1. Introduction
Morphotectonic features such as offset terrace risers, alluvial fans and displaced river channels provide valuable information about shear sense and offset for strike-slip faults (e.g. Sieh & Jahus, Reference Sieh and Jahns1984; Keller & Pinter, Reference Keller and Pinter2002; Talebian & Jackson, Reference Talebian and Jackson2002; Walker & Jackson, Reference Walker and Jackson2002; Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007). Other studies (e.g. Holdsworth, Butler & Roberts, Reference Holdsworth, Butler and Roberts1997; Lacassin, Replumaz & Leloup, Reference Lacassin, Replumaz and Leloup1998) indicate that reactivation of faults is reflected in geomorphic features such as hairpin or bayonet structures.
Although slip-sense inversion along major strike-slip faults is well documented in other parts of the world (e.g. Lacassin, Replumaz & Leloup, Reference Lacassin, Replumaz and Leloup1998; Allen, Alsop & Zhemchuzhnikov, Reference Allen, Alsop and Zhemchuzhnikov2001; Maruyama & Lin, Reference Maruyama and Lin2004; Javadi et al. Reference Javadi, Foroutan, Estrabi Ashtiani, Angel Urbina, Saidi and Faridi2011), the phenomenon, its nature and timing remain undocumented on the Turkish–Iranian plateau and are therefore not considered in tectonic models for the region. In this paper we document the evidence for the kinematic reversal of the Doruneh Fault System (DFS) (Wellman, Reference Wellman1965; also known as the Great Kavir fault by Stocklin & Nabavi, Reference Stocklin and Nabavi1973), which presently indicates a left-lateral slip rate of about 2.4 ± 0.3 mm yr−1 (Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007), along the northern margin of the Central Iranian Microplate (CIM) (Fig. 1b). Furthermore, we constrain the kinematic history of the DFS in the northern CIM using an analysis of field-based fault kinematic data and study of offsets of the bedrock and Quaternary geomorphic features. Interpreting our new data in light of recent published geological slip-rates, we provide a new model for the kinematics of this region that is aimed at providing a more clear insight into the geodynamics of the Arabia–Eurasia collision zone in northern Central Iran.
(a) Tectonic setting of Iran in the Middle East and presentation of major convergence vectors of the region. (b) Main sedimentary-structural zones of Iran (modified from Aghanabati, Reference Aghanabati2004). Major faults discussed in the text are shown. White and black arrows from Sella, Dixon & Mao (Reference Sella, Dixon and Mao2002) and Vernant et al. (Reference Vernant, Nilforoushan, Hatzfeld, Abbassi, Vigny, Masson, Nankali, Martinod, Ashtiani, Bayer, Tavakoli and Chéry2004), respectively. DFS – Doruneh Fault System, MRZF – Main Zagros Reverse Fault, HZF – High Zagros Fault, MFF – Mountain Frontal Fault, ZFF – Zagros Foredeep Fault.
The DFS extends ~900 km between the Anarak area in Central Iran and Herat area in western Afghanistan (Figs 1b, 2a). Furthermore, the DFS can be considered the western continuation of the complicated right-lateral Herat Fault System, which itself extends for >1000 km between the Hindu Kush in the western syntaxis of the Himalaya and Herat in western Afghanistan. These two faults (Herat and DFS) are likely the modified, disarticulated remnants of a once continuous fault system that extended 2000 km through central Asia, probably in the form of an Early Mesozoic suture (Tapponnier et al. Reference Tapponier, Mattauer, Proust and Cassaigneau1981). The present Herat and DFS transverse structures were reactivated in the Cenozoic and have opposing kinematics (the DFS is sinistral and the Herat, dextral).
(a) Combination of Landsat satellite image, SRTM DEM and geological maps showing structure and geology of the northern CIM and Afghanistan region. Eastern, central and western segment of the DFS are shown by yellow, white and red lines, respectively. (b) The eastern segment of the DFS consists of multiple oblique-slip and reverse fault splays that cut through pre-Neogene deposits. (c) The central segment of the DFS passes through dominantly Quaternary deposits to the south and Paleocene–Eocene volcanic and plutonic rocks to the north. The Doruneh magmatic-arc and ophiolitic mélange rocks are well exposed along the central fault segment. (d) The western segment of the DFS runs across the Neogene deposits of the Great Kavir Desert.
The CIM is an important independent continental block in the Arabia–Eurasia collision zone and controls the distribution of strain in the eastern portion of the Turkish–Iranian plateau. The CIM is surrounded by active strike-slip faults and is internally segmented by a network of roughly N–S-oriented dextral strike-slip faults. The major active N–S right-lateral faults of the CIM (Fig. 1b) terminate near the DFS at about 34°N, and accommodate ~15 mm yr−1 N–S right-lateral shear (Vernant et al. Reference Vernant, Nilforoushan, Hatzfeld, Abbassi, Vigny, Masson, Nankali, Martinod, Ashtiani, Bayer, Tavakoli and Chéry2004). This right-lateral shear is suggested by some (Jackson & McKenzie, Reference Jackson and McKenzie1984; Jackson, Haines & Holt, Reference Jackson, Haines and Holt1995; Walker & Jackson, Reference Walker and Jackson2004; Allen et al. Reference Allen, Blanc, Walker, Jackson, Talebian and Ghassemi2006; Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007) to be accommodated by clockwise block rotation on the E–W faults of the DFS and adjacent left-lateral Dasht-e-Bayaz fault. In contrast, Farbod et al. (Reference Farbod, Bellier, Shabanian and Abbassi2011) challenged the rotation model, citing the kinematics of the Doruneh fault segments as inconsistent with a clockwise rotation of this fault.
2. Tectonic setting
The Turkish–Iranian Plateau is an actively deforming zone that has resulted from the collision between the Arabian and Eurasian plates within the greater Alpine–Himalayan collisional belt (Fig. 1b). The collision resulted in major crustal thickening, folding, thrusting and strike-slip faulting. The Iranian portion of the plateau is composed of Cimmerian continental fragments that separated from the northern edge of Gondwanaland, were transported, and accreted to the Eurasian margin during the closure of the Palaeo-Tethys in Late Triassic – Early Jurassic time (Berberian & King, Reference Berberian and King1981). Final amalgamation of the Cimmerian blocks occurred after the closure of the Neo-Tethys Ocean, and the onset of the collision between the Eurasian and Arabian plates in Oligo-Miocene times. After this event, the geodynamics of Iran evolved mainly in response to continuing Arabia–Eurasia convergence (Jackson & McKenzie, Reference Jackson and McKenzie1984, Reference Jackson and McKenzie1988; McQuarrie et al. Reference McQuarrie, Stock, Verdel and Wernicke2003; Guest, Guest & Axen, Reference Guest, Guest and Axen2007).
