Elsevier

Precambrian Research

Volume 352, January 2021, 105980
Precambrian Research

Onset of plate tectonics by the Eoarchean

https://doi.org/10.1016/j.precamres.2020.105980Get rights and content

Highlights

  • Geological archives show that plate tectonics operated since at least the Eoarchean.

  • Accretionary Cycle Plate Tectonics operated from >4.0 Ga–0 Ga, with intraoceanic arcs and accretionary orogens.

  • Continental margin magmatism became stronger by 3.2 Ga–0 Ga with larger accreted arcs and more crust-mantle wedge mixing and melting.

  • Wilson Cycle Plate Tectonics began in a form we can recognize today from ~2.7–2.5 Ga after sufficiently large continents emerged.

  • Two types of plate tectonics operated on early Earth; Accretionary Cycle Plate Tectonics and Wilson Cycle Plate Tectonics.

Abstract

One of the most contentious areas of Earth Science today is when, or whether or not modern-style plate tectonics was in operation in the Archean Eon. In this review we present evidence that the onset of plate tectonics was not at 3.2 Ga, as popularly conceived, but was in operation during the Eoarchean by at least ca. 4.0 Ga. Following a review of the main Eoarchean supracrustal belts of the world, constrained by relevant geochemical/isotopic data, we present evidence that suggests that from at least ca. 4.0 Ga Earth produced considerable juvenile mafic crust and consequent island arcs by Accretionary Cycle Plate Tectonics. From ~3.2 Ga there was a gradual transition in geodynamics to more abundant active continental margin magmatism in the form of voluminous TTGs and sanukitoids. From 3.2 Ga to 2.5 Ga juvenile oceanic crust and arcs continued to form, accompanied by more active continental margin magmatism until ~2.7–2.5 Ga, by which time there were sufficient crustal rocks to amalgamate into incipient large continents, the fragmentation of which started the first complete classical Wilson Cycle Plate Tectonics of breaking apart and re-assembling large continental masses. In other words, there were two types of plate tectonics in operation in the early Earth, Accretionary Cycle Plate Tectonics and Wilson Cycle Plate Tectonics, but Wilson Cycle type plate interactions only became more common after contiguous continental landmass became voluminous and extensive enough around 2.7–2.5 Ga. Failure to realize this dual mechanism of continental growth may lead to erroneous ideas such as “plate tectonics started at 3.2 Ga”, or “mantle plumes generated early Archean magmatic rocks.” We present new geochemical data that together with lithological and structural relationships, negate the various plume-type speculations including stagnant lids, heat pipes, and mushy-lid tectonics. It is interesting to consider that the way Earth’s crust developed in the first Gigayear of the geological record continued later, albeit in more advanced forms, into the Phanerozoic, where we can still recognize Accretionary Cycle Plate Tectonics and orogens still with short boundaries in examples including the Altaids of Central Asia, the Arabian-Nubian Shield, the Japanese Islands, and in incipient form in Indonesia, as well as Wilson Cycle Plate Tectonics that leads inexorably to continental collisions as in the Alpine-Himalayan orogen with its long plate boundaries. We recommend this holistic view of crustal growth and the evolution of continents that leads to a robust, viable, and testable model of Earth evolution.

Introduction

There have been two types of orogens, accretionary and collisional, throughout Earth history (Windley and Condie, 1992, Windley, 1995, Şengör and Natal’in, 1996a) typically expressed as low-grade greenstone-granite belts and high-grade granulite-gneiss belts as defined by Windley and Bridgwater (1971), and most recently by hafnium isotope data in terms of external and internal orogens (Collins et al., 2011).

