RST A 20180169

Early earth geodynamics: cross
examining the geological

rsta.royalsocietypublishing.org
testimony
Anthony I. S. Kemp
Research School of Earth Sciences, University of Western Australia,
Perth 6009, Australia
Cite this article: Kemp AIS. 2018 Early earth
geodynamics: cross examining the geological AISK, 0000-0003-1642-0360
testimony. Phil. Trans. R. Soc. A 376: 20180169.
Many studies link the presence of continents on
http://dx.doi.org/10.1098/rsta.2018.0169
Earth to the operation of plate tectonics. Radiogenic
isotope data have, however, long consigned the bulk
Accepted: 15 August 2018 of crust generation and preservation to the murky
realm of the Precambrian Earth, where the prevailing
geodynamic systems are highly uncertain due to
One contribution of 14 to a discussion meeting
the sparse and complex nature of the geological
issue ‘Earth dynamics and the development of record of these early eons. The purpose of this
plate tectonics’. paper is to examine the nature of this geological
record, considering the biases and artefacts that may
Subject Areas: undermine its fidelity, and to assess what are the
geochemistry, geology, petrology, most robust lines of evidence from which meaningful
plate tectonics geodynamic inferences can be drawn. This is pursued
with reference to Hadean detrital zircons, Archean
Keywords: gneiss complexes and Archean granite–greenstone
terranes, and by considering isotopic proxies of crust–
Archean, continental crust, zircon,
mantle interaction. The evidence reinforces long held
radiogenic isotopes, plate tectonics views that the formation of some of the oldest
continental nuclei involved a distinctive mode of
Author for correspondence: planetary geodynamics that rests uneasily within
Anthony I. S. Kemp definitions of modern style plate tectonics. A detailed
interrogation of the oldest rocks, integrating multi-
e-mail: tony.kemp@uwa.edu.au
scale information from the best preserved whole-
rock and mineral archives, and emphasizing careful
selection at the sampling and analytical stages, will
lead to the most robust input data for petrological and
thermodynamic models of early Earth processes.
This article is part of a discussion meeting issue
‘Earth dynamics and the development of plate
tectonics’.
1. Geodynamics and Earth’s continental crust

More than three decades ago in this journal a case was
made that large volumes of continental crust existed on
2018 The Author(s) Published by the Royal Society. All rights reserved.
the early Earth, and have been recycled en masse through the mantle throughout geological time
2
([1], developed further in [2]). Plate tectonics, via sediment subduction, and the foundering of the
thickened roots of Andean-style magmatic arcs [3], provided an elegant mechanism for this large
rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

........................................................
scale recycling. The bulk composition of the continental crust, resembling that of the andesitic
magmas voluminously erupted along convergent plate boundaries [4] provided notional support
for the role of plate tectonics in building, or at least shaping, the continental crust [5,6]. Numerous
studies have since linked plate tectonics to the processes and rates of continental stabilization
or assembly through time [7–10]. Other studies relate the operation of plate tectonics to the
development of a habitable planet, through formation of buoyant, emergent continents, flux of
essential nutrients into seawater and regulation of the composition of Earth’s atmosphere and
oceans [11–13].
There is, therefore, great interest in determining how far back in time plate tectonic processes
operated on Earth—or at least when plate tectonics became the dominant mode of planetary
dynamics, as it is at the present day. The formidable challenge here is that the accessible record
of the critical early part of Earth evolution is incredibly sparse. It could in fact be argued that
we know more about the early history of the Moon from the few Apollo sample return missions,
than we do about the early history of the Earth. The conventional explanation for this is that
the Earth is a dynamic planet where plate tectonics is operative, generating new crust and
resurfacing the planet, erasing or at least smudging, the earlier record. In this view, the very
process we are attempting to study has effectively camouflaged the geological imprint of its past
operation, as well as that of any geodynamic predecessors—in contrast to the early planetary
history vividly preserved on the surface of Earth’s neighbours Venus and Mars (e.g. [14])—
although it is important to note that plate tectonics may not be the only mechanism to destroy
crust, especially in a hotter early Earth (e.g. [15]). In a bid to mitigate this limitation, scientists
have increasingly resorted to indirect evidence, or chemical proxies of past processes that can
‘see through’ plate tectonic reworking, although some of the geochemical signals used for this
purpose are ambiguous (e.g. [16,17]).
The main aim of the present paper is to consider evidence for geodynamic processes in the
Hadean to Archean Earth from the extant mineral and rock record of these eons. Given the
breadth of this topic, the discussion will be framed around several case studies, and will focus
on evaluating the impact of various biases (preservation, sampling, analytical) on geodynamic
models, and how to identify the best archives from which robust geodynamic inferences can
be made.
2. Terrestrial archives of geodynamic processes

Plate tectonics is a theory, supported by a range of interdisciplinary and observational evidence,
that describes the motion of material in the outer part of the modern Earth. It involves interactions
between a globally linked system of mobile lithospheric plates that are rigid enough to thicken
and deform at the margins, but with sufficient buoyancy contrast to allow one to slide beneath
the other and to sink into the convecting mantle, while retaining physical coherency. Subduction
of oceanic crust, intrinsic to plate tectonics, is a stable and persistent phenomenon on time scales
of tens to hundreds of million years when sustainable plate tectonics is operating.
Given this qualitative working definition, the key question concerns how to evaluate the
operation of plate tectonics in the early part of Earth history. An immediate problem is one of
perspective—our view of the specific mechanisms and products of plate tectonics is very much
prejudiced by what we perceive at the present day. Delving further back into Earth history, with
the likelihood of different thermal and thus mechanical regimes in the crust and mantle, there
is the possibility that these processes and products are also different. For example, lawsonite
blueschists are considered to be the hallmark of ‘cold’ subduction, the only environment on Earth
where the requisite P-T-t conditions for the formation of these rocks are met. The occurrence
of blueschists in the geological record is, therefore, taken as a ‘smoking gun’ for active plate
tectonics. So does the apparent absence of blueschists older than the Neoproterozoic mean that
plate tectonics did not operate prior to that [18], or does it simply mean that a ‘different’ type of
3
plate tectonics operated that was not conducive to blueschist formation (e.g. [19])? If so, does this
overarching process still qualify as plate tectonics? Such considerations risk driving the debate

........................................................
into a semantic cul-de-sac—clearly the priority must be to establish what the processes actually
are, rather than trying to fit a label to them—but it does highlight the need for a consistent,
objective and testable series of criteria for evaluating the prevalent geodynamic mode in the
ancient Earth.
A second issue is that any geodynamic model must be built on robust and unbiased
information that constrains primary processes. But, given the scant geological record of early
Earth history, how can we determine what, if any, information is representative? To answer this
requires knowledge of the nature of biases, and the extent to which models are sensitive to these.
Returning to the blueschist example, could the apparent lack of these distinctive assemblages in
the Archean record just be a reflection of preservation bias (preferential destruction of ancient
accretionary prisms or thermal overprinting), rather than robust evidence against plate tectonics
[20,21]? A third challenge lies in knowing exactly what to look for in the preserved record. In other
words, what is the key evidence to test whether plate tectonics was active? At the present day, the
plate tectonic theory is underlain by a synthesis of observational and geophysical evidence, such
as the shape of continental coastlines, the topography of the seafloor, earthquake foci, magnetic
anomalies on oceanic crust, associations of characteristic rock and mineral assemblages and so
on. Some of these tools are not available further back in time. For this reason, there has been an
increasing search for chemical proxies in robust mineral archives.
To explore early Earth geodynamics, three different strands of evidence are considered in
this paper, namely (i) ancient mineral detritus in sedimentary rocks; (ii) the extant igneous rock
record; and (iii) secular variations in radiogenic isotope compositions. Each of these emphasize
the mineral zircon, a robust repository of microstructural and chemical information that can be
used to constrain aspects of continental evolution, such as thermal histories, deformation styles,
magma sources and, thus by inference, geological processes (e.g. [22]). Continuing advances in
analytical instrumentation and methodology enable us to interrogate the microstructures and
chemical inventory of zircon crystals with ever increasing precision and accuracy, and at greater
spatial resolution, down to the scale of atoms (e.g. [23]). This field has expanded with remarkable
rapidity. But this raises two additional questions, to be considered in the following sections—first,
the question of scale, i.e. how reliably can we translate information obtained from individual
crystals at the micrometre scale to the larger scale of mountain belts and lithospheric plates
(e.g. [24]), and, second, is zircon an informed witness to the operation of plate tectonics?
(a) The Hadean detrital record

Early Earth studies were changed irrevocably by the discovery of water-worn zircon crystals
with greater than 4 Ga dates in otherwise unremarkable quartzites and pebble conglomerates
of the Archean Yilgarn Craton in Western Australia (the Mount Narryer and Jack Hills localities;
[25,26]). This was the first direct evidence for the existence of crust, possibly granitic, on Earth
in the Hadean Eon. The exciting possibility that these tiny, long-orphaned grains provide a
unique window into a formative, but almost completely unknown period of Earth history ushered
in numerous follow up studies to exploit the chemical information locked within these zircon
crystals, and the mineral inclusions they host. Harrison [27] provides an overview of such studies,
and their implications for the Hadean lithosphere, hydrosphere and atmosphere.
The greater than 4 Ga detrital zircons from the Jack Hills, and the tantalizing glimpse into the
Hadean Eon they offer, have captured the imagination of the scientific and wider community.
Visitors to the outcrop containing these venerable crystals are invariably left with the impression
they have been transported back into the geological dark ages. It is simply remarkable that
these tiny crystals survived on Earth for up to 1.5 billion years before being encapsulated in the
sedimentary rocks in which they currently reside. Despite their popularity, the actual significance
of the Jack Hills zircons remains debated. The mere existence of these ancient grains invites
widely diverging opinions. On the one hand, the Hadean detrital zircon archive is miniscule
4
(e.g. [28]). Most greater than 4 Ga zircon grains identified globally come from a ca 2 m2 outcrop in
a thin, fault-bound strip of metasedimentary rock in the Yilgarn Craton. The provenance of the