Palaeomagnetic data has been used to reconstruct the post-Cimmerian history of different blocks of Central Iran (Soffel & Förster, Reference Soffel and Förster1980, Reference Soffel and Förster1984; Schmidt & Soffel, Reference Schmidt and Soffel1984; Soffel et al. Reference Soffel, Davoudzadeh, Rolf and Schmidt1996). Based on these data, slip on pre-existing strike-slip faults in Central Iran accommodates vertical axis rotations of the fault-bounded crustal blocks (e.g. Mattei et al. Reference Mattei, Cifelli, Muttoni, Zanchi, Berra, Mossavvari and Eshraghi2012).
A key component of the eastern Turkish–Iranian plateau is the CIM, which is an active terrane located between the Kopeh-Dagh, Binalud and Alborz mountains in the north (also called the ‘Central Domain’ (CD); Aghanabati, Reference Aghanabati2004); the Zagros Mountains in the west and southwest; and the Helmand Microplate to the east (Fig. 1b). The CIM is bounded by the DFS to the north, the Nehbandan fault zone to the east, and the Dehshir fault to the west and southwest (Fig. 1b). Internally, the CIM is composed of the smaller Lut, Tabas, Posht-e-Badam and Yazd blocks, all of which are separated by linear to curvilinear N–S right-lateral strike-slip faults (Berberian & King, Reference Berberian and King1981; Tirrul et al. Reference Tirrul, Bell, Griffis and Camp1983; Soffel et al. Reference Soffel, Davoudzadeh, Rolf and Schmidt1996; Aghanabati, Reference Aghanabati2004) (Fig. 1b).
The DFS may be considered an intracontinental transform fault between the abovementioned CIM blocks to the south and the CD to the north. The CD and CIM are parts of Central Iran that is bounded by the Palaeo-Tethys suture zone in the north and Neo-Tethys suture zone in the south. Owing to early Cimmerian movements, the CD and CIM separated from the Gondwana plate and migrated from southern Gondwana palaeolatitudes in Early Permian time to subequatorial palaeolatitudes by Middle Permian – Early Triassic times as a consequence of the Neo-Tethys Ocean opening (Muttoni et al. Reference Muttoni, Gaetani, Kent, Sciunnach, Angiolini, Berra, Garzanti, Mattei and Zanchi2009a ,Reference Muttoni, Mattei, Balini, Zanchi, Gaetani, Berra, Brunet, Wilmsen and Granath b ).
Outcrops of Cretaceous – Lower Eocene ophiolite mélanges (southeast of Torbat-e-Heydarieh, north of Taknar fault, west of Doruneh Village and Anarak area) and Paleocene–Eocene flysch-type sediments and Eocene volcanic rocks of a magmatic arc setting along the DFS (Figs 1, 2b–d) may be used to confirm this suggestion. The CIM blocks show N–S to NW–SE structural trends, whereas the CD in the north displays an ENE–WSW structural trend. This may reflect different tectonic and geological histories for these terranes.
Accordingly, generation of the DFS may be ascribed to an old suture zone that resulted from the closure of an undeveloped primitive oceanic basin to the north of the CIM in Early Eocene time (Schmidt & Soffel, Reference Schmidt and Soffel1984).
Zamani, Angelier & Zamani (Reference Zamani, Angelier and Zamani2008) suggested that a change in the structural trends of northeastern Iran, including the DFS, from ESE–WNW in the east to the ENE–WSW in the west is related to the existence of a dual stress regime in the western part of northeastern Iran and the DFS in comparison to the single stress regime in the eastern and middle part of the region. They suggested NNE–SSW and N–S major shortening trends in the western Kopeh-Dagh, eastern Alborz and northwest of the DFS, with ENE–WSW structural trends and a N–S shortening trend for the eastern and central Kopeh-Dagh, Binalud and northeast of the DFS, with a ESE–WNW structural trend that is related to the strain partitioning of this part of the fault (Zamani, Angelier & Zamani, Reference Zamani, Angelier and Zamani2008).
Shabanian et al. (Reference Shabanian, Bellier, Abbassi, Siame and Farbod2010), on the other hand, believed that no strain and/or stress partitioning or systematic block rotations occurred in the Kopeh-Dagh and Allah-Dagh-Binalud deformation domains. They suggested that the modern state of stress, deduced from the kinematic analysis of the youngest faults, shows two distinct strike-slip and compressional tectonic regimes with a regional mean N030 ± 15°E trending horizontal maximum stress axis (σ1), which is adapted to a N023 ± 5°E trending σ1 resulting from the inversion analysis of earthquake focal mechanisms for the region. Moreover the palaeostress state is characterized by a regional transpressional tectonic regime with a mean σ1 trending N140 ± 10°E. They stated that the change from the palaeostress to modern stress states has occurred through an intermediate stress field characterized by a N-trending mean regional σ1.
3. Segments of the DFS
We have distinguished three major segments along the DFS based on geometry and morphology of the fault zone. The NW–SE-striking eastern segment of the DFS extends for 300 km from western Afghanistan to the south of Torbat-e-Heydarieh (Fig. 2a). This eastern segment branches (Fig. 2b) into sub-parallel fault strands exhibiting kinematic indicators consistent with thrust and left-lateral reverse oblique motion. The eastern splays of the DFS cut through folded Oligocene sandstone, conglomerate, siltstone and ophiolitic mélange and Oligocene–Miocene limestone (Fig. 2b).
The E–W-striking central segment of the fault extends ~130 km along the southern flank of the Kuh-Sorkh mountains from Torbat-e-Heydarieh in the east to Anabad in the west (Fig. 2c). This central segment of the DFS consists of sub-parallel, overlapping strike-slip faults exhibiting a reverse component and cuts Upper Neogene – Quaternary deposits along most of its length. Common geomorphic features observed along the central fault segment include left-lateral displacement of Quaternary drainages, alluvial fans and terraces, sag-ponds, pressure ridges, shutter ridges and well-preserved fault scarps in Quaternary deposits. At step-overs between fault strands, active folds and pull-apart basins are developed. The 110 km long, left-lateral Taknar fault lies to the north of the Kuh-Sorkh Mountains, and is considered here to be a branch of the central fault segment (Fig. 2a, c).
