Geoscience Frontiers 8 (2017) 337e346

H O S T E D BY Contents lists available at ScienceDirect

China University of Geosciences (Beijing)

Geoscience Frontiers
journal homepage: www.elsevier.com/locate/gsf

Research paper

Why primordial continents were recycled to the deep: Role of

subduction erosion
S. Azuma a, *, S. Yamamoto b, H. Ichikawa a, c, S. Maruyama a
a
EartheLife Science Institute, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8550, Japan
b
Graduate School of Environment and Information Sciences, Yokohama National University, 79-7 Tokiwadai, Hodogaya-ku, Yokohama 240-8501, Japan
c
Geodynamics Research Center, Ehime University, 2-5 Bunkyo-cho, Matsuyama 790-8577, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Geological observations indicate that there are only a few rocks of Archean Earth and no Hadean rocks on
Received 7 April 2016 the surface of the present-day Earth. From these facts, many scientists believe that the primordial
Received in revised form continents never existed during Hadean Earth, and the continental volume has kept increasing. On the
4 August 2016
other hand, recent studies reported the importance of the primordial continents on the origin of life,
Accepted 10 August 2016
Available online 28 August 2016
implying their existence. In this paper, we discussed the possible process that could explain the loss of
the primordial continents with the assumption that they existed in the Hadean. Although depending on
the timing of the initiation of plate tectonics and its convection style, subduction erosion, which is
Keywords:
Subduction erosion
observed on the present-day Earth, might have carried the primordial continents into the deep mantle.
Hadean Ó 2016, China University of Geosciences (Beijing) and Peking University. Production and hosting by
Plate tectonics Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
Archean licenses/by-nc-nd/4.0/).

1. Introduction
Geological observations indicate that rocks in the Hadean
(4.0e4.6 Ga) do not remain on the current Earth’s surface (Fig. 1).
From these observations, many scientists believe that primordial
continents never existed on Earth (e.g., Condie, 1998; Rino et al.,
2008). In principle, however, there are theoretically two possibil-
ities: (1) the primordial continents once existed, and some
geological processes removed them from the Earth’s surface into
the deep interior, and (2) the primordial continents never existed.
In this paper, we focus on the process of continentalecrust loss and
consider the former possibility. From these discussions, we suggest
new insight about the Hadean environment.
The existence of the primordial continents is considered here to
be the key concept of a habitable planet. The general concept of a
habitable planet has primarily focused on the existence of liquid
water, which basically depends on the distance from the stars (sun)
(Kasting and Calting, 2003; Ikoma and Genda, 2006). Although this
concept is significant, life is not originated only by liquid water. The
origin of life requires not only liquid water, but also plate tectonics
and nutrient-enriched continental crust exposed above the ocean
* Corresponding author.
E-mail address: shintaro.azuma@elsi.jp (S. Azuma).
Peer-review under responsibility of China University of Geosciences (Beijing).
(e.g., Dohm and Maruyama, 2015). Plate tectonics efficiently cools
the mantle and core by the subducted cold slab, helping the con-
vection of the outer core which produces the magnetic field
resulting in protection of life from cosmic rays and outer high-
energy radiation. Plate tectonics also controls the carbon cycle at
plate boundaries through the subduction and eruption. On the
other hand, there is high potential that the existence of primordial
continent is directly related to the origin of life (described below).
In other words, the existence of the primordial continents may be
one of the constraints to find planets with life.
2. Did primordial continents exist in Hadean?
For the origin of life, the existence of continents is quite
important because many essential elements for life (phosphorus,
potassium, calcium, iron, and others) are supplied only from the
following three-kinds of continental rocks: (1) granite, (2) pri-
mordial continents (anorthosite with KREEP basalts, as well as dike
and gabbro as lower mafic crust), and (3) carbonatite (Maruyama
et al., 2013) (Fig. 2). In particular, the continental crust above the
surface of the ocean plays an important role. Weathering and
erosion transforms large rocks into small grains, and these small
grains can efficiently transport the essential elements into the
ocean and lake where life supposedly originated (e.g., Dohm and
Maruyama., 2015) (Fig. 2). On the other hand, the timing of the

http://dx.doi.org/10.1016/j.gsf.2016.08.001
1674-9871/Ó 2016, China University of Geosciences (Beijing) and Peking University. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-
ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
338 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346

Figure 1. Global geological map (after Rino et al., 2008).