........................................................
conglomerate is enigmatic, but it evidently formed by the erosion and re-deposition of older,
mature sedimentary rocks, meaning that the Hadean zircons have been through at least two
sedimentary cycles (see [29]). The original host rock for these zircon grains could have been
destroyed long before these sedimentary rocks were deposited. Searches have failed to locate
greater than 4 Ga rocks among the Eoarchean gneisses flanking the Jack Hills and Mount Narryer
supracrustal belts. It seems logical to infer from the global scarcity of greater than 4 Ga zircons,
in spite of the unprecedented analytical effort devoted to finding such grains, that there cannot
ever have been much zircon-bearing Hadean crust. It has, therefore, been argued that the Jack
Hills zircons crystallized from small melt pockets within an overall mafic to ultramafic (i.e.
not continental) crust [30–33]. In this case, the zircons themselves are not representative of that
bulk crust and manifest local, possibly ad hoc processes that may not provide insight into the
key workings of the Hadean Earth. The opposing, ‘tip of the iceberg’ viewpoint is that given
the hostile conditions expected for the early Earth, with intense bolide bombardment, volcanic
resurfacing and crust–mantle overturns, the fact that any crustal relicts survived at all must mean
that zircon-bearing Hadean crust was once voluminous (e.g. [34]). If the greater than 4 Ga detrital
zircons indeed represent vestiges of granitic continents, then any inferences made from the
chemical inventory of these tiny grains goes to the heart of understanding Hadean crust–mantle
differentiation.
Attempts to use other chemical indices to resolve these contrasting viewpoints are confounded
by another, more fundamental issue, concerning the extent to which the greater than 4 Ga detrital
zircons faithfully preserve their primary compositional information. Disturbance to the U-Pb
isotope system has long been recognized by discordance in both U-Pb and Th-Pb ages [26] and
by variations in 207 Pb/206 Pb ages within individual grains, and even discrete growth domains
[30,35,36]. Concordance is no guarantee of isotopic integrity in these zircons, because ancient Pb
loss (or Pb gain) vectors are parallel to the concordia curve between 4.5 and 3.9 Ga. Such effects
have given rise to uncertainties over the actual crystallization age of the Jack Hills zircons. The
possibility of spurious ages from the Jack Hills zircons prompted Moorbath [37] to comment
‘ . . . I shall try to repress the recurring nightmare that SHRIMP will one of these days come
up with zircon U-Pb ages significantly greater than 4600 Myr, the widely accepted age of the
Earth’. This has since been borne out in spectacular fashion. Ge et al. [38] studied a Jack Hills
zircon grain (14041) that had returned a concordant ion microprobe U-Pb date of ca 4.46 Ga, older
than any other terrestrial mineral. Scanning ion imaging showed that the distribution of 207 Pb is
highly irregular in this crystal, and decoupled from 206 Pb and 238 U. Localized (approx. 5 µm)
‘hot spots’ enriched in ancient radiogenic Pb yield apparent ‘ages’, in this case up to 4.8 Ga,
that are geologically meaningless. The overlapping of these domains by the standard 20 µm
diameter elliptical spot of the ion microprobe is believed responsible for the anomalously old
age previously measured from this grain [38].
Radiogenic Pb mobility compromising U-Pb ages in zircon, the premier geochronometer, is
disconcerting enough (see also [39]). However, if the Pb/U systematics is disturbed, then this
casts a deep shadow over the veracity of other chemical information within the zircon grain,
such as O isotope data, REE patterns, Ti thermometry and even whether mineral inclusions are
the result of secondary processes, especially where igneous microstructures show evidence of
modification or recrystallization [30]. Age uncertainty creates large problems for interpretation
of Lu-Hf isotope data, which relies upon an accurate age framework ([40], and see [41]). Some
studies have established multiple Hf isotope populations within individual Jack Hills zircons,
complicating matters further ([33]; see below).
The above narrative risks portraying a ‘glass half empty’ scenario, and it is important to
acknowledge that approaches have been developed, employing rigorous micro-beam imaging
and analysis, to identify those Hadean zircons most likely to have retained their original
magmatic crystallization age (e.g. [23]). Only a small proportion of greater than 4 Ga zircons
(a) 16 Jack Hills detrital zircon (b) 16
solution Amelin et al. [41] Jack Hills concurrent LA Pb-Hf (n = 96) (Kemp et al. [33]) 5
solution Blichert-Toft & Albarede [44] ‘filtered’ dataset (18 zircons)
12 12
solution Harrison et al. [42]
176Lu/177Hf = 0.020 ± 0.004, TCHUR = 4.46 ± 0.12 Ga (MSWD 4.1)
LA Hf Harrison et al. [42,43]

8 8
........................................................
LA Pb-Hf Harrison et al. [43]
(t) zircon
eHf (t) zircon
4 MORB-DM 4 MORB-DM
CHUR
0 CHUR 0
Hf
e
–4 –4
–8 –8
17 f = 0
7 .020
C 176
Lu/ H
–12 perC –12
lead loss?
Up
3300 3500 3700 3900 4100 4300 4500 3300 3500 3700 3900 4100 4300 4500
(c) 16
Jack Hills ‘least disturbed’ zircon
lunar zircons (Apollo 14) Taylor et al. [45]
12
8
eHf (t) zircon
0 CHUR
–4
–8 0
177 Hf = 0.02
176 Lu/
–12
3300 3500 3700 3900 4100 4300 4500

207Pb/206Pb age (Ma)
Figure 1. The U-Pb age and Hf isotope datasets obtained from the Jack Hills detrital zircon grains by both solution- and laser
ablation (LA)-based techniques (panel a: [42–45]; panel b: [33]). The isotope trajectories of upper continental crust (UCC;
176
Lu/177 Hf = 0.008) and present-day MORB-source depleted mantle are shown for reference on (a) and (b); trajectories that
could result from ancient lead loss are shown as grey arrows on (b). (c) Compares the Hf isotope systematics of a subset of the
‘best preserved’ Jack Hills zircons from Kemp et al. [8] (pink diamonds) with those of lunar zircons separated from Apollo 14
breccias (data from [46]). Reproduced with permission from Kemp & Hawkesworth [47]. (Online version in colour.)
stand up to this careful scrutiny [30]. The critical point is this: interpreting the Jack Hills detrital
zircons, or indeed any detrital zircons, as simple recorders of Earth evolution will inevitably lead
to ambiguous outcomes.
To illustrate this, figure 1 summarizes the contrasting datasets, and interpretations, obtained
by different U-Pb and Hf isotope studies of the Jack Hills zircons. Harrison et al. [43] report a
dataset obtained by digestion of whole zircon grains and by microanalysis, involving separate
U-Pb age (ion microprobe) and Hf (laser ablation) measurements from different volumes of the
crystal. These data show large variations, with both strongly positive and negative εHf values—
complexities also evident in the solution dataset of Blichert-Toft & Albarede [45], but not in
the precursor solution study of Amelin et al. [42] (figure 1). Harrison et al. [43] attribute this
scatter to the formation of continental crust and complementary depleted mantle reservoirs, with
orogenic magmatism and crust–mantle recycling associated with plate tectonics. This large Hf
isotope range was, however, not reproduced by subsequent studies of the Jack Hills zircons
[8,44]. Those studies adopted a different analytical technique, where the 207 Pb/206 Pb date was
measured concurrently with the Hf isotope ratio during laser ablation depth profiling. This
method allowed a more robust integration of age and isotope tracer information, and provided a
way to recognize spurious εHf values produced by mixed sampling of domains with disparate age
and 176 Hf/177 Hf. Moreover, Kemp and co-workers [33] focused on zircons with well-preserved
igneous microstructures and that show minimal evidence for isotope disturbance, in an attempt
to minimize age uncertainty related to ancient Pb loss, which also produces artefacts in εHf –time
plots. The data reported by Kemp et al. [33] define a simple Hf isotope-time array attributed to the
protracted, but small scale, reworking of a primordial mafic crust that separated from chondritic
mantle very soon (perhaps less than 100 Ma) after planetary accretion [42]. This process does not
require the existence of Hadean continents and is hard to reconcile with the operation of plate
6
tectonics (see §2c(ii) and [47], for more detailed discussion).
The stark differences in interpretations of Hadean geodynamics encapsulated in the above

........................................................
example thus reflect differences in analytical methodology, but also in approach, particularly in
the recognition that not all zircons provide reliable constraints on crust–mantle differentiation.
A clear message is that careful pre-screening of grains and analytical sites, rather than an
indiscriminant, en masse analytical approach, is imperative for the study of ancient zircons.
Detrital greater than 4 Ga zircon crystals preserving sharp igneous microstructures and consistent
intra-grain U-Pb ages have proven to be exceedingly rare, which testifies to the environment
in which the original host rocks to these ancient crystals formed, and were subsequently
exposed to.
Controversies notwithstanding, studies such as those of Wilde et al. [48] and Harrison et al.
[43] have greatly raised the profile of research into the early history of the terrestrial planets.
The discipline has, therefore, benefitted substantially from the discovery of these Hadean zircon
grains, and the advances in microanalysis, both in technique and approach (e.g. [49]), that these
have inspired. Nevertheless, the vanishingly small Hadean detrital zircon archive has exerted
a disproportionate weight on models for early Earth evolution. It is interesting to speculate
on the direction that early Earth studies might have taken if these zircons were not found.
Perhaps greater effort would have been spent on studies of igneous basement—the ca 30 000 km2
of Archean gneisses surrounding the Jack Hills and Mount Narryer belts, known to include
components at least as old as 3.75 Ga (and greater than 4 Ga zircon xenocrysts; [50]), could
rightfully sue for neglect. We would certainly know much more about the thermal evolution
of ancient terranes, such as in the Northwest Territories of Canada, Greenland, Labrador or
Antarctica, if analytical effort had instead been directed into these areas. There may well have
been even greater emphasis on using extinct isotope systems, such as 146 Sm–142 Nd and 182 Hf–
182 W, to track early crust–mantle evolution. A stronger case may have been made to revisit the
Moon, given the wealth of Hadean-aged rocks that gleam brightly in Earth’s night skies, and the
possibility that the lunar soil is littered with impact debris of terrestrial crust ejected during heavy
bolide bombardment in the early solar system history.
In any case, the ambiguities surrounding the origin and significance of the greater than
4 Ga Jack Hills zircons highlighted above reflect the fundamental limitation of detrital mineral
studies—that the data must be interpreted in the absence of constraints from co-genetic mineral
populations and the composition and geological context of the host rock. For this reason, we turn
now to the igneous rock record.
(b) Igneous rock record

