当前位置:文档之家› 17. Izuka 2013

17. Izuka 2013

17. Izuka 2013
17. Izuka 2013

Evolution of the African continental crust as recorded by U–Pb,Lu–Hf and O isotopes in detrital zircons

from modern rivers

Tsuyoshi Iizuka a ,b ,?,Ian H.Campbell a ,Charlotte M.Allen a ,James B.Gill c ,

Shigenori Maruyama d ,Fre

′de ′ric Makoka e a

Research School of Earth Sciences,Australian National University,Canberra,ACT 0200,Australia

b

Department of Earth and Planetary Science,University of Tokyo,Hongo 7-3-1,Bunkyo,Tokyo 113-0033,Japan

c

Department of Earth and Planetary Sciences,University of California,Santa Cruz,CA 95060,USA

d

Department of Earth and Planetary Sciences,Tokyo Institute of Technology,O-okayama 2-12-1,Meguro,Tokyo 152-8551,Japan

e

Department of Earth Sciences,University of Kinshasa,B.P.190Kin XI,Democratic Republic of the Congo

Received 9July 2012;accepted in revised form 24December 2012;available online 10January 2013

Abstract

To better understand the evolutionary history of the African continental crust,a combined U–Pb,Lu–Hf and O isotopic study has been carried out by in situ analyses of approximately 450detrital zircon grains from the Niger,Nile,Congo,Zam-bezi and Orange Rivers.The U–Pb isotopic data show age peaks at ca.2.7,2.1–1.8,1.2–1.0,ca.0.8,0.7–0.5and ca.0.3Ga.These peaks,with the exception of the one at ca.0.8Ga,correspond with the assembly of supercontinents.Furthermore,the detrital zircons that crystallized during these periods of supercontinent assembly have dominantly non-mantle-like O and Hf isotopic signatures,in contrast to the ca.0.8Ga detrital zircons which have juvenile characteristics.These data can be inter-preted as showing that continental collisions during supercontinent assembly resulted in supermountain building accompa-nied by remelting of older continental crust,which in turn led to signi?cant erosion of young igneous rocks with non-mantle-like isotopic signatures.Alternatively,the data may indicate that the major mode of crustal development changed dur-ing the supercontinent cycle:the generation of juvenile crust in extensional settings was dominant during supercontinent frag-mentation,whereas the stabilization of the generated crust via crustal accretion and reworking was important during supercontinent assembly.The Lu–Hf and O isotope systematics indicate that terreigneous sediments could attain elevated 18

O/16O via prolonged sediment–sediment recycling over long crustal residence time,and also that reworking of carbonate and chert which generally have elevated 18O/16O and low Hf contents is minor in granitoid magmatism.The highest 18

O/16O in detrital zircon abruptly increased at ca.2.1Ga and became nearly constant thereafter.This indicates that rework-ing of mature sediments increased abruptly at that time,probably as a result of a transition in the dynamics of either granitoid crust formation or sedimentary evolution.To estimate the mantle-extraction age of the reworked crust,we have calculated arc mantle Hf model ages for the detrital zircons using O isotopic data to constrain the Lu/Hf used in the model age calculation.The Hf model age histograms for each period of detrital zircons suggest that a signi?cant amount of the African continental crust was generated in the Paleo-Mesoproterozoic likely by ma?c magmatism,and subsequently reworked into younger gran-itoid crust with varying crustal residence times.ó2012Elsevier Ltd.All rights reserved.

0016-7037/$-see front matter ó2012Elsevier Ltd.All rights reserved.https://www.doczj.com/doc/b413032514.html,/10.1016/j.gca.2012.12.028

?Corresponding author at:Department of Earth and Planetary Science,University of Tokyo,Hongo 7-3-1,Bunkyo,Tokyo 113-0033,

Japan.Tel.:+81358414282:fax:+81358418378.

E-mail address:iizuka@eps.s.u-tokyo.ac.jp (T.Iizuka).

https://www.doczj.com/doc/b413032514.html,/locate/gca

Available online at

https://www.doczj.com/doc/b413032514.html,

Geochimica et Cosmochimica Acta 107(2013)

96–120

1.INTRODUCTION

It is well established that present-day continental crust has an andesitic bulk composition and is vertically strati?ed from lower portions consisting mainly of ma?c rocks to upper portions dominated by granitoids and sedimentary rocks(e.g.,Taylor and McLennan,1985;Christensen and Mooney,1995;Wedepohl,1995;Rudnick and Gao, 2003).Yet there is considerable debate as to when and how it was generated,and how it has evolved to its present form(e.g.,Rudnick,1995;Albare`de,1998;Hawkesworth and Kemp,2006;Rollinson,2008).A major di?culty in the study of the continental crust is that the most accessible rocks,namely granitoids and sedimentary rocks,are not a direct sample of newly generated continental crust.Grani-toids are mostly a product of remelting of pre-existing con-tinental crust and a primary source of continental sediments.Hence,the challenge in deciphering the evolu-tion of continental crust lies with better understanding in-tra-crustal reworking.

Isotopic analyses of detrital zircons from the Earth’s major rivers can be a powerful tool for studying the evolu-tion of continental crust(e.g.,Ledent et al.,1964;Goldstein et al.,1997;Rino et al.,2004,2008;Iizuka et al.,2005,2010; Wu et al.,2007;Wang et al.,2009,2011;Yang et al.,2009; Safonova et al.,2010).Large rivers erode exposed continen-tal crust over an extensive area,and most eroded materials have experienced sediment–sediment recycling(Goldstein et al.,1984;Campbell et al.,2005).Prolonged sediment–sediment recycling results in e?cient mixing of sediments derived from various source rocks,including the parts of granitoid crust that are currently inaccessible.Zircon,an ubiquitous accessory mineral in granitoids,yields precise U–Pb crystallization ages and retains Hf and O isotopic sig-natures of the parental magmas through sedimentary and metamorphic processes.The176Hf/177Hf is a function of residence time and Lu/Hf of the source crustal materials (e.g.,Patchett et al.,1981;Amelin et al.,1999;Bodet and Scha¨rer,2000;Gri?n et al.,2000),whereas the18O/16O is a measure of the fraction of the granitoid source region that had been the Earth’s surface where O isotopic fractiona-tions are large(Valley et al.,2005).Accordingly,an integra-tion of U–Pb,Lu–Hf and O isotopic data of detrital zircons from large riverine systems can potentially allow us to investigate the nature of crustal reworking and possible timing of major crust generation on a continental scale (Kemp et al.,2006;Pietranik et al.,2008;Wang et al., 2009,2011;Lancaster et al.,2011).In addition,because the erosion of igneous rocks should be signi?cantly en-hanced during the formation of giant mountain ranges, U–Pb age distribution of detrital zircons may potentially re-cord major orogenic events associated with building super-continents(Campbell and Allen,2008).

In this study,we present U–Pb,Lu–Hf and O isotopic data for approximately450detrital zircons from the?ve largest rivers in Africa.These data are used to evaluate the nature of crustal reworking through geological history and to link the development of the African continental crust with the supercontinent cycle.We will also discuss the limitations and prospects of using U–Pb,Lu–Hf and O isotope systematics of detrital zircons to constrain the timing of major continental growth.

2.REGIONAL GEOLOGY

The sand samples were collected at or near the mouths of the Niger(NGR1),Nile(NIL),Congo(CNG2),Zambezi (ZMB2)and Orange(ORG2)Rivers,which collectively drain approximately40%of continental Africa(Table1). The African continent comprises several Archean-Paleo-proterozoic cratons,which are rimmed by orogenic belts and partly covered by sedimentary basins(Fig.1).The ma-jor framework of the continent was established during the Neoproterozoic to earliest Paleozoic(Pan-African)orogeny at the center of the supercontinent Gondwana(e.g.,Kro¨ner and Stern,2004;Collins and Pisarevsky,2005;Tohver et al.,2006;Begg et al.,2009).African regional geology is summarized below,with emphasis on the timing and nature of major tectonomagmatic events that would have a?ected drainage areas of the studied rivers.

