Abstract: The formation of the global plate boundary system is a key step in the emergence of plate tectonics, and how plate boundaries of thousands of kilometers are formed in a very short geological time is still unclear. In order to determine the driving force of subduction initiation and plate boundary formation, Dutch scientist van Hinsbergen comprehensively reconstructed a plate boundary of more than 12,000 km in length formed between the Indian and African plates around 105Ma, and analyzed that the mantle column of the same era was the only possible trigger for plate rotation, and the thick craton lithosphere played an important role in the process of mantle column inducing plate rotation and thus triggering subduction initiation. The authors suggest that mantle columns may provide a non-plate tectonic mechanism for plate rotation, and the formation of converged or divergent plate boundaries away from the head of the mantle column may be the root cause of the emergence of modern plate tectonics.
With the formation of plate boundary systems (expanding ridges, transition faults, and subduction zones), early Plate Tectonics of the Earth gradually established. In general, the formation of dilated ridges and transformation faults is a passive process, while the formation of subduction zones is not yet clear. Continental, oceanic plateaus, or volcanic arcs can cause changes in intraplate stress when they reach trenches, leading to gravitational collapse of the ocean lithosphere or changes in subduction zones, a process often used to explain subduction initiation. Numerical simulation studies have concluded that the mantle column is one of the driving forces for the initiation of regional subduction. But while these processes can explain plate tectonics at the regional scale, they cannot explain the geodynamic reasons for geologically short -- (1000 km) plate boundaries, including new subduction zones. Using geological records to study the causes of plate boundary formation and its subsequent plate subduction initiation requires determining the following points: 1) determining whether the subduction initiation of the plate boundary is spontaneous or induced; 2) if induced, determining the time and direction of plate convergence early, and (3) reconstructing the entire plate boundary to identify factors such as collisions, subduction terminations, or mantle columns that may contribute to the stress changes that drive subduction initiation.
The SSZ ophiolite and metamorphic substrate (subduction plate) of the cladding sheet are geological evidence of the early subduction (Agard et al., 2016), and among the multiple SSZ ophiolites in the Neo-Tethys Ocean, there is an SSZ serpentine greenstone belt formed in the Cretaceous period, which extends from the eastern Mediterranean region along northern Arabia to Pakistan (the blue line in Figure 1a). In the metamorphic base plate in Oman and the Eastern Mediterranean region, the burial age of the metamorphic base plate is determined by the age of Lu/Hf garnet growth to ~104Ma, that is, the formation of the metamorphic bottom plate. The age of laminate stretch and SSZ snake greenstone extension was determined by the age of magmatic zircon U/Pb and the cooling age of the homogeneous bottom plate 40Ar–39Ar, which ranged from 96–95 Ma (Pakistan, Oman) to 92–90 Ma (Iran, Eastern Mediterranean region). The time delay of 8–14 Ma between the formation of the metamorphic base plate (plate burial) and the extension of the upper cladding plate (SSZ type hydra formation) provides direct geological evidence for inducing subduction initiation. Through paleomagnetic analysis and detailed kinematic reconstruction of the initial deformation after subduction in the Eastern Mediterranean, Oman and Pakistan, the early convergence of plates in the subduction zone was in the E–W direction. That is, the convergence of the ~E–W towards the plate triggered the start of the 105Ma subduction of the neo-Tethys Ocean.
To find the trigger that led to this subduction, the authors then performed a tectonic reconstruction of the neo-Tethys Ocean and its surrounding continents (Figure 1). The results showed that Africa and India were adjacent plates in the 105Ma Neotesis Ocean subduction zone, and the African plate was mostly surrounded by ridges, with a complex subduction plate boundary in the Mediterranean region. The Indian plate is surrounded by a ridge conversion system to the south and a subduction to the north. The Neo-Tethys Ocean lithosphere between the Arab-Greater Adriatic and Eurasian continents continues to slope northward, and this subduction zone has been present on the southern edge of Eurasia since the Jurassic period (Figures 1b, 1c). Ocean geophysical constraints show that during the rift valley of about 83 Ma before the ocean expansion in the Mascalin Basin began, India rotated counterclockwise relative to Africa around the Euler pole in the western Indian Ocean (Figure 1c), which is associated with east-west convergence spanning hundreds of kilometers of Neoteis (Figure 1d). The boundary of the Cretaceous Neshintete Plate may have converged to and from where it became stretched, and formed a rift valley in the Mascalin Basin, later an expanding ridge separating India from Madagascar (Figure 1b).
