The Oman ophiolite provides a natural laboratory for understanding oceanic lithospheric processes. Previous paleomagnetic and structural investigations have been used to support a model involving rotation of the ophiolite during formation at a mid-oceanic microplate. However, recent geochemical evidence indicates that the ophiolite instead formed in a nascent forearc environment, opening the potential for alternative rotation mechanisms. Central to the conundrum is the contrast between ESE to SE magnetizations and NNW magnetizations from the northern and southern ophiolitic massifs, respectively, attributed previously to either differential tectonic rotations during spreading or complete emplacement-related remagnetization of the southern massifs. Here we report new paleomagnetic data from lower crustal rocks of the southern massifs that resolve this problem. Sampling of a continuous section in Wadi Abyad reveals ENE magnetizations in the dike rooting zone at the top of the lower crust that change systematically downwards to NNW directions in underlying foliated and layered gabbros. This is consistent only with remagnetization from the base upwards, replacing early remanences in layered and foliated gabbros completely but preserving original ENE magnetizations at higher levels. Comparison with new data from Wadi Khafifah provides a positive fold test that shows that this event occurred before late Campanian structural disruption of the regional orientation of the petrologic Moho. These data show that the entire ophiolite experienced large intraoceanic clockwise rotation prior to partial remagnetization, leading to a new tectonic model in which formation, rotation, and emplacement of the ophiolite are all linked to Late Cretaceous motion of Arabia and roll-back of the Oman subduction zone.
Paleomagnetic analyses of ophiolites can provide important constraints on tectonic processes in oceanic lithosphere. In particular, paleomagnetic data from Mesozoic ophiolites of the Alpine-Himalayan orogenic belt have demonstrated the importance of tectonic rotation during phases of intraoceanic and/or emplacement-related deformation (e.g., MacLeod et al., 1990; Morris et al., 2002; Inwood et al., 2009; Maffione et al., 2013; Morris and Maffione, 2016). The ∼500-km-long Oman ophiolite within the Neotethyan suture zone is the largest and most intensively studied ophiolite in the world (e.g., Lippard et al., 1986). It formed at ca. 96 Ma at a fast spreading rate (Nicolas, 1989; MacLeod and Rothery, 1992; Rioux et al., 2012) and was detached soon after formation (Warren et al., 2005), before being emplaced onto the Arabian continental margin by the end of the Cretaceous (Lippard et al., 1986) and subsequently disrupted by gravity-driven extension around basement structural culminations (Cawood et al., 1990). Paleomagnetic studies of the northern massifs of the ophiolite have identified ESE- to southeast-directed magnetizations (Fig. 1), indicating large clockwise rotation of (at least) this part of the ophiolite (Shelton, 1984; Thomas et al., 1988; Perrin et al., 1994, 2000; Weiler, 2000). Systematic differences in declination between early (V1, Geotimes unit) and later (V2, Lasail and Alley units) volcanic units (Perrin et al., 1994, 2000) provide evidence that this rotation began during crustal construction in a pre-emplacement, intraoceanic setting. In contrast, analyses in the southern massifs have identified consistent NNW-directed magnetizations (Fig. 1; Luyendyk et al., 1982; Luyendyk and Day, 1982; Thomas et al., 1988; Feinberg et al., 1999; Weiler, 2000). Together, these data have been used to develop a tectonic model for formation and differential rotation of the ophiolite within a mid-ocean ridge microplate (Boudier et al., 1997) or overlapping spreading center (Weiler, 2000) environment. However, recent geochemical evidence indicates that the ophiolite instead formed in a nascent arc spreading system above a subduction zone immediately following subduction initiation (MacLeod et al., 2013); hence the geodynamic controls driving rotation may be fundamentally different. Furthermore, NNW magnetizations in the southern massifs have also been ascribed to complete late-stage, emplacement-related remagnetization (Feinberg et al., 1999), potentially complicating or even obliterating any earlier magnetic record of intraoceanic rotation in this part of the ophiolite.
Here we present paleomagnetic data from transects through lower crustal rocks of the Oman ophiolite that provide unequivocal evidence that the southern massifs originally held easterly directed magnetizations before experiencing subsequent late-stage remagnetization from the base up. This removes ambiguity in the interpretation of the existing paleomagnetic data from Oman and leads to a new tectonic model involving simple rotation in a suprasubduction zone setting.
LOWER CRUSTAL GEOLOGY AND PALEOMAGNETIC DATA
Lower oceanic crustal gabbros and underlying mantle peridotites dominate the southern massifs of the Oman ophiolite. We focus on sections in Wadi Abyad (Rustaq massif) and Wadi Khafifah (Ibra massif), where long transects from the Moho through typical lower crustal sequences are exposed (Fig. 1A; Pallister and Hopson, 1981; MacLeod and Yaouancq, 2000). The following lithostratigraphic units were sampled in Wadi Abyad (Fig. 1B):
(1) Layered gabbros in the lower half of the section (Fig. DR1 in the GSA Data Repository1), with layering defined principally by modal variations in olivine, clinopyroxene, and plagioclase on a centimeter to meter scale. Layering is consistently subparallel to the NNE-dipping Moho (Fig. DR2), the orientation of which (Table DR1 in the Data Repository) provides a paleohorizontal surface for tilt correction of the paleomagnetic data.