Weathering
Vaporization Pure water

Lake
P Periodical Weathering
P K
P spring
K K
P
K
PK
P
K

Ocean Continental crust P

(Granite, anorthosite with KREEP basalts) K

Oceanic plate

Subduction erosion

Figure 2. The roll of continents and essential elements for life. The essential elements such as potassium and phosphorus are supplied only from continental crust. Weathering and
erosion transform continental crust into small grains which supply the essential elements for life efficiently into the ocean and lakes. The concept suggested to be important for life,
which involves the interaction among nutrient-enriched landmass, ocean, and atmosphere through Sun-driven hydrological cycling, is referred to as Habitable Trinity (Dohm and
Maruyama, 2015).
 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346
birth of life is still controversial. It is suggested from the
carboneisotope composition of the oldest known sediments in Isua
supracrustal belt in west Greenland that life appeared on the Earth
at least before 3.8 Ga (Mojzsis et al., 1996). If this suggestion is
correct, the primordial continents that supplied the essential ele-
ments to life must have existed in the Hadean. However, the re-
mains of a primordial continent have not been found on the surface
of the present-day Earth (Fig. 1). The oldest dated rock is 4.03 Ga
from the Acasta Gneiss Complex of the Slave Craton, northwestern
Canada (Bowring and Williams, 1999), and the oldest supracrustal
rock is reported from the 3.83 Ga Akilia island supracrustal rock in
the Itsaq Gneiss Complex, southern West Greenland (Mojzsis and
Harrison, 2002). Hadean information has been gained only from
zircons in some Proterozoic sedimentary rocks (4.0e4.4 Ga) (e.g.,
Wilde et al., 2001; Iizuka et al., 2007). Many scientists have
researched the geochemical characteristics of the Hadean zircons
(i.e. REE pattern, thermometry, mineral inclusion) to gain Hadean
environmental information (e.g., Maas et al., 1992; Amelin et al.,
1999; Watson and Harrison, 2005). These Hadean zircons suggest
the derivation from TTG magma, providing the evidence for plate
tectonics and an ocean in Hadean (e.g., Harrison et al., 2007;
Hopkins et al., 2008). Plate tectonics with a large volume of water
would induce remelting of oceanic crust and produce the conti-
nental crust (e.g., Campbell and Taylor, 1983). Then, why have Ha-
dean rocks not been discovered? In order to answer this question,
the process of removing continents from the Earth’s surface must
be discussed in addition to the process of increasing continents.
3. Process of continental loss
There are mainly two processes of removing continental crusts.
First, sediments, which are supplied into the ocean by weathering
Kamchatka
Nankai
Japan
Mediterranean Ryukyu
Makran Manila
Philippines
Andaman
Smatra Java
Figure 3. Map showing the type of plate margins in the world, which includes the distribution of erosive (6) and accretionary (:) margins in subduction zones (modified from Von
Huene and Scholl, 1993; Clift and Vannucchi, 2004; Scholl and Von Huene, 2009).
339
and erosion, subduct along with oceanic plate to the mantle (sedi-
ment subduction). Clift and Vannucchi (2004) reported that only
17% of sediments at an oceanic trench is incorporated into the
accretionary prism, and the residual w80% of sediments is carried
into the deep mantle by sediments subduction. Second, the sub-
ducted oceanic plate erodes the overlying continental crust and
carries the crustal material into the mantle (subduction erosion).
Subduction erosion has been confirmed from deep-sea drilling
survey. For instance, Cretaceous granite and accretionary prism are
exposed at the fore-arc margin of East Japan (Von Huene et al., 1978).
The landward slope of Peru trench is composed of Cretaceous gneiss
rock and the accretionary prism does not develop (e.g., Hussong and
Wipperman, 1981). Moreover, analysis of foraminifer in the sedi-
ments that cover the fore-arc margin indicates that the fore-arc
margin subsides (Nasu et al., 1980), suggesting that the subducted
plate erodes the lower base of the continental crust and carries the
crustal material into the deep mantle (subduction erosion). In
particular, basal erosion strongly contributes to the subsidence of
fore-arc margin. If both sediment subduction and subduction
erosion have occurred throughout the Earth’s history, the idea that
the volume of continents steadily has increased through time is
being reconsidered (Armstrong, 1981; Von Huene and Scholl, 1991).
3.1. Characteristics of subduction erosion
The present Earth’s surface is divided into several plates. These
plates collide at the plate margins and one side of the plates
subducts into the mantle (plate tectonics). Subduction erosion
occurs at the plate margins where a plate subducts. Accretionary
prisms may be also formed at subduction zones instead of sub-
duction erosion. In present Earth, subduction erosion is dominant
at 56e75% of the subduction zones (Von Huene and Scholl, 1993;
NE Alaska
SW Alaska Cascadia
Aleutians
Kurile
S.Oregon
Lesser Antilles
(Barbados)
Izu-Mariana Mexico
Guatemara
Costa Rica
New Hebrides
Colombia
Ecuador
Tonga
Peru
Kermadec N.Chile
Hikurangi
S.Chile
MacQuarie
South Sandwitch
340 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346

Clift and Vannucchi, 2004; Scholl and Von Huene, 2007; Clift
et al., 2009; Scholl and Von Huene, 2009; Yamamoto, 2010;
Stern, 2011) (Fig. 3). At the subducting plate margin, subduction
erosion occurs more universally than the formation of accre-
tionary prism.
Then what is the factor to determine whether subduction
erosion or the formation of an accretion wedge is the dominant
process? Previous study has indicated that the most important
120
Erosional margins
Orthogonal convergence
factor is the thickness of sediments (e.g., Clift and Vannucchi,
2004). Fig. 4 shows the convergence rate as a function of thick-
ness of sediments. Open and closed triangles indicate the erosional
margins and accretionary margins, respectively. It is obvious from
Fig. 4 that subduction erosion can be dominant at plate margins
where the convergence rate is large (>50 km/my) and the thickness
of sediments is small (<1 km). Thin sediments cannot fully cover
the roughness (irregular topography including both concavities and
convexities) of an oceanic plate, and therefore the fore-arc wedges
can be eroded by the rough surface. This suggests that the surface
roughness of the subducting plate is important for subduction
erosion.