The upper basin of the Niger River extends from the central to southern part of the West African Craton.The craton comprises a western Archean domain and an eastern Paleoproterozoic domain,which are largely covered by the Neoproterozoic to Paleozoic Taoudeni Basin(Clauer et al., 1982).The Archean basement contains migmatitic orthog-neisses that crystallized at ca.3.5and3.0Ga and were metamorphosed at2.8–2.7and2.1Ga(Potrel et al.,1996; Kouamelan et al.,1997;Thie′blemont et al.,2001).The lat-ter metamorphic event is coincident with a dominant mag-matic event within the eastern domain(Abouchami and Boher,1990).Whole-rock Nd isotopic studies indicated that some of the Archean crust was originally extracted from the mantle as early as 3.6–3.3Ga,whereas the 2.1Ga magmatism formed juvenile continental crust rang-ing in composition from gabbro to granitoid(Boher et al.,1992;Potrel et al.,1996;Kouamelan et al.,1997). During the Pan-African orogeny,the West African Craton collided with the Tuareg and Benin-Nigeria Shields,under-lying the lower basin of the Niger River(Ajibade and Wright,1989;Black et al.,1994;Attoh et al.,1997;A?aton et al.,2000;Jahn,2001).The Tuareg Shield consists mainly of Archean and Paleoproterozoic terranes that were partly remobilized during the Pan-African orogeny,and juvenile oceanic Neoproterozoic terranes(e.g.,Black et al.,1994; Caby,2003).The Archean and Paleoproterozoic terranes host tonalites and trondhjemites emplaced between P3.2 and2.7Ga and granites at ca.2.65,2.5and2.2Ga,and some of them experienced a high-temperature metamor-phism at ca.2.0Ga(e.g.,Ouzegane et al.,2003;Peucat et al.,2003).After a tectonothermally quiescent period from1.6to0.9Ga,syn-tectonic Pan-African granites in-truded into the shield between870and520Ma(Black et al.,1994;Paquette et al.,1998;Caby,2003).The Be-nin-Nigeria Shield contains granitoids with emplacement ages of ca.3.6,3.0,2.7–2.5and2.1–1.8Ga,which were in-truded by and partly reworked into670–580Ma granites (e.g.,Bruguier et al.,1994;Dada,1998;Kro¨ner et al., 2001;Ferre′et al.,2002;Ekwueme and Kro¨ner,2006). The lower basin of the Niger River also covers a part of

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–12097

Table 1

Summary of the samples used in this study.Sample River Sample locality Drainage area (?106km 2) Sediment load (?106t/yr) NGR1Niger 6°0800500N,6°4501900E 1.233NIL Nile Near Cairo City

3.040CNG2Congo 40°1700800S,15°2702900E 3.843ZMB2Zambezi 17°4803300S,35°2303900E 1.435ORG2

Orange

28°3002500S,16°3700700E

0.9

100

Data are from Milliman and Syvitski (1992).

2000 km

0?10?E 10?W 20?W 20?E 30?E 40?E 50?E

0?

0?

10?S

10?N

20?N

30?N

20?S

30?S

0?

10?S 10?N

20?N

30?N

20?S

30?S

10?E

10?W

20?W

20?E

30?E

40?E

50?E

Archean-Paleoproterozoic crust

Archean-Proterozoic crust with Pan-African overprint

Mesoproterozoic orogenic belts Pan-African orogenic belts Cenozoic volcanics

Phanerozoic cover Congo

Craton

Kalahari

Craton West African

Craton

Saharan

Metacraton

Tanzania Craton

AS

Niger River

Congo River

Zambezi River

Orange River

Nile River

ZC

KC KS GCS

BKS

UC

TS

BNS

Central African Fold Belt

KB

IB

LB

GB

ANS

RP

NNB

ZB East African

Orogenic Zone

Simpli?ed geological map of Africa (after Begg et al.,2009and Kampunzu and Popo?,1991),with sampling locations (dots),(blue lines),and drainage basin limits (broken lines)of the Niger,Nile,Congo,Zambezi and Orange Rivers.Abbreviations TS,Tuareg Shield;BNS,Benin-Nigeria Shield;UC,Uganda Craton;ANS,Arabian-Numian Shield;KB,Kibaran Belt;Gabon-Cameroon Shield;BKS,Bomu-Kibalan Shield;AS,Angolan Shield;KS,Kasai Shield;ZC;Zimbabwe Craton;IB,Irumide Belt;LB,Lu?lian Belt;KC,Kaapvaal Craton;RP,Rehobothian Province;NNB,Namaqua-Natal Belt;GB,Gariep interpretation of the references to color in this ?gure legend,the reader is referred to the web version of this article.)

98T.Iizuka et al./Geochimica et Cosmochimica Acta 107(2013)96–120

the Central African Fold Belt(northwest Cameroon), which lies between the Benin-Nigeria Shield and the Congo Craton.The northern part of the Pan-African fold belt comprises medium-to high-grade metamorphic rocks of volcanic and volcano-sedimentary origin(ca.700Ma), pre-,syn-and post-tectonic granitoids emplaced between 660–570Ma and low-grade metamorphic rocks of volcanic and sedimentary origin deposited during the orogeny.The Nd isotopic data indicate that these rocks contain a sub-stantial juvenile component(Toteu et al.,2001;van Schmus et al.,2008,and references therein).

The Nile River,the longest river in the world,drains the northern Tanzania Craton(i.e.,Uganda Craton),eastern Saharan Metacraton,and northern part of the East African Orogenic Zone.The Tanzania Craton is made up mainly of Archean granitoids and greenstone belts(De Waele et al., 2008,and references therein).Volcanism in the greenstone belts is dated at between2.82and2.67Ga,whereas the granitoids have crystallization ages from2.69to2.55Ga (Borg and Krogh,1999;Manya and Maboko,2003;Manya et al.,2006).Whole-rock Nd isotopic data suggest that the granitoids and volcanic rocks lack a component of crust signi?cantly older than 3.2Ga(Manya and Maboko, 2003;Cloutier et al.,2005).The northern part of the craton (Ugandan Craton)contains high-grade metamorphic rocks including gneisses,migmatites,amphibolites,and granulites (Gabert,1990).Some of these metamorphic rocks yielded zircon U–Pb ages of2.9and2.55Ga,which can be inter-preted as metamorphic or protolith crystallization ages (Leggo,1974;Gabert,1990).These basements were sub-jected to intense metamorphism during the Pan-African (ca.650Ma)orogeny(Leggo,1974;Appel et al.,2005). The Saharan Metacraton occupies the north-central part of Africa and extends in the Saharan Desert.The term “metacraton”has been introduced to refer to a craton that has been mobilized during an orogenic event but is still rec-ognizable through its rheological,geochronological and isotopic characteristics(Abdelsalam et al.,2002).The Sah-aran Metacraton is largely covered by Phanerozoic sedi-mentary sequences,and the exposed parts consist dominantly of Neoproterozoic medium-to high-grade gneisses and migmatites.Zircon U–Pb geochronology and isotopic studies(Pin and Poidevin,1987;Toteu et al., 1987;Lie′geois et al.,1994;Stern et al.,1994;Sultan et al., 1994)indicated that their igneous protoliths largely formed during Neoarchean and Paleoproterozoic times,and they were intruded by granites and structurally and thermally overprinted during the Pan-African orogenic event.These three cratons face the East African Orogenic Zone that formed as a result of the consolidation of East and West Gondwana between800and580Ma(Grantham et al., 2003;Johnson and Woldehaimanot,2003;Meert,2003). The northern part of the orogenic zone(the Arabian–Nu-bian Shield)is a collage of Neoproterozoic(870–670Ma) juvenile crust and continental-marginal terranes that locally contain Archean-Paleoproterozoic crustal materials(Stern, 1994;Stein and Goldstein,1996;Teklay,2006).In Ethiopia, this Neoproterozoic crust was overlain by rift-related rhyo-litic and basaltic lavas that erupted at$30Ma,resulting in the formation of the Ethiopian plateau(Kie?er et al.,2004;Wolfenden et al.,2005).A recent Nd-Sr isotopic study (Padoan et al.,2011)indicates that Nile river detritus is de-rived from Precambrian basement,Mesozoic strata and Tertiary volcanic rocks.

The Congo River has the second largest drainage area on Earth(Table1),covering the southern Tanzania Craton, a signi?cant part of the Congo Craton,the Kibaran Belt be-tween the two cratons,and the southern Central African Fold Belt.The southern part of the Tanzania Craton com-prises Paleoproterozoic(ca.2.0Ga)eclogite-facies rocks, syn-and post-tectonic granitoids(2.0–1.8Ga),and metase-dimentary rocks that are sourced from the 2.7–2.4Ga granitoids in the craton(Reddy et al.,2003;Collins et al., 2004;Sommer et al.,2005).The Congo Craton is largely covered by Phanerozoic sedimentary sequences and the basement rocks outcrop in four shields along the edge of the craton.The Gabon-Cameroon Shield exposed at the northwestern edge of the craton is mainly composed of 2.9–2.8Ga tonalite–trondhjemite–granodiorite(TTG)with remnants of older greenstone belt rocks(Nedelec et al., 1990;Mathieu et al.,2001).The TTG and greenstone rocks were intruded by K-rich granites at2.7–2.6Ga,syenites at ca.2.3Ga and dolerite dikes at ca.2.1Ga(Toteu et al., 1994;Tchameni et al.,2001).The Archean-Paleoproterozo-ic granitoids have Nd model ages up to3.2Ga(Tchameni et al.,2001;Shang et al.,2004).The Bomu-Kibalian Shield in the northeastern corner of the craton exposes greenstone belts scattered in granitoid gneiss complexes with protolith ages of3.4–2.9Ga.These rocks were intruded by granitoid plutonism at ca.2.5Ga and development of shear belts around950Ma(Lavreau,1984;Borg and Shackleton, 1997).The Angolan Shield constitutes the southwestern part of the craton and comprises migmatite–granitoid com-plexes yielding Rb–Sr whole rock isochron dates ranging from2.6to1.7Ga,a gabbro–norite–charnockite complex with a Rb–Sr date of 2.8Ga,and metasedimentary se-quences with Rb–Sr dates of>2.2–1.7Ga(Cahen et al., 1984;De Carvalho et al.,2000,and references therein). However,the geological signi?cance of these Rb–Sr dates is still unclear(Hanson,2003).Zircon U–Pb geochronology indicated granitoid intrusions at ca.2.6,2.0and1.8–1.7Ga, anorthosite emplacement at1.4Ga and the occurrence of detrital zircons as old as3.0Ga(Seth et al.,1998;Mayer et al.,2004;McCourt et al.,2004;Kro¨ner et al.,2010). The Kasai Shield exposed in the southeastern margin of the craton is made up of metamorphic and igneous rocks, including a3.0Ga granodiorite and2.55Ga granites(Wal-raven and Rumvegeri,1993;Key et al.,2001).Regionally, porphyritic granites extensively intruded these Archean rocks at2.0Ga(Key et al.,2001).Combined U–Pb and Lu–Hf isotopic analyses of detrital zircons from Congo River tributaries suggest that ca.3.6Ga or even older crus-tal materials had contributed to Neoarchean and Paleopro-terozoic magmatism in the region(Batumike et al.,2009). The Kibaran Belt between the Tanzania and Congo Cra-tons contains two metasedimentary successions,the oldest of which was intruded by a1.38–1.37Ga granite-granodio-rite complex.Both metasedimentary successions were af-fected by a tectonic event accompanied by syan-and post-tectonic granite formation at1.2–1.0Ga(Kokonyangi

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–12099

et al.,2004;Tack et al.,2010).The Congo Craton is bor-dered on the north by the Pan-African Central African Fold Belt,the southern part of which includes gneissic base-ment with igneous protolith ages of ca.2.1Ga,Meso-to Neoproterozoic volcano-sedimentary basins,and630–580Ma granitoids containing a reworked crustal compo-nent in various degree(Toteu et al.,2001;van Schmus et al.,2008).

