a planet in transition: the onset of plate tectonics on ... · kent c. condie department of earth...

Geoscience Frontiers xxx (2016) 1e10

HOSTED BY Contents lists available at ScienceDirect

China University of Geosciences (Beijing)

Geoscience Frontiers

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

Research Paper

A planet in transition: The onset of plate tectonics on Earth between 3and 2 Ga?

Kent C. CondieDepartment of Earth and Environmental Science, New Mexico Tech, Socorro, NM 87801, USA

a r t i c l e i n f o

Article history:Received 5 July 2016Received in revised form16 August 2016Accepted 13 September 2016Available online xxx

Keywords:Plate tectonicsZircon age peaksMantle evolutionStagnant lidContinental crustLIP events

E-mail addresses: [email protected], kentcondie1@Peer-review under responsibility of China University

http://dx.doi.org/10.1016/j.gsf.2016.09.0011674-9871/� 2016, China University of Geosciences (BND license (http://creativecommons.org/licenses/by-n

Please cite this article in press as: Condie, KFrontiers (2016), http://dx.doi.org/10.1016/j.

a b s t r a c t

Many geological and geochemical changes are recorded on Earth between 3 and 2 Ga. Among the moreimportant of these are the following: (1) increasing proportion of basalts with “arc-like” mantle sources;(2) an increasing abundance of basalts derived from enriched (EM) and depleted (DM) mantle sources;(3) onset of a Great Thermal Divergence in the mantle; (4) a decrease in degree of melting of the mantle;(5) beginning of large lateral plate motions; (6) appearance of eclogite inclusions in diamonds; (7)appearance and rapid increase in frequency of collisional orogens; (8) rapid increase in the productionrate of continental crust as recorded by zircon age peaks; (9) appearance of ophiolites in the geologicrecord, and (10) appearance of global LIP (large igneous province) events some of which correlate withglobal zircon age peaks. All of these changes may be tied directly or indirectly to cooling of Earth’s mantleand corresponding changes in convective style and the strength of the lithosphere, and they may recordthe gradual onset and propagation of plate tectonics around the planet. To further understand thechanges that occurred between 3 and 2 Ga, it is necessary to compare rocks, rock associations, tectonicsand geochemistry during and between zircon age peaks. Geochemistry of peak and inter-peak basaltsand TTGs needs to be evaluated in terms of geodynamic models that predict the existence of an episodicthermal regime between stagnant-lid and plate tectonic regimes in early planetary evolution.

� 2016, China University of Geosciences (Beijing) and Peking University. Production and hosting byElsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/

licenses/by-nc-nd/4.0/).

1. Introduction

Although many geologic and geochemical changes are reportedat the end of the Archean (Taylor and McLennan, 1985; Condie andO’Neill, 2010; Condie, 2015a), the timing, duration, and causes ofthese changes are not well known. Geodynamic modeling suggeststhat Earth may have begun as a hot stagnant-lid similar to Io andevolved through an episodic transitional state into plate tectonicsover 1e3 Gyr (O’Neill et al., 2016). Perhaps the changes near the endof the Archean reflect this transitional state. Modeling has shownthat under hotter conditions typical of the early Earth, subductedslabs are weaker andmore likely to detach and sink into the mantle(Van Hunen and Moyen, 2012), and this should lead to episodicsubduction events in a stagnant lid tectonic regime. Also the lowerviscosity of the Hadean mantle would result in lower lithosphericstress levels preventing the initiation of subduction, which couldpromote “bursts” of subduction as mantle cooling progressed(O’Neill et al., 2007; O’Neill and Debaille, 2014).

gmail.com.of Geosciences (Beijing).

eijing) and Peking University. Produc-nd/4.0/).

.C., A planet in transition: Thgsf.2016.09.001

To tie geodynamic models to the geologic record, it is necessaryto have precise ages of global events in the period of time from 4 to2 Ga. We have made progress in this direction in the last 10 yearsfrom advances in zircon geochronology, and our zircon U/Pb agedatabase has grown rapidly for both igneous and detrital zircons(Condie and Aster, 2010; Voice et al., 2011; Hawkesworth et al.,2013; Roberts and Spencer, 2015). In this contribution, I reviewten geologic and geochemical changes that occurred on Earth be-tween 3 and 2 Ga. Any transition of thermal and tectonic regimesduring this time interval must accommodate these changes.

2. Transitional events between 3 and 2 Ga

2.1. Increasing abundance of basalts with “arc-like” mantle sources

Because ratios of elements with similar incompatibilities do notchange much with degree of melting or fractional crystallization ofmantle-derived magmas, they are useful in characterizing compo-sitions of mantle sources of basalts that have not interacted withcontinental crust (Condie, 2003, 2005). Using Nb/Th and Zr/Nbratios in young oceanic basalts, three mantle domains can be

ction and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-

e onset of plate tectonics on Earth between 3 and 2 Ga?, Geoscience

Delta:1_given name

http://creativecommons.org/licenses/by-nc-nd/4.0/


mailto:[email protected]

mailto:[email protected]

www.sciencedirect.com/science/journal/16749871

http://www.elsevier.com/locate/gsf

http://dx.doi.org/10.1016/j.gsf.2016.09.001




K.C. Condie / Geoscience Frontiers xxx (2016) 1e102

identified (Condie, 2015b): enriched (EM), depleted (DM), and hy-drated mantle (HM). DM is characterized by high values of bothratios (Nb/Th > 8; Zr/Nb > 20) due to Th depletion relative to Nb,and to Nb depletion relative to Zr. EM has high Nb/Th ratios buttypically has Zr/Nb ratios <20. In contrast, HM has very low Nb/Th(<8) and variable Zr/Nb ratios. Similar domains are recognized withTh/YbeNb/Yb and La/SmeNb/Th relationships. DM is most wide-spread in the asthenosphere tapped at ocean ridges, but also mayappear in mantle wedges associated with subduction (Hofmann,1997; Stracke, 2012). EM is a common mantle source for basaltserupted in oceanic plateaus and islands, where it is thought tooccur in mantle plumes and as inhomogeneities in the astheno-sphere. As shown in Fig. 1, the proportion of oceanic basalts withHM sources increases with time from values around 20% in theEoarchean to about 80% by the end of the Archean and remains atthe 80e100% level thereafter. The only exceptions are basalts fromKarelia in Finland and Russia around 2 Ga; only 60e70% of theserocks reflect HM sources. Although it is possible that HM basaltscan also be produced in a stagnant lid regime, the rapid increase inthe proportion of basalts with an HM source between 3.5 and 2 Gais consistent with rapid propagation of plate tectonics around theglobe during this time interval.

2.2. Increase in abundance of basalts derived from enriched (EM)and depleted (DM) mantle sources

It is possible that comparison of incompatible element ratios inbasalts from terrestrial oceanic tectonic settings with mare basaltsfrom the Moon and to shergottites from Mars may be usefulin constraining when plate tectonics began on Earth, since thesetwo bodies are examples of stagnant lid planets. Terrestrial oceanicbasalts from greenstones and their corresponding mantle sources(EM, DM, HM) can be tracked to 2.5e2.0 Ga with geologic andgeochemical characteristics (Condie, 2015b) (Figs. 2 and 3). Prior tothis time, however, DM and EM basalts are difficult to distinguish interms of incompatible element ratios, and the two populations

Figure 1. Secular change in the percent of oceanic basalts derived from HM (hydratedmantle) sources with age. Each point is the median value for a 100-Myr movingwindow. Graph does not include samples that are contaminated with older continentalcrust. Data from Condie and Shearer (2016).

Figure 2. Zr/NbeNb/Th graph for oceanic non-arc basalts. (a) EM basalts, 2.5e0.15 Ga;(b) DM basalts, 2.5e0.15 Ga; (c) EM-DM undifferentiated oceanic basalts 3.4e2.5 Ga.PM: primitive mantle from McDonough and Sun (1995); CONT: average continentalcrust from Rudnick and Gao (2014). Data from Condie (2015b) and Condie and Shearer(2016).