The NE–SW-striking western segment of the DFS extends 460 km across the Great Kavir desert (Central Iranian Desert) from Anabad village in the east to the Anarak area in the west (Fig. 2d). The western fault segment is left-lateral and consists of a network of sub-parallel branching fault strands. In the pre-Neogene rocks, bounded between the fault branches, en échelon NE–SW folds are developed that reflect right-lateral motion across the branches (Fig. 7a). An example is present to the north of Doruneh Village where the DFS includes several thrusts, right-lateral fault segments and en échelon folds that strike NE–SW (Fig. 7a).
The western segment of the DFS transforms westward from two strands near Doruneh Village to multiple strands across the Great Kavir desert, and cuts different lithologic units. In the Doruneh area, the northern branch of this segment juxtaposes Upper Cretaceous – Lower Eocene ophiolitic mélanges to the north with Eocene–Oligocene conglomerate, sandstone, marl and tuff to the south. The southern branch juxtaposes Eocene–Oligocene units to the north with the Quaternary alluvial deposits to the south. Towards the southwest, in the Great Kavir desert, the DFS cuts low-lying Neogene deposits of alluvial sediments, clay flats, marly salt diapirs and gypsiferous sandstones of the Middle to Upper Miocene Upper Red Formation (Fig. 2d). The western termination of the DFS is characterized by a multitude of steeply SE-dipping left-lateral reverse fault strands. These strands cut across sandstone, siltstone and conglomerate of Middle to Upper Eocene age. The fault zone bends southwards terminating in Cretaceous – Lower Eocene ophiolites and Palaeogene carbonates, clastic rocks and volcanic rocks as well as the Anarak metamorphic complex that represents an exhumed arc-trench system developed during the Cimmerian orogeny in the interior of the CIM (Zanchi et al. Reference Zanchi, Zancheta, Garzanti, Balini, Berra, Mattei, Muttoni, Brunet, Wilmsen and Granath2009) (Fig. 2d).
4. Kinematic history of the DFS based on structural analysis, Quaternary geomorphology and bedrock displacements
4.a. Fault kinematics and structural observations along the DFS segments
The kinematics of the DFS are controlled by various interplaying factors such as the present-day regional stress state, structural complexities and fault geometry, and kinematic interactions between this fault and other intersecting fault zones (Farbod et al. Reference Farbod, Bellier, Shabanian and Abbassi2011).
Our observations consist of measurements of orientation and sense-of-motion of striated fault planes at 44 sites along the DFS (Fig. 4). Special emphasis was placed on striations directly observed on the major mapped fault planes with mappable geological displacements. We have also accounted for striations observed on smaller faults paralleling a nearby major fault, the assumption being that in a given local stress field, parallel faults exhibit similar kinematics for a given tectonic phase. For each site we plot the trace of the faults and the plunge of striations, with sense-of-movement indicators (Fig. 5). For sites not located along mapped faults, or where the principal mapped fault plane is unexposed, we use striations on the various minor fault surfaces to estimate the orientation of the local stress tensor (e.g. Angelier, Reference Angelier1990). The Plio-Quaternary stress-state was evaluated using fault kinematic observations (mostly striations) from Quaternary or Plio-Quaternary conglomerates. At most sites along the major faults, we have generally found only one set of striations, which we assume to be related to the latest displacements.
The palaeo-stress field determination obtained from fault planes in Quaternary deposits is consistent with those obtained from fault planes in older rocks. This implies either that the stress state was stable through time or that the recent kinematic regime has locally completely overprinted previous regimes. In contrast, some sites exhibit two sets of striations on major fault planes (Fig. 3). In such cases it is assumed that the stress state has changed through time and an effort was made to identify cross-cutting relationships and establish the relative ages of the different kinematic indicators.
Examples of DFS fault planes showing two sets of striation representative of two possibly different stress fields acting on the fault plane. (a) and (c) show fault planes on Oligocene sandstone in the eastern segment, east of Taibad, and Eocene conglomerate in the western segment, northwest of Doruneh, respectively (sites 14 and 39 in Fig. 4a, c). (b) and (d) show striations and kinematics of the fault planes shown in (a) and (c). Insets show relative ages of striations based on their cross-cutting relationships, and sense of relative movement for the existing fault block.
Analysis of fault kinematic data from 25 sites along the eastern segment of the DFS indicates a recent stress field with the σ1 axis trending N70 ± 20°E (Fig. 5). This is consistent with the dominant left-lateral strike-slip and reverse faulting observed along the eastern segment of the DFS (Fig. 4a). A comparison of the general trend of the σ1 axis with the fault zone strike for the eastern segment indicates that the magnitudes of the principal stress components are similar for this segment (inset in Fig. 4a).
Structural maps for the three major segments of the DFS. The eastern, central and western segments are shown in (a), (b) and (c), respectively. Numbers refer to the sites where data for the structural analysis were gathered (see Fig. 5). Insets show the relationship between maximum compressional stress (σ1) and fault strike, and fault-parallel and fault-normal vector components of σ1.
At sites 13, 14 and 22, two sets of cross-cutting striations are observed, which we can use to discriminate between two stress regimes. For example, at site 14 younger striations related to the reverse dip-slip movement during stage 2 cut across the older right-lateral striations of stage 1 (Fig. 3a). For the stage 1 indicators of sites 13, 14 and 22, measured on pre-Pliocene–Quaternary rocks deformed by right-lateral strike-slip faulting, the inversion analysis shows palaeostress tensors with σ1 trending N15 ± 2°W (Figs 4a, 5).
(a) Stereoplots of field measurements and computed data inversion (using the method of Angelier, Reference Angelier1990) in the eastern segments of the DFS. Schmidt nets, lower hemisphere. Numbers refer to sites located in Figure 4. For key see part (c). (a) Stereoplots of field measurements and computed data inversion (using the method of Angelier, Reference Angelier1990) in the eastern and (b) central segments of the DFS. Schmidt nets, lower hemisphere. Numbers refer to sites located in Figure 4. For key see part (c). (b) Stereoplots of field measurements and computed data inversion (using the method of Angelier, Reference Angelier1990) in the central and (c) western segments of the DFS. Schmidt nets, lower hemisphere. Numbers refer to sites located in Figure 4.
