinning of the experimental brittle crust

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

and mantle. The maximum thinning of the lithosphere and brittle crust appears at the same place (Fig. 4d). 3.3. Mantle weakness oblique to the direction of extension; β = 30° 3.3.1. Model-3 (Fig. 5) Surface observation points out that the central part of the model is free of faulting (Fig. 5b). The cross-section, line AB, presented here is such that the centre of MWZ is located to the right of the central axis of the model. It displays two domains of faulting symmetrically disposed with reference to the central axis where rightdipping normal faults dominate the internal fault system. In the cross-section, line CD (the photograph of the

71

section is not presented here), the MWZ is located to the left of the central axis. It exhibits a comparable fault pattern but with the faults dipping to the left. In other words, even if the model appears rather symmetrical in terms of thinning it is not in terms of faulting (Fig. 5c). The system of parallel faults indicates a layer parallel shear into the underlying ductile layer (see arrows in Fig. 5c). Looking at the plot of horizontal distance versus normalised thickness, we may state that the response of the mantle and brittle crust to thinning is quite similar but opposite in direction (Fig. 5d). Another straightforward observation is that the maximum thinning of the experimental lithosphere and brittle crust takes place at the same spot.

Fig. 7. Final deformation stage of model-5 with orthogonal MWZ and oblique DCWZ. (a) Initial set-up; (b) top surface photograph of the model at 55% bulk extension. The line AB specifies the position of the cross-section; (c) cross-section photograph showing that the main normal faults originate along the external DCWZ boundaries; the right side of the mantle anomaly, MWZ, seems to be related to an overlying normal fault; (d) plot of horizontal distance versus normalised thickness. Note the mirror-type symmetry on the response to thinning of the experimental brittle crust and mantle; the maximum thinning of the lithosphere and brittle crust appears at the same place.

72

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

Fig. 8. Layer-parallel shear in the ductile crust related to the presence of a weak zone in the ductile mantle. Upper diagram: The upper crust being more resistant than the underlying lower crust and stretching being stronger in the weak zone than in the surrounding mantle, a horizontal shear top to the weak zone results in the ductile crust. Lower diagram: Reversal of layer-parallel sense of shear around a mantle weak zone trending oblique to the stretching direction. Bc: brittle crust, Dc: ductile crust, Dm: ductile mantle.

3.4. Mantle weakness oblique to the direction of extension with orthogonal weakness in the ductile crust; α = 90° and β = 30° 3.4.1. Model 4 (Fig. 6) The surface fault pattern is characterised by a central zone of deformation consisting of closely spaced faultbounded blocks (Fig. 6b). The cross-sections show two main domains of faulting located close to the extremities of the ductile crust anomaly. The conjugate fault pattern is slightly asymmetric (Fig. 6c). Plotting the horizontal distance versus normalised thickness, two observations can be made. First, the response to thinning of the brittle crust and mantle is in an opposite sense. Second, the maximum thinning of the lithosphere and brittle crust appears at the same place (Fig. 6d). 3.5. Mantle weakness orthogonal to the direction of extension with oblique weakness in the ductile crust; α = 30° and β = 90° 3.5.1. Model-5 (Fig. 7) The surface fault pattern is characterised by one broad zone of faulting with a major fault situated above the right side of the ductile crust anomaly (Fig. 7b). The crosssection, line AB, shows that the two normal fault zones are located above the DCWZ boundaries (Fig. 7c). However,

it must be noted that the right fault zone is also located above the mantle anomaly, MWZ. So, the presence of the oblique DCWZ determines the development of a prominent extensional fault oblique to the extension direction, while the MWZ seems to play a rather limited role on the development of the surface fault pattern. Once again, the plot of horizontal distance versus normalised thickness displays mirror-type symmetry on the response to thinning of the experimental brittle crust and mantle. Again, the maximum thinning of the lithosphere and brittle crust appears at the same places (Fig. 7d). 4. Interpretation of deformation patterns in relation with the presence of weak zones Considering that all models were built with the same materials and deformed at the same rate and with the same bulk extension, their comparison provides interesting information on the effects of location and orientation of weak heterogeneities in the lithosphere. 4.1. Localised versus distributed deformation in the brittle crust The brittle crust deformation patterns vary as a function of the depth of the weak zone(s) as well as their obliquity with reference to the direction of extension.