(i) The challenges of antiquity
It is obvious from global geological maps that Precambrian rocks occupy just a small fraction
of Earth’s exposed continental landmasses. In many cases, Archean crystalline basement is
further obscured by jungle (Amazon Craton), snow and ice (Slave, North Atlantic, Siberian,
Fennoscandian cratons) or recent weathering profiles (Yilgarn Craton). Whether the small amount
of ancient continental crust we see today truly reflects the original volume of ancient crustal
material, or is an indication that the rock record is incomplete, is uncertain. Proponents of the
latter viewpoint cite isotope arguments that younger rocks form by reconstitution of older ones,
and that Archean to Hadean detrital zircons, such as those from the Jack Hills, are relics of large
amounts of crust that have since been destroyed (e.g. [34]), even though the volumes of that
crust are poorly constrained. If this interpretation is correct, it is critical to understand whether
there are factors that bias the nature and volume of what has been preserved. This is so that
an assessment can be made as to whether the surviving crustal fragments are representative
of the composition of the since-vanished crust, and thus whether these provide insights into
petrogenetic and geodynamic processes that were important at the time.
What can cause bias in preservation? The nature of the material itself is an obvious factor—
7
zircons, by virtue of their physiochemical resilience, are much more likely to survive recycling
processes than dense, ultramafic rocks. Tectonic setting is another. Hawkesworth et al. [7,8]

........................................................
developed a model to explain the selective preservation of crust during the supercontinent
cycle. In this model, the survival of rocks formed at the end of the convergent phase (i.e.
juvenile subduction-related magmas) and during the collision phase (crustal melts) is favoured
because these are captured within the interior of the newly assembled supercontinent and are
protected to some extent against subsequent subduction erosion. Survival of backarc sequences
is greatly favoured over forearcs. This selective preservation mechanism is invoked to explain
the conspicuous global peaks in zircon ages (e.g. [51]). Rocks formed before the establishment of
global plate tectonics would not, therefore, be subject to preservation biases associated with the
supercontinent cycle [52].
The debate around preservation bias continues (e.g. [53]). However, an even more fundamental
bias arises from the very nature of the rocks themselves. Geologists working in Precambrian,
and particularly Archean, terranes are commonly confronted by outcrops of great complexity
(e.g. figure 2a–d). This is because Archean gneisses may represent the end product of protracted
thermal and deformational episodes that span hundreds of millions of years (e.g. [56]). The
original protoliths to intensely deformed gneissic rocks may no longer be determinable with
confidence. How do we develop a sampling strategy that optimizes the amount and quality of
information we can retrieve from these precious, but daunting outcrops? How do we recognize
the critical aspects to target, as opposed to the outliers that will only contribute ‘noise’ to
interpretations? What is the best scale on which to observe, sample and analyse these rocks?
Such deliberations are strongly influenced by the scientific questions being asked—researchers
trying to establish the protolith age(s) of the rock shown in figure 2a would sample it differently
from those seeking to understand the development of deformational fabrics, or those aiming to
elucidate the conditions of metamorphism and partial melting. The point here is that decisions
made about how to sample such heterogeneous outcrops, which parts to analyse, and what
analytical methods to employ, introduce biases that are also potentially severe. Three examples
are considered below that deal with this issue.
(ii) The Acasta Gneiss complex, Slave Craton, Canada

The onset of the Earth’s geological record is preserved in the Acasta Gneiss Complex of Canada.
This area harbours the oldest continental rocks dated by the zircon U-Pb method. The landmark
study of Bowring & Williams [34] established that some of the magmatic precursors to tonalitic to
granodioritic gneisses crystallized at ca 4.00–4.03 Ga, with inherited components to 4.06 Ga. Iizuka
et al. [57] recognized an even older (4.2 Ga) xenocrystic zircon in one of the granodioritic gneisses.
The Acasta Gneisses are, however, complicated, highly strained and multicomponent rocks,
where these different components are interlayered at decimetre to centimetre scale [58]. Zircon
microstructures and U-Pb data attest to a multiple metamorphic and deformational overprints.
Hf isotope data from these grains show substantial intra-sample variations [40,59]. The existence
of the Acasta gneisses clearly indicates a terrestrial differentiation process capable of producing
continental-like magmas at ca 4 Ga. But does the composition of these gneisses allow us to
constrain the geodynamic regime under which this occurred?
Reimink et al. [60] report the discovery of a well-preserved tonalitic gneiss unit in a low strain
zone of the Acasta Gneiss Complex that bears directly on this question (figure 2b). These rocks
comprise a small strip a few metres wide and less than 100 m long that was recognized by a
relatively homogeneous outcrop appearance, and by high modal content of garnet. Constituent
zircon crystals exhibit sharp igneous microstructures, and yield a fairly simple spread of U-Pb
ages that indicate crystallization of the tonalitic protolith at ca 4.02 Ga, in line with the findings of
Bowring & Williams [34], but minus some of the complexity.
Two distinctive features of these gneisses are highlighted by Reimink et al. [60]. First, the rocks
show unusual iron enrichment, and, unlike many Archean meta-tonalites, a flat REE pattern and
(a) (b) 8

........................................................
206Pb/238U
0.77
(c) 3660
3620
0.75
3580
3540
0.73
0.71
0.69
zircon, meta-tonalite 207Pb/235U
31.5 32.5 33.5 34.5 35.5

1000
100
10
chondrite normalized
1
La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
206Pb/238U
0.75
3500
amp
0.65
3100
2900
0.55 granodiorite
2700
2500 207Pb/235U
8 12 16 20 24 28 32
Figure 2. Field photos of (a) heterogeneous and strongly deformed Meeberrie Gneiss from the Narryer Terrane, Yilgarn
Craton (Western Australia) in which the different lithological components partly retain their compositional identity, but are
complexly interspersed and infolded with younger granitic leucosomes and pegmatite injections. (b) Low strain outcrop of
the ca 4.02 Ga Idiwhaa Tonalite Gneiss, Acasta Gneiss Complex, Slave Craton (Canada; photograph courtesy of Jesse Reimink),
(c) banded gneiss outcrop in the Narryer Terrane (Yilgarn Craton, Western Australia), to illustrate the geochronology (U-Pb
isotopes in zircon, measured by ion microprobe) and whole rock rare earth element patterns (chondrite normalized abundances)
that have been obtained from the most homogeneous parts of the different components (amp, amphibolite boudin;
Kemp et al. [54]), and (d) partially dismembered anorthosite, represented by trains of plagioclase megacrysts, some with
adhering mafic intercumulus, together with discrete enclaves of leucogabbro and ultramafic rock, strewn through a ca 3.4 Ga
quartzofeldspathic gneiss ([55]; Narryer Terrane, Yilgarn Craton). Ion microprobe U-Pb data (zircon) from a leucogabbro pod,
and a typical well-preserved spinel grain hosted by ultramafic enclaves in the 3.4 gneisses, are also shown [29]. (Online version
in colour.)
(d) zircon, leucogabbro
9
206Pb/238U
0.86

........................................................
0.82 3840
3800
3760
3720
mag 0.78
3680
3640
3600
spl 3560
0.74
0.70
opx
207Pb/235U
32 34 36 38 40 42
see figure 4
Figure 2. (Continued.)
negative Eu anomaly. Second, the oscillatory-zoned igneous zircon shows a small depletion in
18 O compared to zircon in equilibrium with mantle oxygen. This signature in magmatic rocks
is the hallmark of reworking crust that has been hydrothermally altered by interaction with
hot, surface-derived fluids. This combination of features (viz. Fe-rich intermediate magmas, low
zircon δ18 O) is not typical of present-day subduction-related environments. Instead, Reimink
et al. [60] infer formation of the tonalitic precursors by shallow-level fractionation of basaltic
magmas, accompanied by remelting/assimilation of hydrothermally altered volcanic crust,
above an extensional zone of mantle upwelling. This setting is analogous to that presented by
modern Iceland. In this interpretation, the compositional character of the oldest, and possibly
best-preserved, nuclei of continental crust on Earth does not require plate tectonic processes.
The study of Reimink et al. [60] is an excellent example of detailed field investigation,
informing judicious sampling and comprehensive whole rock and integrated zircon
microanalysis. Such targeted approaches provide the best chance of robust data and outcomes
that we can attach weight to in models of early Earth evolution. Given the various selection biases,
10
in the field and in the laboratory, the ‘representativeness’ of these findings can be questioned, but
they nonetheless provide valuable insight into the type of processes that could be operative at the

........................................................
end of the Hadean eon. As stated by Rollinson [61]—‘ultimately a global picture is what we want
to obtain, but this will better come through first understanding processes at regional level’.
(iii) The Narryer Gneiss complex, Yilgarn Craton, Western Australia