In the central segment of the DFS, a N60 ± 25°E trend is calculated for the recent (stage 2) direction of maximum compressional stress (σ1) on the basis of data from 11 sites (Figs 4b, 5). Applying this compression direction to the approximately E–W-striking DFS central segment predicts left-lateral movement with a reverse dip-slip component. The reverse component of the central segment is more minor than that of the eastern segment, and the strike-slip component is more dominant than that predicted for the eastern segment (Fig. 3b, inset). Morphotectonic features, such as left-laterally displaced drainages and alluvial deposits and uplift of the northern block of the fault (Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007), are also consistent with the recent tectonic regime in the central segment of the DFS. Similarly, modern stress analyses by Farbod et al. (Reference Farbod, Bellier, Shabanian and Abbassi2011) indicated the same direction for σ1 (N45 ± 15°E) and left-lateral strike-slip movement on the central segment of the DFS. Although they suggested that the strain resulting from NE–SW compression in the central segment of the DFS is partitioned as pure strike-slip motion on the DFS and thrust faulting on Kuh-Sorkh Mountain, it seems that some of the shortening is taken up as left-reverse-oblique motion on the DFS, which exhibits 20 m fault scarps in Pleistocene deposits (for example, north of the Kashmar and Tigh Ahmad areas) as well as the formation of river terrace flights along the northern, hanging wall reaches of major cross-cutting rivers (such as the Shehs Taraz and Shadmehr rivers) and a general lack of terrace flights on the footwall reaches of the same rivers (see Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007).
Similar to the eastern segment, older striations on fault planes of the central segment indicate a right-lateral mechanism for the DFS. Evidence for stage 1 dextral kinematics is observed at site 37, which indicates a N20°W trend for σ1 (Figs 4b, 5). This result is consistent with calculated palaeostress tensors by Farbod et al. (Reference Farbod, Bellier, Shabanian and Abbassi2011), which yielded the same direction for σ1 (N150 ± 20°E) along the central segment of the DFS.
Since the western segment of the DFS cuts through friable low-lying Neogene deposits along most of its length (Fig. 2d), the fault planes were only observed and measured near its eastern and western terminations. Six sites studied on this segment indicate a N25 ± 20°E trend for the σ1 axis for the stage 2 stress field with the only exception being site 42, stage 2 (Figs 4c, 5). Applying a N25 ± 20°E maximum compressional stress to the NE–SW-striking western segment predicts a dominantly left-lateral strike-slip tectonic regime, contrasting with the two other segments of the DFS (inset in Fig. 4c). The presence of two striation sets indicates that this segment also experienced both the stage 1 and stage 2 deformation episodes (Fig. 4c). For example, at site 39 the older right-lateral slip striations are cross-cut by younger left-lateral slip ones (Fig. 3c).
4.b. The post-Pliocene left-lateral motion along the DFS
Geomorphological indicators along most of the DFS indicate recent left-lateral motion on this fault zone. Prominent geomorphological evidence is well expressed in the central segment, where the fault passes through the Quaternary alluvial deposits. This evidence is clearly visible on satellite images as well as in the field, and is described in detail by Fattahi et al. (Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007) and Farbod et al. (Reference Farbod, Bellier, Shabanian and Abbassi2011). A maximum displacement of about 850 m is suggested for the central segment of the DFS based on the left-lateral offset of the uplifted and abandoned Quch-Palang alluvial fan deposits (Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007) (Fig. 6a).
Comparison between recent left-lateral offsets and remnant ‘fossil’ right-lateral displacement on the DFS. (a) Uplifted and abandoned Quch-Palang alluvial fan deposits show a maximum left-lateral displacement of 850 m across the fault on the Quickbird satellite image (Google Earth). (b) Geological reconstruction (850 m) of the alluvial fan in (a). (c) A maximum of 800 m ‘fossil’ right-lateral displacement of Middle Miocene deposits along an anticline to the west of Doruneh Village which displays 800 m right-lateral offset. (d) Reconstruction of the anticline in (c). (e) ‘Z’ form drag folding marked by yellow dashed line, and smaller residual right-lateral displacement to the west of Doruneh Village. (f) Residual ‘fossil’ right-lateral displacement of the Miocene deposits across the DFS, measured at 110 m using satellite image.
Assuming a constant 2.4 ± 0.3 mm yr−1 slip rate on the central part of the DFS (Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007) provides a minimum age estimate for the Quch-Palang alluvial fan of 315–405 ka (Middle Pleistocene). This contrasts with a 51.4 ± 10.2 ka maximum age estimate for the upper part of this uplifted fan determined by Fattahi et al. (Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007), which would require a slip rate of ~16–18 mm yr−1 to account for the measured offset. This estimate appears to be far too high to be plausible considering that only a small fraction of the ~25 mm yr−1 of Arabia–Eurasia convergence is accommodated in NE Iran (e.g. Vernant et al. Reference Vernant, Nilforoushan, Hatzfeld, Abbassi, Vigny, Masson, Nankali, Martinod, Ashtiani, Bayer, Tavakoli and Chéry2004). We speculate therefore that Fattahi et al. (Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007) sampled and analysed younger deposits that mantle the Quch-Palang fan surface. Clearly, tighter geochronological constraints are required to resolve this contradiction.
4.c. Right-lateral motion along the DFS
4.c.1. Bedrock displacement
Most of the Quaternary alluvial fan deposits, drainages and river channels across the DFS exhibit left-lateral offsets; however, an earlier right-lateral displacement is observed where the fault zone cuts pre-Pliocene rock units (Fig. 6c, e, f). About 40 km to the west of Doruneh Village, the DFS is made up of several sub-parallel fault strands that cut across an alternation of red to grey marl, silty marl, sandstone and limy siltstone. This rock unit is of Middle Miocene age based on fossils assemblages (Ghaemi & Moussavi-Harami, Reference Ghaemi and Moussavi-Harami2007). An anticline formed in this unit along the southern branch of the fault zone is displaced about 800 m in a right-lateral sense (Fig. 6c). Assuming a maximum cumulative left-lateral displacement of about 850 m along the fault (Fig. 6a), a total post-Middle Miocene right-lateral displacement of about 1650 m is required. To the west, the DFS splays into three sub-parallel fault strands. The southern strand juxtaposes Quaternary alluvial deposits to the south with Miocene sedimentary rocks in the north, whereas the northern and central strand cut across Miocene bedrock (Fig. 6e). The rocks bounded between the northern and southern strand are folded and show Z-shaped dextral drag folds. Lastly, the central fault strand exhibits a ~110 m fossil right-lateral displacement (Fig. 6f).
4.c.2. Fault-related en échelon folds
To the north of Doruneh Village, several thrust faults as well as NE–SW-striking en échelon folds oriented parallel to the strike-slip faults are exposed (Fig. 7). Owing to the earlier NW–SE-trending right-lateral transpressional tectonic regime, several large-scale NE–SW-striking en échelon folds are developed in the Eocene–Oligocene conglomerate and sandstone bounded between the fault strands, north and west of Doruneh Village (Figs 7a–e, 8). The folds are clearly oblique to the fault traces and are interpreted here as structural evidence of the formerly right-lateral transpressional regime. Furthermore, NE–SW- and ENE–WSW-striking fold axial planes and thrust faults that merge with the DFS (Fig. 7a, e) are consistent with a NW–SE shortening direction and a dextral transpressional tectonic regime, consistent with the post-Oligocene right-lateral movement on the DFS (Fig. 7b).