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

73

Fig. 9. Bulk lithospheric thinning in relation to the position of the weak zone(s). The curves represent the final normalised thickness of the experimental lithosphere. The maximum bulk lithospheric thinning (indicated by numbers) is directly related to the position of the weak zone in the ductile crust.

Weak zones located in the lower crust exert a more pronounced influence on localising deformation in the brittle crust, whereas those located in the upper mantle tend to distribute deformation over a wider area. In all models, the mechanical effect of mantle weak zones is damped and/or converted by the ductile lower crust. Models 1 and 3 that have no lower crust weak zones differ only by the orientation of the mantle weak zone. When the weak zone is perpendicular to the stretching direction the deformation is partitioned in three zones of conjugate faults in the brittle layer. When the weak zone trends oblique to model boundaries, only one or two fault zones are observed with a strong asymmetry in crosssection. This effect directly depends on the location of the weak zone in a given section (Fig. 8). At the onset of deformation, when the upper brittle layer is not yet faulted, because the weak zone undergoes more intense stretching than the surrounding mantle the lower crustal layer undergoes a horizontal shear, whose top is directed toward the weak zone. Consequently, if the shear couple is strong enough, faulting occurs with parallel faults whose sense of shear is similar to the one of the underlying horizontal shear (Fig. 8a). In 3D, because the position of the weak zone changes across the model the sense of horizontal shear responds accordingly (Fig. 8b).

4.2. Lithospheric thinning In all models bulk lithosphere thinning is heterogeneous and in direct relation with the position of weak zones (Fig. 9). However, thinning is more distributed in models that have only one weak zone in the mantle (Fig. 9a and c). The presence of a weak zone in the lower crust always contributes to localisation of thinning (Fig. 9b, d and e). The analysis of individual layer thinning shows that the location of thinning localisation also corresponds to the strain localisation in the brittle crust and smaller thinning in the mantle (Figs. 3–7). This indicates that lateral variations in the brittle crust thinning are partially compensated by mantle thinning (i.e. maximum brittle crust thinning corresponds to minimum mantle thinning, and vice versa) such as to minimise the lateral gradients of body forces. It is therefore important to recognise the role of weak zones located in the lower crust as they can significantly obscure the role of mantle weak zones. Moreover, when they are not superposed vertically to the mantle weak zones (Fig. 9d and e), they contribute strongly not only to locating the zone of maximum lithosphere thinning but also to controlling the largescale thinning asymmetry.

74

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

4.3. Conjugate versus parallel fault patterns In most models, cross-sections display fault patterns involving conjugate faults that define either grabens (Figs. 4, 6 and 7)) or large zones of narrow spaced faulting (Fig. 3). Only one model displays a dominant system of parallel faults (Fig. 5). Previous experiments illustrated that conjugate normal faults and graben structures are favoured by low strain rates and weak brittle–ductile coupling. On the contrary parallel faulting and block tilting is the result of strong brittle–ductile coupling (Allemand, 1990; Brun, 1999). Additionally, a conjugate fault pattern could also indicate that finite strain is of pure shear type whereas parallel fault systems result from a non-coaxial progressive deformation, due to a combination of pure and simple shear (Brun, 1999). In the present series of experiments, where the rheology of the materials and the bulk rate of extension was kept the same for all models, we cannot invoke variations in strain rate and therefore in the degree of brittle–ductile coupling. It is therefore most likely that the variations in the asymmetry of fault patterns directly reflect variations in strain regimes.

The sense of layer-parallel shear in the ductile crust depends directly on the position of the weak zones (Fig. 8). However, it must also be realised that the shear intensity is itself dependent upon the same parameter. As the sense of shear reverses across the model the intensity of shear decreases with the distance to the axis of shear sense reversal. Conjugate fault patterns would be favoured in areas where shear intensity is low while parallel fault systems will be favoured by shear intensity increase. 4.4. Oblique faulting Among the models that contain an oblique weak zone, oblique faults develop in the brittle crust only when the weak zone is located in the ductile crust (Fig. 10a). In the two other models where the oblique weak zone is located in the mantle the fault pattern is regularly oriented and does not show significant variation in the fault trend (Fig. 10b and c). Oblique faults located above a weak zone boundary (Fig. 10a) accommodate large displacement and control the surface width of the main grabens. In mechanical terms, weak zones located in the mantle have a “far field” effect as they contribute to distribute the

Fig. 10. Summary of the surface fault pattern in relation to the position and orientation of the weak zones. (a) Faults oblique to the direction of extension develop in the brittle crust only when the oblique weak zone is located in the ductile crust. Oblique faults located above a weak zone boundary accommodate large displacement and control the surface width of the main grabens. On the contrary, when the oblique weak zone occurs in the mantle, the faults trend orthogonal to the direction of extension and no significant oblique faults develop (b, c). Arrows indicate the direction of extension, while white lines indicate the main faults.