The Zambezi River?ows through the southeastern Con-go Craton,the northern Kalahari Craton(i.e.,Zimbabwe Craton),the Mesoproterozoic Irumide Belt,and Neoprote-rozoic orogenic belts among the Tanzania,Congo and Kal-ahari Cratons.The Irumide Belt along the southern margin of the Tanzania Craton is made up of2.7and2.0–1.9Ga basement rocks and overlying1.85Ga supracrustal units, which were locally intruded by granites at 1.6Ga and underwent extensive metamorphism accompanied by gran-ite intrusions at1.1–1.0Ga(Ring et al.,1999;Hanson, 2003;Johnson et al.,2005;De Waele and Fitzsimons, 2007).The southern Irumide Belt comprises arc-related terranes having a peak of magmatic as well as metamorphic activity at1.1–1.0Ga(De Waele and Fitzsimons,2007; Johnson et al.,2007a).The magmatic rocks within the belt yield Nd model ages up to3.3Ga(Johnson et al.,2007a;De Waele et al.,2008).The Zimbabwe Craton represents greenstone belts formed at ca.2.9and2.7Ga and grani-toids(gneisses)with emplacement ages between3.57and 2.7Ga(Horstwood et al.,1999;Jelsma and Dirks,2002; Prendergast,2004,and references therein).These rocks were overprinted by syn-tectonic granitoids at ca.2.6Ga and cut by the ma?c–ultrama?c Great Dyke at2.57Ga (Armstrong and Wilson,2000;Zeh et al.,2009).Evidence of Eoarchean crustal materials in the craton is provided by$3.8Ga detrital zircons(Dodson et al.,1988).The northwestern margin of the craton contains a volcanic and sedimentary sequence intruded by granites and meta-morphosed at 2.0–1.8Ga(Treloar and Kramers,1989; Majaule et al.,2001).The Neoproterozoic Zambezi and Lu?lian Belts are rift-related volcanic and sedimentary se-quences deposited on the passive margins of the Congo and Kalahari Cratons.The Zambezi Belt consists of Meso-proterozoic(ca.1.1Ga)basement,unconformably overlain by880Ma volcanic rocks and passive margin sediments that are cut by820Ma granites(Hanson,2003;Johnson et al.,2007b).The basement of the Lu?lian Belt is a com-plex mass of Paleoproterozoic(2.0–1.8Ga)granitoids, gneisses and sedimentary rocks that were intruded by 880Ma granites(Armstrong et al.,2005;Rainaud et al., 2005).The basement was unconformably overlain by vol-cano-sedimentary and passive margin sequences between 880and$635Ma(Wendor?,2005;De Waele et al., 2008).Both the Zambezi and Lu?lian Belts were over-printed by metamorphism at ca.530Ma as a result of crus-tal thickening(John et al.,2004;Johnson et al.,2007b).

The Orange River basin is underlain by the southern Kalahari Craton(i.e.,Kaapvaal Craton),Rehobothian Province,Namaqua-Natal and Gariep Belts.The Kaapvaal Craton consists mainly of granitoids and gneisses,inter-spersed with greenstone belts and overlain by Neoarchean volcanic and sedimentary sequences.Early crustal genera-tion took place from3.7to3.1Ga,followed by accretion of crustal fragments,granitoid plutonism and deposition of the volcanic and sedimentary sequences between 3.1 and 2.6Ga(de Wit et al.,1992;Brandl and de Wit, 1997).The western margin of the craton is further covered by Paleoproterozoic volcanic and sedimentary rocks,which are cut by1.75Ga ma?c sills and ca.1.2Ga felsic dykes (Cornell et al.,1998).Recent zircon U–Pb and Lu–Hf iso-topic studies(Zeh et al.,2009,2011)indicated that Paleo-and Mesoarchean crust provinces in the craton mostly have chondrite-like Hf isotopic compositions,whereas Neoarch-ean ones exhibit near-chondritic to highly sub-chondritic compositions,suggesting reworking of the Paleo-and Mes-oarchean crustal materials into some of the Neoarchean crust provinces.The Kaapvaal Craton is mantled by two Mesoproterozoic terranes:the Rehobothian Province to the west and the Namaqua-Natal Belt to the south.The Rehobothian Province is an assemblage of Mesoproterozo-ic arc-related igneous and sedimentary sequences,in which orthogneisses and migmatites showing zircon U–Pb ages from1.85to1.3Ga are overlain by1.25–1.1Ga volcanic and sedimentary rocks and cut by1.25–1.0Ga granodio-ritic to dominantly granitic intrusions(Hanson,2003;Beck-er et al.,2005).The Namaqua-Natal Belt comprises1.3–1.2Ga juvenile crust and1.2–1.0Ga reworked crust with Paleoproterozoic(2.0–1.8Ga)microcontinents,which were accreted to the Kalahari Craton by$1.03Ga during assem-bly of the Rodinia supercontinent(Robb et al.,1999;Sch-mitz and Bowring,2004;Eglinton,2006).The accretion was followed by rift-related magmatism at ca.850Ma and low-grade Pan-African(ca.500Ma)metamorphism (Robb et al.,1999;Frimmel et al.,2001;Raith et al., 2003).To the west,the Pan-African Gariep Belt represents a collage of rift-related sedimentary rocks(780–740Ma), oceanic crusts(ca.650Ma)incorporated during collisional orogeny at ca.550Ma,and syn-to post-orogenic granitic-syenitic intrusives dated at550–520Ma(Allsopp et al., 1979;Ha¨lbich and Alchin,1995;Frimmel and Frank, 1998).

3.METHODS

The zircons were concentrated from the river sand sam-ples using conventional magnetic and heavy liquid separa-tion techniques.Aliquots of zircon concentrates from the same samples were previously used for zircon U–Pb dating and Hf isotopic studies by Rino et al.(2008)and Iizuka et al.(2010).The zircons analyzed in this study were new separates from the same samples.All analyses were carried out in situ at the Research School of Earth Sciences,the Australian National University.The zircons studied here were treated in two di?erent ways with di?erent goals.First, zircons were hand-picked from the aliquots of zircon con-centrates and mounted in a conventional electron micro-scope-probe epoxy mount(25mm diameter)only to determine the zircon U–Pb age distributions of the river sand samples.To minimize preferential selection of zircon grains during mounting and analyses,we tried to pick all recognized zircons in a sub-population of the aliquots and analyzed them regardless of appearance.Second,zircons

100T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120

with minimum dimensions of50l m were selected and soaked in concentrated HF acid overnight at room temper-ature to remove metamict domains that may have compro-mised their isotope systematics(King et al.,1998).These zircons were mounted on tape and analyzed for U–Pb dat-ing by the rim-piercing method described in Campbell et al. (2005).This method provides a continuous pro?le of the age variation from the surface of the grain towards its cen-ter for a total depth penetration of ca.20l m.Based on the results,we identi?ed zircon grains suitable for O and Hf isotopic analyses.The grains selected for O and Hf isotopic analyses are concordant or<10%discordant for U–Pb iso-topes,are free of age zonation,and have few inclusions. These grains were mounted in an epoxy megamount (35mm diameter;Ickert et al.,2008)and were imaged using cathodoluminescence(CL)spectroscopy to reveal any inter-nal structure(Fig.2).The CL images are particularly useful to identify the presence of a small inherited core and thick over-growths which might be overlooked by the rim-pierc-ing method.Subsequently,we analyzed these grains for?rst O and then Hf isotopes.The second approach is designed especially for O isotopic analyses with high precision and accuracy,which require?at sample surface with minimal relief(Ickert et al.,2008;Kita et al.,2009).The analytical methods used in this study are given in Appendix A.