100
Accretionary margins 3.2. Mechanism of subduction erosion

Three models have been proposed to explain the mechanisms of

rate (km/my)

80
subduction erosion: (1) horst and graben model (Hilde, 1983); (2)
model involving subducted seamounts and ridges (Von Huene and
Lallemand, 1990); and (3) hydrofracture model (Von Huene and
60
Culotta, 1989; Von Huene et al., 2004) (Fig. 5). These models,
which are discussed in the following, are advocated from geological
40 survey, with high potential that all three models can be simulta-
neously applied to subduction zones.
In the horst and graben model, the sediments that are supplied
20 from continents are trapped in the graben of the subducted slabs
and carried into the deep mantle (Hilde, 1983). On the other hand,
the horst of the subducted slab erodes the basal crust, and the
0 eroded crustal materials are also trapped in graben (basal erosion)
0 1 2 3 4 5 6 7 8 9 (Tsuru, 2004) (Fig. 5). However, if the volume of sediments exceeds
the capacity of the graben, the irregularity of the surface of the
Sediment thickness (km) oceanic plate is completely filled, and the sediments are accreted to
Figure 4. Orthogonal convergence rate as a function of sediment thickness (after Clift
continents (forming an accretionary prism). This model can well
and Vannucchi, 2004) for erosional (open triangles) and accretionary (closed triangles) explain the relationship between the thickness of sediments and
margins. erosional margins (or accretionary margins).

Mechanism of subduction erosion in Subduction zone Sediment subduction

Weathering

Subsidence Oceanic crust

Upper crust
ion
n eros
d uctio Mantle lithosphere
Sub

Lower crust
ust ure
n ic cr p ress
Oce
a Pore
al
Norltm
unt en
a mo t Grab fau
Se Hors

Erosion by seamounts Bucket model Hydrofracture model

and ridges
Figure 5. Mechanisms-models of subduction erosion suggested in previous studies (Hilde, 1983; Von Huene and Culotta, 1989; Von Huene and Lallemand, 1990).
 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346
Subduction erosion by seamounts and ridges of the second
model is considered as a main mechanism to erode the massive
fore-arc crust (Ranero and Von Huene, 2000). This is supported by
the observation of the subduction of seamounts at the Middle
America Trench and PerueChile Trench by seismological tech-
niques. This model can explain the “intermittent” basal erosion and
subsidence of a fore-arc region. However, the subducted seamounts
and ridges may not be able to induce the “continual” basal erosion
because of their intermittent subduction.
To explain the continual subsidence of a fore-arc region (basal
erosion), a hydrofracture model was proposed, in which case the
dehydrated fluids from the subducted slab fracture the overlying
crust (Von Huene and Culotta, 1989; Von Huene et al., 2004). In this
model, the subducted oceanic crust and sediments dehydrate in
high temperature and pressure conditions, and therefore high pore
pressure is produced. The high pore pressure makes the basal
portion of continental crusts weak; this includes the promotion of
fracturing as indicated both theoretically and experimentally (e.g.,
Paterson and Wong, 2005). Then the fractured crust is carried into
the mantle with the oceanic plate. The high pore pressure in-
fluences the rock strength chemically and physically. In particular,
as a physical effect, the pore pressure Pp decreases the effective
stress Pe, and the rock strength s decreases as follows (Eq. 1):


s ¼ s0 þ fc Pe ¼ s0 þ fc Pc  aPp
where fc is the frictional coefficient, Pc is confining pressure, and s0
and a are the constants. This decrease of the effective stress can
promote the basal erosion because friction and fracture strength
depends on the effective stress (Eq. 1). Actually the progress of basal
Figure 6. The timing of initiation of plate tectonics and the suggested evidences for operating plate tectonics. References cited are: Farquhar et al. (2002), Cawood et al. (2006),
Nutman et al. (2002), Furnes et al. (2007), Komiya et al. (2015), Hopkins et al. (2008), Wilde et al. (2001), Harrison et al. (2005), Hamilton (2011), Stern (2005).
341
erosion has been observed from seismic reflection profiles at the
Middle America Trench, Costa Rica, and Ecuador Trench (Ranero
and Von Huene, 2000; Zhu et al., 2009). Vannucchi et al. (2008)
found a subduction channel, which was a pathway of the frac-
tured rocks and sediments into the mantle (Cloos and Shreve,
1988a, b; Ichikawa et al., 2013, 2015), at the past plate boundary
in the Northern Apennines, Italy, suggesting that subduction
erosion had occurred in the past.
The review and consideration of these mechanisms of subduc-
tion erosion tell us that the subduction erosion must have occurred
throughout the Earth’s history. If plate tectonics works, subduction
erosion is a universal phenomenon. In particular, the basal erosion
can occur even in accretionary margin. Therefore, the timing of the
initiation of plate tectonics is a key to understand the past sub-
duction erosion and the existence of primordial continents.
4. Plate tectonics in early Earth
4.1. When did plate tectonics begin?
Based on various geological evidences, the timing of the
beginning of plate tectonics has been discussed, while remaining
controversial. The geological evidences that have been used as an
(1) indicator of plate tectonics are ophiolite, duplex structure, ultra
high pressure (UHP) metamorphism including coesite and dia-
mond, geochemical analysis of zircons, and others (e.g., Wilde
et al., 2001; Harrison et al., 2005; Stern, 2005; Furnes et al.,
2007; Hopkins et al., 2008; Hamilton, 2011; Komiya et al., 2015)
(Fig. 6). For instance, it has been suggested from the blueschists
342 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346