A short walk from the greenschist facies metasedimentary successions of the Jack Hills lie
scattered low exposures of upper amphibolite facies quartzofeldspathic gneisses (‘Meeberrie
Gneiss’ [62]) that exhibit in abundance all of the bewildering intricacies of Archean gneiss
complexes. The mixed parentage of these gneisses is clearly evident from outcrop alone, where
a variety of different lithological components, dominantly granitic but including amphibolite,
ultramafic bodies, supracrustal material and fragments of anorthosite-leucogabbro, are intimately
interlayered (figure 2a,c,d). The pervasive gneissic fabrics developed during intense late Archean
strain, metamorphism and anatexis, but there is also evidence for a prior episode of granulite
facies metamorphism at ca 3.3 Ga [29]. As expected from the field characteristics, these Meeberrie
gneisses contain a range of different zircon age populations, from ca 3.73, 3.68, 3.62, 3.6,
3.3 and 2.65 Ga; individual samples commonly contain two or more of these age groupings.
Kinny & Nutman [56] point out examples where different lithological components within the
heterogeneous gneiss outcrops, where lower strain zones permit these to be isolated and sub-
sampled, contain the same, multiple zircon age populations (although this is not always true,
figure 2c). They also find that this multiplicity of zircon ages is retained even in apparently simple
layered gneisses. These observations show that different lithological components of disparate
age have been intimately intercalated on a grain scale. It is little wonder, given this complexity,
that researchers retreat to the relative sanctuary of the Jack Hills pebble conglomerates, with
lithological simplicity and the allure of greater than 4 Ga zircons.
Retrieving meaningful process orientated information from heterogeneous lithological
assemblages like the Meeberrie Gneiss is difficult. Several workers have noted the dominantly
monzogranitic composition of these rocks, thereby inferring a mature continental setting. It is,
however, instructive to consider the bulk composition of the ‘granitic’ portion of these gneisses
in tandem with their zircon populations. In cases where the Meeberrie gneisses yield a complex
spread of zircon ages, they tend to have abundant K-feldspar, low to moderate Na/K ratios and
pronounced negative Eu anomalies (figures 2c and 3). Where these rocks are dominated by a
single population of greater than 3.6 Ga zircon, however, they tend to be tonalitic, with high
Na/K, and steep REE patterns without distinct negative Eu anomalies (figures 2c and 3). This
suggests that the ‘granitic’ character of many samples reflects the injection and tectonic blending
of younger, pegmatitic and granitic components, not the melting at 3.6 Ga of differentiated
continental sources. The greater than 3.62 Ga tonalitic gneisses, with single component igneous
precursors, would, therefore, appear to be the best candidates to explore further, although it is
unclear whether such compositions are sufficiently distinctive to diagnose a particular tectonic
setting in which the tonalitic magma, as opposed to its source rocks, formed [17].
Study of other enclaves contained within the heterogeneous quartzofeldspathic gneisses
may bear riper fruit. In places, the gneisses enclose fragments of metamorphosed gabbro,
anorthosite and ultramafic rock that are thought to be derived from layered igneous intrusions
that were dismembered first magmatically, by the granitic protolith to the host gneisses, and then
structurally, by subsequent intense deformation (figure 2d; [29,62]. These enclaves range in size
from a few centimetres, the scale of individual plagioclase crystals, to lenticular bodies up to
several hundred metres long. Most rocks have been pervasively recrystallized to an equigranular
metamorphic texture, although relict igneous textures and mineralogical layering are locally
preserved. Zircon U-Pb ages of ca 3.73 Ga have been determined for leucogabbros and anothosites
[29,55] (figure 2d), making these the oldest terrestrial anorthosites. Greater geodynamic insight
comes from the associated ultramafic pods. These are primarily talc-anthophyllite schists, but
1.2
>3.6 Ga, single age component 11
1.0 <3.6 Ga, or multiple age component

........................................................
0.8
Eu/Eu*
0.6
0.4
0.2
0
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Na2O/K2O
Figure 3. Plot of whole rock Na2 O/K2 O versus Eu anomaly (Eu/Eu*) for Meeberrie gneisses, comparing data from tonalitic
samples dominated by a single 3.6–3.62 Ga zircon age component (squares), with data from granodioritic-granitic samples
with mixed zircon age populations, or with zircon ages less than 3.6 Ga (diamonds) (adapted from Kemp et al. [54]). (Online
version in colour.)
in exceptional instances, as in the central portions of boudin-like bodies, they retain a cumulate
igneous texture dominated by coarse orthopyroxene [29]. Unaltered chromian spinel contained
within relict olivine and orthopyroxene in these rocks has an aluminous composition that, of
the common magma types in the modern Earth, only resembles spinel formed in ultramafic
cumulates of thickened oceanic plateaus ([63]: figure 4), in accord with the scenario proposed
for the Acasta Gneisses by Reimink et al. [60]). This does not rule out plate tectonics at 3.73 Ga
(formation of less Cr-rich spinel could, for example, have been favoured in a more fertile
Eoarchean mantle wedge), but provides no evidence for it either. Chromite-bearing ultramafic
pods are an exceedingly minor component of the Narryer Terrane, but arguably provide best
archive from which geodynamic inferences about the early evolution of the Yilgarn Craton can be
made.
(iv) The East Pilbara Terrane, Pilbara Craton, Western Australia

Often referred to as the archetypal Archean granite-greenstone terrane [65], the East Pilbara
Terrane preserves a 700 million record of Earth history, from 3.53 to 2.83 Ga. It consists of large
(35–120 km diameter), dome-like granitic complexes flanked by curvilinear belts of metavolcanic
and sedimentary rocks (greenstones). Unlike the Narryer Terrane of the Yilgarn Craton, the East
Pilbara Terrane exposes relatively shallow crustal levels at low metamorphic grade (regionally
lower greenschist) and lacks an intense Neoarchean deformational imprint. As a result, volcanic
features of the eruptive rocks are locally exquisitely preserved, and the primary geological
architecture of the terrane is largely unmodified, providing an opportunity to understand the
large-scale processes that built and stabilized this sizeable tract of continental crust.
Any geodynamic model for the East Pilbara Terrane needs to explain (1) the dome and basin
geometry and (2) the strikingly pulsed eruptive history recorded by greenstone belts, where
volcano-sedimentary cycles of 5–30 Myr duration accumulated over 300 Myr; the role of mantle
plumes in such sustained and craton-scale periodicity has been vigorously debated (e.g. [16,21]).
Regarding the first point, a salient question concerns whether the domal map pattern of the East
Pilbara granitic complexes reflect diapirism, ‘normal’ batholith structures, core complexes or fold
interferences patterns. Two simple observations can be made. First, the ovoid geometry of the East
Pilbara granitic complexes contrasts with the linear batholithic bodies characteristic of convergent
margins (figure 5), indicating a very different physical control on magma emplacement. The
granitic complexes represent structural domes, whereas arc batholiths are commonly sheet-like
(a) (b)
mt
12
500 mm opx opx opx
spl spl

........................................................
ol spl
opx ol
opx ol
spl opx
spl
ol opx
ol
ol opx
opx cpx
mt
500 mm
(c) Fe3+ Fe3+
Narryer ultramafic
Oceanic spinel
Plateau
Ultramafic Island Arc
Cumulates Tholeiites
Cr3+ Al3+ Cr3+ Al3+

Fe3+ Fe3+
Ocean Island Boninites

Tholeiites
Cr3+ Al3+ Cr3+ Al3+
Figure 4. (a) Photomicrograph of a well preserved ultramafic cumulate rock (meta-harzburgite) from the Narryer Terrane,
containing olivine (ol), orthopyroxene (opx) and chromian spinel (spl, outlined in dashed white hexagons). Note magnetite
concentrations outlining original cumulate olivine grain boundaries. (b) Back-scattered electron image of the same sample as in
(a), showing spinel euhedra enclosed within orthopyroxene. (c) Comparison of spinel compositions from the Narryer ultramafic
rocks with spinels from mafic igneous rocks formed in different tectonic settings (from Rowe & Kemp [63]). The dark grey and
light grey shading represents 50th and 90th percentile fields, after Barnes & Roeder [64]. (Online version in colour.)
bodies controlled by arc-normal extension and contraction, and arc-parallel translation (e.g.
[67]). The Pilbara granitic complexes also display a distinctive internal organization, where
older components are displaced towards the peripheries and deformed and migmatized by
successively younger magmatic additions [65]. Secondly, the granitic complexes were constructed
by multiple magmatic pulses over several hundred million years, implying a geographically
(b) 35°
13
NT
QLD
WA
SA
NSW

Canberra
........................................................
(a) VIC
TAS
Moruya
Batholith
36°
50 km
CG
Kosciuszko
Batholith Berridale Bega
ME Batholith Batholith
Bega
37°
TASMAN SEA
0 50 km
BASS STRAIT
149° 150°
Figure 5. Comparison of the northern portion of the Pilbara Craton, Western Australia (a, Landsat image, courtesy U.S.
Geological Survey; ME, Mt Edgar Granitic Complex), with a simplified map of the Lachlan Fold Belt, southeastern Australia,
shown at the same scale (b, modified from Kemp et al. [66]). Shaded patterns represent granitic intrusions. (Online version in
colour.)
fixed (but episodic) thermal anomaly, and that successively younger magmas exploited the same
structural conduits as older magma batches. The Mount Edgar Granitic Complex was assembled
by magmatic pulses at 3.47–3.46 Ga, 3.45–3.42 Ga, 3.33–3.28 Ga, 3.47–3.2 Ga and 2.85 Ga, with a
record of coeval felsic volcanism in the enveloping greenstone belts [65]. Other granitic complexes
in the East Pilbara Terrane also record (to varying degrees) these, or other, episodes of granitic
emplacement (the adjacent Shaw Granitic Complex has large tracts of granite emplaced at 2.95–
2.93 Ga). The extreme magmatic longevity of up to 650 Myr recorded by individual batholitic
bodies is not matched along Phanerozoic convergent plate margins. For example, subduction-
related processes may have been active along parts of the proto-Andean margin since the early
Paleozoic, but magmatic episodes tend to comprise discrete belts and intrusive structures (e.g.
[68]). It is difficult to conceive of any plate tectonic process that could reproduce the spatial
pattern and temporal scale of felsic magmatism as recorded by the East Pilbara Terrane. Indeed,
the 3.52–3.2 Ga basaltic lavas in the greenstone belts do not show arc-like geochemical features,
although such features may appear (transiently) in eruptive sequences in the adjacent West
Pilbara Superterrane at around 3.1 Ga [69]. For this reason, models involving convective crustal
overturn or sagduction have long been proposed for the Pilbara Craton (e.g. [70]).
A lot of work remains to fully characterise the architecture of the Pilbara Craton, particularly
to image the mid- to lower crustal structure to assess whether this is compatible with crustal
overturn models. Knowledge of the metamorphic evolution of the supracrustal belts is also
incomplete; this can reveal the P-T conditions and time scales of material transport in the
crust, which may be compared to the results of thermodynamic modelling. François et al. [71]
employed this approach, in tandem with other geological constraints, to support operation of a
sagduction process for the East Pilbara Terrane. In summary, the first order geological features of
the Paleoarchean East Pilbara Terrane highlighted above are difficult to reconcile with modern-
style plate tectonic processes. We return to this point below, in considering isotope data for the
Pilbara Craton igneous rocks.
(c) Secular variations in isotope composition 14
(i) Archean Andean analogues?