(2) Foliated gabbros (Fig. DR3), with preferred mineral orientations defining a consistently steep, near-Moho-perpendicular foliation and a subvertical lineation.
(3) Varitextured gabbros (Fig. DR4), marked by high variability in grain size, texture, and composition over short distances (and locally cut by discrete dikes), that represent the fossil axial melt lens (MacLeod and Yaouancq, 2000).
(5) The dike rooting zone, which has a gradational contact with the underlying varitextured gabbros and consists of microgabbro and basalt dikes separated by gabbro screens (Fig. DR5).
Layered and foliated gabbros with the same characteristics were also sampled in Wadi Khafifah (Fig. DR6), but here the Moho dips to the south (Table DR1), allowing a regional-scale fold test to be performed between localities.
Oriented paleomagnetic cores collected at 19 sites along Wadi Abyad and at 13 sites along Wadi Khafifah were analyzed using standard paleomagnetic laboratory techniques (see the Data Repository). Following removal of occasional low-stability, present-field viscous components, demagnetization data define either linear, single components of magnetization or, in the case of varitextured gabbro samples in Wadi Abyad, curvilinear trajectories leading to linear components at high treatment levels (Fig. DR7). In all cases, laboratory unblocking temperatures close to 580 °C and supporting rock magnetic experiments indicate that these remanences are carried by nearly stoichiometric magnetite (Figs. DR8 and DR9). All sites yielded statistically well-defined mean magnetization vectors (Table DR1) that have a consistent normal polarity (positive inclinations) after tilt correction. In both Wadi Abyad and Wadi Khafifah, layered and foliated gabbros overall display NNW-directed magnetizations (Fig. 1C; Fig. DR6). This is consistent with data reported previously from one layered gabbro site in each of these sections (Weiler, 2000). However, sampling of the complete crustal section shows that site mean magnetizations in the Wadi Abyad foliated gabbros vary systematically from NNW directions in the lowermost sites to north directions in the uppermost sites, passing upwards into varitextured gabbros with distributed northeast-directed magnetizations, and finally into the dike rooting zone marked by ENE remanences (Fig. 1C).
ORIGIN AND TIMING OF MAGNETIZATIONS
East-northeast–directed, normal-polarity magnetization directions at the top of the lower crust in Wadi Abyad require clockwise vertical-axis rotation of the section. The distribution of site directions beneath, however, can only be explained by remagnetization from the base upwards via production of secondary magnetite during fluid-mediated alteration, with pervasive overprinting of the base of the section decreasing progressively upwards to preserve original ENE magnetizations at the top. This is supported by the following evidence:
(1) Curvilinear demagnetization paths in the varitextured gabbros (Fig. DR7) suggest significant overlap in unblocking temperature or coercivity spectra of two different magnetization components, consistent with the presence of a magnetic overprint superimposed on an original thermoremanence.
(2) Demagnetization data at lower treatment levels in the varitextured gabbros define great-circle paths that trend toward the magnetization direction of the foliated and layered gabbros beneath (Fig. 1C). The magnetic overprint in these rocks hence shares the same direction as the NNW-directed characteristic magnetization seen lower in the section.
(3) Petrographic evidence shows a dominance of secondary magnetite in the layered and foliated gabbros, mainly associated with serpentinization of olivine, and mostly primary magnetite in the varitextured and dike-rooting-zone gabbros (Yaouancq and MacLeod, 2000; Meyer, 2015). This can account for observed variations in magnetization directions in both the foliated and varitextured gabbros, which may be attributed to between-site variations in the relative proportions of primary and secondary magnetite and in the degree of overlap of their blocking temperature or coercivity spectra.
Remagnetization of the southern blocks of the ophiolite has been invoked previously by Feinberg et al. (1999) on the basis of comparison of paleomagnetic data from seven sites in mantle peridotites and two sites in layered gabbros against those obtained from five sites in high-grade continental metabasites underlying the ophiolite. NNW magnetizations with shallow inclinations in the metabasites were interpreted as thermochemical remanences acquired during exhumation and cooling from peak metamorphic conditions between 80 and 70 Ma. Observation of identical directions of magnetization in overlying serpentinized peridotites and layered gabbros led Feinberg et al. (1999) to infer contemporaneous remagnetization of the ophiolite via alteration associated with upward advection of metamorphic fluids in the faulted ophiolite. This interpretation is entirely consistent with our new data, which independently demonstrate that remagnetization of the ophiolite occurred from the base upwards. Preservation of ENE magnetizations in the Wadi Abyad dike rooting zone defines the upper limit of fluid advection at this locality, although NNW remanences reported from dikes of the Ibra massif (Fig. 1; Luyendyk et al., 1982) suggest that remagnetization elsewhere extended further up into the sheeted dike complex.