Please cite this article in press as: Condie, K.C., A planet in transition: The onset of plate tectonics on Earth between 3 and 2 Ga?, GeoscienceFrontiers (2016), http://dx.doi.org/10.1016/j.gsf.2016.09.001

Figure 3. Zr/Nb versus age in oceanic non-arc basalts. Arrowheads are median valueswith 1s uncertainty as vertical purple bars. DM: depleted mantle; PM: primitivemantle; EM: enriched mantle; undifferentiated basalts include only those with oceanicaffinities. Other information given in caption of Fig. 2.

Figure 4. Zr/NbeNb/Th graph for lunar and martian basalts. (a) Apollo 11, 12 and 15lunar basalts; (b) Martian shergottites and nakhilite. Classification of shergottites(enriched, intermediate, depleted) from Greenough and Ya’acoby (2013). Data fromCondie and Shearer (2016), other information given in caption of Fig. 2.

K.C. Condie / Geoscience Frontiers xxx (2016) 1e10 3

seem to collapse into one group with a Zr/Nb ratio of about 20(Fig. 3). This distribution is similar to the distribution of most marebasalts from the Moon sampled by the Apollo missions, except forthe high-Ti basalts from Apollo 11 and 17 both of which areenriched in Nb due to ilmenite fractionation. The primitive mantlegrouping is especially evident for Apollo 11, 12 and 15 basalts inwhich samples cluster tightly around primitive mantle (PM)composition on Zr/NbeNb/Th (Fig. 4), Th/YbeNb/Yb and La/SmeNb/Th diagrams (Condie and Shearer, 2016). Shergottites fallinto three groups (enriched, intermediate and depleted; Greenoughand Ya’acoby, 2013), which form a scattered population on theincompatible element ratio diagrams as illustrated by the Zr/NbeNb/Th diagram (Fig. 4b). Black Beauty is a shergottite brecciathat may be a sample of mixed mantle sources and it falls nearprimitive mantle (PM) composition. The other shergottites have Zr/Nb ratios both below and above primitive mantle. Even moreimportant is the fact that none of the lunar and very few of themartian basalts fall in the terrestrial DM or EM fields on theincompatible element diagrams.

The fact that prior to 2.5 Ga, Earth’s non-arc-like oceanic basaltshave similar incompatible element signatures suggests they comefrom a relatively unfractionated mantle source similar to primitivemantle in composition. Such a source may be characteristic of astagnant lid planet like Mars or a satellite like the Moon, and if so,this suggests that Earth also may have been in a stagnant lid regimeprior to 2.5 Ga. An onset of plate tectonics on Earth around 3 Gamay have led to the production and growth of DM and EM mantlesources after this time. From studies of terrestrial mantle xenoliths(Griffin et al., 2003) and from geodynamic modeling (Komiya,2004), it has also been shown that the composition of the uppermantle changed rapidly during the Proterozoic.

2.3. Onset of a Great Thermal Divergence in the mantle

Based onmajor element distributions in oceanic basalts throughtime, Condie et al. (2016a,b) estimated magma generation tem-peratures (approximately equal to mantle potential temperatures[Tp]) of the three types of mantle HM, DM and EM. Young EM ba-salts yield an average Tp about 150 �C higher than the ambientupper mantle temperature as recorded by DM basalts, whereas inthe Archean the calculated Tps from DM and EM basalts areindistinguishable (ca. 1500 �C) (Fig. 5). Unlike the calculated Tps

Please cite this article in press as: Condie, K.C., A planet in transition: ThFrontiers (2016), http://dx.doi.org/10.1016/j.gsf.2016.09.001

from DM and HM basalts, there is very little change in the averageTp of EM basalts, which remains close to 1500 �C through time. Thismeans that Tp of EM sources was indistinguishable from ambientmantle (DM-HM) before 2.5e2.0 Ga, and during this time intervalthe temperature of these sources began to diverge, which isreferred to as a “Great Thermal Divergence” in the mantle (Condieet al., 2016a) (Fig. 5). This divergence is interpreted to reflectcooling of ambient mantle from 2.5 Ga onwards, and prior to 2.5 Ga,EM and DM-HM are indistinguishable. Perhaps this cooling couldreflect the onset and propagation of subduction between 3 and2 Ga. DM and HM also show an increase in Mg# after 2.5 Ga, whichis consistent with progressive depletion of ambient mantle afterthis time (Condie et al., 2016a). Komatiites come from a high-temperature mantle source (probably from plumes generated inthe D00 layer at the base of the mantle), which may have cooledslightly at the end of the Archean. Although this source is recog-nized throughout time, it is rarely tapped after the end of theArchean, when komatiites decrease rapidly in abundance.


Man

tle P

oten

tial T

empe

ratu

re (°

C)

Age (Ga)

42.50

AmbientMantle

PLATE TECTONICS

1600

1500EM

DM+HM

TransitionalTectonics

STAGNANT LID TECTONICS

Komatiite plumes

EM type Plumes

1350

1700KM

Figure 5. Mantle potential temperature as calculated from major element composi-tions of oceanic basalts showing a Great Thermal Divergence beginning at the end ofthe Archean. Mantle sources: KM, komatiite; EM, enriched; DM, depleted; HM, hy-drated. Modified from Condie et al. (2016a).


2.4. Decrease in degree of melting of the mantle

Keller and Schoene (2012) documented a decrease in compat-ible element contents and a corresponding increase in incompat-ible elements in mafic igneous rocks through time (Fig. 6). They

Ni (

ppm

)

100

300

500

1000

1400

Cr (

ppm

)

200

600

12

15

MgO

(wt.

%)

6

9

App

aren

t Per

cent

Mel

t

10

20

30

40

1000 2000 3000 4000

Age (Ma)

0ArcheanProterozoicPhanero-

zoic

Figure 6. Secular changes in MgO, Ni and Cr in mafic igneous rocks and calculatedchange in percent melting of the mantle. Error bars are 2 standard errors of the mean.Modified from Keller and Schoene (2012), courtesy of Brenhin Keller.


show that this trend cannot be related to changes in the degree offractional crystallization, but is probably related to a decreasingdegree of melting of the mantle through time as Earth cooled. Theirresults show a progressive decrease in average melt fraction inmantle sources from about 0.35 in the Archean to 0.1 today.Although as pointed out by Keller and Schoene (2012), this rapiddecrease in degree of melting is at least partly due to non-linearmelting of the mantle and exhaustion of clinopyroxene in thesource, the cooling rate might also be enhanced by increasing thevolume of cool slabs subducted into the mantle. Most notably is therapid decrease in cooling rate between 3 and 2 Ga (Fig. 6). Again,this is consistent with the rapid spread of plate tectonics during thistime interval.

2.5. Beginning of large lateral plate motions

Perhaps the most robust argument for plate tectonics is paleo-magnetic evidence for large lateral plate motions in the geologicpast. Although robust paleomagnetic data clearly indicate that by3.5 Ga Earth had a stable geomagnetic field probably undergoingmagnetic reversals (Usui et al., 2009; Biggin et al., 2011), evidencefor large lateral plate motions during the Archean remainscontroversial. To unambiguously support large lateral plate mo-tions, it is critical to have demagnetization data that includes vectorsubtraction, precise dating of the magnetization, an adequatenumber of samples, field tests to constrain the magnetization age,and tectonic coherence within a given block or craton (Van der Voo,1993; Pisarevsky et al., 2014). The oldest lateral plate motions ofthousands of kilometers for which such high-quality paleomag-netic evidence is available is around 2 to 2.5 Ga. Mitchell et al.(2014) showed partial apparent polar wander paths for the Slaveand Superior cratons that indicate large separation of these cratonsbefore final collision and amalgamation around 2.0e2.2 Ga. Like-wise, the paleomagnetic results of Bispo-Santos et al. (2014)showed large lateral motions of the Guiana and West African cra-tons leading to their linkage into a single craton at 1.98e1.96 Ga.Using angular speeds of plates calculated from paleomagnetic data,Condie et al. (2015a) showed that average plate speed has increasedwith time from about 30�/100 Myr at 2 Ga to 50�/100 Myr in thelast 100 Myr (Fig. 7). Hence, it would appear that at least by2.5e2.0 Ga, plate tectonics was operative on Earth, althoughwhether or not it covered the entire planet is still uncertain.