........................................................
Radiogenic isotope data are widely held to reflect the age and composition of magma sources.
It follows that changes in isotope compositions through geological time, which reflect changes
in the availability of materials to melt and the exchange of matter between Earth’s geochemical
reservoirs, potentially track geodynamic processes [72]. For example, Kemp et al. [73] report whole
rock Sm-Nd and zircon Hf-O isotope data for Paleozoic–Mesozoic igneous rocks in southeastern
Australia. The emplacement of these was related to long-lived subduction–accretion processes
along the active paleo-Pacific Gondwana margin [74]. The zircon and whole-rock isotope data
define striking spatial temporal trends that reflect the dynamics of the outboard subduction
zone [73]. Igneous rocks become systematically more juvenile with decreasing age and with
distance from the craton margin, and define cyclic secular trends that correlate with the pattern
of compressional and extensional episodes. Abrupt, periodic decreases in εNd –εHf precede pulses
of felsic magmatism and follow intense, short-lived (less than 5 Ma) contractional events; these
isotopic dips reflect enhanced reworking of metasedimentary rocks in tectonically thickened
crust. Shifts towards higher εNd and zircon εHf reflect a transition to I-type magmatism during
extension-dominated periods associated with slab rollback. Links between igneous rocks with
high zircon εHf , arc extension and crustal thinning have also been observed in the Peruvian
Eastern Cordillera [75]. DeCelles et al. [76] draw attention to cyclic Nd-isotope time trends shown
by igneous rocks of the North American Cordillera and central Andes, which are linked to high
flux magmatic events and periodic foundering of over-thickened arc roots.
The above-mentioned studies imply that inflected isotope-time trends defined by igneous
rocks are the signature of convergent plate margin processes. Are such trends present in
Archean terranes? figure 6 compares zircon Hf isotope-time patterns defined by igneous rocks
in Greenland and the Pilbara Craton. In Greenland, Eoarchean samples disperse along a sloping
array from ca 3.89 to 3.60 Ga, which is consistent with the remelting of mafic crust [77]. This simple
trend towards increasingly unradiogenic Hf appears to continue to about 3.2 Ga, but is then
interrupted by an abrupt switch back to mantle-like isotope compositions. The hafnium isotope
compositions of igneous rocks then oscillate between depleted mantle-like and unradiogenic
values for the next 200 Myr, reflecting juvenile input and crust reworking, in a pattern reminiscent
of Phanerozoic accretionary orogenic systems. The Pilbara igneous rocks define a contrasting
Hf isotope secular trend. Here, zircon Hf isotope ratios from igneous rocks are approximately
chondritic from 3.65 to 3.12 Ga, suggesting they represent largely juvenile additions to the cratonic
crust. After 3.1 Ga, a timing that coincides with the onset of convergent deformation attributed to
the collision of the East and West Pilbara Cratons [65], isotope compositions deviate towards
progressively less radiogenic values. This signifies reworking of ancient basement by magmas
emplaced after 3.1 Ga.
The isotope-time patterns defined by igneous rocks of the Greenland and Pilbara Craton
differ in detail, but both have initial segments that depict a simple evolution for greater than
500 Myr followed by an inflection at ca 3.2–3.1 Ga that reflects the onset of crust-mantle interaction
(Greenland) and enhanced melting of older crust after crustal thickening (Pilbara). These isotopic
switches, therefore, reflect substantial changes in felsic magma sources that were conceivably
caused by a change in tectonic style. Whether this change was ‘global’ awaits additional isotopic
characterization of other Archean Cratons, although it is notable that other studies also invoke
fundamental changes in geodynamic regimes at around 3 Ga [79].
Another point that emerges from figure 6 is that the oldest rocks in both cratons have
approximately chondritic Hf isotope compositions, thus providing no evidence for inputs from
older crust or from an incompatible element depleted mantle reservoir. This is a consistent
feature of global Hf isotope datasets from early Archean meta-igneous rocks [28,40,47,59,80,81]
and argues against widespread and early incompatible element depletion of the mantle, and,
therefore, the existence of large volumes of complementary stabilized crust at 3.8 Ga (cf. [10]).
(a)
15
5
MOR
B-DM

........................................................
4
crust–mantle
2 interaction
CHUR
eHf (t) zircon
–2
ng
ju ki
–4
an ven or
d c ile ew
lr
–6
ru ad
st di u sta
rew tio cr detrital zircon
or n TTG gneisses [77]
kin
–8 g TTG gneisses published
–10
2750 2950 3150 3350 3550 3750 3950
(b) 8
MORB- felsic igneous rocks
6 DM
4
juvenile ad
2 dition
CHUR
eHf (t) zircon
–2
–4
–6
–8 reworking of
ancient crust
–10
–12
2700 2800 2900 3000 3100 3200 3300 3400 3500 3600 3700
age (Ma)
Figure 6. Zircon Hf isotope versus time plots for Archean meta-igneous rocks of (a) Greenland (modified after [77]) and the
(b) Pilbara Craton (data from [28,78]). Each datum for an igneous sample represents the mean zircon εHf (error bars are ±2
s.d. of the mean). MORB-DM is the mean of present day radiogenic MORB projected to εHf = 0 at 4.5 Ga. Vertical orange bars
in (a) highlight time slices of prominent crust–mantle interaction. The labelled arrows refer to the dominant inferred process
operating during different stages in the evolution of these terranes. (Online version in colour.)
Should this be correct, the present-day extent of early Archean crust (excluding that reworked into
younger rocks) may actually be a reasonable proxy for the original volume of material that was
stabilized above the Moho. Perhaps the ancient geological record is a reasonably representative
geodynamic archive after all—if so this would further emphasize the importance of findings such
as those of Reimink et al. [60] that are based on the preserved geology. A caveat on this is that
radiogenic isotopes only constrain the volume of crust that survives long enough to evolve away
10
16
5 Kaapvaal

........................................................
Craton
0
eHf (t) zircon –5

2
77 f
= 0.0 Jack Hills
–10 176 Lu
/1 H
detrital zircons
–15 176Lu/177Hf = 0.0022
–20
1900 2100 2300 2500 2700 2900 3100 3300 3500
age (Ma)
Figure 7. Hf isotope evolution diagram comparing zircon data from Laurent & Zeh [83] (blue circles, Kaapvaal Craton) with the
Jack Hills detrital zircon dataset of Kemp et al. [33] (red squares, scaled along the age axis so as to fit on the plot). The regression
line of Laurent & Zeh [83], with 176 Lu/177 Hf = 0.0022, is shown in grey. The dashed line through the Jack Hills data, corresponding
to 176 Lu/177 Hf = 0.020, is the best estimate of the crustal source to the melts from which the Jack Hills zircons crystallised, based
on a subset of the ‘best preserved’ zircon crystals (from [33], and figure 1). Almost the entire Jack Hills dataset fits in the 630 Myr
gap in the ‘array’ of Laurent & Zeh [83]. (Online version in colour.)
from mantle isotope compositions—large volumes of continental crust could have been generated
and destroyed rapidly, leaving no isotopic record of its existence, or mode of formation [2].
(ii) The perceived perils of isotope-time arrays

It is pertinent to return to the simple Hf isotope-time array defined by the Jack Hills detrital
zircons. Does this give us clues about how the Hadean Earth evolved? Several authors have
advised caution in interpreting such arrays. Gerdes & Zeh [82] show how steep linear trends
in εHf —time space can result from alteration and resetting of zircon ages (see also [41]). In such a
process, the slope of the array corresponds to the 176 Lu/177 Hf ratio of zircon (approx. 0), however
that from the Jack Hills array is much higher (ca 0.018–0.020, [33]), and reflects substantial
evolution in 176 Hf/177 Hf of a crust-like source reservoir. Laurent & Zeh [83] posit that linear
εHf —time arrays can arise by ‘complex geodynamic processes and crust mantle interactions’
rather than protracted reworking of the same crustal source (but were content to invoke the
latter process to explain the Hf isotope composition of their granitic samples, from the Kaapvaal
Craton). Figure 7 examines this scenario. Even if one accepts the regression line suggested for
the 2.94–2.68 Ga Kaapvaal samples by Laurent & Zeh [83], extrapolating this to the youngest
sample (2.05 Ga) spans a magmatic gap of greater than 600 Myr, an interval close to the entire
magmatic evolution of the Pilbara Craton, and most of the age range covered by the Jack Hills
detrital zircons shown in figure 1. Inferring commonality of source and process for the formation
of any rocks (or zircon grains) separated by such a time gap would be egregious. Moreover, the
regression line to the Kaapvaal samples yields a source 176 Lu/177 Hf (0.0022) that is much lower
than that defined by the Hadean detrital zircons, and typical Archean igneous suites (e.g. [84]),
and is anomalous in terms of the basaltic source compositions usually inferred for Archean felsic
igneous rocks. The general point made by Laurent & Zeh [83] is, however, valid, in that care
is required not to conflate samples formed in different petrogenetic or geodynamic systems on
isotope-time plots, if the goal is to use these to deduce larger crustal evolutionary processes
associated with a particular orogenic system. Detrital zircons are particularly prone to this
uncertainty, although well-defined Hf isotope-time arrays identified from detrital populations
17
are unlikely to be fortuitous. On the other hand, forcing linear ‘reworking’ trends through clouds
of scattered Hf (or Nd) isotope data is likely to be a poor representation of the processes occurring