4.RESULTS

The U–Pb isotopic and trace element data for zircons are shown in Electronic Annex Table EA-1.The U–Pb iso-topic data for detrital zircons randomly picked from the samples are also graphically presented on concordia dia-grams(Fig.3)and U–Pb age histograms(Fig.4)for each river.207Pb/206Pb ages have been used for zircons having the ages older than1.5Ga,and206Pb/238U ages for younger grains.The O and Lu–Hf isotopic data for the dated zircons are listed in Electronic Annex Tables EA-2and EA-3, respectively.All O isotopic data are presented as the d18O notation,expressed as deviations from Vienna standard mean ocean water(VSMOW,18O/16O=0.0020052,Baert-schi,1976)in parts per thousand.The U–Pb and O isotope systematics are illustrated in Fig.5as plots of d18O versus U–Pb age,together with the d18O range for mantle zircon (5.3±0.6&,Valley et al.,2005).The U–Pb and Lu–Hf iso-tope systematics are summarized in Fig.6as plots of e Hf(t) against U–Pb age.The present-day chondritic parameters reported by Bouvier et al.(2008)were used to calculate e Hf(t).Fig.6also shows the e Hf(t)range for the arc mantle (Appendix B),which covers the e Hf(t)?eld of382young is-land arc basalts(+13.3+3.9/à8.7e at95%con?dence level) and extrapolate back to e Hf(4.56Ga)=0.

The randomly picked zircon grains from the Niger River (NGR1)gave U–Pb ages between3.2and0.2Ga with a dominant population at0.7–0.5Ga(Figs.3a and4a).The O and Lu–Hf isotopic analyses reveal that zircons dated at0.7–0.5Ga have mantle-like to markedly elevated d18O values(5–11&)and highly negative to positive e Hf(t)values (à22to+7e)(Figs.5a and6a).Two ca.2.1Ga zircon grains have positive e Hf(t)values and d18O close to the range for mantle zircon,suggesting that they derived from Paleoproterozoic juvenile crust in the West African Craton. Zircons dated at ca.1.0and0.2Ga also have nearly mantle-like or slightly higher d18O values.The e Hf(t)of the ca.

cathodoluminescence images for(a–c)igneous zircons(oscillatory-and sector-zoned)and(d)

sites for O and Lu–Hf isotopes are shown,together with d18O and e Hf(t)values.

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120101

1.0Ga zircons range fromà24to+8e,whereas those of the ca.0.2Ga zircons are betweenà9andà3e.

The U–Pb age population of the Nile River zircons (NIL)indicate age peaks at1.1–0.9,0.85–0.7,and0.7–0.55Ga and small groups at ca.2.6and2.0Ga(Figs.3b and4b).Zircons dated at0.85–0.7and0.2Ga are charac-terized by positive e Hf(t)and near mantle-like d18O values (Figs.5b and6b).The0.85–0.7Ga zircons can be inter-preted as coming from Neoproterozoic juvenile arc crust in the Arabian–Nubian Shield.Zircons having U–Pb ages of1.1–0.9and0.7–0.5Ga show wide variations in d18O (5–11and4–8&,respectively)and e Hf(t)(à25to+8and à30to+10e,respectively).The e Hf(t)of Archean-Paleo-proterozoic zircons scatter from clearly subchondritic to chondritic(from ca.à15to0e).Their d18O are generally lower than7.5&,but a few ca.2.0Ga grains have higher values up to10&.

Sample CNG2from the Congo River contains a signif-icantly higher proportion of Archean and Paleoproterozoic zircons than the other samples(Figs.3c and4c).The zircon U–Pb age population exhibits?ve groups:2.7–2.5,2.1–1.8, 1.2–0.9,0.7–0.5and ca.0.3Ga.Zircons dated at1.2–0.9 and0.7–0.5Ga display wide ranges of e Hf(t)from highly negative to mantle-like values(from ca.à25to+10e)and d18O values that vary from mantle-like to strongly elevated values(from ca.5&to10&)(Figs.5c and6c).A similar O isotopic variation is observed for2.2–1.7Ga zircons,but their e Hf(t)values show a smaller range(from ca.à16to +4e).Most P2.5and60.4Ga zircons exhibit mantle-like or weakly elevated d18O values and nearly chondritic to weakly negative e Hf(t)values,although there are only six 60.4Ga zircons in this study.These U–Pb and Lu–Hf iso-topic data are in excellent agreement with those obtained by Iizuka et al.(2010).

The U–Pb age data for the Zambezi River zircons (ZMB2)show prominent peaks at 1.1–1.0,ca.0.8and 0.6–0.5Ga and small groups at ca. 2.7and 2.0Ga (Figs.3d and4d).Zircons with U–Pb ages between650

102T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120

and450Ma exhibit wide variations in both d18O(5.0–12.8&;Fig.6d)and e Hf(t)(à23to+6e;Fig.5d).The e Hf(t) of1.1–0.9and ca.0.8Ga zircons are scattered from super-chondritic to highly subchondritic,but their d18O values are clustered around7.5&and5.0&,close to the mantle value. Paleoproterozoic(ca.2.0Ga)zircons gave chondritic to moderately subchondritic e Hf(t)and d18O slightly higher than the mantle value.Archean zircons yielded near-chon-dritic e Hf(t)and mantle-like d18O values.

The Orange River zircons(ORG2)de?ne two major U–Pb age groups at1.3–1.0and0.7–0.5Ga(Figs.3e and4e). Most ORG2zircons younger than1.5Ga have elevated d18O values(up to ca.10&)relative to mantle zircon (Fig.5e).Notably,a few middle-late Proterozoic zircons ex-hibit signi?cantly low d18O values down toà1.2&(Fig.5e). Such low d18O zircons have been found from sub-volcanic granites in the British Tertiary Igneous Province and Yel-lowstone and from metamorphosed granites in Dabie-Sulu orogenic belt in China.The low d18O values are attributed either to remelting of country rocks altered by meteoric water or to metamorphic zircon crystallization/recrystalli-zation assisted by low d18O hydrothermal?uids(Gilliam and Valley,1997;Bindeman and Valley,2001;Monani and Valley,2001;Rumble et al.,2002;Chen et al.,2003, 2011;Zheng et al.,2004).Our low d18O zircons are charac-terized by very high U abundances(ca.1500ppm)and low Th/U(ca.0.1).The latter are frequently observed in meta-morphic zircons(Hoskin and Ireland,2000;Rubatto,2002; Chen et al.,2011).Furthermore,the CL images(Fig.2d)re-veal that fractures are developed along high-U metamict zones within the zircon crystals,which can signi?cantly en-hance hydrothermal alteration.These observations suggest that the low d18O Orange River zircons derived their low d18O values from secondary hydrothermal alteration,rather than from the parental magmas.Therefore,we have not used the data obtained from these zircons for the following discussion.The Lu–Hf isotopic analyses reveal that the1.3–1.0Ga zircons have e Hf(t)values ranging fromà18to +10e,whereas those of the zircons younger than1.0Ga are betweenà11and+4e(Fig.6e).The1.2–1.0Ga detrital zircons with non-mantle-like isotopic signatures may be de-rived from the reworked crust in the Namaqua-Natal Belt (Eglinton,2006).

The U–Pb age distributions for all samples studied here (Fig.4)are broadly similar to those observed by Rino et al. (2008),but in detail this study resolved more prominent sharp peaks in the Neoproterozoic and Phanerozoic era. This is most likely because we utilized206Pb/238U ages for the young zircons,whereas Rino et al.(2008)used impre-cise207Pb/206Pb ages,resulting in much broader peaks for the young zircon populations.

5.DISCUSSION

Fig.7compares the timing of the supercontinent assem-bly with the accumulated U–Pb age population of the detri-tal zircons from the?ve rivers.To avoid bias by giving undue weight to rivers with the highest number of analyses,the probability was calculated by accumulating the age propor-tions rather than the number of grains of detrital zircons from the?ve rivers:the relative probability of zircons having U–Pb ages between t and t-50Ma is an average of percent-ages of the zircons for each river sand sample.All the O and Lu–Hf isotopic data for unaltered detrital zircons from the?ve rivers are summarized in Fig.8a and b,as functions of U–Pb age.In addition,Fig.8c shows a plot of d18O versus D e Hf(t)AM,the deviation of e Hf(t)between detrital zircons and the arc mantle.Zircon D e Hf(t)AM is a function of resi-dence time of the reworked crust,whereas zircon d18O is a sensitive record of reworking of supracrustal materials that had undergone water–rock interactions at low temperatures (Valley et al.,2005).Hence,integration of U–Pb,Lu–Hf and O isotopic studies of zircons will provide constraints on the nature and antiquity of the reworked crust(e.g.,Zheng et al.,2006;Kemp et al.,2007;Bolhar et al.,2008;Hiess et al.,2009).We will now use these plots to link the super-

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120103

continent cycle with African crustal development and to dis-cuss reworking and generation of the continental crust.

5.1.Supercontinent cycle and crustal development

All samples studied here show prominent peaks at0.7–0.5Ga in the detrital zircon U–Pb age histograms (Fig.7).Furthermore,some of the samples exhibit clear peaks at1.2–1.0and ca.0.8Ga and small groups at ca.