2.6. Appearance of eclogite inclusions in diamonds

Although Archean eclogites are known both in outcrop and asmantle or crustal xenoliths, eclogites do not become commonuntil after the end of the Archean (Rollinson, 1997; Barth et al.,2001). Mafic crust will invert to eclogite at high pressures,beginning at about 50 km depth and should sink into the mantle,which is less dense. Another way to track recycling of eclogite intothe mantle is with mineral inclusions in diamonds that formed inthick Archean lithospheric roots. Diamonds tend to protect theinclusions from later metasomatic effects and alteration. Shireyand Richardson (2011) recognized both peridotitic and eclogiticinclusions in diamonds from kimberlite pipes, the latter of whichappear rather suddenly around 3 Ga. This provides compellingevidence that from 3 Ga onwards eclogite was captured andpreserved in the subcontinental lithosphere, perhaps linked todeep subduction. The absence of eclogite inclusions in Archeandiamond populations is even more important in than it suggeststhat subduction did not occur prior to this time. The major caveatto this conclusion is that most of the data come from the Kaapvaalcraton in southern Africa, and hence there could be a geographicbias.


Figure 7. Secular changes in craton collision frequency and average area-weightedplate speed (deg/100 Myr) (modified from Condie et al., 2015a). Collision frequencybetween cratons is expressed as number of orogen segments per 100-Myr bin movingin 100 Myr increments. Lines are linear regression analysis: n ¼ 8.68 e 0.00224a,r ¼ 0.287; s ¼ 9.927 - 0.00223a, r ¼ 0.393 (n, number of orogens; s, plate speed dividedby five; a, age). Also shown are supercontinent assembly (blue stripes) and breakup(pink stripes) times.

4.01.0 2.0 3.0Age (Ga)

20

40

60

80

Volu

me

Cont

inen

tal C

rust

(%)

50%

12

3

Figure 8. Models for cumulative growth of continental crust. Curves 1 and 2 are basedon Hf isotope model ages of zircons and curve 3 is based on whole-rock Nd isotopesand zircon ages. 1, Dhuime et al. (2012); 2, Belousova et al. (2010); 3, Condie et al.(2016b).


2.7. Appearance and rapid increase in frequency of collisionalorogens

A collisional orogen is an orogen produced by the collision oftwo cratons, such as the modern Himalayan and Alpine orogens.The earliest well documented appearance of collisional orogens isnear the end of the Archean: Majorquq 2.6 Ga in West Greenland(Dyck et al., 2015), MacQuoid 2.56 Ga in Western Canada (Pehrssonet al., 2013), Commonwealth Bay 2.5 Ga in Antarctica (Menot et al.,2005), Nito Rodrigues 2.48 Ga in Uruguay (Hartmann et al., 2004),and Sleafordian 2.47 Ga in South Australia and the Mawson cratonin Antarctica (Hand et al., 2007). Although there are numerousorogens in the Archean, most appear to accretionary orogens,although some investigations have reported terrane collisionswithin Archean accretionary orogens. However, terrane collisionsdo not require plate tectonics as they may occur in response toconvective flow in the asthenosphere (Harris and Bedard, 2014). By2.0 Ga, craton-craton collisions are common in the geologic recordwith a large peak in frequency at about 1.9 Ga (Fig. 7) (Condie et al.,2015a). This dramatic increase in frequency of craton collisions inthe Paleoproterozoic would seem to demand widespread propa-gation of plate tectonics.

2.8. Rapid increase in volume of continental crust

In recent years Hf isotopes have been the preferred method toconstrain continental growth rates (Condie and Aster, 2010, 2013;Dhuime et al., 2011, 2012; Hawkesworth et al., 2013). However, asdiscussed in detail by Roberts and Spencer (2015), there are manyuncertain assumptions in using Hf model ages and hence growthrates based on Hf (or other) isotopic model ages remain


controversial and uncertain. Because both 3Hf and 3Nd distribu-tions with age show prominent vertical data arrays, growthmodels based on mixing of mantle and crustal endmembers avoidthe model age uncertainties (Condie et al., 2016b). Also, any modelfor the growth of continental crust must address the episodicity ofages shown by both outcrop and detrital zircon populations(Condie and Aster, 2010, 2013; Hawkesworth et al., 2013). Whole-rock Nd isotopes in convergent-margin-related felsic igneousrocks show the fraction of juvenile continental crust producedduring zircon age peaks in the last 2 Gyr is similar to that pro-duced during age valleys (Condie et al., 2016b). This suggests thatthe mechanisms of continental growth are similar in age peaksand valleys, most probably related to subduction. Unlike crustalgrowth models based on Hf model ages, the Nd end-membermixing model shows a step-like cumulative growth curve (curve3, Fig. 8) for continental crust with rapid growth at 3.0e2.4,2.0e1.3 and 0.6e0.45 Ga. The dominance of juvenile continentalcrust in accretionary orogens does not require selective preser-vation during craton-craton collisions (as suggested byHawkesworth et al., 2009) as an explanation of global zircon agepeaks, and thus favors the interpretation of these peaks as crustalproduction peaks (Condie et al., 2016b). The Nd-based continentalgrowth pattern differs significantly from results based on Hfmodel ages, which show most continental growth between 4 and2 Ga and lacks the step-like pattern of the Nd model (Belousovaet al., 2010; Dhuime et al., 2012) (curve 3, Fig. 8). The Nd modelshows about 50% of continental growth by 1.9 Ga, whereas the Hfmodels show 50% growth by 3.5e3.0 Ga. The rapid growth ofcontinental crust based on the Nd isotope model at 0.6e0.45 Gacorresponds to the assembly of Pangea-Gondwana, the 2.0e1.3 Gagrowth partly overlaps the assembly of Nuna at 1.7e1.5 Ga, andthe 3.0e2.4 Ga growth parallels supercraton assembly in theNeoarchean at 2.7e2.5 Ga (Condie et al., 2016b). Even detritalzircons preserve a message of change in the Paleoproterozoic. Thed18O of detrital zircons shows a large increase in the amount ofreworked sediment between 2.7 and 2 Ga, probably reflecting therapid growth of continental crust (Spencer et al., 2014). However,for all models of continental growth, the amount of materialrecycled back into the mantle is uncertain, so the cumulativegrowth curves (Fig. 8) are minima in terms of absolute crustalgrowth.

Numerous geochemical trends in the continental crust showincreasing rates beginning in the Archean, and these are consistent


Figure 9. Secular distribution of LIPs (large igneous provinces). Data from Condie et al.(2015b with updates).


with both increasing rates of growth of continental crust and withthe widespread propagation of plate tectonics between 3 and 2 Ga.For instance, rapid decreases in Ni/Co and Cr/Zn in shales, whichtrack MgO distribution in the sources, suggest that the uppercontinental crust changed from largelymafic before 3 Ga to felsic by2.5 Ga (Tang et al., 2016). This compositional change involved afivefold increase in the mass of the upper continental crust due toaddition of felsic igneous rocks, consistent with an onset of globalplate tectonics around 3.0 Ga. In another study, Dhuime et al. (2015)show a dramatic increase in Rb/Sr ratio in igneous rocks between 3and 2 Ga. They interpret this change to reflect rapid growth ofcontinental crust associated with the onset and propagation ofplate tectonics. Significant increases in large-ion lithophile andhigh-field strength elements in continental crust also occur be-tween 3 and 2 Ga (Condie and O’Neill, 2010; Condie, 2015a). De-creases in (La/Yb)n and Sr/Y and an increase in K2O/Na2O ingranitoids after 2.5 Ga reflect greater degrees of fractionation offelsic magmas at shallower depths. Also attesting to a change in thecomposition of upper continental crust frommafic (and komatiitic)to felsic is a drop in the Ni/Fe ratio in banded iron formations be-tween 2.7 and 2.0 Ga (Konhauser et al., 2009).