and UHP metamorphism that plate tectonics originated at 1.0 Ga,
because they are absent before 1.0 Ga (Stern, 2005). However,
blueschists and UHP metamorphism may not be strong evidence
for plate tectonics because the style of plate tectonics was possibly
different between the early and present-day Earth. In addition,
slab melting had occurred at a relatively shallow portion because
the potential temperature of early Earth was higher than that of
present-day. Even if blueschist-facies metamorphic conditions are
attainable in subduction zones of early Earth, blueschists could not
be formed from hot, early Earth’s mantle (Palin and White, 2016).
Palin and White (2016) indicated that oceanic crust formed from
the hotter mantle would be rich in magnesium oxide, therefore
greenschists instead of blueschists would have been formed from
MgO-rich oceanic crust. On the other hand, Furnes et al. (2007)
reported the evidence of oceanic crustal accretion in the
w3.80 Ga Isua Supracrustal Belt (ISB) in southwest Greenland and
contended that plate tectonics was operating at least before
3.80 Ga. This ancient phase of plate tectonics is supported by the
discovery of oceanic plate stratigraphy (OPS) and duplex struc-
tures within the Saglek Block in northern Labrador, Canada by
Komiya et al. (2015), which indicates that plate tectonics had been
in operation before 3.95 Ga. The strongest and oldest evidence for
the operation of plate tectonics is Hadean zircons. Mineral in-
clusions and oxygen isotopes in these Hadean zircons suggest their
derivation from TTG magma, indicating that plate tectonics began
at 4.4 Ga (e.g., Wilde et al., 2001). These previous studies suggested
that some kind of recycling process of crustal material had worked
in the Hadean, and that plate subduction, which is similar to plate
tectonics of the present-day Earth, had existed at least before
3.95 Ga. The operation of subduction erosion strongly depended
on the style of plate tectonics in the past Earth. For instance, drip
tectonics (e.g., Foley et al., 2014) could cause the subduction of
crustal material, however subduction erosion would not occur in
this style of tectonics.
Neoproterozoic blueschists indicates higher global rates of subduc-
tion erosion in the remote past because of higher plate convergence
rates. On the other hands, some scientists have suggested that the
plate tectonics in the early Earth is characterized by a low plate
velocity compared with that of the present-day plate tectonics (e.g.,
Korenaga, 2013). However, even if plate velocity in the past Earth
was lower compared with that of the present-day Earth, the dif-
ference between them is not expected to be so large. The average
velocity of plates during the early Earth has been estimated to be
1.75 cm/yr (Foley et al., 2014). In addition, plate size is considered to
be smaller than that of present-day. The potential temperature of
the ancient mantle was higher than today (Komiya, 2004), and
therefore, a higher Rayleigh number, resulting in small convection
cells and many plate boundaries (e.g., Yanagisawa and Yamagishi,
2005). Thus, the Archean Earth is inferred to have been covered by
several hundreds of plates instead of dozens of plates of present-day
Earth. The several hundreds of plates would have resulted in more
active subduction erosion, which includes more active subduction of
island arcs (e.g., Yamamoto et al., 2009a, b).
The number and size of plate boundaries also depends on the
plate strength (e.g., Tackley, 1998, 2000). If the plate strength of the
past Earth was smaller than that of the present-day Earth, it would
have been easier to develop the plate boundaries. The plate
strength is strongly dependent on the thermal structure of the
plate. We have calculated the thermal structure of the oceanic plate
at early Archean using a model of transient half-space cooling by
Turcotte and Schubert (2002) at given mantle and surface tem-
perature, respectively 1500  C (Komiya, 2004) and 100  C (Komiya
et al., 1999) (Fig. 7b). The corresponding temperatures for present-
1600
Present day (a)
1400
Temperature (ºC)

1200
4.2. Convection style in the early Earth
1000
Previous studies have pointed out that the styles of plate tec-
800
tonics are different between the past (HadeaneArchean) and
10 Myr
present-day Earth (e.g., van Hunen and van den Berg, 2008; 600 20 Myr
Korenaga, 2011; Foley et al., 2014). This difference could be
caused by higher potential temperature of the past Earth (e.g., 400 40 Myr
Komiya, 2004), which produced the thicker crust and slab melting. 60 Myr
200
Recent numerical calculation incorporating the grain-damage, a
grain size-feedback mechanism (Foley et al., 2014) showed that the 0
style of subduction process in the Hadean was similar to the drip
0 50 100 150 200
tectonics, not to the plate tectonics of the present-day Earth. It is Depth (km)
inferred that the subduction of supracrustal material was initiated 1600
by drip tectonics w100 Myr later from the solidification of magma Archean (b)
ocean. This interpretation is consistent with the recycle of crust in 1400
Temperature (ºC)

Hadean (e.g., Watson and Harrison, 2005) supposed by the analysis

of Hadean zircons. On the other hand, Komiya et al. (2015) found 1200
that the accretionary prism and the oceanic plate stratigraphy 1000
(OPS) had already formed at 3.95 Ga, suggesting that the style of
subduction at that time was similar to that of the present Earth. If 800
the proposed drip tectonics was dominant just after the solidifi- 10 Myr
600 20 Myr
cation of magma ocean in the early Earth, the subduction style
should have changed in the HadeaneEoarchean to the style similar 400 40 Myr
to that of present-day Earth. In other words, the 60 Myr
HadeaneEoarchean Earth satisfied the necessary conditions for 200
subduction erosion. 0
The plate velocity during HadeaneArchean is expected to have 0 50 100 150 200
been larger than that of the present-day Earth because of higher Depth (km)
potential mantle temperature (e.g., Komiya, 2004). For instance, Figure 7. Calculated thermal structures of oceanic plate in (a) the present-day Earth
Stern (2011) suggested that the paucity of ancient pre- and (b) the past (Archean).
 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346 343

Present day Present day

Moho Moho Mo o

OH =500 ppm H/Si) =0.38, COH=500 ppm H/Si) (λ=0.38, COH=500 ppm H/Si)

Figure 8. Rheological structures of the oceanic plate of the present-day Earth (10, 20, 40 Ma).