(a) En échelon folds to the north and west of Doruneh Village. All of these folds show a general NE–SW trend. Inset is a simplified map, which shows the structures developed during the NW–SE right-lateral transpression in the Doruneh area. (b–e) Stereographic projections of poles of bedding planes in the folds b, c, d and e. Axis and axial plane of the folds b, c, d and e, respectively, are [2/251; N71E/89NW], [8/241; N60E/85NW], [11/039; N31E/79NW], [6/44; N44E/87NW] (See (a) for location of the folds).
En échelon fold b in Fig. 7a, developed in Eocene–Oligocene rocks north of Doruneh Village.
4.c.3. Fault-zone fabrics
In addition to the macro-scale sigmoidal S-shaped folds, meso-scale structures and fabrics along the DFS also provide evidence for an earlier right-lateral kinematic history for the fault zone. In some places, where the fault strands cut across Eocene clastic and pyroclastic rocks (Fig. 9a), semi-cohesive fault gouge exhibits cataclasitic S-C fabric (Fig. 9b, d). A complex history of motion is recorded across the fault zone, in which a later left-lateral displacement, documented by precisely measured fault striations, tension fractures, tool marks and lunate structures, is superimposed over earlier right-lateral S-C fabric kinematic indicators (Fig. 9c, e). This evidence is also observed at site 39 (Fig. 4c), which has older right-lateral strike-slip and more recent left-lateral mechanisms based on two different striation sets (Fig. 3c).
Field evidence for older right-lateral motion across the Doruneh fault zone, northwest of Doruneh Village. (a) Sub-parallel fault strands to the northwest of Doruneh Village. (b) S-form semi-cohesive S-C fabrics formed in the Eocene sandstone associated with fault planes of the middle strand. Length of the hammer is approximately 35 cm. (c) Stereographic projection of the S-C fabric, which indicates a right-lateral shear sense while the striation on the fault plane shows left-lateral movement. (d) Brittle S-C fabric formed in semi-cohesive cataclasite across the southern fault. Length of the pen is approximately 15 cm. (e) Stereographic projection of the S-C fabric and fault plane kinematic analysis, which show a right-lateral and left-lateral shear sense, respectively.
5. Taknar right-lateral imbricate fan
To the north of the central segment of the DFS, the Taknar fault juxtaposes Cambrian sandstones (Lalun Formation), Cretaceous limestones and ophiolite mélanges to the north with the Eocene clastic and pyroclastic rocks to the south (Fig. 10). Lindenberg & Jacobshagen (Reference Lindenberg and Jacobshagen1983) documented evidence for Palaeogene right-lateral slip on the Taknar fault. In this area, the Taknar fault manifests as a wide zone of cataclastic fault rocks, composed of slivers of Cretaceous limestone embedded in Eocene fault gouge (Fig. 10d). The recent left-lateral motion on the Taknar fault is reflected in fault-related structures and river offsets as well as its seismic activity (Figs 10a–c, 11a, b). The recent left-lateral slip of the Taknar fault is compatible with the fault orientation along the NE–SW-trending σ1 direction. However, S-C fabrics in the semi-cohesive cataclasitic rocks along the Taknar fault indicate a more complex kinematic history (Fig. 10d). The S-C fabrics and clasts with monoclinic symmetry in these fault rocks are evidence of former right-lateral motion on the Taknar fault.
(a) The Taknar fault in the contact between Cretaceous limestone (to the north) and Eocene clastic rocks (to the south). (b) Close-up view of the Taknar fault. (c) Fault surface striations, steps and lunate fractures, which indicate its recent left-lateral movement. (d) S-C fabrics showing older right-lateral kinematics acted on the fault rocks.
(a) The Taknar right-lateral imbricate fan in the north of the central segment of DFS. Inset shows the fan that formed following right-lateral motion on the Taknar fault as the master fault in pre-Quaternary time that caused the formation of second- and third-order faults (green thick and black thin lines, respectively) as well as displacement of rock units in this area. (b) Stereographic projection of fault planes along the Taknar master fault showing left-lateral mechanisms. (c) Stereographic projection of the second- and third-order fault surfaces, mechanisms of which were not distinguished. This set has not been reactivated after the final development of the duplex in the older, right-lateral regime. (d) Stereographic projection of the second- and third-order fault surfaces, mechanisms of which are recognizable after the reactivation. Most of them show a reverse mechanism with a left-lateral component or left-lateral strike-slip with a reverse component.
Right-lateral motion on the Taknar fault is also documented by a right-lateral imbricate fan exposed to the north of the master fault trace (Fig. 11a). This fan consists of an imbricated zone of many en échelon, fault-bounded and folded horses that are bound southward by the continuous, large-displacement Taknar fault zone. In the Taknar imbricate fan, second-order faults splay from the master fault (Taknar fault) and accommodate a greater displacement of geological units (Fig. 11a), while third-order faults splay from second-order ones. Strike-slip imbricate fans are observed as both active and ancient examples, and usually stop forming when displacement ceases on their parent major fault (Woodcock & Fischer, Reference Woodcock and Fischer1986). It appears that this has occurred to the Taknar imbricate fan, probably owing to inversion from right-lateral to left-lateral kinematics. There is kinematic evidence of left-lateral slip on the Taknar fault where it bounds the imbricate fan (Fig. 11b), but there is no evidence of recent activity on most of the second- and third-order faults (Fig. 11c). However, some of these higher order faults do exhibit reverse and left-lateral strike-slip kinematics, indicating some reactivation (Fig. 11d).
6. Superimposed folding
Folds are superimposed on the Palaeogene and Neogene deposits to the northwest of Bardaskan, 35 km north of the DFS central segment. These folds locally generate both types 1 and 3 of Ramsay's (Reference Ramsay1967) interference patterns (Fig. 12). In this region there are two generations: (1) a NE–SW generation, which is observed in the Lower Eocene conglomerate, sandstone and marl and (2) a NW–SE generation, which is observed deforming Neogene conglomerate and marl as well as older strata. The NE–SW axial traces of the F1 (the first-generation folds; black on Fig. 12) indicate a NW–SE direction of shortening, while the NW–SE-striking F2 axial traces (the second-generation folds; white on Fig. 12) are consistent with the present-day kinematics shown by our analysis on the DFS, as well as the general shortening direction across the Iranian Plateau revealed by GPS data (Vernant et al. Reference Vernant, Nilforoushan, Hatzfeld, Abbassi, Vigny, Masson, Nankali, Martinod, Ashtiani, Bayer, Tavakoli and Chéry2004). The applications of large-scale superimposed folding for kinematic inversion along strike-slip faults was used by Allen, Alsop & Zhemchuzhnikov (Reference Allen, Alsop and Zhemchuzhnikov2001). They interpreted the dome and basin interference patterns as an indication of transpressive inversion caused by switching of σ1 and σ3 stress axes owing to reactivation of an earlier transpressional system formed in a different stress field. In their model, this process generates upright folds directed at 45° to the dominant strike-slip system, which are then overprinted by a second set of folds forming at 45° in the opposing sense, similar to what we document for the DFS.