2.9. Appearance of ophiolites in the geologic record

The term ophiolite has been used and misused in the literaturesuch that the meaning is ambiguous (Condie, 2015a,b). Some in-vestigators have gone so far as the call almost all Archean green-stones “ophiolites” (Furnes et al., 2015). Herein, I use ophiolite torefer to a preserved fragment of crust (�mantle) formed at anocean ridge (associated with or not associated with subduction).This means that arc volcanics without evidence that theyformed at ocean ridges should not be called ophiolites. Ophiolitesshould preserve one or more of the definitive components(such as sheeted dykes, podiform chromite, and serpentinizedharzburgite). Although ophiolites have been reported in theArchean, none of them has passed this rigorous definitional filter.The first appearance of ophiolites is around 2 Ga, and we have atleast six well-described ophiolites ranging in age from 2.1 to1.85 Ga (Table 1). Only two of these contain complete ophiolitestratigraphy: Purtuniq and Jormua (Scott et al., 1992; Peltonenet al., 1996). A review of the greenstone literature, however, sug-gests that many dismembered ophiolites occur in this time rangepossibly extending to 2.5 Ga. Clearly, the first robust evidence wehave of ocean ridges appears in the Paleoproterozoic, and thusrequires plate tectonics.

2.10. Appearance of global LIP events

Global LIP (large igneous province) events, defined as occurringon six or more cratons, are generally thought to be the result ofwidespread mantle plume events (Condie et al., 2015b). They arefirst recorded beginning about 3 Ga with six significant eventsbetween 3 and 2 Ga (3.0, 2.7, 2.45, 2.2, 2.05, and 1.9 Ga; Fig. 9).Although older LIP events are known, they appear to have only local

Table 1Examples of Paleoproterozoic ophiolites.

Location Age (Ma) Reference

Kandra, India 1850 Kumar et al. (2010)Nagssugtoquidian, Greenland 1900e1850 Garde and Hollis (2010)Amisk, Flin Flon, Canada 1900 Stern et al. (1995)Jormua, Finland 1995 Peltonen et al. (1996)Purtuniq, Eastern Canada 1998 Scott et al. (1992)Central Karelia 2100 Stepanova et al. (2014)


or atmost regional geographic extent, and of the LIP events<1.9 Ga,only the 100 Ma event has been widely sampled and discussed.Why should so many global LIP events occur in the 3 to 2 Ga timeinterval? Possibly rapid accumulation of subducted slabs at thebase of the mantle was responsible for triggering widespreadmantle plume events.

3. Discussion

Although it has long been recognized that significant changesoccur at the end of the Archean, it is now clear that most, if not all,of these changes were not sudden, but occurred more graduallyover�1 Gyr. They all seem to be tied directly or indirectly to coolingof Earth’s mantle and corresponding changes in convective styleand the strength of the lithosphere, and they may record thegradual onset and propagation of plate tectonics around the planet.Although many questions should be addressed to better under-stand these changes, I would like to focus on four questions relatedto the ten transitions discussed above.

3.1. Can felsic crust be produced in a stagnant lid tectonic regime,and if so, in what volume can it be produced?

Do large volumes of TTG (tonalite-trondhjemite-granodiorite)magma require convergent margins, with either thickened maficcrust or/and descending slabs as TTG sources? Or can theserocks also be produced in a stagnant lid setting by mechanismssuch as heat-pipe cycling or mantle plumes (Bédard, 2006;Van Kranendonk, 2010; Moore and Webb, 2013; Johnson et al.,2014)? If we examine modern oceanic plateaus as the closestanalog to a stagnant lid, Iceland is the only example where felsicigneous rocks occur in a significant volume (�20% of the igneouscomponent). Most of these rocks are felsic volcanics and do notshow the geochemical characteristics of typical continental crust(e.g., Nb-Ta depletion) and even the zircons from such rocks do nothave trace element signals of heavy-REE depleted Archean TTG(Carley et al., 2014). An exception is the �3.9 Ga component in theArchean Acasta gneisses which resembles some Iceland felsic


volcanics in terms of heavy REE (Reimink et al., 2016). But perhapswe are just seeing the surficial felsic components in Iceland and atgreater depths, yet to be exposed, there are better examples ofArchean-like TTG. Experimental results indicate that heavy-REEdepleted TTGs can be produced by partial melting of wet maficrocks in oceanic plateaus at depths �50 km (Johnson et al., 2014;Hastie et al., 2015). But the question remains as to what volumeof TTG magma can be produced in such a setting (Condie, 2014)?

The sources for TTG magmas are wet eclogites and garnet am-phibolites in the deep crust (�50 km) or upper mantle (Moyen andMartin, 2012; Condie, 2014). However, only if wet eclogitescontinue to sink into the mantle can a continuous supply of TTG beproduced (Vlaar et al., 1994; Zegers and van Keken, 2001; Bédard,2006; Van Kranendonk, 2010; Johnson et al., 2014). As mafic andfelsic crust are produced, a growing volume of depleted harzburgite(olivine þ orthopyroxene) is left behind, and because it lacksgarnet, it will not readily sink but accumulates beneath the crust(Herzberg and Rudnick, 2012; Sizova et al., 2015). Unless eclogitecan sink through the growing harzburgite root, TTG productionshould soon come to an end with only small volumes produced, notenough to account for the vast volumes we see in Archean crustoften being produced for over 100 Myr in some locations. Thegrowing thick harzburgite rootmay also impede the transfer of newbasaltic magma from deep fertile mantle sources to the surface.Another question is that of how water is continuously supplied tothe sinking mafic crust, a necessity for continuous production ofTTG magma by partial melting (Rapp et al., 1991). If oceans covereda Hadean stagnant lid, a continual supply of sinking hydrated ba-salts could result in prolonged TTG production (Nagel et al., 2012;Zhang et al., 2013), as permitted for instance by the heat-pipemodel of Moore and Webb (2013). Water or not, however, eclo-gite would need to continually sink through the thick harzburgiteroot, and we need more models that focus on this question. Withour current understanding, subduction is the only known way toproduce large volumes of TTG over time periods of �100 Myr.

3.2. Are there two sources for plumes in the mantle after 2.5 Ga?

As pointed out by Burke (2011), most LIPs (large igneousprovinces) and many hotspots <300 Ma correlate with andprobably come from LLSVPs (large low S-wave velocity provinces)in the deep mantle beneath Africa and the South Pacific. However,other plumes, and especially those in the Northern Hemisphere,appear to come from the D00 layer at the base of the mantle. Hasthis always been so in earlier periods of time, and when and howdid the LLSVPs come into existence? Reexamination of the ther-mal history of the mantle based on basalt major elementgeochemistry (Condie et al., 2016a) suggests that plumes that giverise to basalts from enriched mantle (EM) sources appeared onlyafter the end of the Archean. Before this time, mantle potentialtemperatures for oceanic basalts were similar (around 1500 �C)and incompatible element ratios show that the enriched (EM) anddepleted (DM) mantle sources were not yet well developed(Condie, 2015b) (Figs. 2 and 3). It is probable, however, thatArchean komatiites came from plume sources that were consid-erably hotter (�1700 �C) than ambient mantle (Herzberg andAsimow, 2008). These observations suggest that prior to 2.5 Ga,there may have been only one source of mantle plumes (the D00

layer), and that it was not until after this time that lower tem-perature (w1500 �C) plumes could survive and make it to the baseof the lithosphere. If the lower temperature plumes come fromLLSVPs, this means the LLSVPs did not come into existence untilafter the Archean. Today, high-T plumes are rare and Gorgona isthe only known example producing komatiites (low-Mg typesonly), and its mantle potential temperature is similar to LLSVP-


derived plumes (Herzberg and Asimow, 2008; Herzberg andGazel, 2009). In contrast, in the Archean high-T plumes mayhave been widespread. To more fully test this idea, we need toanalyze more oceanic basalts especially in the age range of2.5e2.0 Ga, when thermal regimes were probably changing in themantle, possibly in response to widespread propagation of platetectonics. In turn, geodynamic modeling of mantle plume pro-duction needs to focus on this same period of time. One particu-larly important question is when did low-T plumes come intoexistence, and what are the geodynamic changes in the mantlethat led to their production? Another important question is whenand how rapidly did the LLSVPs grow in the deep mantle (Garneroet al., 2016)? If they grew as a result of rapid accumulation ofsubducted slabs, they may record the timing and rate of devel-opment of plate tectonics. Clearly we need more geodynamicmodels that focus on growth rate of the LLSVPs through geologictime.