Archean

Moho Mo o Mo o

Dry
Wet
=0.38, COH=500 ppm H/Si) (λ (λ=0.38, COH=500 ppm H/Si)

Figure 9. Rheological structures of oceanic plate of the early Archean Earth (10, 20, 40 Ma).

day Earth are 1330  C and 0  C (Iwamori et al., 1995; Komiya et al.,
1999). Based on calculated thermal structures, we determined the
rheological structure of the past and present-day Earth. In the
calculation of the rheological structure, pressures are assumed to
be equal to the lithostatic pressure determined from a crustal
density of 2900 kg/m3 and a mantle density of 3300 kg/m3. The
crustal thickness is assumed to be 7 km for present-day Earth and
20 km for Archean Earth (Komiya, 2004). Whereas the rock
strength in the brittle-deformation region is calculated from Bye-
rlee’s law (Byerlee, 1978), the rock strength in the ductile-
deformation region is determined by a flow law of each material.
The flow laws of plagioclase are applied for the crustal strength
(Rybacki and Dresen, 2000), and the rock strength of the mantle is
determined from the flow law of olivine (Karato and Jung, 2003).
Power-law creep, in which the strain rate is proportional to a power
of stress, is generally used to infer rheological structure (e.g.,
Kohlstedt et al., 1995); this type of flow law is commonly applied to
high-temperature creep. However, the Peierls mechanism becomes
dominant at low temperatures and high stresses (Tsenn and Carter,
1987; Katayama and Karato, 2008; Demouchy et al., 2013). In this
mechanism, the strain rate is exponentially proportional to the
applied stress. In our calculations, the Peierls mechanism is also
applied. The strain rate is assumed to be 1015/s. Figs. 8 and 9 show
the calculated rheological structures of the oceanic plate of the
present-day Earth and the Archean Earth, respectively. The devel-
opment of plate boundaries requires a moderate strength of the
oceanic plate (<200e300 MPa) (e.g., Tackley, 1998, 2000). The plate
strength under dry conditions far exceeds the threshold in both
Archean and present-day Earth (Figs. 8 and 9), suggesting that the
existence of an ocean is the minimum necessary condition for plate
344 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346
8E+09
7E+09
Volume of lost continent (km3)
6E+09
5E+09
4E+09
3E+09
2E+09
1E+09
0 Africa
0 0.5
Time (Gyr)
Figure 10. The accumulated volume of crustal losses as a function of time. Difference of the decrease and increase rate in volume is assumed to be 0.5, 1.0, 3.0, and 5.0 km3/yr. The
volumes of the present-day continent are indicated as a rough standard.
tectonics and therefore subduction erosion. In addition, these fig-
ures clarify that the plate strength of Archean Earth is w70e80% of
that of the present-day Earth (The plate strength of Hadean Earth is
calculated in Maruyama et al. in this volume). This difference of the
plate strength between Archean and present-day Earth supports a
hypothesis that more plate boundaries were produced in Archean
and Hadean. The greater number of plate boundaries must have
caused extensive subduction erosion and direct subduction of
oceanic island arcs.
5. Estimated volume of continent losses in Hadean and
Archean
Although it is expected that extensive subduction erosion
occurred in the Hadean and Archean, it is difficult to quantitatively
estimate the volume of crustal losses in these eons. Some previous
studies evaluated production and loss rate of continental crust in
present-day Earth. Clift et al. (2009) and Stern and Scholl (2010)
classified geological process into additions or losses of continen-
tal crust, and evaluated the generation and loss in volume of con-
tinental crust. Clift et al. (2009) suggested a generation rate of
3.2e5.0 km3/yr and a loss rate of 4.85 km3/yr in volume of conti-
nental crust. Stern and Scholl (2010) suggested a generation rate of
3.2 km3/yr and loss rate of 3.2 km3/yr. In both evaluations, the
generation rate balances the loss rate in volume of continental
crust. However, the conclusion by Stern and Scholl (2010) is an
estimate for just Phanerozoic time, and they also pointed out that
the volume of continental crust loss due to lower crust delamina-
tion and subduction at collision zones are difficult to constrain.
Therefore, it is possible that the rate of the continental loss can far
exceed the rate of continental generation (e.g., the loss and gen-
eration rate of continental crust in the Tethyan region). If the
generation falls below the loss, how much volume of continental
crust does the Earth’s surface lose? In discussing this topic, plate
tectonics that is similar to that of the present-day Earth is assumed
to initiate in HadeaneEoarchean (Harrison et al., 2005; Furnes
World
r
3 /y
km
5
3 /yr
m
3k
Eurasia
3 yr
1 km /
0.5 km /yr
3
1.0 1.5
et al., 2007; Hopkins et al., 2008; Komiya et al., 2015). Fig. 10
shows the volume of crustal losses as a function of time. In
Fig. 10, the difference between loss and generation rate in volume of
continental crust is assumed to be 0.5e5.0 km3/yr, and the volumes
of the each present-day continent are indicated as a rough stan-
dard. In the case of crustal losses of 0.5 km3/yr, the volume of the
North America continent would have disappeared from Earth’s
surface in 1.5 billion years. In the case of 1 km3/yr, the volume equal
to Eurasia continent would have been eroded. In the case of 3 km3/
yr, all of the continents on the surface of present-day Earth would
have been lost during the Archean (w1.5 Gyr) and the volume of
the Eurasia continent would have disappeared during the Hadean
(w500 Myr). If the crustal generation rate falls below the loss rate,
all of the continental crust at Earth’s surface can be carried into the
mantle by subduction erosion.
6. Conclusion
We discussed the plate tectonics and subduction erosion in
Hadean and Archean to consider why the primordial continent
cannot be found on the present-day surface of the Earth. The
conclusions are as follows: (1) subduction erosion is a natural
consequence of plate tectonics when the thickness of sediments is
small (<1 km); (2) plate size during the Hadean and Archean was
relatively small when compared with present-day Earth, and thus
extensive subduction erosion had occurred; (3) primordial conti-
nents possibly existed and were eroded by subduction erosion. The
eroded primordial continents were carried into the deep mantle
(e.g., Kawai et al., 2009; Ichikawa et al., 2017).
Acknowledgement
We thank James Dohm for editing English in this paper. This
research was supported by JSPS KAKENHI (Grand-in-Aid for Sci-
entific Research (S)) Grant No. 23224012 (Growth of the second
continent and mantle).
 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346 345