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

deformation in the brittle crust whereas those located in the ductile crust have more localising effects. 4.5. Ductile doming Below zones of intense brittle crustal thinning the ductile layers are uplifted into smooth domes. In the present experiments the amount of stretching has not been large enough to allow their complete exhumation. But obviously, stronger extension would bring not only the ductile crust but also the ductile mantle up to the surface. In other words, extension would not ultimately produce core complexes in the classical sense – i.e., made of continental crust materials – but would mostly exhume the lithospheric mantle like in passive margin formation and continental break up (Brun and Beslier, 1996). 5. Discussion 5.1. Comparison with the necking of a strong and cold lithosphere The mechanical behaviour of cold lithospheres strongly depends on the presence of a high strength sub-Moho mantle that plays a major role in the process of lithospheric-scale necking as this layer must undergo necking and boudinage to allow further extension of multiple lithospheric layers (Buck, 1991; Brun and Beslier, 1996; Nagel and Buck, 2004; Ziegler and Cloetingh, 2004). The necking and rupture of the high strength mantle provides a localisation at lithosphere scale. If weak zones – e.g., magmas – are present in the upper part of the lithospheric mantle the high strength sub-Moho mantle is interrupted. As an example, such a situation can take place above mantle hot spots. Extension is therefore localised from the earlier stages of lithospheric extension, leading to rapid continental break-up, a process that is directly relevant to the formation of volcanic passive margins (Callot et al., 2001, 2002). Other types of inherited structures, like ancient suture zones or large-scale strike–slip shear zones (which also exist in the upper part of the lithospheric mantle) will also play an important role in lithospheric extension, as illustrated in many natural examples. The break-up of Gondwana occurred along old mobile belts (Dunbar and Sawyer, 1989). Rifting in the Central Atlantic (between Africa and North America) followed the trend of the Permo-Carboniferous Alleghanian–Hercynian orogenic belt (Vogt and Tucholke, 1989); in the southern part of the North Atlantic, rifting and oceanic opening followed approximately the

75

Variscan orogeny (Manatschal and Bernoulli, 1999). Further north between Norway and East Greenland, extension of the lithosphere and continental break-up was almost completely confined to the Palaeozoic mobile zones between old cratonic blocks (e.g., Hill, 1991; Skogseid and Eldholm, 1988; Knott et al., 1993). The rifting of Antarctica from Africa was largely confined to Late Proterozoic mobile belts (Hill et al., 1991). In such tectonic situations the direction of extension is not necessarily orthogonal to the inherited structures leading to so-called oblique rifting where the normal faults do not initiate at right angle to the direction of extension (Tron and Brun 1991). In summary, the extension of a strong and cold lithosphere may lead narrow continental rifts to become passive margins. This process is characterised by a strong strain localisation at lithosphere scale – i.e., necking instability – and a high sensitivity to inherited zones of weakness, in particular those interrupting the high strength sub-Moho mantle. The series of experiments presented here illustrates that the extension of hot lithosphere, without a high strength sub-Moho mantle, produces specific deformation patterns that significantly differ from those that result from the necking of a strong and cold lithosphere. The models do not exhibit necking-type behaviour despite the presence of weak zones in either the lithospheric mantle or the ductile crust. The only model that could possibly be interpreted in terms of lithosphere necking is model-2 that contains two weak anomalies in the mantle and ductile crust (Fig. 4). However, considering that the amount of extension is higher than 50%, the observed effect of necking remains moderate and cannot be compared with what occurs in volcanic passive margins (e.g. Callot et al., 2001). All other models display almost no necking effects. Models containing an oblique weak zone do not show significant effects of oblique rifting. Only the model in which the weak zone is located in the ductile crust shows some oblique normal faults in the brittle crust but, even in this model, faults orthogonal to the direction of extension dominate. In summary, the extension of such soft and hot lithospheres appears unable to develop a significant necking instability, despite the presence of weak zones, and rather leads to a wide rifting-type of deformation. The experiments illustrate that such soft lithospheres remain rather insensitive to inherited zones of weakness. 5.2. Comparison with the extension of hot and thick lithosphere The extension of hot and thick lithospheres, corresponding to domains that are thermally relaxed