Superimposed folding in the north of the DFS. NE–SW axial traces (black lines) of F1 are consistent with a NW–SE shortening event due to a right-lateral transpression. NW–SE F2 axial traces (white lines) are consistent with the recent regional shortening direction.
7. Discussion
7.a. Kinematic inversion on the DFS
7.a.1. Mechanism of slip-sense inversion
A major reorganization of tectonic deformation occurred in the Arabia–Eurasia collision zone around 5 ± 2 Ma (e.g. Wells, Reference Wells1969; Quennell, Reference Quennell, Dixon and Robertson1984; Westaway, Reference Westaway1994; Axen et al. Reference Axen, Lam, Grove, Stockli and Hassanzadeh2001; Allen, Jackson & Walker, Reference Allen, Jackson and Walker2004; Allen et al. Reference Allen, Kheirkhah, Emami and Jones2011) following the reorganization of the Dead Sea transform at the start of the westward motion of the Turkish Plate and activation of the North and East Anatolian faults at this time (Westaway, Reference Westaway1994), and the onset of oceanic spreading in the Red Sea (Joffe & Garfunkel, Reference Joffe and Garfunkel1987). This reorganization was associated with (1) rapid south Caspian subsidence (Nadirov et al. Reference Nadirov, Bagirov, Tagiyev and Lerche1997), (2) cooling, exhumation and uplift of the west–central Alborz (Axen et al. Reference Axen, Lam, Grove, Stockli and Hassanzadeh2001; Ballato et al. Reference Ballato, Uba, Landgraf, Strecker, Sudo, Stockli, Friedrich and Tabatabaei2011), (3) coarse molasse deposition, and onset of folding in the Zagros foreland (Dewey et al. Reference Dewey, Pitman, Ryan and Bonin1973; Beydoun, Hughes Clarke & Stoneley, Reference Beydoun, Hughes Clarke and Stoneley1992), as late as the Pliocene (Falcon, Reference Falcon and Spencer1974; Hessami, Reference Hessami2002), and (4) reactivation of CIM strike-slip faults (Walker & Jackson, Reference Walker and Jackson2004). However, the 5 ± 2 Ma age range is much shorter than the overall age of the collision, which began in Early Miocene time (16–23 Ma) or even earlier (Hempton, Reference Hempton1987; Yilmaz, Reference Yilmaz1993; Robertson, Reference Robertson, Bozkurt, Winchester and Piper2000; Guest et al. Reference Guest, Axen, Lam and Hassanzadeh2006; Shabanian et al. Reference Shabanian, Bellier, Siame, Arnaud, Abbassi and Cochemé2009a ,Reference Shabanian, Siame, Bellier, Benedetti and Abbassi b , Reference Shabanian, Bellier, Abbassi, Siame and Farbod2010; Ballato et al. Reference Ballato, Uba, Landgraf, Strecker, Sudo, Stockli, Friedrich and Tabatabaei2011). It seems that the collision zone has not accommodated the overall plate convergence consistently throughout its history.
Our present geological, tectonic and geomorphological data indicate that the DFS developed under a NW–SE-oriented right-lateral transpressional tectonic regime since Early Eocene time and switched from right-lateral to left-lateral motion in Late Miocene – Middle Pleistocene time. We suggest that right-lateral kinematics dominated during convergence between the CIM and the CD; the dip-slip kinematic component manifested as reverse faulting, which uplifted the northern mountains and facilitated the deposition of thick foreland sequences of Miocene sediments (Fig. 2) on the southern footwall fault blocks.
We recognize three kinematic phases on the DFS: (1) Early Eocene to Late Miocene (~55.8–5.33 Ma) right-lateral movement, (2) a Late Miocene to Early Pleistocene (~5.33–1.8 Ma) kinematic reversal interval, and (3) post-Early Pleistocene (~1.8 Ma to Recent) left-lateral motions.
Following the anticlockwise rotation of the CIM (e.g. Wensink, Reference Wensink1970; Soffel et al. Reference Soffel, Davoudzadeh, Rolf and Schmidt1996; Mattei et al. Reference Mattei, Cifelli, Muttoni, Zanchi, Berra, Mossavvari and Eshraghi2012) and its NW migration in Palaeogene time (Schmidt & Soffle, Reference Schmidt and Soffel1984) (after the accretion of the CIM to the CD in Late Cretaceous – Early Eocene time) right-lateral transpression dominated along the northern margin of the microplate. Strain partitioning owing to this right-lateral transpression resulted in NE–SW-trending contractional structures such as folds and thrust faults to the north of the DFS.
Transpressive inversion may thus represent a widely applicable mechanism for generating large-scale fold interference patterns in zones of major crustal deformation (Allen, Alsop & Zhemchuzhnikov, Reference Allen, Alsop and Zhemchuzhnikov2001). Fold interference patterns to the north of the central segment of the fault (Fig. 12) also show two different directions of shortening. The NE–SW first phase of folding was produced by right-lateral transpression, while the NW–SE second phase formed owing to a left-lateral transpression that dominated after the Arabian–Iranian final collision and tectonic reversal in the north of Central Iran.
Post-Pliocene left-lateral transpression is limited predominantly to the left-lateral faulting concentrated along the DFS. We speculate that the slip-sense inversion along the DFS from right-lateral in the Tertiary to left-lateral in the Quaternary took place owing to a diminishing anticlockwise rotation of the CIM, and a change in far field effects of the Arabian–Iranian collision in Late Miocene time. After the collision, the NE–SW convergence vectors predominated in Iran and resulted in a left-lateral movement on the E–W DFS.