2.7,2.1–1.8,and ca.0.3Ga.Importantly,these ages except ca.0.8Ga correspond well with the timing of superconti-nent assembly(e.g.,Bleeker,2003;Veevers,2004;Zhao et al.,2004;Bradley,2011).The correspondence between peaks in U–Pb zircon ages and the timing of supercontinent assembly has been also observed in detrital zircons from other regions(Rino et al.,2004,2008;Campbell and Allen, 2008;O’Reilly et al.,2008;Hawkesworth et al.,2010;Iizuka et al.,2010),and led authors to propose models linking the supercontinental cycle to crustal development in di?erent ways:(i)the supercontinent formations are coupled with episodic major igneous events resulting either from sub-ducted slab avalanches and mantle instabilities(Condie, 1998;Rino et al.,2004,2008)or from signi?cant crustal remelting during continent–continent collisions(Campbell and Allen,2008;Bradley,2011;Voice et al.,2011);or(ii) the preservation potential of formed crust is greater in con-tinental collisional settings than subduction-and extension-zones(Hawkesworth et al.,2010;Condie et al.,2011;Lan-caster et al.,2011).Here,we argue that the detrital zircon record is likely to be biased by varying rate of crustal ero-sion associated with the supercontinent cycle.

It has been shown that crystallization age distributions of detrital zircons from the Mississippi River and Austra-lian sediments are broadly consistent with those of the base-ment rocks in the source regions(Rino et al.,2004; Hawkesworth et al.,2010).The observations suggest that the U–Pb age peaks of detrital zircons from large riverine systems are essentially representative of major magmatic

104T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120

events in the drainage area,even though there is a general tendency of young basement rocks to be more readily sam-pled(Alle`gre and Rousseau,1984;Cawood et al.,2003;Dhuime et al.,2011b).However,our data reveal that the prominent Pan-African aged peaks are observed even for the rivers in which Pan-African igneous rocks are minor in the drainage basins(i.e.,the Congo and Orange Rivers). This indicates that the detrital zircons with Pan-African crystallization ages have been recycled from older sedimen-tary rocks that were deposited in older basins which drained areas that were di?erent to those drained by the modern river basins,unless a signi?cant amount of unrec-ognized Pan-African granitoids exists in the drainage ba-sins.Furthermore,considering that the Niger River basin mainly comprises Archean and Paleoproterozoic terranes (Section2and Fig.1),the Pan-African age peak is clearly over represented in the U–Pb age histogram(Fig.4a).

The largest mountain ranges in geological history were built during continent–continent collisions associated with supercontinent amalgamation(Squire et al.,2006;Camp-bell and Allen,2008).Because the rate of crustal erosion strongly depends on the basin relief(Pinet and Souriau, 1988;Milliman and Syvitski,1992;Summer?eld and Hulton,1994;Montgomery and Brandon,2002)and also

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120105

because colliding continents should have rims composed of young igneous rocks as a result of continental arc magmatism and/or accretion of oceanic arcs followed by collisional magmatism,rapid erosion of young igneous rocks is an inevitable consequence of supermountain for-mation whereby thick sedimentary rocks envelop the super-continent and its o?shore basins.Subsequently,these sedimentary rocks would be recycled into younger sedi-ments especially when they are tectonically uplifted.Such sediment–sediment recycling should proceed much faster than the erosion of igneous/metamorphic basement(Veizer and Jansen,1979;Goldstein et al.,1984;McLennan,1988), because sedimentary rocks,the most abundant rock type on the continental surface,are disaggregated by wind and water more readily than igneous basements.Accordingly, preferential erosion of young igneous basements during the periods of supercontinent assembly followed by sedi-ment–sediment recycling over geological history is likely represented as the U–Pb age peaks of detrital zircons from the modern rivers.The impact of mountain building on detrital zircon U–Pb age population is indicated by the observation that most detrital zircons with various U–Pb ages have U–Th–He or?ssiontrack and,by extension, exhumation ages similar to the timing of young collisional orogenies in the sedimentary basins(Carter and Moss, 1999;Rahl et al.,2003;Campbell et al.,2005).Thus,the geological signi?cance of quantitative age distributions of detrital zircons should be interpreted with caution(see also Dickinson,2008;Yang et al.,2012).

The models linking supercontinent cycle and crustal development can be further evaluated using the detrital zir-con O and Lu–Hf isotopic data.Detrital zircons crystallized during the periods of supercontinent assembly have wide variations in d18O and e Hf(t)values,and most grains show non-mantle-like isotopic signatures.In contrast,most ca.

0.8Ga detrital zircons have mantle-like isotopic signatures (Fig8).A similar pattern is observed in a global zircon U–Pb and Lu–Hf isotopic database(Belousova et al.,2010; Condie et al.,2011;Roberts,2012).These features can be attributed to supermountain building during superconti-nent assembly,because continental collisions generate granitoids having non-mantle-like isotopic signatures via crustal melting(e.g.,Deniel et al.,1987;France-Lanord and Le Fort,1988;Inger and Harris,1993).Subsequent up-lift of these granitoids leads to their preferential erosion.

Alternatively,if the detrital zircon record is representa-tive of the major magmatic events in the drainage basin rather than biases in sedimentary processes,the observed trends can be interpreted to indicate that crustal remelting was signi?cant during periods of supercontinent amalgam-ation,whereas juvenile crust generation was dominant at ca.0.8Ga as well as at$1.7–1.2and ca.0.4Ga(Condie et al.,2011;Roberts,2012).These contrasting modes of crust formation may be linked to a change in the major tec-tonic regime associated with supercontinent cycle(Collins et al.,2011).Recent isotopic studies of extensional accre-tionary orogens(Kemp et al.,2009;Holm-Denoma and Das,2010;Phillips et al.,2011)have indicated that the e?ect of juvenile magmatic input was enhanced during episodes of extensional rifting as a result of the diminishing avail-ability of older crust for arc magma genesis,whereas crustal remelting dominated during contractional periods.Crustal remelting would result in the formation of evolved buoyant crust as well as residual crust that might be recycled back into the mantle due to its high density(Kay and Kay, 1993;Kelemen,1995;Rudnick,1995;Gao et al.,2004; Zandt et al.,2004;Lee et al.,2007).As a consequence, the supercontinent cycle could regulate crustal development in two di?erent modes:(i)during supercontinent fragmen-

106T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120

tation o?shore subduction zones globally developed and generated more juvenile crust(Scholl and von Huene, 2007;Silver and Behn,2008),versus(ii)during superconti-nent assembly the generated crust stabilized through accre-tion and reworking(Cawood and Buchan,2007; Hawkesworth et al.,2010;Roberts,2012).The change in crust formation mode associated with the supercontinent cycle is comparable with the models either of preferential crustal preservation(Hawkesworth et al.,2010;Condie et al.,2011)or rapid granitoid magmatism by crustal remelting during the supercontinent formations(Campbell and Allen,2008).Signi?cant crustal remelting during the supercontinent formations is also supported by the obser-vation that the main periods of granulite facies ultrahigh temperature metamorphism coincide with supercontinent assembly(Brown,2007).However,the detrital zircon isoto-pic data are inconsistent with the hypothesis that the forma-tion of supercontinents were associated with slab avalanches and mantle instabilities(Condie,1998;Rino et al.,2004,2008),because major magmatic events due to mantle instabilities would generate juvenile magmas with mantle-like,or chondritic,isotopic signatures(Stein and Hofmann,1994;Guitreau et al.,2012).

All sediment samples studied here yielded a few detrital zircons older than2.7Ga but none older than3.5Ga.The low proportions of Archean detrital zircon,despite the sig-ni?cant extent of Archean cratons in the drainage areas, can be attributed to the low erosion rate of low topography Archean basement and the probability that ancient detrital zircon grains were preferentially dissolved during sediment–sediment recycling due to the accumulation of radiation damage in the grains.In addition,considering that sedi-ment–sediment recycling has essentially taken place in con-tinental settings(Veizer and Jansen,1985;Iizuka et al., 2010),the lack of>3.5Ga detrital zircons may be attrib-uted to the absence of continents before ca.3.5Ga.This view is consistent with the geological evidence that Archean cratons were generally stabilized through accretionary orogeny in the Mesoarchean(de Wit et al.,1992;Moyen et al.,2006;Van Kranendonk et al.,2010;Shirey and Rich-ardson,2011),which set the scene for subsequent sediment–sediment recycling.

5.2.Reworking of continental crust

5.2.1.Insights into the nature of reworked crust

Many zircons,even those with Archean crystallization ages,have e Hf(t)that lie well below the arc mantle evolu-tion line(Fig.8).In detail,only21%of the zircons plot within the range for Hf isotopic evolution of the arc man-tle as shown in Fig.8.Note that the Hf isotopic ratios of the Gri?n et al.(2000)depleted mantle evolution curve, which is widely used in zircon Lu–Hf isotopic studies (e.g.,Wu et al.,2007;Yang et al.,2009;Belousova et al.,2010),lie above the arc mantle curve at any given time.The results indicate that parental magmas of most detrital zircons formed,at least partly,by remelting pre-existing crust.In other words,granitoid magmatism has played a fundamental role in the di?erentiation of the con-tinental crust since the Archean era,and although the pri-mary agent of growth of the continental crust may be ma?c magmatism,this type of magmatism yields few zircons.

It is evident in Fig.8c that all zircons with mantle-like Hf isotopic compositions(D e Hf(t)AM>à4)have mantle-like or moderately elevated d18O values(4.2–7.5&),which contrasts with the wide variation in D e Hf(t)AM of the zir-cons having mantle-like d18O values.The lack of zircons having mantle-like e Hf(t)and high d18O indicates little con-tribution of rocks having high d18O and either mantle-like e Hf(t)or extremely low Hf contents.Typical rocks with ele-vated d18O and very low Hf concentrations are carbonate and chert(Eiler,2001).In contrast to the chemical sedimen-tary rocks,most terreigneous sedimentary rocks contain substantial amounts of Hf.Moreover,mature terreigneous sedimentary rocks such as pelite tend to have higher d18O than immature sediments such as graywacke(Longsta?e and Schwarcz,1977;Shieh and Schwarcz,1978;Eiler, 2001;Valley et al.,2005).This suggests that the high d18O of terreigneous sedimentary rocks is a result of sediment–sediment recycling involving a high fraction of clay miner-als.Hence,the lack of zircons having high d18O and mantle-like e Hf(t)suggests that sediment–sediment recycling asso-ciated with the d18O elevation has taken place over crustal residence times long enough to produce signi?cant change in Hf isotopic ratios relative to CHUR.