........................................................
over the zircon crystallization interval.
How geologically tenable are protracted crustal reworking models? The formation of a
reworking array does not require that melt be continually extracted from the same source rock
volume—this would mean loss of fertility to melt production (noted by [85]), and could change
source Lu/Hf, militating against a linear Hf isotope evolution being defined by successive
magmatic extracts. Rather, these reworking arrays can be explained by melting of a similar bulk
source composition (in terms of Lu–Hf systematics) in different places at different times. Input
from, and interaction with, mantle derived magmas is important for crustal reworking along
Phanerozoic convergent margins, as borne out by the complex isotope-time patterns highlighted
above [73,76], but perhaps less so for the early Earth given the potential for internal radiogenic
heat production to promote crustal fusion (e.g. [86]). Such a process, intermittently generating
small melt increments from an enriched source that is hot and near its solidus, may be apposite for
the Jack Hills zircons [33]. Taylor et al. [46] document an identical Hf isotope-time array from the
Apollo 14 lunar zircons, for which plate tectonic processes cannot be implicated (figure 1). Such
patterns tend to be interpreted in terms of ‘stagnant lid’ tectonics (e.g. [87]), although the cause of
the isotope arrays is unclear. Do they reflect episodic squeezing of melts from depth during bolide
impacts (e.g. [88,89], periodic overstepping of solidi in response to radiogenic heat flux, or the rate
of foundering of enriched, hydrated basaltic protocrust during volcanic resurfacing episodes [86]?
The geodynamic contrast between such scenarios and modern plate tectonics is stark.
3. Conclusion
Armstrong [2, p. 614] encapsulated the challenges of unravelling the earliest terrestrial evolution
in stating ‘Short of the invention of time machines, it is difficult to prove that any inference of
Earth history from Earth today is true’. Although this invention remains some way off, there have
been great strides in developing methods for time travel by proxy. It is now possible to measure
minute variations in isotopic abundance that reflect the decay of short-lived radionuclides in the
first 10–100 s of millions of years of solar system history, providing a remarkable chronology of
the formation of planetary cores and silicate mantles. There is growing recognition that studies of
the neighbouring rocky bodies in the solar system, which preserve the visible legacy of Hadean
magma oceans, bolide impacts and intense volcanism, may provide useful analogies for the
infant Earth (e.g. [90,91]). Formulating tests for recognizing the imprint of such processes in the
terrestrial geological record is the logical next step. Zircon grains in terrestrial rocks are time
capsules, albeit not always impervious to younger events, that capture thermal histories and the
composition of the medium (melt or fluid) from which these crystals grew. The ability of scientists
to extract and interpret chemical information from zircons and other accessory minerals is rapidly
escalating—this has involved the development of new tools (e.g. [23,92]) and sharpening of
existing ones. Further improvements in this field are predicted to come with better screening
procedures to identify the most ‘pristine’ zircon crystals that best preserve their chemical cargo
(e.g. [38]), and in tying that information to data from coexisting minerals and the host rock.
What of early Earth geodynamics? It must be acknowledged that, in the case of plate tectonics,
the absence of evidence does not constitute evidence for absence, given that the process itself is
efficient at disguising its geological ‘fingerprints’. Nonetheless, ‘smoking guns’ for the operation
of modern-style plate tectonics before 3.2 Ga are yet to be recognized. Perhaps these have long
since been destroyed, although radiogenic isotope data from (meta)igneous rocks suggests that
stabilization of significant continental volumes was only underway by 3.8–3.6 Ga, and that the
record of Archean continental crust may be less fragmented than generally considered. The
radiogenic isotopic record contains no legacy from crust that was rapidly recycled, however,
and thus the actual volume, composition and geodynamic mode of formation of crustal material
present on Earth’s surface at any given time remains difficult to appraise. Nevertheless, the basic
geological features of the best-preserved Archean terranes, such as in the Pilbara Craton, attest to
18
a distinctive mode of planetary evolution that does not operate on the planet today. Attempts to
view these terranes through the prism of modern day plate tectonics will do more to cloud the

........................................................
understanding of early Earth dynamics than it will to shed light on the formative, but enigmatic
Archean to Hadean eons.
Data accessibility. This article has no additional data.
Competing interests. I declare I have no competing interests.
Funding. This study was partly supported by ARC fellowship FT10010059.
Acknowledgements. The patience of the editors is admired and appreciated. Discussions with Jeff Vervoort, and
rapid and perceptive reviews by Chris Hawkesworth, Peter Cawood and Nigel Harris greatly improved the
presentation and helped to clear the mist around a number of points.
References
1. Armstrong RL. 1981 Radiogenic isotopes: the case for crustal recycling on a near-steady-
state no-continental-growth Earth. Phil. Trans. R. Soc. Lond. A 301, 443–472. (doi:10.1098/
rsta.1981.0122)
2. Armstrong RL. 1991 The persistent myth of crustal growth. Australian J. Earth Sci. 38, 613–630.
(doi:10.1080/08120099108727995)
3. Kay RW, Kay SM. 1993 Delamination and delamination magmatism. Tectonophysics 219,
177–189. (doi:10.1016/0040-1951(93)90295-U)
4. Taylor SR. 1967 The origin and growth of continents. Tectonophysics 4, 17–34. (doi:10.1016/
0040-1951(67)90056-X)
5. Tatsumi Y. 2005 The subduction factory: how it opersates in the evolving Earth. GSA Today 15,
4–10. (doi:10.1130/1052-5173(2005)015[4:TSFHIO]2.0.CO;2)
6. Davidson JP, Arculus RJ. 2006 The significance of Phanerozoic arc magmatism in generating
continental crust. In Evolution and differentiation of the continental crust (eds M Brown,
T Rushmer), pp. 135–172. Cambridge, UK: Cambridge University Press.
7. Hawkesworth CJ, Cawood P, Kemp AIS, Dhuime B, Storey CD. 2009 A matter of preservation.
Science 323, 49–50. (doi:10.1126/science.1168549)
8. Hawkesworth CJ, Dhuime B, Pietranik AB, Cawood P, Kemp AIS, Storey CD. 2010
The generation and evolution of the continental crust. J. Geol. Soc. London 167, 229–248.
(doi:10.1144/0016-76492009-072)
9. Collins WJ, Belousova EA, Kemp AIS, Murphy JB. 2011. Two contrasting Phanerozoic
orogenic systems revealed by hafnium isotope data. Nat. Geosci. 4, 333–337. (doi:10.1038/
ngeo1127)
10. Dhuime B, Hawkesworth C, Cawood P, Storey C. 2012 A change in the geodynamics of
continental growth 3 billion years ago. Science 335, 1334–1336. (doi:10.1126/science.1216066)
11. Squire RJ, Campbell IH, Allen CM, Wilson CJL. 2006 Did the Transgondwanan
Supermountain trigger the explosive radiation of animals on Earth? Earth Planet. Sci. Lett.
250, 116–133. (doi:10.1016/j.epsl.2006.07.032)
12. Maruyama S, Ikoma M, Genda H, Hirose K, Yokohama T, Santosh M. 2013. The naked planet
Earth: most essential pre-requisite for the origin and evolution of life. Geoscience Frontiers 4,
141–165. (doi:10.1016/j.gsf.2012.11.001)
13. Cox GM, Lyons TW, Mitchell RN, Hasterok D, Gard M. 2018 Linking the rise of atmospheric
oxygen to growth in the continental phosphorous inventory. Earth Planet. Sci. Lett. 489, 28–36.
(doi:10.1016/j.epsl.2018.02.016)
14. Hansen VL, López I. 2010 Venus records a rich early history. Geology 38, 311–314. (doi:10.1130/
G30587.1)
15. Bédard J. 2006 A catalytic delamination-driven model for coupled genesis of Archean
crust and sub-continental lithospheric mantle. Geochim. Cosmochim. Acta 70, 1188–1214.
(doi:10.1016/j.gca.2005.11.008)
16. Bédard J. 2013 How many arcs can dance on the head of a plume?: A ‘comment’ on: A critical
assessment of Neoarchean ‘plume only’ geodynamics: evidence from the Superior Province,
by Derek Wyman. Precambrian Res. 229, 189–197. (doi:10.1016/j.precamres.2012.05.004)
17. Moyen J-F, Laurent O. 2018 Archean tectonic systems: a view from igneous rocks. Lithos
19
302–303, 99–125. (doi:10.1016/j.lithos.2017.11.038)
18. Stern RJ. 2005 Evidence from ophiolites, blueschists and ultrahigh pressure metamorphic