Accretionary orogens (Cawood et al., 2009) form by the accretion of diverse components from the ocean floor, such as oceanic crust formed at mid-ocean ridges and associated hydrothermal vent deposits, epidosites, ophiolites, ocean plate stratigraphic sections, seamounts, and oceanic plateaus, as well as accretionary prisms, and island arcs with their forearcs and back-arcs. However, we only see today minor relicts of these components in accretionary orogens, particularly because of preferential subduction erosion during off-scraping/underplating/subduction type accretion (Stern, 2011, Scholl and von Huene, 2007, Scholl and von Huene, 2009) as in the subducting margins of eastern Asia and other parts of the circum-Pacific (Kimura and Ludden, 1995, Kusky and Young, 1999, Wakita et al., 2013) where ocean-derived ultramafic rocks, gabbros, sheeted dykes and most of the basalts are typically subducted, leaving only a small portion of the upper basalts and the overlying sediments to be accreted. This is the main reason why the largest subduction-accretion orogen in Japan contains only one ocean-derived, fore-arc ophiolite (the 580 Ma O-eyama ophiolite), but over 1000 small accreted fragments of seamounts, oceanic crust, oceanic plateaus, and arcs (Isozaki, 1996) appropriately termed ‘ophirags’ by Şengör and Natal’in, 1996b) and Şengör and Natal'in (2004). These fragmentary relations in Phanerozoic accretionary orogens bear heavily on the occurrence and preservation of comparable relicts of oceanic lithosphere in Archean accretionary orogens/belts (e.g., Kusky, 1989, Kusky, 2020, Kusky and Polat, 1999, Kusky et al., 2001, Kusky et al., 2018).

In contrast, continent–continent collisional orogens typified by the Alpine-Himalayan Chain form by the collision of two large continental blocks like India/Afro-Arabia and Eurasia (Treloar and Searle, 2019). Such collisional orogens can be traced back in Earth history with high degrees of confidence only to the Paleoproterozoic, when there is substantial evidence for the first existence of major continents (Wise et al., 1974, Windley, 1977, Hynes, 2001, Kasting et al., 2006, Flament et al., 2008, Pope et al., 2012, Johnson and Wing, 2020) and perhaps even the first supercontinent (Pehrsson et al., 2013). The Wilson Cycle, as traditionally viewed (Wilson et al., 2019), could only have started to operate when there were continental blocks large enough to rift, enabling the separation and drift of individual continents, which were then susceptible to subduction on their margins leading to the formation of Andean-type or active magmatic, granite-dominated orogens, and eventually to collision with another continent (Şengör, 1990, Wilson et al., 2019).

The essential point arising out of the above differences in understanding of orogens in Earth history is that we accept the geological evidence for the presence of accretionary orogens throughout the geological record into the Eoarchean (Condie, 2007, Cawood et al., 2009, Kusky et al., 2018), and therefore for the existence of plate tectonics back to at least 3.9–4.0 Ga or beyond (Harrison, 2009, Harrison, 2020, Turner et al., 2020), because accretionary orogens necessarily form by subduction and accretion via horizontal motion of plates with associated sea-floor spreading, transform motions between plates, and convergence and subduction. In contrast, since no large continental blocks have been demonstrated to have formed until perhaps the latest Archean (Pehrsson et al., 2013), the Wilson Cycle of rifting and re-assembling of large continents was only able to start near the Archean-Proterozoic boundary and therefore large continent–continent collisional orogens were only able to initially form to a significant degree in the Paleoproterozoic. These relations were all described and explained at some length in The Evolving Continents (Windley, 1995).

Today it is well recognized that subduction-accretion-erosion has played a major role in the differential preservation of the geological record (Kimura and Ludden, 1995, Isozaki, 1996, Scholl and von Huene, 2007, Scholl and von Huene, 2009, Stern, 2011). However, the possibility that there has been so much removal of the geological record that it would make it impossible for us to make a viable assessment of the differences seen today across the Archean-Proterozoic boundary palls into almost insignificance in comparison with the vast differences between the geological record of the late Archean and the early Proterozoic (Windley, 1995, Condie and O’Neill, 2010, and hundreds of other publications); as Goodwin (1996) put it “The early Proterozoic era represents a major watershed in Earth’s crustal history.” Accordingly, models that do not take account of the details and implications of the structural characteristics, field relationships, lithotectonic associations, and derivative geochemical-isotopic character of accretionary orogens, compared with Wilson Cycle orogens, only provide a partial analysis of the available data from the preserved record, and must be abjured. These relations and their consequences will become apparent in the detailed documentation below.