76

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

after continental collision (e.g. England and Thompson, 1986), is characterised by wide rifting and core complex modes of deformation (Wernicke, 1990; Buck, 1991; Brun, 1999; Bertotti et al., 2000; Wijns et al., 2005) leading to large scale extensional systems such as the Basin and Range of western US (Sonder et al., 1987; Wernicke, 1992; Snow and Wernicke, 2000). Dynamic modelling (e.g. Block and Royden, 1990; Wdowinski and Axen, 1992; Tirel et al., 2004, 2006) shows that the structure and development of core complexes requires a ductile crust viscosity low enough to flow rapidly, allowing the surface and Moho to remain relatively flat during dome rise and continuing extension. The present experiments also consider a hot lithosphere but non-previously thickened. Consequently, the ductile crust is not thick enough to allow the easy lateral flow that could promote the development of core complexes, keeping the Moho flat. On the contrary, the experiments show that below the zones where the brittle crust was strongly stretched the Moho is bent upwards in broad ductile domes. At higher values of stretching, these domes would reach the surface allowing the mantle to exhume. Such an exhumation of lithospheric mantle does not happen in wide domains of extension like the Basin and Range where, after 40 Ma of extension, the crust still has a thickness of around 30 Km. This emphasises the need to recognise the difference between thickened and non-thickened lithosphere. Whereas both types have in common a wholly ductile mantle, and therefore can be characterised by wide rifting when submitted to extension, they differ in their source of those deep rocks, either from the crust or mantle, that can be exhumed in zones of intense brittle crust stretching. 5.3. Applicability to back-arc-type tectonic environments The physical conditions simulated in our models fit those commonly observed in back-arc-type tectonic environments (Currie et al., 2004, Hyndman et al., 2005; Cloetingh et al., 2005), namely, high surface heat flow – i.e., at least 70–75 mW/m2 – and 30–45 km thick crust. From this point of view the deformation features of our models are interesting to compare with Mediterranean back-arc domains, in particular the Aegean and the Tyrrhenian, that have undergone large-scale extension related to more than 500 km of trench retreat (Jolivet and Faccenna, 2000). In the Aegean, the first phase of the core complex stage of extension, during Oligocene– Miocene times, brought the crust that was previously thickened during Cretaceous–Eocene times to a mean thickness of 30 km. Then, during Pliocene–Quaternary times, grabens developed at the scale of the whole

domain, in a still hot environment as demonstrated by high surface heat flow (Jongsma, 1974; Göktükler et al., 2003). The models presented here have a strong potential to explain the development of these late Tertiary grabens, in particular related to deep seated weak zones inherited from previous tectonic events. It is especially interesting to note that grabens are concentrated in and around the previously exhumed metamorphic core complexes (Rhodope: Brun and Sokoutis, 2004, Cyclades: Gautier and Brun, 1994, Menderes: Bozkurt, 2003). This is more likely related to the presence below the core complexes of a ductile crust hotter and weaker than in surrounding areas. This is in agreement with still very high present day heat flow values, reaching magnitudes higher than 100 mW/m2 in the Rhodope (Kolios et al., 2005) and in the Menderes (Göktükler et al., 2003). In the Tyrrhenian Sea, oceanic rifting occurred as the end product of back arc extension related to subduction rollback (e.g. Faccenna et al., 1996). In this region, that previously underwent extension with mid-crustal detachment faults (Jolivet et al., 1998), it is especially significant, as predicted by our models, that extreme thinning of continental crust lead to mantle exhumation in the Vavilov basin (Kastens et al., 1987; Bonatti et al., 1990). 6. Conclusions The current series of experimental models was designed to explore the consequences of weak heterogeneities in the extension of a soft and hot lithosphere with no significant previous crustal thickening. Such conditions should in particular apply to backarc type extensional systems. The experimental tests examined the effects of i) the location of heterogeneities in the ductile crust and/or in the lithospheric mantle, and ii) the orientation of weaknesses which may be either perpendicular or oblique to the direction of extension. Model results suggest the following main conclusions: 1) The observed deformation patterns are all relevant to the so-called “wide rifting” mode of extension. 2) Weak zones located in the ductile crust exert a more pronounced influence on localisation of deformation in brittle layer than those located in the lithospheric mantle. In particular, the former localise faulting in the brittle crust whereas the latter tend to distribute faulting over a wider area. This latter behaviour depends in turn upon the decoupling provided by the ductile crust. 3) Whatever the depth location and the orientation of weak zones, maximum thinning domains in the