........................................................
terranes that the modern episode of subduction tectonics began in neoproterozoic time.
Geology 33, 557–560. (doi:10.1130/G21365.1)
19. Brown M, Johnson T. 2018 Invited Centennial Article: secular change in metamorphism and
the onset of global plate tectonics. Am. Mineral. 103, 181–196. (doi:10.2138/am-2018-6166)
20. Veizer J, Jansen SL. 1985 Basement and sedimentary recycling—2. Time dimension to global
tectonics. J. Geol. 93, 625–643. (doi:10.1086/628992)
21. Wyman D. 2013 A critical assessment of Neoarchean ‘plume only’ geodynamics: evidence
from the Superior Province. Precambrian Res. 229, 3–19. (doi:10.1016/j.precamres.2012.01.010)
22. Harley SL, Kelly NM, Moller A. 2007 Zircon behaviour and the thermal history of mountain
chains. Elements 3, 25–30. (doi:10.2113/gselements.3.1.25)
23. Valley JW et al. 2014 Hadean age for a post-magma-ocean zircon confirmed by atom-probe
tomography. Nat. Geosci. 7, 219–223. (doi:10.1038/ngeo2075)
24. Moser DE, Bowman JR, Wooden J, Valley JW, Mazdab F, Kita N. 2008 Creation of a continent
recorded in zircon zoning. Geology 36, 239–242. (doi:10.1130/G24416A.1)
25. Froude DO, Ireland TR, Kinny PD, Williams IS, Compston W, Williams IR, Myers JS. 1983
Ion microprobe identification of 4100–4200 Myr-old terrestrial zircons. Nature 304, 616–618.
(doi:10.1038/304616a0)
26. Compston W, Pidgeon RT. 1986 Jack Hills, evidence of more very old detrital zircons in
Western Australia. Nature 321, 766–769. (doi:10.1038/321766a0)
27. Harrison TM. 2009 The Hadean crust: evidence from >4 Ga zircons. Annu. Rev. Earth Planet.
Sci. 37, 479–505. (doi:10.1146/annurev.earth.031208.100151)
28. Kemp AIS, Hickman AH, Kirkland C, Vervoort JD. 2015 Hf isotopes in detrital and inherited
zircons of the Pilbara Craton provide no evidence for Hadean continents. Precambrian Res. 261,
112–126. (doi:10.1016/j.precamres.2015.02.011)
29. Kemp AIS, Wilde S, Spaggiari C. 2018 The Narryer Terrane, Yilgarn Craton, Western Australia:
review and recent developments. In Earth’s oldest rocks (eds MJ Van Kranendonk, VC Bennett,
JE Hoffmann), 2nd edn. Amsterdam, The Netherlands: Elsevier.
30. Nemchin AA, Pidgeon RT, Whitehouse MJ. 2006 Re-evaluation of the origin and evolution of
>4.2 Ga zircons from the Jack Hills metasedimentary rocks. Earth Planet. Sci. Lett. 244, 218–233.
(doi:10.1016/j.epsl.2006.01.054)
31. Kamber BS. 2007 The enigma of the terrestrial protocrust: evidence for its former existence and
the importance of its complete disappearance. In Earth’s oldest rocks (eds MJ Van Kranendonk,
RH Smithies, VC Bennett), pp. 75–89. Amsterdam, Netherlands: Elsevier.
32. Shirey SB, Kamber BS, Whitehouse MJ, Mueller PA, Basu AR. 2008 A review of the isotopic
and trace element evidence for mantle and crustal processes in the Hadean and Archean:
implications for the onset of plate tectonic subduction. In When Did Plate Tectonics Begin on
Planet Earth? (eds KC Condie, V Pease). Geological Society of America Special Paper 440,
1–29.
33. Kemp AIS, Wilde SA, Hawkesworth CJ, Coath C, Nemchin AA, Pidgeon RT, Vervoort JD,
DuFrane SA. 2010 Hadean crustal evolution revisited: new constraints from Pb-Hf isotope
systematics of the Jack Hills zircons. Earth Planet. Sci. Lett. 296, 45–56. (doi:10.1016/j.epsl.
2010.04.043)
34. Bowring SA, Williams IS. 1999 Priscoan (4.00–4.03 Ga) orthogneisses from northwestern
Canada. Contrib. Mineral. Petrol. 134, 3–16. (doi:10.1007/s004100050465)
35. Cavosie AJ, Wilde SA, Liu D, Weiblen PW, Valley JW. 2004 Internal zoning and U-Th-Pb
chemistry of Jack Hills detrital zircons: a mineral record of early Archean to Mesoproterozoic
(4348–1576 Ma) magmatism. Precambrian Res. 135, 251–279. (doi:10.1016/j.precamres.2004.
09.001)
36. Trail D, Mojzsis SJ, Harrison TM. 2007 Thermal events documented in Hadean zircons
by ion microprobe depth profiles. Geochim. Cosmochim. Acta 71, 4044–4065. (doi:10.1016/
j.gca.2007.06.003)
37. Moorbath S. 1986 The most ancient rocks revisted. Nature 321, 725. (doi:10.1038/321725a0)
38. Ge R, Wilde SA, Nemchin AA, Whitehouse MJ, Belluci JJ, Erickson TM, Frew A, Thern ER.
2018 A 4463 Ma apparent zircon age from the Jack Hills (Western Australia) resulting from
ancient Pb mobilization. Geology 46, 303–306. (doi:10.1130/G39894.1)
39. Kusiak M, Whitehouse M, Wilde SA, Nemchin AA, Clark C. 2013 Mobilisatiobn of radiogenic
20
Pb in zircon revealed by ion imaging: implications for early Earth geochronology. Geology 41,
291–294. (doi:10.1130/G33920.1)

........................................................
40. Amelin Y, Lee D-C, Halliday AN. 2000 Early-middle Achaean crustal evolution deduced
from Lu-Hf and U-Pb isotopic studies of single zircon grains. Geochim. Cosmochim. Acta 64,
4205–4225. (doi:10.1016/S0016-7037(00)00493-2)
41. Vervoort JD, Kemp AIS. 2016 Clarifying the zircon Hf isotope record of crust-mantle
evolution. Chem. Geol. 425, 65–75. (doi:10.1016/j.chemgeo.2016.01.023)
42. Amelin Y, Lee D-C, Halliday AN, Pidgeon RT. 1999 Nature of the Earth’s earliest crust from
hafnium isotopes in single detrital zircons. Nature 399, 252–255. (doi:10.1038/20426)
43. Harrison TM, Blichert-Toft J, Muller W, Albarède F, Holden P, Mojzsis SJ. 2005 Heterogeneous
Hadean Hafnium: Evidence for Continental Crust at 4.4 to 4.5 Ga. Science 310, 1947–1950.
(doi:10.1126/science.1117926)
44. Harrison TM, Schmitt AK, McCulloch MT, Lovera OM. 2008 Early (4.5 Ga) formation of
˛ 18 O, and Ti thermometry results for Hadean zircons. Earth Planet.
terrestrial crust: Lu–Hf, ïAd’
Sci. Lett. 268, 476–486. (doi:10.1016/j.epsl.2008.02.011)
45. Blichert-Toft J, Albarède F. 2008 Hafnium isotopes in Jack Hills zircons and the formation of
the Hadean crust. Earth Planet. Sci. Lett. 265, 686–702.
46. Taylor DJ, McKeegan KD, Harrison TM. 2009 Lu-Hf zircon evidence for rapid lunar
differentiation. Earth Planet. Sci. Lett. 279, 157–164. (doi:10.1016/j.epsl.2008.12.030)
47. Kemp AIS, Hawkesworth CJ. 2014 Growth and differentiation of the continental crust
from isotope studies of accessory minerals. In Treatise on geochemistry (2nd edition), vol. 12
(ed. RL Rudnick). The Crust. Oxford: Elsevier Science.
48. Wilde SA, Valley JW, Peck WH, Graham CM. 2001 Evidence from detrital zircons for the
existence of continental crust and oceans on the Earth 4.4 Gyr ago. Nature 409, 175–178.
(doi:10.1038/35051550)
49. Holden P, Lanc P, Ireland TR, Harrison TM, Foster JL, Bruce Z. 2009 Mass spectrometric
mining of Hadean zircons by automated SHRIMP multi-collector and single-collector U/Pb
zircon age dating: the first 10,000 grains. Int. J. Mass spectrom. 286, 53–63. (doi:10.1016/
j.ijms.2009.06.007)
50. Nelson DR, Robinson BW, Myers JS. 2000 Complex geological histories extending for
≥4.0 Ga deciphered from xenocryst zircon microstructures. Earth Planet. Sci. Lett. 181, 89–102.
(doi:10.1016/S0012-821X(00)00186-2)
51. Campbell IH, Allen CM. 2008 Formation of supercontinents linked to increases in atmospheric
oxygen. Nat. Geosci. 1, 554–558. (doi:10.1038/ngeo259)
52. Hawkesworth CJ, Cawood P, Dhuime B, Kemp AIS. 2017 Earth’s continental lithosphere
through time. Annu. Rev. Earth Planet. Sci. 45, 169–198. (doi:10.1146/annurev-earth-
063016-020525)
53. Condie KC, Arndt N, Davaille A, Puetz SJ. 2017 Zircon age peaks: production or preservation
of continental crust? Geosphere 13, 227–234. (doi:10.1130/GES01361.1)
54. Kemp et al. In preparation.
55. Kinny PD, Williams IS, Froude DO, Ireland TR, Compston W. 1988 Early Archean zircon
ages from orthogneisses and anorthosites at Mount Narryer. West. Aust.: Precambrian Res. 38,
325–341. (doi:10.1016/0301-9268(88)90031-9)
56. Kinny PD, Nutman AP. 1996 Zirconology of the Meeberrie gneiss, Yilgarn Craton,
Western Australia: an early Archean migmatite. Precambrian Res. 78, 165–178.
(doi:10.1016/0301-9268(95)00076-3)
57. Iizuka T, Komiya T, Maruyama S, Hirata T, Hidaka H, Windley BF. 2006 4.2 Ga zircon
xenocryst in an Acasta gneiss from northwestern Canada: evidence for early continental crust.
Geology 34, 245–248. (doi:10.1130/G22124.1)
58. Iizuka T, Komiya T, Maruyama S. 2007 The early Archean Acasta Gneiss Complex: geological,
geochronological and isotopic studies and implications for early crustal evolution. In Earth’s
oldest rocks (eds MJ Van Kranendonk, RH Smithies, VC Bennett), pp. 127–147. Amsterdam,
The Netherlands: Elsevier.
59. Bauer AM, Fisher CM, Vervoort JD, Bowring SA. 2017 Coupled zircon Lu–Hf and U–Pb
isotopic analyses of the oldest terrestrial crust, the >4.03 Ga Acasta Gneiss Complex. Earth
Planet. Sci. Lett. 458, 37–48. (doi:10.1016/j.epsl.2016.10.036)
60. Reimink JR, Chacko T, Stern RA, Heaman LM. 2014 Earth’s earliest evolved crust generated
21
in in Iceland-like setting. Nat. Geosci. 7, 529–533. (doi:10.1038/ngeo2170)
61. Rollinson H. 2017 There were no large volumes of felsic continental crust in the early Earth.