All studied African zircons older than2.1Ga or younger than0.4Ga have d18O values less than7.6&and most are mantle-like(4.0–6.5&)(Fig.8),despite having negative

D e Hf(t)AM values down to-25e.The zircon d18O value of

7.5&corresponds to d18O value of$9&for felsic magmas and such magmas can be generated either by remelting of supracrustal rocks with moderately elevated d18O such as altered basalts and clastic sediments or by restricted con-tamination by mature sediments(e.g.,Cavosie et al., 2005;Kemp et al.,2007;Nebel et al.,2011).The detrital zir-con data therefore suggest that in African granitoid mag-matism reworking of igneous rocks dominated over reworking of mature sedimentary rocks before2.1Ga and after0.4Ga.By contrast,about40%of the2.1–0.4Ga zir-cons with e Hf(t)below the range for the arc mantle have d18O values higher than7.5&,indicating signi?cant reworking of both igneous and mature sedimentary rocks in granitoid formation during that period.

5.2.2.Secular change in detrital zircon d18O and its geological signi?cance

The restricted range of Archean d18O values(Fig.8)also appeared in a compilation of zircon O isotopic data by Val-ley et al.(2005)and was interpreted as re?ecting a limited contribution of a crustal component relative to a juvenile magma component,with an implication of high rates of continental crust generation throughout the Archean.This interpretation is,however,inconsistent with the highly neg-ative D e Hf(t)AM values of Archean zircons(Fig.8).Recon-ciling the zircon O and Lu–Hf isotopic data requires a signi?cant contribution of old igneous and/or immature supracrustal rocks to the source of Archean granitoid mag-mas.Valley et al.(2005)also suggested a continuous in-crease in zircon d18O values through the Proterozoic era.

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120107

Since the compilation by Valley et al.(2005),however,a signi?cant number of O isotopic data have been newly ob-tained for magmatic zircons especially for the early-middle Proterozoic(e.g.,Wang et al.,2009,2011;Li et al.,2012; Yin et al.,2012;this study),and the new data present a dif-ferent picture.There is a step increase in the maximum d18O in the early Proterozoic,from a restricted range in d18O throughout the Archean,followed by a wide range in zir-cons younger than 2.1Ga without signi?cant secular change in the maximum values thereafter(Fig.8a).This ?nding suggests that reworking of sediments with elevated d18O in granitoid crust formation increased abruptly rather than gradually at ca.2.1Ga.Note that one2.1Ga zircon grain has a markedly elevated d18O(12.9&)as compared to the other African detrital zircons with similar crystalliza-tion ages,but the value is comparable to the highest d18O of $2.0Ga detrital zircons from Greater Russian Rivers (Wang et al.,2011).Note also that even if the outlying data for the2.1Ga zircon grain are ignored,there is still a step increase in d18O at ca.2.1Ga.

There are two plausible mechanisms capable of abruptly changing the contribution of high d18O sediments to gran-itoid magmatism.The?rst is a change in the formation pro-cess of the majority of the granitoids.The style of major granitoid genesis has been considered to change around the Archean-Proterozoic boundary,because there are prominent geochemical di?erences between typical Archean TTG and most post-Archean granitoids:the former show REE patterns having HREE depletion and little Eu anom-alies,while the latter have negative Eu anomalies(e.g.,Tay-lor and McLennan,1985;Martin,1986).These geochemical features can be explained if Archean TTG magmas were generated at depths in equilibrium with residual garnet (Petford and Atherton,1996;Smithies,2000;Whalen et al.,2002;Kleinhanns et al.,2003;Rapp et al.,2003;Be′-dard,2006)possibly by partial melting of subducted slabs (Martin,1986;Drummond and Defant,1990;Foley et al., 2002;Hiess et al.,2009),whereas during post-Archean times most granitoid magmas formed at crustal depths in the presence of plagioclase as a residual or fractionating phase(Wyllie,1984;Taylor and McLennan,1985).Consid-ering that immature terrigenous sediment with restricted d18O is a major component of subducted sediments(Veizer and Jansen,1985;Plank and Langmuir,1998;Iizuka et al., 2010)and also that metapelites occur commonly in granu-lite terranes and xenoliths(Rudnick and Gao,2003),the change in granitoid formation process from slab melting to deep continental crust melting may account for the rise of magmatic d18O values at ca.2.1Ga.

An alternative mechanism is O isotopic evolution of sed-imentary rocks.There is a tendency for older sediments to contain a higher fraction of graywackes than pelites(Veizer and Mackenzie,2003).The likely cause is that dominant Archean sediments were immature volcanoclastic sediments or were directly derived from young igneous basements (Hessler and Lowe,2006;Squire et al.,2010),whereas the dominant process in young sedimentary formation is sedi-ment–sediment recycling which could result in elevated sed-imentary d18O values(Veizer and Mackenzie,2003). Furthermore,it has been shown that Archean sediments experienced aggressive weathering likely due to higher atmospheric CO2concentrations and surface temperatures (Hessler and Lowe,2006).Such aggressive weathering would e?ciently dissolve chemically labile minerals such as plagioclase and micas,leaving highly quartz-rich sedi-ments that have restricted d18O values and are sterile as magma sources(Clemens and Vielzeuf,1987;Lowe and Tice,2004;Valley et al.,2005;Hessler and Lowe,2006). Hence,the formation and remelting of high d18O sedimen-tary rocks would be inhibited on the Archean Earth.The step increase in zircon d18O at ca.2.1Ga may indicate either the enhancement of sediment–sediment recycling dur-ing the formation of the supercontinent Columbia(Zhao et al.,2004;Hou et al.,2008)or the decrease in atmospheric CO2concentrations and surface temperatures at that time. The latter may be linked to the great oxygenation of the atmosphere and oceans at 2.45–2.22Ga(e.g.,Holland, 2002;Bekker et al.,2004;Can?eld,2005).

5.3.Implications for the timing of major continental growth

The parental magmas of most detrital zircons contain reworked pre-existing crust.Assuming that the reworked crust with a certain176Lu/177Hf was extracted from the arc mantle(solid line in Fig.8b),their formation ages can be constrained by calculating two-stage Hf model ages. However,if the parental magma is a mixture of di?erent age components,then the Hf model age provides a hybrid age of the constituent components(Arndt and Goldstein, 1987).The mixing of di?erent age components can geolog-ically take place when the parental magma contains both a juvenile component and an old reworked crustal compo-nent,and when the reworked crustal component itself in-volves mixed-age source rocks such as sedimentary rocks. In addition,because the model age calculation depends on the176Lu/177Hf of the reworked crust,large uncertainty in the ratio leads to imprecise age determination of the juve-nile crust generation event.

Given that reworking of sediment into the zircon paren-tal magma is represented by elevated zircon d18O,we can evaluate the extent to which the sedimentary reworking af-fects Hf model ages by comparing the model age distribu-tions of detrital zircons having mantle-like and non-mantle-like d18O values(Kemp et al.,2006).Furthermore, since more geochemically evolved crust tends to have higher d18O and lower Lu/Hf(Vervoort and Patchett,1996),the zircon d18O values allow us to reduce uncertainty in the choice of176Lu/177Hf for the model age calculations(Wang et al.,2009,2011).We have considered that zircons with near-mantle-like d18O(<6.5&)crystallized from a melt de-rived by melting of ma?c crust(176Lu/177Hf=0.021,Kemp et al.,2006),whereas those with d18O higher than10.0& crystallized from a melt generated by melting of upper con-tinental crust(176Lu/177Hf=0.0083,Rudnick and Gao, 2003).For zircons with d18O lying between these extremes, 176Lu/177Hf values have been determined by linear interpo-lation using an equation:

176Lu=177Hf?0:0446à0:0036?d18O

108T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120

The calculated Hf model ages are listed in Electronic Annex Table EA-3.For grains with positive D e Hf(t)AM values,the U–Pb ages are used as the Hf model ages.