3.3. Of what significance are global zircon age peaks and LIPevents?

Although global age peaks in detrital zircon populations arewell established, the significance of these peaks is still not wellunderstood and remains a topic of continued controversy anddiscussion (Hawkesworth et al., 2009; Condie et al., 2016b). Atleast part of the answer to this question lies with the interpreta-tion of the timing of global LIP events, which are probably theproduct of widespread mantle plume events. Some major LIPevents correlate well with zircon age peaks (e.g., 2.7, 2.2, and1.9 Ga; Fig. 9), whereas most do not have zircon age peak coun-terparts. Whether the enhanced peak heights of LIP events be-tween 3 and 2 Ga (Fig. 9) are real or a product sampling bias is notyet known. An understanding of these events is also tied to abetter understanding of the supercontinent cycle and the age andorigin of the LLSVPs mentioned above. The studies of Jellinek andManga (2004) suggest that plate tectonic events during whichslabs collapse to D00 in short times may initiate mantle plumeevents. If the collapsing slabs are swept into the LLSVPs by con-vection as suggested by Burke (2011), the frequency of LLSVP-derived plumes may also increase during the slab collapses. Incontrast, Arndt and Davaille (2013) have suggested that pre-1.5 Gazircon age peaks reflect crustal growth peaks caused by globalmantle plume events, which are the products of thermal-chemicalinstabilities at the base of the mantle. In this scenario, it is themantle plume events that increase the rate of subduction leadingto increases in continental growth, and thus to zircon age peaks.Geodynamic modeling has not yet yielded an unambiguous so-lution to this problem. Although most models agree that a stag-nant lid should transition into a plate tectonic regime with burstsof plate tectonic activity (O’Neill et al., 2007), the timing of thesebursts in various models ranges from a few million years (VanHunen and Moyen, 2012; Sizova et al., 2015) to billions of years(Weller et al., 2015). Statistical correlation of global LIP and zirconage peaks shows a strong periodicity in both around 270 Myr(Isley and Abbott, 2002; Puetz et al., 2016), and this agrees wellwith the experimental studies of Davaille et al. (2005) which showdestabilization in the D00 layer giving rise to mantle plume eventsevery 100e300 Myr. The question still remains, however, of whatcauses destabilizations in D00 and especially what caused the firstlarge destabilization a 2.7 Ga? Do these events require cata-strophic collapse of subducted slabs into D00 or could they be theresult of poorly understood events in the core? If the first “burst”of plate tectonics began about 3 Ga, perhaps arrival of newly-produced slabs in D00 was responsible for the first global mantleplume event at 2.7 Ga. Alternatively, could catastrophic growth of


Figure 10. Frequency of ages of detrital zircons from modern rivers in the time periodof 2600 to 1950 Ma. Data from Puetz et al. (2016).


the inner core or intense turbulence in the outer core (Andraultet al., 2016) have been responsible for destabilization of D00 at2.7 Ga? In this case, it would have been events in the core thatpromoted rapid continental growth at 2.7 Ga. Again, better con-strained geodynamic modeling is required to adequately addressthese questions.

3.4. Did Earth transition from stagnant lid to plate tectonicsbetween 3 and 2 Ga?

Geodynamic modeling indicates that rocky planets shouldevolve from stagnant lid to plate tectonics through an episodicregime (O’Neill et al., 2007). Can we recognize this episodic regimein the geologic record? This is perhaps themost difficult question toaddress because we really cannot yet identify stagnant-lid rockpackages. Many greenstone belts are characterized by extensive“mafic plains” composed of submarine basalts and associatedshallow intrusives in the early stages of development, often over-lain by “arc-like” successions (Wyman, 1999; Benn and Moyen,2008; Turner et al., 2014). Could these mafic plains represent areturn to stagnant-lid tectonics? As pointed out above, incompat-ible element ratios indicate the DM and EM basalts are rare before2.5 Ga and thus it is geochemically permissible for the Archeanmafic-plain basalts to be erupted in a stagnant-lid regime. Theoverlying “arc-like” sequences could be the results of short-lived“bursts” in subduction activity.

Evidence clearly shows that recycling of crustal material into thedeep mantle occurred prior to 3 Ga. For instance, sulfur isotopes insulfides in young basalts show that crust must have been recycledinto the mantle before the Great Oxidation Event at 2.5e2.4 Ga.Also the preservation of 142Nd and 182Wanomalies in some Archean(and younger) basalts appears to reflect an early crust-formingevent during which the crust was recycled into the deep mantleand later, remnants of this crust and/or the depleted restite servedas sources for some basalts (Caro et al., 2003; Bennett et al., 2007;Touboul et al., 2012). However, recycling of crust into the mantledoes not necessarily require subduction, and it may be possible forsuch recycling to occur in stagnant-lid regimes such as heat-pipesand mantle plumes (Bédard, 2006; Moore and Webb, 2013;Sizova et al., 2015).

It has been suggested that the last return to a stagnant-lidregime occurred during a possible crustal age gap at 2.4e2.2 Ga(Condie et al., 2009). The problemwith this idea is that the crustalage gap is rapidly disappearing as more data are becomingavailable from greatly undersampled geographic regions. We nowhave an increasing inventory of both detrital and outcrop zirconages in the 2.4e2.2 Ga “age gap” (Fig. 10). Three orogens containexposed TTG-greenstone associations in this time window: theArrowsmith orogen in Western Canada (Berman et al., 2013;Partin et al., 2014), the orogen(s) around the Sao Franciscocraton in Brazil (Klein et al., 2009; Teixeira et al., 2015), and theTrans-North China orogen (Luo et al., 2008; Yang and Santosh,2015). Other less-well defined orogens that contain igneousrocks produced in 2.4e2.2 Ga time window have also been iden-tified in Australia, Antarctica, West Africa and other locations inSouth America (Gasquet et al., 2003; Swain et al., 2005; Wanget al., 2016). There is also evidence in the detrital zircon popula-tion for an increasing peak at about 2350 Ma (Fig. 10). Hence, ifEarth transitioned from stagnant-lid to plate tectonics between 3and 2 Ga, we need to be able to identify rocks formed during thestagnant-lid periods, which means more intense sampling andanalysis of rocks formed in the valleys between the zircon agepeaks. From the information at hand, there is nothing that standsout as being different in rocks formed between from those formedduring the age peaks.


3.5. What the future holds?

To make progress in a better understanding of the evolution ofEarth between 4 and 2 Ga, we need to address the four questionsdiscussed above. It is necessary to compare rocks, rock associationsand tectonics during and between zircon age peaks. In particular,incompatible element ratios in age peak and inter-peak basalts andTTGs need to be compared. In particular, we need more geologicand geochemical data in the time window of 1.6e0.8 Ga. One po-tential problem is that the width of zircon age peaks overlaps thevalleys between the peaks and the position and width of peaks andvalleys varies with geographic location. We need to know whetherthe changes we see in Earth history between 3 and 2 Ga are due tothe onset and propagation of plate tectonics or to some other causethat may have followed the onset of plate tectonics. On the otherhand, if stagnant-lid tectonics transitioned to plate tectonics be-tween 4 and 3 Ga, we should find corresponding changes in thegeologic and geochemical record, which so far have not beenidentified. The first global pulse of crustal generation is at 2.7 Ga asreflected by the large zircon age peak at this time, whereas earlierpulses may be of only regional extent. To address and eventuallyanswer these questions, we need to constrain geodynamic modelssuch that they better accommodate geologic and geochemical da-tabases, especially focusing on the time period of 4 to 2 Ga. This is acritical time in Earth history as internal radiogenic heat sourcesdecayed and the mantle correspondingly cooled. Many review pa-pers have appeared in the last few years on the transition fromstagnant-lid to plate tectonics, and these have served to focus ourattention on specific problems where collaboration is needed be-tween investigators in different disciplines (examples are Cawoodet al., 2006; Korenaga, 2008, 2013; Van Hunen and Moyen, 2012;Condie and Aster, 2013; Hawkesworth et al., 2013; O’Neill andDebaille, 2014; Roberts and Spencer, 2015; O’Neill et al., 2016).What we do not need more of at the present time are more reviewpapers on this topic; rather we need investigators to return to thedrawing board an come up with some new, innovative models thataccommodate existing and new data from a variety of scientificdisciplines.