While applying a uniformitarian view in the interpretation of early Archean rock sequences, or applying the null hypothesis in which it is assumed that plate tectonics operated until it can be “proven guilty through negligence” and shown to not have operated (e.g., Kusky et al., 2018, Harrison, 2020), one must also be astutely aware of the effects of secular cooling of the planet's interior and changes in atmospheric chemistry on the constituents and composition of rocks from ancient terranes (Wang et al., 2020). While some models (e.g., Dhuime et al., 2015, Tang et al., 2016, Chen et al., 2019) have assumed a significantly more mafic early continental crust, and others have assumed from that, that plate tectonic processes were not operating early on, Keller and Harrison (2020) recently demonstrated that incompatible and compatible elements that are commonly used as proxies to estimate crustal compositions have significantly different concentrations and relationships, at constant silica values, for progressively higher mantle temperatures with time. Thus, studies that use trace element ratios as proxies to infer drastic changes in composition of continental crust with time (e.g., Dhuime et al., 2015) are “significantly undermined” (Keller and Harrison, 2020) by not accounting for secular changes in mantle temperature, because the same results can be produced at higher temperatures but with constant Si contents. At the same time, there remains a significant body of evidence that the early crust was slightly more enriched in mafic elements (Mg, Fe) and Cr, and that these have gradually evolved through time to their current values, whereas K2O has gradually increased with time (Condie, 1993). Other secular changes on the planet including atmospheric oxidation and biological activity also have significant feedback effects that can affect estimates of early crustal composition and volume (Harrison, 2020, Guo and Korenaga, 2020).

The Earth’s continents contain the fragmentary record of its tectonic evolution and crustal growth (not the same) from 4.0 Ga to the Present, the first third of which constitutes the Archean Eon that extends from ca. 4.0 Ga to 2.5 Ga, preceded by the Hadean, for which there is no preserved rock record (Harrison, 2020). Multi-disciplinary estimates suggest that ca. 85% of the present-day volume of continental crust had formed by the end of the Archean at an average growth rate of 11.7 Pg/a (Dewey and Windley, 1981) or 65% by 3.0 Ga (Cawood et al., 2013), which is comparable to that of the present-day crust (Scholl and von Huene, 2009). When considering models of crustal growth and composition however, we must consider the possibility of selective preservation of terranes that have compositions that would resist deformation, whereas those prone to deformation such as having high concentrations of heat-producing elements or being located along margins that could be subducted, or on the upper plates of collisions, may be preferentially under-represented (Morgan, 1985, Morgan, 1989, Harrison, 2009, Kusky et al., 2018, Hildebrand et al., 2018).

With the above in mind, we note that the Archean continental crust preserves segments of 1: upper crust represented by volcanic-dominated, low-grade, greenstone-granite belts (Williams et al., 1991, Polat and Kerrich, 2001a, Kusky and Polat, 1999, Percival et al., 2012), which show evidence of deposition at or near the ambient sea-level in the Neoarchean of the Superior and Slave Provinces in Canada (Percival and Helmstaedt, 2004), but at abyssal ocean depths (2.5–3.8 km) in others such as the Paleoarchean of the Eastern Pilbara (Kitajima et al., 2001a), and 1.7–4 km in the Paleoarchean (3.56–3.33 Ga) Barberton belt of South Africa (de Wit et al., 2011), and 2: lower crust represented by TTG-dominated granulite-gneiss belts that were metamorphosed in the high amphibolite-granulite facies, as in West Greenland (Nutman et al., 2004a, Windley and Garde, 2009), the North China Craton (Kusky et al., 2016), and the Barberton Greenstone Belt (Furnes et al., 2013, de Wit et al., 2018). Sedimentary rocks are present in the Eoarchean belts but form a subordinate component, including generally chert, banded iron formation (BIF), sandstones, conglomerates, and mafic shales (Eriksson et al., 2007). In the 4.0–3.2 Ga greenstone belts, the volcanic units greatly predominate in abundance, but occur with associated generally mafic volcaniclastic sediments, along with BIF's, cherts, pelites, syn-orogenic turbidites, conglomerates, sandstones, and even less-common carbonate remnants, and stromatolitic evaporites (Fedo et al., 2001). Mesoarchean-Neoarchean upper crustal greenstone belts show a greater preponderance of distal to proximal graywacke-turbidite deposits such as comprise much of the Slave Province (Kusky, 1989, Kusky, 1991, Kusky and De Paor, 1991), along with the appearance of high-gradient alluvial fan deposits, low-sinuosity river deposits, and shallow marine deposits formed near storm base (Windley, 1995, Eriksson et al., 2007, Fedo et al., 2001) with the oldest well-preserved tidalites from the 3.1 Ga Mozaan Group of the Pongola rift on the Kaavvaal Craton (Burke et al., 1985, Luskin et al., 2019). The first widespread appearance of passive margin type deposits are preserved from the Neoarchean of the Superior, Slave, Kaapvaal, and Pilbara cratons (Bradley, 2008). Some generally low-grade cratons such as the Superior Province also contain belts of high-grade gneisses as in the Pikwitonei belt and Kapuskasing structural zone (Bell, 1978, Percival et al., 2012), and conversely some high-grade Archean cratons as in West Greenland (Windley and Garde, 2009) contain strips of amphibolite facies, volcanic-dominated, upper crustal rocks such as the Eoarchean Isua belt (Komiya et al., 2004, Nutman and Friend, 2009), and the Mesoarchean Ivisaartoq belt (Polat et al., 2007), the Qussuk, Bjorneøen and Storø belts (Szilas et al., 2016), and the Tartoq belt (Kisters et al., 2012, Polat et al., 2016; Szilas et al., 2013a). These dual relations are important because they show us how the upper and lower crust were intercalated by sub-horizontal thrusting, and also how both crustal levels contain the same diagnostic lithologies, although in vastly different quantities.