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

brittle crust and ductile crust and mantle are in opposition. 4) Localised thinning in the brittle crust is accompanied by ductile doming of both lower crust and mantle. It may be locally amplified by strain localisation above the rheological boundaries of ductile crust weak zones, becoming a potential locus for triggering mantle exhumation. Such patterns of wide rift-type extension are mainly dependent on the lack of a high strength sub-Moho mantle. Lateral differences in brittle crust thinning are accommodated by lateral flow in the ductile crust and mantle. This contrasts with “cold and strong” lithospheres whose high strength lithospheric mantle triggers a necking instability at lithospherescale. This also differs from the extension of hot and soft lithospheres that result from the thermal relaxation of a previously thickened crust whose ductile crust is thick enough to give rise to metamorphic core complexes. The mechanics of some characteristic structures of back-arc-type extension, like those observed in the Mediterranean, could be examined in the light of our experimental results. This would in particular concern the development of narrow spaced grabens above previously exhumed metamorphic core complexes in the Aegean, and mantle exhumation in the Tyrrhenian. Acknowledgements This manuscript greatly benefited from careful comments provided by the Journal Reviewers F.O. Marques and N. Sleep. F. Gueydan and R. Stephenson are thanked for pertinent and constructive criticism on the manuscript. D. Sokoutis, S. Cloetingh and T. Mauduit kindly acknowledge research grants by ISES and NWO. Research partly supported by CNR Funds (RSTL no. 105 “Evoluzione della parte Nord del rift Afroarabico e distribuzione regionale delle georisorse”) assigned to M. Bonini. The Authors thank Prof. C.J. Talbot for the use of the centrifuge apparatus at the Hans Ramberg Tectonic Laboratory of Uppsala University. References Alfonso, J.C., Ranalli, G., 2004. Crustal and mantle strengths in continental lithosphere: is the jelly sandwich model obsolete? Tectonophysics 394, 221–232. Allemand, P., 1990. Approche expérimentale de la mécanique du rifting continental. Mémoires de Géosciences Rennes 38. Babuška, V., Plomerová, J., Vecsey, L., Granet, M., Achauer, U., 2002. Seismic anisotropy of the French Massif Central and predisposition