References Kohlstedt, D.L., Evans, B., Mackwell, S.J., 1995. Strength of lithosphere: constraints
imposed by laboratory experiments. Journal of Geophysical Research 100,
17587e17602.
Amelin, Y., Lee, D.-C., Halliday, A.N., Pidgeon, R.T., 1999. Nature of the Earth’s earliest
Komiya, T., 2004. Material circulation model including chemical differentiation
crust from hafnium isotopes in single detrital zircons. Nature 399, 252e255.
within the mantle and secular variation of temperature and composition of
Armstrong, R.L., 1981. Radiogenic isotopes: the case for crustal recycling on a near-
mantle. Physics of the Earth and Planetary Interiors 146, 333e367.
steady-state no continental-growth Earth. Philosophical Transactions of the
Komiya, T., Maruyama, S., Nohda, S., Masuda, T., Hayashi, M., Okamoto, S.,
Royal Society London A 301, 443e472.
1999. Plate tectonics at 3.8 to 3.7 Ga Field evidence from the Isua accre-
Bowring, S.A., Williams, I.S., 1999. Priscoan (4.00e4.03 Ga) orthogneisses from
tionary complex, southern West Greenland. The Journal of Geology 107,
northwestern Canada. Contributions to Mineralogy and Petrology 134, 3e16.
515e554.
Byerlee, J.D., 1978. Friction of rocks. Pure and Applied Geophysics 116, 615e626.
Komiya, T., Yamamoto, S., Aoki, S., Sawaki, Y., Ishikawa, A., Tashiro, T., Koshida, K.,
Campbell, I.H., Taylor, S.R., 1983. No water, no graniteseno oceans, no continents.
Shimojo, M., Aoki, K., Collerson, K.D., 2015. Geology of the Eoarchean, >3.95 Ga,
Geophysical Research Letters 10, 1061e1064.
Nulliak supracrustal rocks in the Saglek Block, northern Labrador, Canada: the
Cawood, P.A., Kroner, A., Pisarevsky, S., 2006. Precambrian plate tectonics: criteria
oldest geological evidence for plate tectonics. Tectonophysics 662, 40e66.
and evidence. GSA Today 16, 4e11.
Korenaga, J., 2011. Thermal evolution with a hydrating mantle and the initiation of
Clift, P.D., Vannucchi, P., Morgan, J.P., 2009. Crustal redistribution, crust-mantle
plate tectonics in the early Earth. Journal of Geophysical Research 116, B12403.
recycling and Phanerozoic evolution of the continental crust. Earth-Science
Korenaga, J., 2013. Initiation and evolution of plate tectonics on Earth: Theories and
Reviews 97, 80e104.
observations. Annual Review of Earth and Planetary Sciences 41, 117e151.
Clift, P., Vannucchi, P., 2004. Controls on tectonic accretion versus erosion in sub-
Maas, R., Kinny, P.D., Williams, I.S., Froude, D.O., Compston, W., 1992. The Earth’s
duction zones: implications for the origin and recycling of the continental crust.
oldest known crust: a geochronological and geochemical study of 3900e4200
Reviews of Geophysics 42, RG2001. http://dx.doi.org/10.1029/2003RG000127.
Ma old detri- tal zircons from Mt. Narryer and Jack Hills, Western Australia.
Cloos, M., Shreve, R., 1988a. Subduction-channel model of prism accretion, melange
Geochimica et Cosmochimica Acta 56, 1281e1300.
formation, sediment subduction, and subduction erosion at convergent plate
Maruyama, S., Ikoma, M., Genda, H., Hirose, K., Yokoyama, T., Santosh, M., 2013. The
margins: 1. Background and description. Pure and Applied Geophysics 128,
naked planet Earth: most essential pre-requisite for the origin and evolution of
455e500.
life. Geoscience Frontiers 4, 141e165.
Cloos, M., Shreve, R., 1988b. Subduction-channel model of prism accretion, melange
Maruyama, S., Santosh, M., Azuma, S., 2016. Initiation of plate tectonics in Hadean:
formation, sediment subduction, and subduction erosion at convergent plate
Eclogitization triggered by ‘‘ABEL’’ bombardment. Geoscience Frontiers (this
margins: 2. Implications and discussion. Pure and Applied Geophysics 128,
volume).
501e545.
Mojzsis, S.J., Arrhenius, G., McKeegan, K.D., Harisson, T.M., Nutman, A.P.,
Condie, K.C., 1998. Episodic continental growth and supercontinents: a mantle
Friend, C.R.L., 1996. Evidence for life on Earth before 3,800 million years ago.
avalanche connection? Earth and Planetary Science Letters 163, 97e108.
Nature 384, 55e59.
Demouchy, S., Tommasi, A., Ballaran, T.B., Cordier, P., 2013. Low strength of Earth’s
Mojzsis, S.J., Harrison, T.M., 2002. Establishment of a 3.83 Ga magmatic age for the
uppermost mantle inferred from tri-axial deformation experiments on dry
Akilia tonalite (southern West Greenland). Earth and Planetary Science Letters
olivine crystals. Physics of the Earth and Planetary Interiors 220, 37e49.
202, 563e576.
Dohm, J.M., Maruyama, S., 2015. Habitable trinity. Geoscience Frontiers 6, 95e101.