A distinctive lithology in many Archean granulite-gneiss belts, as in West Greenland (e.g. Windley and Garde, 2009), South India (Subramaniam, 1956), South Canada (Bell, 1978), West Australia (Myers, 1988), and the Superior Province (Fig. 1), occurs in strips up to 2 km wide of layered and cumulate megacrystic calcic anorthosite-leucogabbro-gabbro-ultramafic complexes, which are not considered in most recent Archean crustal-tectonic syntheses (e.g. Cawood et al., 2013, Cawood et al., 2018); and yet they are considerable in volume and importance, amounting for example to at least 500 km strike-length in West Greenland. Eoarchean to Neoarchean anorthosites share the geochemical and petrological characteristics of Tethyan ophiolite-hosted anorthosites, suggesting that Archean anorthosite-bearing layered intrusions largely formed at intra-oceanic convergent plate margins (Sotiriou and Polat, 2020).

Section snippets

Models of Archean tectonics

There are many models to explain the formation of Early Archean tectonic belts (Fig. 1), which we will discuss later, but here we outline the two most important trends in thought:

Nulliak, Labrador, Canada

Following early work organised by Ken Collerson on the early Archean rocks in the Saglek area of northern Labrador (Bridgwater et al., 1975, Collerson et al., 1976, Komiya et al., 2015; Komiya et al., 2017) made a very detailed and comprehensive study, in five different areas, of the Nulliak volcano-sedimentary belt (Fig. 1, Fig. 2). Protoliths of the Uivak orthogneisses, which have U-Pb zircon ages of >3.95 Ga (Shimojo et al., 2013, Shimojo et al., 2016), were intruded into and

Components and rock associations

Here we describe the most important components that make-up Eoarchean orogens in terms of understanding their origin, and we discuss whether the components and structures are consistent with a plate tectonic origin, or require the operation of some alternative mechanism of planetary heat loss in the Eoarchean.

Interpretation of the trace element compositions in Archean igneous rocks

The geochemical characteristics of Eoarchean to Neoarchean volcanic and intrusive rocks and their geodynamic implications were discussed in detail in many recent studies (Dilek and Furnes, 2011, Polat et al., 2011a, Arndt, 2013, Turner et al., 2014, Furnes et al., 2014, Blichert-Toft et al., 2015, Wang et al., 2015, Grosch and Slama, 2017, Hildebrand et al., 2018; Hastie and Fitton, 2019; Nakamura et al., 2020, Van de Löcht et al., 2020, Sobolev et al., 2019, Lowrey et al., 2019, Gamal El Dien