Seismic records confirm the left-lateral nature of the present kinematic regime on the DFS (Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007; Farbod et al. Reference Farbod, Bellier, Shabanian and Abbassi2011). Historical and instrumental seismic records for the central part of the DFS indicate seismic activity along this fault (e.g. Berberian & Yeats, Reference Berberian and Yeats1999, Reference Berberian and Yeats2001; Fattahi et al. Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007; Hollingsworth et al. Reference Hollingsworth, Fattahi, Walker, Talebian, Bahroudi, Bolourchi, Jackson and Copley2010) (Fig. 13). Out of eight focal mechanisms shown in Figure 13 that are located close to the central part of the fault, the 9 December 1979 and 30 July 2010 events occurred on the DFS and give left-lateral kinematic solutions. Moreover GPS measurements by Tavakoli (Reference Tavakoli2007) evaluated the present-day velocity field in several places along the DFS and showed that the velocities along the DFS decrease from its central part (longitude ~58°, 2.5 ± 2 mm yr−1) towards the eastern segment (longitude ~60.2°, 1 ± 2 mm yr−1), which is compatible with our interpretations of the maximum horizontal stress components on the DFS segments (see Section 4.a). Similarly, Tavakoli (Reference Tavakoli2007) demonstrated that the decrease in velocities along the DFS from west to east is consistent with a spatial strike-slip velocity variation along the fault and increase in shortening towards the eastern end of the fault.
Map of the central segment of the DFS with the fault-plane solutions of medium- to large-magnitude earthquakes. Solutions determined by body-waveform modelling (from McKenzie, Reference McKenzie1972) and Harvard CMT solutions in black and grey, respectively. White circles show the cities and red stars display instrumental earthquakes with Mw > 5.
7.a.2. Time of slip-sense inversion
We suggest that the timing of kinematic reversal along the DFS may be as early as Middle Miocene or as late as Early Pleistocene, because the youngest right-laterally offset rocks and the oldest left-laterally offset feature are the Middle Miocene units (Ghaemi & Moussavi-Harami, Reference Ghaemi and Moussavi-Harami2007) and the Middle Pleistocene alluvial fan deposits (Quch-Palang alluvial fan), respectively. Alternatively, we obtain an age of 315–405 ka (Middle Pleistocene) for the onset of left-lateral shear if we use the 2.4 ± 0.3 mm yr−1 slip rate presented by Fattahi et al. (Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007). This slip-sense inversion along the DFS is also consistent with the onset of strike-slip motion within the Kopeh-Dagh and a change in the boundaries of the Kopeh-Dagh and Binalud deformation zones at ~4 Ma (Shabanian et al. Reference Shabanian, Bellier, Siame, Arnaud, Abbassi and Cochemé2009 a,b). It is also consistent with the change in the regional state of stress during Pliocene–Quaternary time (since ~5 Ma) (Shabanian et al. Reference Shabanian, Bellier, Abbassi, Siame and Farbod2010).
In addition, as mentioned above, our time estimation for slip-sense inversion along the DFS is consistent with the timing of a widespread tectonic reorganization across the Arabia–Eurasia collision zone. For example, a major reorganization is seen in the sedimentation and deformation in both the Alborz and South Caspian basin at 5–6 Ma (Devlin et al. Reference Devlin, Cogswell, Gaskins, Isaksen, Pitcher, Puls, Stanley and Wall1999; Axen et al. Reference Axen, Lam, Grove, Stockli and Hassanzadeh2001; Jackson et al. Reference Jackson, Priestly, Allen and Berberian2001; Allen, et al. Reference Allen, Jones, Ismail-Zadeh, Simmons and Anderson2002; Guest et al. Reference Guest, Axen, Lam and Hassanzadeh2006; Ritz et al. Reference Ritz, Nzazari, Ghassemi, Salamati, Shafei, Solaymani and Vernant2006; Ballato et al. Reference Ballato, Uba, Landgraf, Strecker, Sudo, Stockli, Friedrich and Tabatabaei2011) and at 5 Ma in the Kopeh-Dagh (Lyberis & Manby, Reference Lyberis and Manby1999). Also, although the onset of folding along the Zagros is both poorly constrained and likely to be diachronous, previous work has suggested the onset of widespread folding across much of the belt may be as late as the Pliocene (Falcon, Reference Falcon and Spencer1974; Hessami, Reference Hessami2002). Also a Mio-Pliocene change in kinematics occurred to the south of the Zagros around the Hormoz strait. This change may indicate the real set up of the transform zone between the Zagros and Makran, and may be closely related to the change from thin-skinned to thick-skinned tectonics at the same time in the Zagros (Regard et al. Reference Regard, Hatzfeld, Molinaro, Aubourg, Bayer, Bellier, Yamini-Fard, Peyret, Abbassi, Leturmy and Robin2010). Thus, the most likely timing for the kinematic shift on the DFS should be the Plio-Pleistocene, when the Middle East in general underwent a regional kinematic shift (Hempton, Reference Hempton1987; Westaway, Reference Westaway1994; Axen et al. Reference Axen, Lam, Grove, Stockli and Hassanzadeh2001; Allen, Jackson & Walker, Reference Allen, Jackson and Walker2004). Lastly, Farbod et al. (Reference Farbod, Bellier, Shabanian and Abbassi2011) pointed to post-Miocene left-lateral movement on the DFS and considered the Pliocene (≤5 Ma) to be a maximum age for the initiation of left-lateral faulting on the fault. Our results for the Late Miocene – Early Pleistocene slip-sense inversion along the DFS are in agreement with this estimate.
7.b. Total displacement and slip rate of the DFS
The small drainages flowing across the DFS display left-lateral displacements, while the deeply incised large rivers exhibit more complex right-lateral displacements. These larger features probably delineate long-term cumulative displacement on the DFS. We suggest that the left-lateral drainage offsets are related to the Recent, active movements along the fault, whereas right-lateral deflections record pre-Quaternary fossil displacement of the fault zone. Left-lateral displacements of geomorphic features, such as river channels, alluvial fans and river terraces, and the lack of evidence for major left-lateral displacement of bedrock features indicates that left-lateral displacement along the DFS is a Recent phenomenon.
Bedrock offsets yield valuable information about the long-term cumulative displacement of the DFS. Although Wellman (Reference Wellman1965) reported two sets of left-laterally displaced streams offset by 75 m and 200 m along the DFS, Fattahi et al. (Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007) reported 800–850 m and 200–400 m left-lateral displacement on the Quch-Palang alluvial fan deposits and Tigh Ahmad anticline, respectively. From their observations, Fattahi et al. (Reference Fattahi, Walker, Khatib, Dolati and Bahroudi2007) estimated a slip rate of 2.4 ± 0.3 mm yr−1 for the central segment of the DFS.