There is no viable tectonic-related driver for the formation of reconstructed ~105Ma plate boundaries: there are many remnants of microcontinents that have been proliferating in subduction zones since the Paleozoic period in southern Eurasia, but they do not occur at 105Ma. Continental subduction and collisions are underway in the central Mediterranean, but it is unclear whether subduction dynamics along the E-W trend in southern Eurasia will lead to convergence in the E-W direction of the Neo-Tethys Ocean. In the eastern neo-Tethys Ocean, the intra-ocean Woyla arc collides with the continental edge of Sundaland, resulting in a subduction polarity reversal and begins to subduct eastward below Sundaland, creating a plate pull along with the formation of the Indian/Neotethic plate boundary (Figure 1c). However, the eastward pull of the plates below Sundaland could not drive the east-west convergence of the neo-Tethys Ocean, and the stretching of the plates is likely to be a manifestation of the rotation of the Indian plate rather than a trigger. As a result, the mantle column, the only remaining non-tectonic tectonic plate movement driver in the region, could play a key role. Indo-Madagascar land fracture is widely believed to be related to the formation of the Morondava Igneous Province (LIP) in Madagascar and southwestern India at ~94 Ma. After the initial plate boundary was formed ~ 10 Myr, the igneous province began to form.
Figure 1 Plate kinematic reconstruction of the Neo-Tethys Ocean and its surrounding continents (Van Hinsbergen et al., 2021)
In order to study the relationship between mantle columns and igneous provinces and plate boundary formation, the authors conducted a torque balance simulation of mantle column-lithosphere interaction, semi-analytical calculation of The Indian and African plates at ~105Ma, and evaluated the influence of the Craton lithosphere on the location of the Indo-African Euler pole (Figure 2). In experiments without a thick craton lithosphere, the push of the mantle column below Madagascar/India caused India to rotate counterclockwise relative to Africa, but euler poles, located north of the Arabian Peninsula (Figure 2a), did not cause convergence in the E-W direction of the Neo-Tethys Ocean. However, in experiments with the presence of claton-thick lithospheres in India and Africa, the calculated Euler pole position moved south toward the Indian continent, converging in the E-W direction along the boundary of most of the plates of the Neo-Tethys Ocean (Figure 2b). Moreover, if the injection and induced flow of mantle column material are confined to a 200 km thick weak asthenosphere, it can trigger aggregation of up to hundreds of kilometers, which is enough to induce self-sustained subduction.
Fig. 2 Modeling results of mantle-lithosphere interaction torque balance with/without craton keel (Van Hinsbergen et al., 2021)
Assuming that there is no frictional resistance at the boundaries of other plates, the ascent force obtained from the head of the mantle column with a diameter of 1000 km and a density of 30 kgm-3 is about ~1.5×1020 N. If half of this force acts on the Indian plate and the lever arm is 4000km, it corresponds to a torque of 3×1026 Nm. This torque is balanced at the convergence boundary (length of about 5000 km, plate thickness of about 100 km) and generates a stress of 240 MPa, which is much greater than the friction resistance between subduction and overburden plates. The mantle column induced compressive stress may have increased pre-existing compressive stress (particularly due to the ridge thrust generated around the African and Indian plates), and this additional compressive stress may help move the Euler column further south, closer to where it was rebuilt in Figure 1. Numerical models show that after induced convergence of 50 to 100 km, it can begin from continuous subduction, which is equivalent to an Indo-African rotation of about 1° between about 105 Ma and ~96–92 Ma. Subsequent east- and west-leaning subductions (figure 1) may have contributed to and accelerated the Indo-African/Arab rotation, driving the southward spread of the Euler pole (Figures 2a, 2c).
Previous numerical simulation experiments have shown that the mantle column may trigger the onset of a circular subduction around the head of the mantle column (Gerya et al., 2015). Based on the geological record of the Cretaceous > 12,000 km plate boundary, the authors perform kinetic reconstruction to provide the first evidence of the role of the mantle column in the formation of subduction zones, and prove that the rise of the mantle column is a key driver of subduction plate boundary formation. Mantle columns are also common to planets without plate tectonic motion, such as Mars and Venus, which provides new perspectives for understanding the evolution of different planets.
Key References (Swipe up and down to view)
Agard P, Yamato P, Soret M, et al. Plate interface rheological switches during subduction infancy: Control on slab penetration and metamorphic sole formation[J]. Earth and Planetary Science Letters, 2016, 451: 208-220.
Arculus R J, Ishizuka O, Bogus K A, et al. A record of spontaneous subduction initiation in the Izu–Bonin–Mariana arc[J]. Nature Geoscience, 2015, 8(9): 728-733.
Gerya T V, Stern R J, Baes M, et al. Plate tectonics on the Earth triggered by plume-induced subduction initiation[J]. Nature, 2015, 527(7577): 221-225.
Stern R J. Subduction initiation: Spontaneous and induced[J]. Earth and Planetary Science Letters, 2004, 226(3-4): 275-292.
Van Hinsbergen D J J, Peters K, Maffione M, et al. Dynamics of intraoceanic subduction initiation: 2. Suprasubduction zone ophiolite formation and metamorphic sole exhumation in context of absolute plate motions[J]. Geochemistry, Geophysics, Geosystems, 2015, 16(6): 1771-1785.
Van Hinsbergen D J J, Steinberger B, Guilmette C, et al. A record of plume-induced plate rotation triggering subduction initiation[J]. Nature Geoscience, 2021, 14(8): 626-630.（原文链接）
Editor: Chen Feifei
Proofreader: Zhang Tengfei Jiang Xuejiao