Archean orogens and modern analogues

If one is going to consider tectonic development in terms of ancient and modern analogues, it is essential to have a clear view of the variability of modern tectonic settings, and which are most applicable to serve as analogues of possible Archean tectonic environments. This of course must be done with an appreciation of any changes caused by the secular cooling of the planet, and other changes in rock types influenced by changes in biology, oceanic and atmospheric chemistry, proximity of the

Conclusions

The period of transition in the geochemical record of geodynamic growth of continental crust and mantle at ca. 3.2–3.0 Ga (Fig. 18) was the result of the change from the dominant melting of the mantle source of juvenile oceanic and arc magmas from ca. >4.0 to 3.3 Ga to the rise of melting of enriched mantle wedges and of increasing active continental margins from ca. 3.2 to 3.0 Ga, which continued to well beyond the end of the Archean. As Halla et al. (2017) pointed out, these changes were

Funding

This work was supported by the National Natural Science Foundation of China (Grant Numbers: 41890834, 91755213, 41961144020, and 41602234), Chinese Ministry of Education (BP0719022), the Chinese Academy of Sciences (QYZDY-SSWDQC017), the MOST Special Fund (MSF-GPMR02-3), and the Open Fund (GPMR201704) of the State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences (Wuhan), and the Fundamental Research Fund (CUGL 180406) from the China University of

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We wish to express our thanks to our many colleagues with whom, over the years, we have spent many months in the field mapping different Archean terranes, some of which are discussed in this manuscript. This paper is dedicated to the memory of Alfred Kröner, whose seminal contributions to the field of Archean geology and geochronology paved the way for many advances, debates, and theories on the development of the early Earth. Without his high-precision zircon geochronology from numerous

References (468)

  • D. Bercovici et al.

    The generation of plate tectonics from mantle dynamics

  • R. Bolhar et al.

    Characterisation of early Archean chemical sediments by trace element signatures

    Earth Plane. Sci. Lett.

    (2004)
  • R. Bolhar et al.

    A trace element study of siderite-jasper banded iron formation in the 3.45 Ga Warrawoona Group, Pilbara Craton – formation from hydrothermal fluids and shallow seawater

    Precambr. Res.

    (2005)
  • H. Bouhallier et al.

    Diapirism, bulk homogeneous shortening and transcurrent shearing in the Archaean Dharwar craton: the Holenarsipur area, southern India

    Precambr. Res.

    (1993)
  • H. Bouhallier et al.

    Strain patterns in Archaean dome-and-basin structures: the Dharwar Craton (Karnataka, South India)

    Earth Planet. Sci. Lett.

    (1995)
  • D.C. Bradley

    Passive margins through Earth history

    Earth Sci. Rev.

    (2008)
  • D.C. Bradley

    Secular trends in the geological record and the supercontinent cycle

    Earth Sci. Rev.

    (2011)
  • N.L. Cates et al.

    Pre-3750 Ma supracrustal rocks from the Nuvvuagittuq supracrustal belt, northern Québec

    Earth Planet. Sci. Lett.

    (2007)
  • D. Chardon et al.

    Strain patterns, décollemont and incipient sagducted greenstone terrains in the Archaean Dharwar craton (South India)

    J. Struct. Geol.

    (1996)
  • D. Chardon et al.

    Sinking of the Dharwar basin (South India): Implications for Archaean tectonics

    Precambr. Res.

    (1998)
  • D. Chardon et al.

    Flow of hot-orogens: a review from the Precambrian, clues for the Phanerozoic

    Tectonophysics

    (2009)
  • P. Chowdhury et al.

    Peel-back controlled lithospheric convergence explains the secular transitions in Archean metamorphism and magmatism

    Earth Planet. Sci. Lett.

    (2020)
  • K.C. Condie

    Chemical composition and evolution of the upper continental crust: Contrasting results from surface samples and shales

    Chem. Geol.

    (1993)
  • G.F. Davies

    Gravitational depletion of the early Earth’s upper mantle and the viability of early plate tectonics

    Earth Planet. Sci. Lett.

    (2006)
  • H. Deng et al.