........................................................
Geosphere 13, 235–246. (doi:10.1130/GES01437.1)
62. Myers JS. 1988 Early Archean Narryer Gneiss Complex, Yilgarn Craton, Western Australia.
Precambrian Res. 38, 297–307. (doi:10.1016/0301-9268(88)90029-0)
63. Rowe ML, Kemp AIS. Submitted. Origin of the Eoarchaean Manfred Complex: primitive
mantle source & oceanic plateau magmatism. Lithos.
64. Barnes SJ, Roeder PL. 2001 The range of spinel compositions in terrestrial mafic and ultramafic
rocks. J. Petrol. 42, 2279–2302. (doi:10.1093/petrology/42.12.2279)
65. Van Kranendonk MJ, Smithies RH, Hickman AH, Champion DC. 2007 Secular tectonic
evolution of Archean continental crust: interplay between horizontal and vertical processes.
Terra Nova 19, 1–38. (doi:10.1111/j.1365-3121.2006.00723.x)
66. Kemp AIS, Whitehouse MJ, Hawkesworth CJ, Alarcon MK. 2005 The implications of zircon
U-Pb isotope systematics for the genesis of metaluminous granites in the Lachlan Fold Belt,
southeastern Australia. Contrib Mineral Petrol 150, 230–249.
67. Richards SW, Collins WJ. 2004 Growth of wedge-shaped plutons at the base of active half-
grabens. Trans. R. Soc. Edinburgh: Earth Sci. 95, 309–317. (doi:10.1017/S0263593300001097)
68. Chew DM, Schaltegger U, Košler J, Whitehouse MJ, Gutjahr M, Spikings RA, Miškovíc A.
2007 U-Pb geochronologic evidence for the evolution of the Gondwanan margin of the north-
central Andes. Geol. Soc. Am. Bull. 119, 697–711. (doi:10.1130/B26080.1)
69. Smithies RH, Van Kranendonk MJ, Champion DC. 2005 It started with a plume- early
Archaean basaltic proto-continental crust. Earth Planet. Sci. Lett. 238, 284–297. (doi:10.1016/
j.epsl.2005.07.023)
70. Collins WJ, Van Kranendonk MJ, Teyssier C. 1998 Partial convective overturn of Archean crust
in the east Pilbara Craton, Western Australia: driving mechanisms and tectonic implications.
J. Struct. Geol. 20, 1405–1424. (doi:10.1016/S0191-8141(98)00073-X)
71. François C, Philipott P, Rey P, Rubatto D. 2014 Burial and exhumation during Archean
sagduction in the East Pilbara granite-greenstone terrane. Earth Planet. Sci. Lett. 396, 235–251.
(doi:10.1016/j.epsl.2014.04.025)
72. Allegre CJ. 1987 Isotope geodynamics. Earth Planet. Sci. Lett. 86, 175–203. (doi:10.1016/
0012-821X(87)90220-2)
73. Kemp AIS, Hawkesworth CJ, Collins WJ, Gray CM, Blevin PL. 2009 Isotopic evidence for
rapid continental growth in an extensional accretionary orogen: The Tasmanides, eastern
Australia. Earth Planet. Sci. Lett. 284, 455–466. (doi:10.1016/j.epsl.2009.05.011)
74. Cawood PA. 2005 Terra Australis Orogen: Rodinia breakup and development of the Pacific
and Iapetus margins of Gondwana during the Neoproterozoic and Paleozoic. Earth Sci. Rev.
69, 249–279. (doi:10.1016/j.earscirev.2004.09.001)
75. Miskovic A, Schaltegger U. 2009 Crustal growth along a non-collisional cratonic margin: a
Lu-Hf isotopic survey of the Eastern Cordilleran granitoids of Peru. Earth Planet. Sci. Lett. 279,
303–315. (doi:10.1016/j.epsl.2009.01.002)
76. DeCelles PG, Ducea MN, Kapp P, Zandt G. 2009 Cyclicity in Cordilleran orogenic systems.
Nat. Geosci. 2, 251–257. (doi:10.1038/ngeo469)
77. Naeraa T, Schersten A, Rosing MT, Kemp AIS, Hofmann JE, Kokfelt TF, Whitehouse MJ. 2012
Hafnium isotope evidence for a transition in the geodynamics of continental growth after
3.2 Ga. Nature 485, 627–630. (doi:10.1038/nature11140)
78. Kemp AIS, Vervoort JD, Smithies RHS, Hickman AH, Van Kranendon MJ. 2011. Recoupling
the Nd-Hf isotope record of the early Earth? Evidence from the Pilbara Craton, Western
Australia. In AGU Fall Meeting Abstracts.
79. Shirey SB, Richardson SH. 2011 Start of the Wilson Cycle at 3 Ga shown by diamonds from
subcontinental mantle. Science 333, 434–436. (doi:10.1126/science.1206275)
80. Hiess J, Bennett VC. 2016 Chondritic Lu/Hf in the early crust-mantle system as recorded by
zircon populations from the oldest Eoarchean rocks of the Yilgarn Craton, West Australia
and Enderby Land, Antarctica. Chem. Geol. 427, 125–143. (doi:10.1016/j.chemgeo.2016.
02.011)
81. Fisher CM, Vervoort JD. 2018 Using the magmatic record to constrain the growth of
continental crust – the Eoarchean zircon Hf record of Greenland. Earth Planet. Sci. Lett. 488,
79–91. (doi:10.1016/j.epsl.2018.01.031)
82. Gerdes A, Zeh A. 2009 Zircon formation versus zircon alteration - new insights from combined
22
U-Pb and Lu-Hf in-situ LA-ICP-MS analyses, and consequences for the interpretation of
Archean zircon from the Central Zone of the Limpopo Belt. Chem. Geol. 261, 230–243.

........................................................
(doi:10.1016/j.chemgeo.2008.03.005)
83. Laurent O, Zeh A. 2015 A linear Hf isotope-age array despite different granitoid sources and
complex Archean geodynamics: example from the Pietersburg block (South Africa). Earth
Planet. Sci. Lett. 430, 326–338.
84. Fisher CM, Vervoort JD. 2018 Using the magmatic record to constrain the growth of
continental crust—the Eoarchean zircon Hf record of Greenland. Earth Planet. Sci. Lett. 488,
79–91.
85. Payne J, McInerney DJ, Barovich KM, Kirkland CL, Pearson N, Hand M. 2016 Strengths and
limitations of zircon Lu-Hf and O isotopes in modeling crustal growth. Lithos 248–251, 175–
192. (doi:10.1016/j.lithos.2015.12.015)
86. Kamber BS, Whitehouse MJ, Bolhar R, Moorbath S. 2005 Volcanic resurfacing and the early
terrestrial crust: zircon U-Pb and REE constraints from the Isua Greenstone Belt, southern
West Greenland. Earth Planet. Sci. Lett. 240, 276–290. (doi:10.1016/j.epsl.2005.09.037)
87. O’Neill C, Debaille V, Griffin W. 2013 Deep Earth recycling in the Hadean and constraints on
surface tectonics. Am. J. Sci. 313, 912–932. (doi:10.2475/09.2013.04)
88. Nemchin AA, Pidgeon RT, Whitehouse MJ, Vaughan JP, Meyer C. 2008 SIMS U–Pb study of
zircon from Apollo 14 and 17 breccias: implications for the evolution of lunar KREEP. Geochim.
Cosmochim. Acta 72, 668–689. (doi:10.1016/j.gca.2007.11.009)
89. Marchi S et al. 2014 Widespread mixing and burial of Earth’s Hadean crust by asteroid
impacts. Nature 511, 578–582. (doi:10.1038/nature13539)
90. Hansen VL. 2007 Venus: A thin-lithosphere analog for early Earth? In Earth’s oldest rocks
(eds MJ Van Kranendonk, RH Smithies, VC Bennett), pp. 971–983. Amsterdam, Netherlands:
Elsevier.
91. Hansen VL. 2015 Impact origin of Archean Cratons. Lithosphere 7, 563–578. (doi:10.1130/
L371.1)
92. Watson EB, Harrison TM. 2005 Zircon thermometer reveals minimum melting conditions on
earliest Earth. Science 308, 841–844. (doi:10.1126/science.1110873)

RST A 20180169

Uploaded by

Copyright:

Available Formats

RST A 20180169

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

RST A 20180169

Uploaded by

Copyright:

Available Formats

Early earth geodynamics: cross

examining the geological

1. Geodynamics and Earth’s continental crust

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

2. Terrestrial archives of geodynamic processes

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

(a) The Hadean detrital record

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

3300 3500 3700 3900 4100 4300 4500

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

(b) Igneous rock record

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

(ii) The Acasta Gneiss complex, Slave Craton, Canada

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

31.5 32.5 33.5 34.5 35.5

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

(iii) The Narryer Gneiss complex, Yilgarn Craton, Western Australia

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

(iv) The East Pilbara Terrane, Pilbara Craton, Western Australia

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

(c) Fe3+ Fe3+

Cr3+ Al3+ Cr3+ Al3+

Ocean Island Boninites

Cr3+ Al3+ Cr3+ Al3+

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

eHf (t) zircon –5

–15 176Lu/177Hf = 0.0022

(ii) The perceived perils of isotope-time arrays

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 376: 20180169

You might also like