References

Andrault, D., Monteux, J., Le Bars, M., Samuel, H., 2016. The deep Earth may not becooling down. Earth Planetary Science Letters 443, 195e203.

Arndt, N.T., Davaille, A., 2013. Episodic Earth evolution. Tectonophysics 609,661e674.

Barth, M.G., Rudnick, R.L., Horn, I., McDonough, W.F., Spicuzza, M.J., Valley, J.W.,Haggerty, S.E., 2001. Geochemistry of xenolithic eclogites fromWest Africa, PartI: a link between low MgO eclogites and Archean crust formation. Geochimicaet Cosmochimica Acta 65 (9), 1499e1527.

Bédard, J.H., 2006. A catalytic delamination-driven model for coupled genesis ofArchaean crust and sub-continental lithospheric mantle. Geochimica et Cos-mochimica Acta 70, 1188e1214.

Belousova, E.A., Kostitsyn, Y.A., Griffin, W.L., O’Reilly, S.Y., Pearson, N.J., 2010. Thegrowth of the continental crust: constraints from zircon Hf-isotope data. Lithos119, 457e466.

Benn, K., Moyen, J.-F., 2008. The late Archean Abitibi-Opatica terrane, superiorprovince: a modified oceanic plateau. Geological Society of America, SpecialPaper 440, 173e197.

Bennett, V.C., Brandon, A.D., Nutman, A.P., 2007. Coupled 142Nd-143Nd isotopic ev-idence for Hadean mantle dynamics. Science 318, 1907e1910.

Berman, R.G., Pehrsson, S., Davis, W.J., Ryan, J.J., Qui, H., Ashton, K.E., 2013. TheArrowsmith orogeny: geochronological and thermobarometric constraints onits extent and tectonic setting in the Rae craton, with implications for pre-Nunasupercontinent reconstruction. Precambrian Research 232, 44e69.

Biggin, A.J., de Wit, M.J., Langereis, C.G., Zegers, T.E., Voute, S., Dekkers, M.J.,Drost, K., 2011. Palaeomagnetism of Archaean rocks of the Onverwacht Group,Barberton Greenstone Belt (southern Africa): evidence for a stable and poten-tially reversing geomagnetic field at ca. 3.5 Ga. Earth and Planetary ScienceLetters 301, 314e328.

Bispo-Santos, F., Agrella-Filho, M.S., Janikian, L., Reis, N.J., Trindade, R.I.F.,Reis, M.A.A.A., 2014. Towards Columbia: paleomagnetism of 1980-1960 MaSurumu volcanic rocks, northern Amazonian craton. Precambrian Research 244,123e138.

Burke, K., 2011. Plate tectonics, the Wilson Cycle, and mantle plumes: geodynamicsfrom the top. Annual Review of Earth and Planetary Sciences 39, 1e29.

Carley, T.L., Miller, C.F., Wooden, J.L., Padilla, A.J., Schmitt, A.K., Economos, R.C.,Bindeman, I.N., Jordan, B.T., 2014. Iceland is not a magmatic analog for theHadean: evidence from the zircon record. Earth and Planetary Science Letters405, 85e97.

Caro, G., Bourdon, B., Birck, J.-L., Moorbath, S., 2003. 146Sm-142Nd evidence from Isuametamorphosed sediments for early differentiation of the Earth’s mantle. Na-ture 423, 428e432.

Cawood, P.A., Kroner, A., Pisarevsky, S., 2006. Precambrian plate tectonics: criteriaand evidence. GSA Today 16 (7), 4e11.

Condie, K.C., 2003. Incompatible element ratios in oceanic basalts and komatiites:tracking deep mantle sources and continental growth rates with time.Geochemistry, Geophysics, Geosystems 4 (1), 1005. http://dx.doi.org/10.1029/2002GC000333.

Condie, K.C., 2005. High field strength element ratios in Archean basalts: a windowto evolving sources of mantle plumes. Lithos 79, 491e504.

Condie, K.C., 2014. How to make a continent: thirty-five years of TTG research. In:Dilek, Y., Furnes, H. (Eds.), Evolution of Archean Crust and Early Life, ModernApproaches to Solid Earth Sciences, vol. 7. Springer Science, Dordrecht,pp. 179e193.

Condie, K.C., 2015a. Earth as an Evolving Planetary System. Elsevier, Academic Press,Amsterdam, 418 pp.

Condie, K.C., 2015b. Changing tectonic settings through time: indiscriminate use ofgeochemical discriminant diagrams. Precambrian Research 266, 587e591.

Condie, K.C., Aster, R.C., 2010. Episodic zircon age spectra of orogenic granitoids: thesupercontinent connection and continental growth. Precambrian Research 180,227e236.

Condie, K.C., Aster, R.C., 2013. Refinement of the supercontinent cycle with Hf, Ndand Sr isotopes. Geoscience Frontiers 4, 667e680.

Condie, K.C., O’Neill, C., 2010. The Archean-Proterozoic boundary: 500 My of tec-tonic transition in Earth history. American Journal of Science 310, 775e790.

Condie, K.C., Shearer, C.K., 2016. Tracking the Evolution of Mantle Sources withIncompatible Element Ratios in Stagnant Lid and Plate Tectonic Planets (sub-mitted for publication).

Condie, K.C., O’Neill, C., Aster, R., 2009. Evidence and Implications for a widespreadmagmatic shutdown for 250 My on Earth. Earth and Planetary Science Letters282, 294e298.

Condie, K.C., Pisarevsky, S.A., Korenaga, J., Gardoll, S., 2015a. Is the rate of super-continent assembly changing with time? Precambrian Research 259, 278e289.

Condie, K.C., Davaille, A., Aster, R.C., Arndt, N., 2015b. Upstairs-downstairs: super-continents and large igneous provinces, are they related? International GeologyReview 57 (11e12), 1341e1348.

Condie, K.C., Aster, R.C., van Hunen, J., 2016a. A great thermal divergence in themantle beginning 2.5 Ga: geochemical constraints from greenstone basalts andkomatiites. Geoscience Frontiers 7, 543e553.

Condie, K.C., Arndt, N., Davaille, A., Puetz, S.J., 2016b. Zircon age peaks: productionor preservation of continental crust? Geosphere (in press).

Davaille, A., Stutzmann, E., Silveira, G., Besse, J., Courtillot, V., 2005. Convective patternunder the Indo- Atlantic “box”. Earth and Planetary Science Letters 239, 233e252.


Dhuime, B., Hawkesworth, C.J., Cawood, P., 2011. When continents formed. Science331, 154e155.

Dhuime, B., Hawkesworth, C.J., Cawood, P.A., Storey, C.D., 2012. A change in thegeodynamics of continental growth 3 billion years ago. Science 335, 1334e1336.

Dhuime, B., Wuestefeld, A., Hawkesworth, C.J., 2015. Emergence of modern conti-nental crust about 3 billion years ago. Nature Geoscience 8, 552e555.

Dyck, B., Reno, B.L., Kokfelt, T.F., 2015. The Majorqaq Belt: a record of Neoarchaeanorogenesis during final assembly of the North Atlantic Craton, southern WestGreenland. Lithos 220-223, 253e271.

Furnes, H., Dilek, Y., de Wit, M., 2015. Precambrian greenstone sequences representdifferent ophiolite types. Gondwana Research 27, 649e685.