The timing of remagnetization of the southern massifs may be constrained by comparing paleomagnetic data from Wadis Abyad and Khafifah, as variations in the orientation of the Moho between these locations allow a paleomagnetic fold test to be performed. Upon restoration to the paleohorizontal, site mean directions from layered and foliated gabbros at these localities converge (Figs. DR10A and DR10B). A bootstrap statistical test (Tauxe and Watson, 1994) indicates that the tightest grouping occurs at close to 100% untilting (94%–116%; Fig. 1D; Fig. DR10C), constituting a positive fold test. Hence remanences were acquired prior to structural disruption of the regional orientation of the Moho that occurred during gravitational sliding away from emerging basement structural highs in the Campanian (Cawood et al., 1990). Combined with the preservation of demonstrably older (and presumed primary, ridge-related) remanences in the Wadi Abyad dike rooting zone, this is consistent with remagnetization during emplacement of the structurally intact ophiolite sheet onto the Arabian margin.
The lowermost three layered gabbro sites in Wadi Abyad (WA09–WA11; Table DR1) have tilt-corrected magnetizations that differ from the rest of the data from this section (Fig. 1). However, their in situ magnetizations are indistinguishable from tilt-corrected directions observed in layered and foliated gabbros higher up the section. This implies that these sites acquired their magnetizations later in the deformation history than the rest of the section, potentially indicating that the lowest part of the sequence became remagnetized during the waning stages of fluid migration, after significant tilting of the section.
IMPLICATIONS FOR THE TECTONIC EVOLUTION OF THE OMAN OPHIOLITE
Identification in Wadi Abyad of two significantly different remanence directions that both pre-date deformation of the Moho provides critical constraints on the rotation history of the southern massifs and their relationship to the northern massifs. Comparison of the mean tilt-corrected magnetization of the foliated and layered gabbros (excluding sites WA09–WA11; Table DR1) with a reference direction derived from the 70 Ma African apparent polar wander path (Torsvik et al., 2012) requires 22° ± 7° of counterclockwise rotation after remagnetization (Demarest, 1983). Back-stripping the effect of this rotation from the pre-remagnetization direction of the dike rooting zone then indicates an earlier clockwise rotation of 87° ± 11°. This demonstrates that the southern massifs experienced a clockwise rotation along with the northern massifs (Shelton, 1984; Thomas et al., 1988; Perrin et al., 1994, 2000; Weiler, 2000) and removes the need for large differential rotations within the ophiolite that otherwise require complex tectonic models (Weiler, 2000).
Given paleomagnetic evidence for active rotation during crustal accretion (Perrin et al., 1994, 2000) and geochemical evidence for formation of the ophiolite above an intraoceanic subduction zone (e.g., MacLeod et al., 2013), the recognition of ophiolite-wide, pre-remagnetization clockwise rotation is consistent with a simple tectonic model (Fig. 2) involving: (1) impingement of the Arabian continental margin with a young intraoceanic subduction zone that begins to roll back, resulting in suprasubduction zone spreading and early rotation of newly formed crust; (2) continued northward movement of Arabia and accompanying roll-back of the Oman subduction zone system, leading to further rotation and eventually to emplacement of the rotated ophiolite onto the Arabian margin; (3) impingement of the southern massifs with basement structural highs (Saih Hatat), triggering a wave of orogenic fluids, near-wholesale remagnetization from the base up, and subsequent back-rotation; and (4) extensional collapse generating the present-day configuration of the ophiolite.
This model is consistent with a conceptual framework for rotation in modern subduction systems undergoing collision with buoyant indenters proposed by Wallace et al. (2005) on the basis of plate kinematics in the Pacific region. In these systems a combination of indentation and subduction roll-back generates a torque on the upper plate, leading to rapid rotation of forearc regions (Wallace et al., 2005). In this context, we note that clockwise rotation of the Oman ophiolite mirrors the ∼90° counterclockwise rotation of the Troodos and Hatay ophiolites in the southern Neotethys ocean to the west of the Arabian margin (Clube et al., 1985; Inwood et al., 2009). Rotation of these ophiolites similarly occurred prior to emplacement, in response to impingement of the Arabian indenter with the southern Neotethyan subduction zone.
We thank Mohamed Alaraimi (Sultanate of Oman Ministry of Commerce and Industry, Directorate General of Minerals) for permission to undertake field sampling in Oman. The bootstrap fold test was performed using PmagPy software (Tauxe et al., 2010). We thank Gabriel Gutiérrez-Alonso, Phil McCausland, and Eric Ferre for constructive reviews.
- Received 19 July 2016.
- Revision received 3 October 2016.
- Accepted 3 October 2016.
- © 2016 Geological Society of America