77

of Cenozoic rifting and volcanism by Variscan suture hidden in the mantle lithosphere. Tectonics 21. doi:10.1029/2001TC901035. Bertotti, G., Podladchikov, Y., Daehler, A., 2000. Dynamic link between the level of ductile crustal flow and style of normal faulting of brittle crust. Tectonophysics 320, 195–218. Block, L., Royden, L.H., 1990. Core complex geometries and regional scale flow in the lower crust. Tectonics 9, 557–567. Bonatti, E., Seyler, M., Channell, J., Girardeau, J., Mascle, G., 1990. Peridotites drilled from the Tyrrhenian Sea, ODP Leg 107. In: Kastens, K.A., Mascle, J., et al. (Eds.), Proceedings of the ODP, Scientific Results, vol. 107, pp. 37–48. Bonini, M., Sokoutis, D., Mulugeta, G., Boccaletti, M., Corti, G., Innocenti, F., Manetti, P., Mazzarini, F., 2001. Dynamics of magma emplacement in centrifuge models of continental extension with implications for flank volcanism. Tectonics 20, 1053–1065. Bozkurt, E., 2003. Origin of NE trending basins in western Turkey. Geodynamica Acta 16, 61–81. Braun, J., Beaumont, C., 1987. Styles of continental rifting: results from dynamics models of lithospheric extension zones. In: Beaumont, C., Tankard, A.J. (Eds.), Sedimentary basins and basins-forming mechanisms. Canadian Society Petroleum Geology, vol. 12, pp. 241–258. Braun, J., Beaumont, C., 1989. Dynamical models of the role of crustal shear zones in asymmetric continental extension. Earth and Planetary Science Letters 93, 405–423. Buck, W.R., 1991. Modes of continental lithospheric extension. Journal of Geophysical Research 96, 20161–20178. Brun, J.-P., 1999. Narrow rifts versus wide rifts: inferences for the mechanics of rifting from laboratory experiments. Philosophical Transaction of the Royal Society London 357, 695–710. Brun, J.-P., Beslier, M.O., 1996. Mantle exhumation at passive margins. Earth and Planetary Science Letters 142, 161–173. Brun, J.-P., Sokoutis, D., 2004. North Aegean extension: from the Rhodope core complex to Neogene basins. Proceedings of the 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, 14–20 April 2004, vol. 1, pp. 49–52. Brun, J.-P., Tron, V., 1993. Development of the North Viking Graben: inferences from laboratory modelling. Sedimentary Geology 86, 31–51. Brun, J.-P., Wenzel, F., and the Ecors-Dekorp team, 1991. Crustal-scale structure of the Southern Rhinegraben from Ecors-Dekorp seismic reflection data. Geology 19, 758–762. Brun, J.-P., Sokoutis, D., Van Den Driessche, J., 1994. Analogue modelling of detachment fault systems and core complexes. Geology 22, 319–322. Callot, J.-P., Grigné, C., Geoffroy, L., Brun, J.-P., 2001. Development of volcanic passive margins: Two-dimensional laboratory models. Tectonics 20, 148–159. Callot, J.-P., Geoffroy, L., Brun, J.-P., 2002. Development of volcanic passive margins: Three-dimensional laboratory models. Tectonics 21, 1052. doi:10.1029/2001TC901019. Cheadle, M.J., McGeary, S., Warner, M.R., Mathews, D.H., 1987. Extensional structures on the western UK continental shelf: a review of evidence from deep seismic profiling. In: Coward, et al. (Ed.), Continental Extensional Tectonics. Geol Soc. Spec. Publ., vol. 28, pp. 445–465. Cloetingh, S., Ziegler, P.A., Beekman, F., Andriessen, P.A.M., Matenco, L., Bada, G., Garcia-Castellanos, D., Hardebol, N., Dezes, P., Sokoutis, D., 2005. Lithospheric memory, state of stress and rheology: neotectonic controls on Europe's intraplate continental topography. Quaternary Science Reviews 24, 241–304. Collins, W.J., 2002. Hot orogens, tectonic switching, and creation of continental crust. Geology 30, 535–538.