As outlined above(Section5.1),the U–Pb age distribu-tion of detrital zircons(Fig.7)is likely to be biased by either sedimentary processes or preferential crust preserva-tion associated with the supercontinent cycle.It follows that their Hf model age distribution would be biased too(Lan-caster et al.,2011),making it di?cult to estimate the gener-ation rate of juvenile continental crust.Despite this,periods of major continental crust generation can still be retrieved from Hf model age distributions of each generation of detrital zircons(Fig.9).All Hf model age histograms for the1.3–0.9,0.7–0.45and<0.45Ga zircon populations have a peak between1.8and1.2Ga and no trend toward older model age peaks is observed in older detrital zircon popula-tions(Fig.9a–d;https://www.doczj.com/doc/b413032514.html,ncaster et al.,2011).Moreover,the detrital zircons with mantle-like d18O also de?ne a major Hf model age group at that time(Fig.9h).These observa-tions strongly suggest that a signi?cant amount of juvenile crust was generated in the Paleo-Mesoproterozoic and that some of the generated crust was subsequently remelted to form younger granitoid crust with various crustal residence times,while some crust was reworked into younger granit-oid crust via sedimentary processes resulting in elevated d18O,even though the possibility that the Hf model ages are results of mixing between>1.8Ga reworked crustal melts and<1.2Ga juvenile magmas cannot been ruled out.The signi?cance of juvenile crust generation in the Pro-terozoic is supported by the observation that Os model ages of African mantle xenoliths from o?-cratonic localities, which re?ect the timing of mantle depletion events,cluster between2.0and0.8Ga(e.g.,Carlson and Pearson,2005). Furthermore,the1.8–1.2Ga period corresponds with the timing of supercontinent fragmentation(Fig.7),when the generation of juvenile crust might be signi?cant(Sec-tion5.1).The lack of1.8–1.2Ga detrital zircons with man-tle-like isotopic signatures may highlight that the juvenile crust was dominantly ma?c and therefore contained few zircons.

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120109

6.CONCLUSIONS

U–Pb,Lu–Hf and O isotope systematics of detrital zir-cons from the Africa’s largest rivers reinforce the interplay between the supercontinent cycle and crustal development. The supercontinent cycle could regulate crustal develop-ment in two di?erent ways.First,juvenile crust was signif-icantly generated in extensional settings during supercontinent fragmentation,and subsequently the gener-ated juvenile crust was stabilized through crustal accretion and reworking during supercontinent assembly.Second, continental collision during supercontinent assembly built supermountains consisting largely of the young igneous rocks,resulting in e?cient erosion and cycling of the igne-ous rocks into sedimentary system.As a consequence,it is likely that the detrital zircon record is biased either by sed-imentary processes or preferential crust preservation in association with the supercontinent cycle.Nevertheless,ma-jor African continental growth in the Paleo-Mesoprotero-zoic has been retrieved from Hf model age histograms for each age population of the detrital zircons.We found that terreigneous sediments could acquire their heavy O isotopic compositions over long crustal residence times via pro-longed sediment–sediment recycling.Reworking of older igneous rocks and/or immature sedimentary rocks is impor-tant in granitoid magma genesis throughout geological his-tory,whereas the role of rocks with high d18O and low Hf contents such as chert and carbonate is minor.The e?ect of reworking of mature sedimentary rocks has been signi?cant since ca.2.1Ga.This suggests an increase either in the availability of mature sedimentary rocks for granitoid mag-ma genesis or in the production of mature sedimentary rocks through sediment–sediment recycling.

ACKNOWLEDGMENTS

We are grateful to L.Kinsley,P.Holden,I.Williams and M. Fanning for analytical support.We thank S.Rino,N.Hammond and G.Davies for assistance in?eldwork and K.Ozawa for discus-sion.Constructive comments by T.Kemp,B.Dhuime and J.Kra-mers were helpful for improving the manuscript.This work was ?nanced by grants from the Ministry of Education,Culture, Sports,Technology and Science,Japan to S.M.,from the Japan Society for the Promotion of Science to S.M.and T.I.,and from the Australian Research Council to I.H.C.(DP0556923).T.I. acknowledges support by Overseas Internship Program for Out-standing Young Earth and Planetary Researchers from the Univer-sity of Tokyo.

APPENDIX A.ANALYTICAL PROCEDURES

A.1.U–Pb isotopic dating

U–Pb isotopic dating was performed on an Agilent 7500a quadrupole inductively coupled plasma mass spec-trometry(ICPMS)attached with a HelEx ArF excimer la-ser ablation(LA)system.Abundances of Ti,La,Ce,Sm, Eu,Dy,Lu,Hf and Th were also determined concurrently with the U–Pb isotopic dating.We utilized the integration time of40ms for206,207,208Pb,235,238U and232Th,and 10ms for all other isotopes(29Si,31P,49Ti,91Zr,139La, 140Ce,147Sm,153Eu,163Dy,175Lu,177Hf).The data were ac-quired for a period of60s including a$25s baseline mea-surement at the beginning,with a time resolved analytical procedure(TRA),in which signal intensities for each mass and isotopic ratios are displayed as a function of time dur-ing the https://www.doczj.com/doc/b413032514.html,ser ablation sampling was carried out with a beam diameter of40l m and repetition rate of 5Hz in a helium atmosphere in a small-volume sample cell ($2.5cm3e?ective volume)(Eggins et al.,1998).The re-sponse time of the ablation cell,which we de?ne as the time the signal takes to decay by a factor of10,is$2s.The ra-pid response time allows us to evaluate down-hole age zon-ing and the presence of inclusions.Note that failure to recognize age zoning would result in reporting of meaning-less mixed ages,which in turn lead to arti?cial e Hf(t)(Har-rison et al.,2005)and Hf model ages.The presence of inclusions was checked by monitoring signal intensities of 31P and49Ti,as these elements are major constitutes of min-erals frequently included by zircon such as apatite and rutile.

Depth-dependent inter-elemental U–Th-Pb fraction-ation was corrected by reference to multiple measurements of standard zircon TEMORA(Black et al.,2004).The amount of common Pb was determined by the208Pb meth-od(Compston et al.,1984)and the calculated amount sub-tracted,assuming a common Pb composition from the age-dependent Pb model of Cumming and Richards(1975). Note that common206Pb was less than1%of total206Pb in most cases so the di?erence between the common Pb cor-rected and uncorrected ages was generally small.A com-mon Pb-corrected age was selected over the uncorrected one only if the correction brought the analysis closer to concordia.Analytical uncertainties of each spot analysis combine the internal precisions and the reproducibility of the TEMORA standard zircon analyses,added in quadra-ture.For the trace element analyses,depth-dependent ele-mental fractionation was normalized against NIST610 SRM,and concentrations were calculated using stoichiom-etric Si as the internal standard.More detailed data pro-cessing procedures are described in Campbell et al.(2005).

The precision and accuracy of our zircon U–Pb isotopic dating were evaluated by analyzing the standard zircon R33 as an unknown sample,which has an ID-TIMS age of 418.9±0.4Ma(Black et al.,2004).The LA-ICPMS U–Pb isotopic data were obtained from209spots during the analytical sessions of this study,and yielded mean 206Pb/238U and207Pb/235U ages of420±16and422±31Ma(2s.d.),respectively(Electronic Annex Table EA-4).These results are in good agreement with the literature value.

A.2.O isotopic analyses

The O isotopic compositions of the dated zircons were determined with the SHRIMP II multiple collector ion microprobe.A$3.5nA primary133Cs+beam was acceler-ated to15keV,and focused into a spot$25l m diameter on the sample surface coated with10nm Al.This generates approximately250pA of Oàsecondary ions from most sil-

110T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120

icate samples.Surface charge was neutralized by a moder-ate-energy(1.1keV)electron beam focused to$100l m with a45°incident angle.The Oàsecondary ions were accelerated to10kV with extraction lenses and separated into di?erent isotopes with electrostatic analyzer and mag-netic sector.The separated18O and16O were simulta-neously detected by dual Faraday cups with1011and 1010X resistors.We utilized a mass resolution of$2500 at1%peak height which is su?cient to separate potential isobaric interferences by17OH and16OD on18O.These operating conditions resulted in typical signal intensities of$0.1V for18O and$5V for16O,respectively.Data acquisition comprises two sets of six measurements,each with10s integration time,leading to total count times of $120s.A180s pre-sputter and secondary auto-tuning pre-cede each isotopic ratio measurement.Background count rates were measured at the start of each analytical session.

Because the e?ect of mass discrimination in ion micro-probe highly depends on sample-matrix,a matrix-matched standard reference should be used for the isotopic ratio nor-malization.In this study,analyses of in-house zircon stan-dard TEMORA-3(18O/16O=0.0020204Valley et al., unpublished data obtained by the laser?uorination tech-nique)embedded in each epoxy mount are interspersed with every three or four unknown sample measurements.The measured18O/16O on the standard were used to correct for instrumental drift and instrumental fractionation including mass discrimination and relative gain between electrometers.Analytical uncertainties of each unknown sample analysis combine the internal run errors(2s.e.,typ-ically0.1–0.4&)and the reproducibility of the standard zir-con analyses(2s.d.after the instrumental drift correction), added in quadrature.Analytical accuracy and precision are indicated by the results from the zircon standard R33func-tioning as a secondary reference material(Electronic Annex Table EA-5).Our data across all analytical sessions in this study gave a mean d18O=5.37±0.73&(2s.d.,n=81 which is consistent with the reported value of 5.55±0.08&(2r,Valley,2003).More detailed instrumen-tal setting and analytical procedures are described in Ickert et al.(2008).

A.3.Lu–Hf isotopic analyses

The Lu–Hf isotopic analyses were performed using the HelEx ArF excimer LA system coupled with a Thermo Fin-igan Neptune multiple collector(MC)-ICPMS.The data were acquired from$50l m ablation pits with a laser repe-tition rate of5Hz and$60s ablation times.To correlate Lu–Hf isotopic data with the O isotopic compositions, the laser ablation site for Lu–Hf isotopic analysis was placed within similar internal structure and close to the ori-ginal pit for O isotopic analysis.Helium gas was used for ?ushing the ablation pit.Furthermore,ca.3–4ml/min N2 was mixed into the Ar sample carrier gas to enhance the sig-nal intensity(Iizuka and Hirata,2005).