Nasu, N., Von Huene, R., Ishiwada, Y., Langseth, M., Burns, T., Honza, E., 1980. Sci-
Farquhar, J., Wing, B.A., McKeegan, K.D., Harris, J.W., Cartigny, P., Thiemens, M.H.,
entific Party, Initial Reports of the Deep Sea Drilling Project 489. Interpretation
2002. Mass-independent sulfur of inclusions in diamond and sulfur recycling
of Multichannel Seismic Reflection Data, Legs 56 and 57, Japan Trench Transect,
on early Earth. Science 298, 2369e2372.
Deep Sea Drilling Project, pp. 489e504.
Foley, B.J., Bercovici, D., Elkins-Tanton, L.T., 2014. Initiation of plate tectonics from
Nutman, A.P., Friend, C.R.L., Bennett, V.C., 2002. Evidence for 3650e3600 Ma as-
post-magma ocean thermochemical convection. Journal of Geophysical
sembly of the northern end of the Itsaq Gneiss Complex, Greenland: implica-
Research: Solid Earth 119, 8538e8561. http://dx.doi.org/10.1002/2014JB011121.
tion for early Archaean tectonics. Tectonics 21. http://dx.doi.org/10.1029/
Furnes, H., de Wit, M., Staudigel, H., Rosing, M., Muehlenbachs, K., 2007. A vestige of
2000TC001203, 5e1e5-28.
Earth’s oldest ophiolite. Science 315, 1704e1707.
Palin, R.M., White, R.W., 2016. Emergence of blueschists on Earth linked to secular
Hamilton, W.B., 2011. Plate tectonics began in Neoproterozoic time, and plumes
changes in oceanic crust composition. Nature Geoscience 9, 60e64. http://
from deep mantle have never operated. Lithos 123, 1e20.
dx.doi.org/10.1038/ngeo2605.
Harrison, T.M., Blichert-Toft, J., Muller, W., Albarede, F., Holden, P., Mojzsis, S.J., 2005.
Paterson, M.S., Wong, T.-F., 2005. Experimental Rock Deformationethe Brittle Field.
Heterogeneous Hadean hafnium: evidence of continental crust at 4.4 to 4.5 Ga.
SpringereVerlag, New-York, pp. 155e163.
Science 310, 1947e1950.
Ranero, C.R., Von Huene, R., 2000. Subduction erosion along the Middle America
Harrison, T.M., Watson, E.B., Aikman, A.B., 2007. Temperature spectra of zircon
convergent margin. Nature 404, 748e752.
crystalliza- tion in plutonic rocks. Geology 35, 635e638.
Rino, S., Kon, Y., Sato, W., Maruyama, S., Santosh, M., Zhao, D., 2008. The Grenvillian
Hilde, T.W.C., 1983. Sediment subduction versus accretion around the Pacific. Tec-
and Pan-African orogens: world’s largest orogenies through geologic time, and
tonophysics 99, 381e397.
their implications on the origin of superplume. Gondwana Research 14, 51e72.
Hopkins, M., Harrison, T.M., Manning, C.E., 2008. Low heat flow inferred from .4 Gyr
Rybacki, E., Dresen, G., 2000. Dislocation and diffusion creep of synthetic anorthite
zircons suggests Hadean plate boundary interactions. Nature 456, 493e496.
aggregates. Journal of Geophysical Research 105, 26017e26036.
Hussong, D.M., Wipperman, L.K., 1981. Vertical movement and tectonic erosion of
Scholl, D.W., Von Huene, R., 2009. Implications of estimated magmatic additions
the continental wall of the Peru-Chile Trench near 11300 S latitude. In:
and recycling losses at the subduction zones of accretionary (non-collisional)
Kulm, L.D., Dasch, E.J., Hussong, D.M. (Eds.), Nazca Plate: Crustal Formation and
and collisional (suturing) orogens. Geological Society, London, Special Publi-
Andean Convergence, vol. 154. Geological Society of America Memoir,
cations 318, 105e125.
pp. 519e524.
Scholl, D.W., Von Huene, R., 2007. Crustal recycling at modern subduction zones
Ichikawa, H., Greaux, S., Azuma, S., 2017. Subduction of the primordial crust into the
applied to the past-issues of growth and preservation of continental basement
deep mantle. Geoscience Frontiers 8, 347e354.
crust, mantle geochemistry, and supercontinent reconstruction. Geological
Ichikawa, H., Kawai, K., Yamamoto, S., Kameyama, M., 2013. Supply rate of conti-
Society of America Memoirs 200, 9e32.
nental materials to the deep mantle through subduction channels. Tectono-
Stern, C.R., 2011. Subduction erosion: rates, mechanisms, and its role in arc mag-
physics 592, 46e52.
matism and the evolution of the continental crust and mantle. Gondwana
Ichikawa, H., Kawai, K., Yamamoto, S., Kameyama, M., 2015. Effect of water on
Research 20, 284e308.
subduction of continental materials to the deep Earth. In: Khan, A.,
Stern, R.J., 2005. Evidence from ophiolites, blueschists, and ultrahigh-pressure
Deschamps, F. (Eds.), The Earth’s Heterogeneous Mantle. Springer International
metamorphic terranes that the modern episode of subduction tectonics
Publishing, pp. 275e299.
began in Neoproterozoic time. Geology 33, 557e560.
Iizuka, T., Komiya, T., Ueno, Y., Katayama, I., Uehara, Y., Maruyama, S., Hirata, T.,
Stern, R.J., Scholl, D.W., 2010. Yin and yang of continental crust creation and