78

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79

Corti, G., 2004. Centrifuge modelling of the influence of crustal fabrics on the development of transfer zones: insights into the mechanics of continental rifting architecture. Tectonophysics 384, 191–208. Corti, G., 2005. Dynamics of periodic instabilities during stretching of the continental lithosphere: view from centrifuge models and comparison with natural examples. Tectonics 24, TC2008. doi:10.1029/2004TC001739. Corti, G., Bonini, M., Conticelli, S., Innocenti, F., Manetti, P., Sokoutis, D., 2003a. Analogue modelling of continental extension: a review focused on the relations between the patterns of deformation and the presence of magma. Earth-Science Reviews 63, 169–247. Corti, G., Van Wijk, J., Bonini, M., Sokoutis, D., Cloetingh, S., Innocenti, F., Manetti, P., 2003b. Transition from continental break-up to punctiform seafloor spreading: how fast, symmetric and magmatic. Geophysical Research Letters 30, 1604. doi:10.1029/2003GL017374. Corti, G., Bonini, M., Sokoutis, D., Innocenti, F., Manetti, P., Cloetingh, S., Mulugeta, G., 2004. Continental rift architecture and patterns of magma migration: a dynamic analysis based on centrifuge models. Tectonics 23, TC2012. doi:10.1029/2003TC001561. Currie, C.A., Wang, K., Hyndman, R.D., He, J., 2004. The thermal effects of slab-driven mantle flow above a subducting plate: the Cascadia subduction zone and backarc. Earth and Planetary Science Letters 223, 35–48. doi:10.1016/j.epsl.2004.04.020. Daly, M.C., Chorowicz, J., Fairhead, J.D., 1989. Rift basin evolution in Africa: the influence of reactivated steep basement shear zones. In: Cooper, M.A., Williams, G.D. (Eds.), Inversion Tectonics. Geological Society Special Publications, vol. 44, pp. 309–334. Dunbar, J.A., Sawyer, D.S., 1988. Continental rifting at pre-existing lithospheric weaknesses. Nature 242, 565–571. Dunbar, J.A., Sawyer, D.S., 1989. How pre-existing weaknesses control the style of continental break-up. Journal of Geophysical Research 94, 7278–7292. England, P.C., Thompson, A., 1986. Some thermal and tectonic model for crustal melting in continental collision zones. In: Cowards, M.P., Ries, A.C. (Eds.), Collision Tectonics. Geological Society Special Publication, vol. 19, pp. 83–94. Faccenna, C., Davy, P., Brun, J.-P., Funiciello, R., Giardini, D., Mattei, M., Nalpas, T., 1996. The dynamics of back-arc extension: an experimental approach to the opening of the Tyrrhenian sea. Geophysical Journal International 126, 781–795. Faugere, E., Brun, J.-P., 1984. Modélisation expérimentale de la distension continentale. Comptes Rendus de l'Académie des Sciences Paris 299, 365–370. Gautier, P., Brun, J.-P., 1994. Ductile crust exhumation and extensional detachments in the Central Aegean (Cyclades and Evvia Islands). Tectonophysics 238, 399–424. Göktükler, G., Salk, M., Sari, C., 2003. Numerical modeling of the conductive heat transfer in Western Anatolia. Journal of the Balkan Geophysical Society 6, 1–15. Harry, D.L., Sawyer, D.S., 1992. A dynamic model of extension in the Baltimore Canyon Through region. Tectonics 8, 420–436. Hill, R.I., 1991. Starting plumes and continental break-up. Earth and Planetary Science Letters 104, 398–416. Hill, R.I., Campbell, I.H., Griffiths, R.W., 1991. Plume tectonics and the development of stable continental crust. Exploration Geophysics 22, 185–188. Hopper, J.R., Buck, R.W., 1996. The effect of ductile crust flow on continental extension and passive margin formation. Journal of Geophysical Research 101, 20175–20194. Hyndman, R.D., Currie, C.A., Mazzotti, S.P., 2005. Subduction zone backarcs, mobile belts, and orogenic heat. GSA Today 15 (2), 4–10.