Garde, A.A., Hollis, J.A., 2010. A buried Palaeoproterozoic spreading ridge in thenorthern Nagssugtoqidian orogen, West Greenland. Geological Society of Lon-don, Special Publication 338, 213e234.

Garnero, E.J., McNamara, A.K., Shim, S.-H., 2016. Continent-sized anomalous zoneswith low seismic velocity at the base of Earth’s mantle. Nature Geoscience 9,481e489.

Gasquet, D., Barbey, P., Adou, M., Paquette, J.L., 2003. Structure, SreNd isotopegeochemistry and zircon UePb geochronology of the granitoids of theDabakala area (Côte d’Ivoire): evidence for a 2.3 Ga crustal growth eventin the Palaeoproterozoic of West Africa? Precambrian Research 127,239e354.

Greenough, J.D., Ya’acoby, A., 2013. A comparative geochemical study of Mars andEarth basalt petrogenesis. Canadian Journal of Earth Sciences 50, 78e93.

Griffin, W.L., O’Reilly, S.Y., Abe, N., Aulbach, S., Davies, R.M., Pearson, N.J., Doyle, B.J.,Kivi, K., 2003. The origin and evolution of Archean lithospheric mantle. Pre-cambrian Research 127, 19e41.

Hand, M., Reid, A., Jagodzinski, L., 2007. Tectonic Framework and evolution of theGawler Craton, Southern Australia. Economic Geology 102, 1377e1395.

Harris, L.B., Bedard, J.H., 2014. Crustal evolution and deformation in a non-plate-tectonic Archaean Earth: comparisons with Venus. In: Dilek, Y., Furnes, H.(Eds.), Evolution of Archean Crust and Early Life. Springer, Dordrecht,pp. 215e291.

Hartmann, L.A., Philipp, R.P., Liu, D., Wan, Y., Wang, Y., Santos, J.O.,Vasconcellos, M.A.Z., 2004. Paleoproterozoic magmatic provenance of detritalzircons, porongos complex quartzites, Southern Brazilian Shield. InternationalGeology Review 46, 127e157.

Hastie, A.R., Fitton, J.G., Mitchell, S.F., Neill, J., Nowell, G.M., Millar, I.L., 2015. Canfractional crystallization, mixing and Assimilation processes be responsible forJamaican-type Adakites? Implications for generating Eoarchaean continentalcrust. Journal of Petrology 56 (7), 1251e1284.

Hawkesworth, C., Cawood, P., Kemp, T., Storey, C., Dhuime, B., 2009. A matter ofpreservation. Science 323, 49e50.

Hawkesworth, C., Cawood, P., Dhuime, B., 2013. Continental growth and the crustalrecord. Tectonophysics 609, 651e660.

Herzberg, C., Asimow, P.D., 2008. Petrology of some oceanic island basalts: PRI-MELT2.XLS software for primary magma calculation. Geochemistry GeophysicsGeosystems 8, Q09001. http://dx.doi.org/10.1029/2008GC002057.

Herzberg, C., Gazel, E., 2009. Petrological evidence for secular cooling in mantleplumes. Nature 458, 619e622.

Herzberg, C., Rudnick, R., 2012. Formation of cratonic lithosphere: an integratedthermal and petrological model. Lithos 149, 4e15.

Hofmann, A.W., 1997. Mantle geochemistry: the message from oceanic volcanism.Nature 385, 219e229.

Isley, A.E., Abbott, D.H., 2002. Implications of the temporal distribution of high-Mgmagmas for mantle plume volcanism through time. Journal of Geology 110 (2),141e158.

Jellinek, A.M., Manga, M., 2004. Links between long-lived hot spots, mantle plumes,D00 , and plate tectonics. Reviews in Geophysics 42. RG3002/2004.

Johnson, T.E., Brown, M., Kaus, B.J.P., VanTongeren, J.A., 2014. Delamination andrecycling of Archaean crust caused by gravitational instabilities. Nature Geo-science 7, 47e52.

Keller, C.B., Schoene, B., 2012. Statistical geochemistry reveals disruption in secularlithospheric evolution about 2.5 Gyr ago. Nature 485, 490e493.

Klein, E.L., Luzardo, R., Moura, C.A.V., Lobato, D.C., Brito, R.S.C., Armstrong, R., 2009.Geochronology, Nd isotopes and reconnaissance geochemistry of volcanic andmetavolcanic rocks of the São Luís Craton, northern Brazil: implications for tectonicsettingandcrustal evolution. JournalofSouthAmericanEarthSciences27,129e145.

Komiya, T., 2004. Material circulation model including chemical differentiationwithin the mantle and secular variation of temperature and composition of themantle. Physics of the Earth and Planetary Interiors 146, 333e367.

Konhauser, K.O., Pecoits, E., Lalonde, S.V., Papineau, D., Nisbet, E.G., Barley, M.E.,Arndt, N.T., Zahnle, K., Kamber, B.S., 2009. Oceanic nickel depletion and amethanogen famine before the great oxidation event. Nature 458, 750e754.

Korenaga, J., 2008. Plate tectonics, flood basalts and the evolution of Earth’s oceans.Terra Nova 20 (6), 419e439.

Korenaga, J., 2013. Initiation and evolution of plate tectonics on Earth: theories andobservations. Annual Review of Earth and Planetary Sciences Letters 41,117e151.

Kumar, K.V., Ernst, W.G., Leelanandam, C., Wooden, J.L., Grove, M.J., 2010. FirstPaleoproterozoic ophiolite from Gondwana: geochronologicegeochemicaldocumentation of ancient oceanic crust from Kandra, SE India. Tectonophysics487, 22e32.

Luo, Y., Sun, M., Zhao, G., Li, S., Ayers, J.C., Xia, X., Zhang, J., 2008. A comparison ofUePb and Hf isotopic compositions of detrital zircons from the North and South


http://refhub.elsevier.com/S1674-9871(16)30127-X/sref2





























































http://dx.doi.org/10.1029/2002GC000333

http://dx.doi.org/10.1029/2002GC000333















































































































http://dx.doi.org/10.1029/2008GC002057






















































Liaohe groups: constraints on the evolution of the Jiao-Liao-Ji Belt, North ChinaCraton. Precambrian Research 163, 279e306.

McDonough, W.F., Sun, S.-S., 1995. Composition of the Earth. Chemical Geology 120,223e253.

Menot, R.-P., Pecher, A., Rolland, Y., Peucat, J.-J., Pelletier, A., Duclaux, G., Guillot, S.,2005. Structural setting of the Neoarchean terrains in Commonwealth Bay are(143-145�E), Terre Adelie craton, East Antarctica. Gondwana Research 8 (9),1e9.

Mitchell, R.N., Bleeker, W., Breemen, O.V., Lecheminant, T.N., Peng, P.,Nilsson, M.K.M., Evans, D.A.D., 2014. Plate tectonics before 2.0 Ga: evidencefrom paleomagnetism of cratons within supercontinent Nuna. American Jour-nal of Science 314, 878e894.

Moore, W.B., Webb, A.G., 2013. Heat-pipe Earth. Nature 501, 501e505.Moyen, J.-F., Martin, H., 2012. Forty years of TTG research. Lithos 148, 312e336.Nagel, T.J., Hoffmann, J.E., Munker, C., 2012. Generation of Eoarchean tonalite-

trondhjemite-granodiorite series from thickened mafic arc crust. Geology 40,375e378.

O’Neill, C., Debaille, V., 2014. The evolution of Hadean-Eoarchaean geodynamics.Earth and Planetary Science Letters 406, 49e58.

O’Neill, C.O., Lenardic, A., Moresi, L., Torsvik, T.H., Lee, C.-T.A., 2007. Episodic pre-cambrian subduction. Earth and Planetary Science Letters 262, 552e562.

O’Neill, C.O., Lenardic, A., Weller, M., Moresi, L., Quenette, S., Zhang, S., 2016.Awindow for plate tectonics in terrestrial planet evolution? Physics of the Earthand Planetary Interiors 255, 80e92.