Johnson, S.P., Dunkley, D., 2007. Geology and zircon geochronology of the Acasta
destruction by plate tectonic processes. International Geology Review 52, 1e31.
Gneiss Complex, northwestern Canada: new constraints on its tectonothermal
Tackley, P.J., 1998. Self-consistent generation of tectonic plates in three-dimensional
history. Precambrian Research 153, 179e208.
mantle convection. Earth and Planetary Science Letters 157, 9e22.
Ikoma, M., Genda, H., 2006. Constraints on the mass of a habitable planet with
Tackley, P.J., 2000. Self-consistent generation of tectonic plates in time-dependent,
water of nebular origin. The Astrophysical Journal 648, 696e706.
three-dimensional mantle convection simulations 1. Pseudoplastic yielding,
Iwamori, H., McKenzie, D., Takahashi, E., 1995. Melt generation by isentropic mantle
Geochemistry, Geophysics. Geosystems 1, 1021. http://dx.doi.org/10.1029/
upwelling. Earth and Planetary Science Letters 134, 253e266.
2000GC000036.
Karato, S., Jung, H., 2003. Effects of pressure on high-temperature dislocation creep
Tsenn, M.C., Carter, N.L., 1987. Upper limits of power law creep of rocks. Tectono-
in olivine. Philosophical Magazine A83, 401e414.
physics 136, 1e26.
Kasting, J.F., Calting, D., 2003. Evolution of a habitable planet. Annual Review of
Tsuru, T., 2004. A new phase of tectonic erosion study in the Japan Trench
Astronomy and Astrophysics 41, 429e463.
convergent margin. Chikyu Monthly 26, 672e679 (Japanese Journal).
Katayama, I., Karato, S., 2008. Low-temperature, high-stress deformation of olivine
Turcotte, D.L., Schubert, G., 2002. Geodynamics 2nd Edition. Cambridge Univ. Press,
under water-saturated conditions. Physics of the Earth and Planetary Interiors
New York, pp. 132e144.
168, 125e133.
Vannucchi, P., Remitti, F., Bettelli, G., 2008. Geological record of fluid flow and seis-
Kawai, K., Tsuchiya, T., Tsuchiya, J., Maruyama, S., 2009. Lost primordial continents.
mogenesis along an erosive subducting plate boundary. Nature 451, 699e703.
Gondwana Research 16, 581e586.
346 S. Azuma et al. / Geoscience Frontiers 8 (2017) 337e346
van Hunen, J., van den Berg, A.P., 2008. Plate tectonics on the early Earth: limitations
imposed by strength and buoyancy of subducted lithosphere. Lithos 103, 217e235.
Von Huene, R., Culotta, R., 1989. Tectonic erosion at the front of the Japan Trench
convergent margin. Tectonophysics 160, 75e90.
Von Huene, R., Lallemand, S., 1990. Tectonic erosion along the Japan and Peru
convergent margins. Geological Society of America Bulletin 102, 704e720.
Von Huene, R., Nasu, N., Arthur, M., Cadet, J.P., Carson, B., Moore, G.W., Honza, R.,
Fujioka, K., Barron, A.J., Keller, G., Reynolds, R., Shaffer, B.L., Sato, S., Bell, G., 1978.
Japan Trench transected. Geotimes 23, 16e20.
Von Huene, R., Ranero, C.R., Vannucchi, P., 2004. Generic model of subduction
erosion. Geology 32, 913e916.
Von Huene, R., Scholl, D.W., 1991. Observations at convergent margins concerning
sediment subduction, subduction erosion, and the growth of continental crust.
Reviews of Geophysics 29, 279e316.
Von Huene, R., Scholl, D.W., 1993. The return of sialic material to the mantle indi-
cated by terrigeneous material subducted at convergent margins. Tectono-
physics 219, 163e175.
Watson, E.B., Harrison, T.M., 2005. Zircon thermometer reveals minimum melting
conditions on earliest Earth. Science 308, 841e844.
Wilde, S.A., Valley, J.W., Peck, W.H., Graham, C.M., 2001. Evidence from detrital
zircons for the existence of continental crust and oceans on the Earth 4.4 Gyr
ago. Nature 409, 175e178.
Yamamoto, S., 2010. Tectonic erosion: new perspectives on Pacific-type orogeny and
continental growth models. Journal of Geography 119, 963e998.
Yamamoto, S., Nakajima, J., Hasegawa, A., Maruyama, S., 2009a. Izu-Bonin arc
subduction under the Honshu island, Japan: evidence from geological and
seismological aspect. Gondwana Research 16, 572e580.
Yamamoto, S., Senshu, H., Rino, S., Omori, S., Maruyama, S., 2009b. Granite sub-
duction: arc subduction, tectonic erosion and sediment subduction. Gondwana
Research 15, 443e453.
Yanagisawa, T., Yamagishi, Y., 2005. Rayleigh-Bénard convection in spherical shell
with infinite Prandtl number at high Rayleigh number. Journal of the Earth
Simulator 4, 11e17.
Zhu, J., Kopp, H., Flueh, E.R., Klaeschen, D., Papenberg, C., Planert, L., 2009. Crustal
structure of the central Costa Rica subduction zone: implications for basal
erosion from seismic wide-angle data. Geophysical Journal International 178,
1112e1131.