    Geochemistry of Neoarchean mafic volcanic rocks and late mafic dikes in the Zanhuang Complex, Central Orogenic Belt, North China Craton: Implications for geodynamic setting

    Lithos

    (2013)
  • M.J. de Wit et al.

    Earth’s earliest continental lithosphere, hydrothermal flux and crustal recycling

    Lithos

    (1993)
  • M.J. de Wit et al.

    Geology and tectonostratigraphy of the Onverwacht Suite, Barberton greenstone belt, South Africa

    Precambr. Res.

    (2011)
  • M. de Wit et al.

    Paleoarchean bedrock lithologies across the Makhonjwa Mountains of South Africa and Swaziland linked to geochemical, magnetic and tectonic data reveal early plate tectonic genes flanking subduction margins

    Geosci. Front.

    (2018)
  • C.V. Dharma Rao et al.

    Chromite-silicate chemistry of the Neoarchean Sittampundi Complex, southern India: implications for subduction-related arc magmatism

    Precambr. Res.

    (2013)
  • N. Dygert et al.

    Plate tectonic cycling modulates Earth’s 3He/22Ne ratio

    Earth Planet. Sci. Lett.

    (2018)
  • L.J. Duck et al.

    Microbial remains and other carbonaceous forms from the 3.24 Ga Sulphur Springs black smoker deposit, Western Australia

    Precambr. Res.

    (2007)
  • C.M. Fedo et al.

    Depositional setting and paleogeographic implications of Earth's oldest supracrustal rocks, the > 3.7 Ga Isua greenstone belt, West Greenland

    Sed. Geol.

    (2001)
  • N. Flament et al.

    A case for late-Archean continental emergence from thermal evolution model and hypsometry

    Earth Planet. Sci. Lett.

    (2008)
  • I.R. Fletcher et al.

    Sm-Nd, Pb-Pb and Rb-Sr geochronology of the Manfred Complex, Mount Narryer, Western Australia

    Precambr. Res.

    (1988)
  • J. Adam et al.

    Hadean greenstones from the Nuvvuagittuq fold belt and the origin of the Earth’s early continental crust

    Geology

    (2012)
  • S. Agrawal et al.

    Tectonic discrimination of basic and ultrabasic volcanic rocks through log-transformed ratios of immobile trace elements

    Int. Geol. Rev.

    (2008)
  • R. Agrusta et al.

    Strong plates enhance mantle mixing in early Earth

    Nat. Commun.

    (2018)
  • Alt, J.C., 1995. Subseafloor processes in mid-ocean ridge hydrothermal systems. In: Humphris, S., Lupton, J.,...
  • J.C. Alt et al.

    Alteration and mineralization of an oceanic forearc and the ophiolite-ocean crust analogy

    J. Geophys. Res.

    (1998)
  • L. André et al.

    Early continental crust generated by reworking of basalts variably enriched by seawater

    Nat. Geosci.

    (2019)
  • C.C. Appel et al.

    Complex chromite textures reveal the history of an early Archaean layered ultramafic body in West Greenland

    Mineral. Mag.

    (2002)
  • N.T. Arndt

    The formation and evolution of the continental crust

    Geochem. Perspect.

    (2013)
  • N.T. Arndt et al.
  • N.T. Arndt et al.

    Komatiites

    (1982)
  • L.D. Ashwal
  • J.M. Barton et al.

    The geology, age, and tectonic setting of the Messina Layered Intrusion, Limpopo mobile belt, southern Africa

    Am. J. Sci.

    (1979)
  • A.M. Bauer et al.

    Hafnium isotopes in zircons document the gradual onset of mobile-lid tectonics

    Geophys Perspect Lett

    (2020)
  • A.P. Beall et al.

    Formation of cratonic lithosphere duing the initiation of plate tectonics

    Geology

    (2018)
  • J.S. Beard

    Characteristic mineralogy of arc-related cumulate gabbros: implications for the tectonic setting of gabbroic plutons and for andesite genesis

    Geology

    (1986)
  • C.K. Bell

    Geology, Wekusko Lake map-area, Manitoba

    Geol. Surv. Canada Memoir

    (1978)
  • Cited by (156)

    View all citing articles on Scopus
    View full text