All of the palaeomagnetic observations in the CIM were carried out on Jurassic to Eocene rock units from the Tabas, Yazd and Lut blocks, south of the DFS (Wensink, Reference Wensink1982; Bina et al. Reference Bina, Bucur, Prevot, Meyerfeld, Daly, Cantagrel and Mergoil1986; Soffel et al. Reference Soffel, Davoudzadeh, Rolf and Schmidt1996). These results showed significant anticlockwise rotation when compared with the direction expected from the coeval Eurasian palaeopoles of Besse & Courtillot (Reference Besse and Courtillot2002). Furthermore, these data are in agreement with recent work by Mattei et al. (Reference Mattei, Cifelli, Muttoni, Zanchi, Berra, Mossavvari and Eshraghi2012), and support the hypothesis that the CIM underwent significant anticlockwise vertical axis rotations during the Cenozoic.
It has long been suggested that the CIM experienced 35° anticlockwise vertical axis rotation in response to indentation between India and Eurasia since Eocene time (Soffel & Förster, Reference Soffel and Förster1980, Reference Soffel and Förster1984; Davoudzadeh, Soffel & Schmidt, Reference Davoudzadeh, Soffel and Schmidt1981; Bagheri, Reference Bagheri2007; Bagheri & Stampfli, Reference Bagheri and Stampfli2008). This CIM rotation is confirmed by long-term right-lateral movement along the approximately N–S-trending Nehbandan and NW–SE-trending Dehshir faults defining the eastern and western edges of the CIM, respectively (Figs 1, 14) (Meyer & Le Dortz, Reference Meyer and Le Dortz2007). Furthermore Mattei et al. (Reference Mattei, Cifelli, Muttoni, Zanchi, Berra, Mossavvari and Eshraghi2012) showed a significant amount of anticlockwise rotation of the CIM (35° ± 18° for the Anarak area and 19.8° ± 8° for the Tabas area) during the Late Cenozoic.
Schematic reconstruction of the right-lateral motion along the DFS based on ~35° of anticlockwise rotation of the CIM since Eocene time. The CIM rotation is assumed to have occurred along the DFS according to the present curved geometry of the fault (details in text).
To develop a kinematic model, we consider the curved geometry of the DFS, comparing the fault zone with a semicircle on which we have reconstructed post-Late Cretaceous – Early Eocene anticlockwise rotation of the CIM (Fig. 14). In this reconstruction, the town of Jandaq, shown as a geographical reference point located on the fault trace, shifts approximately 280 km to the east. Assuming that the CIM has rotated about 35° anticlockwise along the DFS with the present curved geometry (Fig. 14), and considering 55.8 Ma (beginning of the Eocene period) as the time of onset for right-lateral movement on the fault and a 5.33 to 1.8 Ma time interval for kinematic reversal, we estimate a former right-lateral slip rate of ~5.2–5.55 mm yr−1.
7.c. Geotectonic evolution of the northern CIM
We have tried to constrain the post-Neogene geotectonic evolution of the CIM using our analyses of palaeomagnetic and geological data, and our detailed field work along the northern margin of the DFS.
The palaeomagnetic and palaeogeographical reconstructions indicate the development of 3–4 km deep troughs in a horst–graben pattern during the fragmentation of the Iranian Plate into the CIM and CD. These two domains consist of the Sabzevar zone and Alborz Mountains, as well as the formation of the ‘Doruneh–Sabzevar oceanic basin’, which resulted from an early Cimmerian dextral transtension tectonic regime (Fig. 1) (Soffel & Förster, Reference Soffel and Förster1980; Davoudzadeh, Soffel & Schmidt, Reference Davoudzadeh, Soffel and Schmidt1981; Şengör, Reference Şengör1984). Subduction of the Neo-Tethys beneath the CIM was initiated in Cretaceous time along the SW margin of the CIM. This late Cimmerian event caused a switch from right-lateral transtension to a transpressional tectonic regime, which drove vertical axis rotation of the CIM (Schmidt & Soffel, Reference Schmidt and Soffel1984).
Further convergence between the CIM and the CD during Early Eocene to Late Miocene time, produced a ~35° anticlockwise rotation of the CIM. This convergence between the CIM with the CD also caused Early Eocene obduction of ophiolite mélanges and the emplacement of voluminous Upper Cretaceous–Eocene alkaline magmatic rocks (Schmidt & Soffel, Reference Schmidt and Soffel1984). To the north of the DFS, intensive folding and faulting of the Taknar region (Figs 2c, 12a) and the Cretaceous–Tertiary unconformity are owing to this Early Eocene event (Lindenberg & Jacobshagen, Reference Lindenberg and Jacobshagen1983). Regional right-lateral transpression and the associated strain partitioning to the north of the DFS were responsible for folds with NE–SW-trending axes and right-lateral movement on the Taknar fault. The primary episode of right-lateral activity of the Taknar fault formed the dextral Taknar strike-slip imbricate fan, which has subsequently been inverted after the kinematic reversal to left-lateral motion. The kinematic inversion of the right-lateral Taknar imbricate fan requires a different regional shortening direction in this part of the Iranian Plateau prior to the Late Miocene to Early Pleistocene interval.
8. Conclusion
The DFS developed under a NW–SE-oriented right-lateral transpressional tectonic regime since Early Eocene time as a primary right-lateral strike-slip fault. Right-lateral kinematics dominated during convergence between the CIM and the CD. We recognize three kinematic phases on the DFS: (1) Early Eocene to Late Miocene right-lateral movement, (2) a Late Miocene to Early Pleistocene kinematic reversal interval, and (3) post-Early Pleistocene left-lateral motions.
The slip-sense inversion along the Doruneh fault from right-lateral to left-lateral took place owing to a diminishing anticlockwise rotation of the CIM, and a change in far field effects of the Arabian–Iranian collision in Late Miocene time. Assuming that the CIM has rotated about 35° anticlockwise along the DFS with the present curved geometry, and considering 55.8 Ma (beginning of the Eocene period) as the time of onset for right-lateral movement on the fault and a 5.33 to 1.8 Ma time interval for kinematic reversal, we estimate a former right-lateral slip rate of ~5.2–5.55 mm yr−1.
Regional right-lateral transpression and the associated strain partitioning to the north of the DFS were responsible for folds with NE–SW-trending axes and right-lateral movement on the Taknar fault. In addition, interaction between older right-lateral transpression produced NE–SW-folding in the north of the fault, while recent left-lateral transpression has superimposed a NW–SE second phase of folds on the first generation.
Acknowledgements
This research was supported by the Geological Survey of Iran (GSI) and the Research Institute for Earth Sciences. We acknowledge the GSI for providing funds to undertake this research, especially M. T. Korehie and M. Ghorashi for their financial and logistic support. We thank Phil Leat, editor of Geological Magazine, and two anonymous reviewers for constructive and detailed reviews.