Partin, C.A., Bekker, A., Sylvester, P.J., Wodicka, N., Stern, R.A., Chacko, T.,Heaman, L.M., 2014. Filling in the juvenile magmatic gap: evidence for unin-terrupted Paleoproterozoic plate tectonics. Earth and Planetary Science Letters388, 123e133.

Pehrsson, S.J., Berman, R.G., Eglington, B., Rainbird, R., 2013. Two Neoarchean su-percontinents revisited: the case for a Rae family of cratons. PrecambrianResearch 232, 27e43.

Peltonen, P., Kontinen, A., Huhma, H., 1996. Petrology and geochemistry of meta-basalts from the 1.95 Ga Jormua Ophiolite, Northeastern Finland. Journal ofPetrology 37, 1359e1383.

Pisarevsky, S.A., Elming, S.-A., Pesonen, L.J., Li, Z.-X., 2014. Mesoproterozoicpaleogeography: supercontinent and beyond. Precambrian Research 244,207e225.

Puetz, S.J., Condie, K.C., Pisarevsky, S., Davaille, A., Schwarz, C.J., de Araujo, G., 2016.Quantifying the evolution of the continental and oceanic crust. Earth-ScienceReviews (in press).

Rapp, R.P., Watson, E.B., Miller, C.F., 1991. Partial melting of amphibolite/eclogite andthe origin of Archean trondhjemites and tonalites. Precambrian Research 51,1e25.

Reimink, J.R., Chacko, T., Stern, R.A., Heaman, L.M., 2016. The birth of a cratonicnucleus: lithogeochemical evolution of the 4.01-2.94 Ga Acasta Gneiss Com-plex. Precambrian Research 281, 453e472.

Roberts, N.M.W., Spencer, C.J., 2015. The zircon archive of continent formationthrough time. Geological Society of London, Special Publication 389, 197e225.

Rollinson, H., 1997. Eclogitic xenoliths in west African kimberlites as residues fromArchaean granitoid crust formation. Nature 389, 173e176.

Rudnick, R.L., Gao, S., 2014. Composition of the Continental Crust. Treatise onGeochemistry, second ed. Elsevier, pp. 1e51.

Scott, D.J., Helmstaedt, H., Bickle, M.J., 1992. Purtuniq ophiolite, Cape Smith belt,northern Quebec, Canada: a reconstructed section of Early Proterozoic oceaniccrust. Geology 20, 173e176.

Shirey, S.B., Richardson, S.H., 2011. Start of the Wilson cycle at 3 Ga shown by di-amonds from subcontinental mantle. Science 333, 434e436.

Sizova, E., Gerya, T., Stuwe, K., Brown, M., 2015. Generation of felsic crust in theArchean: a geodynamic modeling perspective. Precambrian Research 271,198e224.

Spencer, C.J., Cawood, P.A., Hawkesworth, C.J., Raub, T.D., Prave, A.R.,Roberts, N.M.W., 2014. Proterozoic onset of crustal reworking and collisional


tectonics: reappraisal of the zircon oxygen isotope record. Geology 42,451e454.

Stepanova, A.V., Samosonov, A.V., Salnikova, E.B., Puchtel, I.S., Larionova, Y.O.,Larionov, A.N., Stepanov, V.S., Shapolvalov, Y.B., Egorova, S.V., 2014. Palae-oproterozoic continental MORB-type Tholeiites in the Karelian Craton:petrology, geochronology, and tectonic setting. Journal of Petrology 55,1719e1751.

Stern, R.A., Syme, E.C., Bailes, A.H., Lucas, S.B., 1995. Paleoproterozoic (1.90e1.86 Ga)arc volcanism in the Flin Flon Belt, Trans-Hudson Orogen, Canada. Contribu-tions to Mineralogy and Petrology 119, 117e141.

Stracke, A., 2012. Earth’s heterogeneous mantle: a product of convection-driveninteraction between crust and mantle. Chemical Geology 330-331,274e299.

Swain, G., Woodhouse, A., Hand, M., Barovich, K., Schwarz, M., Fanning, C.M., 2005.Provenance and tectonic development of the late Archaean Gawler Craton,Australia; UePb zircon, geochemical and SmeNd isotopic implications. Pre-cambrian Research 141, 106e136.

Tang, M., Chen, K., Rudnick, R.L., 2016. Archean upper crust transition from mafic tofelsic marks the onset of plate tectonics. Science 351, 372e374.

Taylor, S.R., McLennan, S.M., 1985. The Continental Crust: its Composition andEvolution. Blackwell Scientific Publishers, Oxford, 312 pp.

Teixeira, W., Avila, C.A., Dussin, I.A., Neto, A.V.C., Bongiolo, E.M., Santos, J.O.,Barbosa, N.S., 2015. A juvenile accretion episode (2.35e2.32 Ga) in the Mineirobelt and its role to the Minas accretionary orogeny: zircon UePbeHf andgeochemical evidences. Precambrian Research 256, 148e169.

Touboul, M., Puchtel, I.S., Walker, R.J., 2012. 182W evidence for long-term preser-vation of early mantle differentiation products. Science 335, 1065e1069.

Turner, S., Rushmer, T., Reagan, M., Moyen, J.-F., 2014. Heading down early on? Startof subduction on Earth. Geology 42, 139e142.

Usui, Y., Tarduno, J.A., Watkeys, M., Hofmann, A., Cottrell, R.D., 2009. Evidence for a3.45-billion-year-old magnetic remanence: hints of an ancient geodynamofrom conglomerates of South Africa. Geochemistry, Geophysics, Geosystems 10.http://dx.doi.org/10.1029/2009GC002496.

Van der Voo, R., 1993. Paleomagnetism of the Atlantic, Tethys and Iapetus Oceans.Cambridge University Press, Cambridge, UK, 412 pp.

Van Hunen, J., Moyen, J.-F., 2012. Archean subduction: fact or fiction? Annual Re-view of Earth and Planetary Sciences 40, 195e219.

Van Kranendonk, M.J., 2010. Two types of Archean continental crust: plume andplate tectonics on early Earth. American Journal of Science 310, 1187e1209.

Vlaar, N.J., van Keken, P.E., van den Berg, A.P., 1994. Cooling of the Earth in theArchaean. Earth and Planetary Science Letters 121, 1e18.

Voice, P.J., Kowalewski, M., Eriksson, K.A., 2011. Quantifying the timing and rate ofcrustal evolution: global compilation of radiometrically dated detrital zircongrains. Journal of Geology 119 (2), 109e126.

Wang, H., Wu, Y.-B., Gao, S., Qin, Z.-W., Zhao-Chu, H., Zheng, J.-P., Yang, S.-H., 2016.Continental growth through accreted oceanic arc: zircon HfeO isotope evi-dence for granitoids from the Qinling orogen. Geochimica et CosmochimicaActa 182, 109e130.

Weller, M.B., Lenardic, A., O’Neill, C., 2015. The effects of internal heating and largescale climate variations on tectonic bi-stability in terrestrial planets. Earth andPlanetary Science Letters 420, 85e94.

Wyman, D.A., 1999. A 2.7 Ga depleted tholeiite suite: evidence of plume-arc inter-action in the Abitibi Greenstone Belt, Canada. Precambrian Research 97, 27e42.

Yang, Q.-Y., Santosh, M., 2015. Paleoproterozoic arc magmatism in the North ChinaCraton: no Siderian global plate tectonic shutdown. Gondwana Research 28,82e105.

Zegers, T.E., van Keken, P.E., 2001. Middle Archean continent formation by crustaldelamination. Geology 29, 1083e1086.

Zhang, C., Holtz, F., Koepke, J., Wolff, P.E., Ma, C., Bedard, J., 2013. Constraints fromexperimental melting of amphibolite on the depth of formation of garnet-richrestites, and implications for models of Early Archean crustal growth. Pre-cambrian Research 231, 206e217.




































































































































http://dx.doi.org/10.1029/2009GC002496









































a planet in transition: the onset of plate tectonics on ... · kent c. condie department of earth...

Documents