Jolivet, L., Faccenna, C., 2000. Mediterranean extension and the Africa–Eurasia collision. Tectonics 19, 1095–1106. Jolivet, L., Faccenna, C., Goffe, B., Mattei, M., Rossetti, F., Brunet, C., Storti, F., Funiciello, R., Cadet, J.P., D'Agostino, N., 1998. Midcrustal shear zones in post-orogenic extension: Example from the northern Tyrrhenian Sea. Journal of Geophysical Research 103, 12,123–12,160. Jongsma, D., 1974. Heat flow in the Aegean Sea. Geophysical Journal of the Royal Astronomical Society 37, 337–346. Kastens, K.A., Mascle, J., Auroux, C., et al., 1987. Proceedings of the Ocean Drilling Program (Part A—Initial Reports, Sites 650–656), vol. 107. College Station, Texas. 772 pp. Klemperer, S., Mathews, D., 1987. Iapetus suture located beneath the North Sea by BIRPS deep seismic reflection profiling. Geology 15, 195–198. Knott, S.D., Burchell, M.T., Jolley, E.J., Fraser, A.J., 1993. Mesozoic to Cenozoic plate reconstructions of the North Atlantic and hydrocarbon plays of the Atlantic margins. In: Parker, J.R. (Ed.), Petroleum Geology of Northwest Europe. Proceedings of the 4th Conference. The Geological Society, London, pp. 953–974. Kolios, N., Koutsinos, S., Arvanitis, A., Karydakis, G., 2005. Geothermal situation in Northeastern Greece. Proceedings of the World Geothermal Congress (14 pp). Manatschal, G., Bernoulli, D., 1999. Architecture and tectonic evolution of non-volcanic margins: present-day Galicia and ancient Adria. Tectonics 18, 1099–1119. Mulugeta, G., 1988. Squeeze-box in the centrifuge. Tectonophysics 148, 323–335. Mulugeta, G., Ghebreab, W., 2001. Modeling heterogeneous stretching during episodic or steady rifting of the continental lithosphere. Geology 29, 895–898. Nagel, T.J., Buck, W.R., 2004. Symmetric alternative to asymmetric rifting models. Geology 32, 937–940. Ranalli, G., 2000. Rheology of the crust and its role in tectonic reactivation. Journal of Geodynamics 30, 3–15. Ramberg, H., 1981. Gravity, deformation and the Earth's crust. Academic Press, London. (452 pp). Skogseid, J., Eldholm, O., 1988. Early Cainozoic evolution of the Norwegian volcanic passive margin and the formation of marginal highs. In: Morton, A.C., Parson, L.M. (Eds.), Early Tertiary volcanism and the opening of the NE Atlantic. Geological Society London Special Publications, vol. 39, pp. 49–56. Sibson, R.H., 1995. A note on fault reactivation. Journal of Structural Geology 7, 751–754. Snow, K.J., Wernicke, B.P., 2000. Cenozoic tectonism in the central basin and range: magnitude, rate, and distribution of upper crustal strain. American Journal of Science 300, 659–719. Sonder, L.J., England, P.C., Wernicke, B.P., Christiansen, R.L., 1987. A physical model for Cenozoic extension of western North America. In: Coward, M.P., Dewey, J.F., Hancock, P.L. (Eds.), Continental extensional tectonics. Blackwell Scientific Publications, pp. 187–201. Tirel, C., Brun, J.-P., Burov, G., 2004. Thermo-mechanical modeling of extensional gneiss domes. In: Withney, D.L., Teyssier, C., Siddoway, C.S. (Eds.), Gneiss domes in orogeny. Geological Society of America Special Paper, vol. 380, pp. 67–78. Tirel, C., Brun, J.-P., Sokoutis, D., 2006. Extension of thickened and hot lithospheres, inferences from laboratory modelling. Tectonics 25, TC1005. doi:10.1029/2005TC001804. Tommasi, A., Vauchez, A., 2001. Continental rifting parallel to ancient collisional belts: an effect of the mechanical anisotropy of the lithospheric mantle. Earth and Planetary Science Letters 185, 199–210.

D. Sokoutis et al. / Tectonophysics 444 (2007) 63–79 Tron, V., Brun, J.-P., 1991. Experiments on oblique rifting in brittleductile systems. Tectonophysics 188, 71–84. Van Wijk, J.W., 2005. Role of weak zone orientation in continental lithosphere extension. Geophysical Research Letters 32, L02303. doi:10.1029/2004GL022192. Vauchez, A., Tommasi, A., Barruol, G., 1998. Rheological heterogeneity, mechanical anisotropy and deformation of the continental lithosphere. Tectonophysics 296, 61–86. Vogt, P.R., Tucholke, B.E., 1989. North Atlantic Ocean Basin: aspects of geological structure and evolution. In: Bally, A.W., Palmer, A.R. (Eds.), The Geology of North America — an overview. Geological Society of America, Washington D.C., pp. 53–79. Wdowinski, S., Axen, G.J., 1992. Isostatic rebound due to tectonic denudation: A viscous flow model of a layered lithosphere. Tectonics 11, 303–315.

79

Wernicke, B., 1990. The fluid crustal layer and its implications for continental dynamics. In: Salisbury, M.H., Fountain, D.M. (Eds.), Exposed Cross-Sections of the Continental Crust. Springer, New York, pp. 509–544. Wernicke, B., 1992. Cenozoic extensional Tectonics of the U.S. Cordillera, in The Geology of North America. In: Burchfield, C., Lipman, P.W., Zoback, M.L. (Eds.), The Cordilleran Orogen: Conterminous U.S. Geol. Soc. of Am., Boulder, Colorado, vol. G-3, pp. 553–581. Wijns, C., Weinberg, R., Gessner, K., Moresi, L., 2005. Mode of crustal extension determined by rheological layering. Earth and Planetary Science Letters 236, 120–134. Ziegler, P.A., Cloetingh, S., 2004. Dynamic processes controlling evolution of rifted basins. Earth-Science Reviews 64, 1–50.