The Finnigan Neptune MC-ICPMS contains a move-able array of nine Faraday collectors,four on either side of a?xed axial position.Eight Faraday cups were used to monitor isotopes of171Yb,173Yb,174(Hf+Yb),175Lu,176(Hf+Yb+Lu),177Hf,178Hf and179Hf.Because the re-sponse of Faraday ampli?ers is not fast enough to follow

rapid changes in signal intensity,we employed empirical

corrections for the di?erent response times of individual

ampli?ers,based on the linear correlation between

measured isotopic ratios and rate of signal intensity change

(Iizuka et al.,2011).The contribution of the isobaric inter-

ferences by176Yb and176Lu on176Hf was corrected by mea-

suring173Yb and175Lu before the mass bias correction for 176Hf/177Hf.For the interference correction,the true

176Yb/173Yb value of0.78696(Thirlwall and Anczkiewicz, 2004)and176Lu/175Lu value of0.026549(Chu et al., 2002)were employed.All mass discrimination e?ects were corrected using an exponential law.The mass bias factor for Yb was calculated by normalizing the measured 173Yb/171Yb to 1.12346(Thirlwall and Anczkiewicz, 2004),whereas that for Lu was assumed to be identical to that for Hf(Iizuka and Hirata,2005).The exponential mass bias factor for Hf isotopic ratios was calculated by normal-izing the measured179Hf/177Hf to0.7325(Patchett et al., 1981).In addition,to allow accurate comparison with liter-ature values,Hf isotopic ratios corrected for mass bias were further normalized to reference values178Hf/177Hf= 1.467168and176Hf/177Hf=0.282507for the Mud Tank zircon standard(which is normalized to176Hf/177Hf=

0.282160for JMC-Hf475;Woodhead and Hergt,2005),

i.e.,the o?set values were determined from the mean Hf iso-

topic ratios obtained for this standard on any given analyt-

ical session(Electronic Annex Table EA-6).Mud Tank

zircon was chosen for the normalization as it has very

low heavy rare earth elements(HREE)/Hf ratios,thereby

requiring relatively minor correction for isobaric interfer-

ences of176Yb and176Lu on176Hf(176Yb/177Hf of$0.001

and176Lu/177Hf of$0.00005).

We report analytical errors on the initial176Hf/177Hf,

rather than the present-day176Hf/177Hf,for single spot

measurements,because resolvable variations in the pres-

ent-day176Hf/177Hf due to radiogenic in-growth of176Hf

may exist within ancient zircon grains having heteroge-

neous Lu/Hf.The analytical errors combine the internal

run errors(2s.e.)and the reproducibility of the Mud

Tank standard zircon analyses(2s.d.),added in quadra-

ture.The calculation of the initial176Hf/177Hf used the 176Lu decay constant of 1.867?10à11yrà1(So¨derlund et al.,2004).

To evaluate the accuracy and precision of data obtained

by the present technique,we have analyzed the zircon stan-

dards91500,R33and TEMORA-2with the same analytical

conditions as this study.Our data obtained during the

course of this study yielded mean initial176Hf/177Hf of

0.282292±42(2s.d.,n=47)for91500,0.282738±43

(2s.d.,n=45)for R33,and0.282669±41(2s.d.,n=33)

for TEMORA-2(Electronic Annex Table EA-6).These re-

sults are in excellent agreement with the solution-MC-

ICPMS determinations of the initial176Hf/177Hf

(0.282299±6for91500,0.282753±18for R33,and

0.282677±8for TEMORA-2,Woodhead and Hergt,

2005;Blichert-Toft,2008;Vervoort,2010).This indicates

that our analytical protocols are robust under analysis of

various HREE/Hf zircon.

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120111

APPENDIX B.HAFNIUM ISOTOPIC EVOLUTION OF

THE ARC MANTLE The widely used Hf isotopic evolution curve for the de-pleted mantle proposed by Gri?n et al.(2000)would be inappropriate for estimating the timing of continental crust generation,because the evolution curve is based on 176

Hf/177Hf of mid-ocean ridge basalts (MORB)including a high fraction of North Atlantic MORB with unusually high 176Hf/177Hf.In contrast,new continental crust forms principally at subduction zones,not mid-ocean ridges.Hence,the best way to estimate Hf isotopic evolution of the mantle from which continental crust is now being ex-tracted would be to use the mean value for island arc volca-nic rocks (Dhuime et al.,2011a ).We have compiled 176

Hf/177Hf of 382Neogene island arc basalts and their dif-ferentiates from eleven arc provinces (Kermadec,Tonga,Vanuatu,New Britain,Indonesia,Luzon,Mariana,Izu,NE Japan,Aleutian,and South Sandwich arcs:Pearce et al.,1999,2007;Jicha et al.,2004;Mu ¨nker et al.,2004;Marini et al.,2005;Wade et al.,2005;Barry et al.,2006;Hanyu et al.,2006;Stern et al.,2006;Tamura et al.,samples thought to be a?ected by crustal-level contamina-tion (e.g.,Martinique,Antilles;Serua,Banda)or anoma-lous mantle (e.g.,north Tonga)were excluded.Third,all data were normalized to a common value for SRM JMC475,and we used a common,speci?ed CHUR value when calculating e Hf.

Fig.10shows a histogram of e Hf for the 382arc rocks used in our compilation.Note that it is strongly skewed to-wards low values.A cumulative probability plot reveal two populations,both of which approximate to normal distri-butions.The larger of the populations (88%)has a mean of 0.283190(e Hf =14.3)and smaller standard deviation of ±0.000085(2s.d)whereas the smaller population (12%)has a mean of 0.283020(e Hf =8.3)and larger stan-dard deviation of ±0.000270(2s.d.)We interpret the larger population to represent magmas derived from mantle that contains little Hf from subducted oceanic crust and sedi-ment,and the smaller population to represent magma from a mantle source containing signi?cant slab-derived Hf.The uncertainty for the combined asymmetric population for the average 176Hf/177Hf of modern arc mantle is given by the 95%con?dence interval on a cumulative frequency plot.Hf values for 382modern island arc basalts shown in red.The insert shows the there are two distinct populations as indicated by the change in slope.The black shown in blue,and not included in the average,are e Hf for 21juvenile detrital zircons the Virgin Islands.Zircon grains with non-mantle-like O isotopic compositions zircons is 13.8±0.6,13.3±0.4for the Virgin Islands,and 13.6±0.3(2s.e.)for the references to color in this ?gure legend,the reader is referred to the web version 112T.Iizuka et al./Geochimica et Cosmochimica Acta 107(2013)96–120

The Hf isotopic evolution line and error envelop for the arc mantle shown in Figs.6and8were obtained by extrap-olating the mean and95%con?dence limits for modern arc magmas back to the CHUR value at4.56Ga.The arc man-tle evolution is generally consistent with e Hf(t)of juvenile rocks with crystallization ages up to3.4Ga as identi?ed by Vervoort and Blichert-Toft(1999),after recalculating to the176Lu decay constant of1.867?10à11yrà1(So¨derl-und et al.,2004).Note also that the arc mantle evolution is consistent with Lu–Hf isotope systematics of Archean arc lavas(Polat and Mu¨nker,2004)and komatiites(mantle sources with176Lu/177Hf of0.0375;Blichert-Toft and Puch-tel,2010).

APPENDIX C.SUPPLEMENTARY DATA Supplementary data associated with this article can be found,in the online version,at https://www.doczj.com/doc/b413032514.html,/10.1016/ j.gca.2012.12.028.

REFERENCES

Abdelsalam M.G.,Lie′geois J.P.and Stern R.J.(2002)The Saharan Metacraton.J.Afr.Earth Sci.34,119–136. Abouchami W.and Boher M.(1990)A major2.1Ga event of ma?c magmatism in West Africa:an early stage of crustal accretion.

J.Geophys.Res.95,17605–17629.

A?aton P.,Kro¨ner A.and Seddoh K. F.(2000)Pan-African granulite formation in the Kabye Massif of northern Togo (West Africa):Pb–Pb zircon ages.Int.J.Earth Sci.88,778–790.

Ajibade A.C.and Wright J.B.(1989)The Togo-Benin-Nigeria Shield:evidence of crustal aggregation in the Pan-African belt.

Tectonophysics165,125–129.

Albare`de F.(1998)The growth of continental crust.Tectonophysics 296,1–14.

Alle`gre C.J.and Rousseau D.(1984)The growth of the continent through geological time studied by Nd isotope analysis of shales.Earth Planet.Sci.Lett.67,19–34.

Allsopp H.L.,Ko¨stlin E.O.,Welke H.J.,Burger A.J.,Kro¨ner A.

and Blignault H.J.(1979)Rb–Sr and U–Pb geochronology of Late Precambrian–Early Paleozoic igneous activity in the Richtersveld(South Africa)and southern South West Africa.

Trans.Geol.Soc.S.Afr.82,185–204.

Amelin Y.,Lee D.-C.,Halliday A.N.and Pidgeon R.T.(1999) Nature of the Earth’s earliest crust from hafnium isotopes in single detrital zircons.Nature399,252–255.

Appel P.,Schemk V.and Schumann A.(2005)P–T path and metamorphic ages of politic schists at Murchison Falls,NW Uganda:evidence for a Pan-African tectonometamorphic event in the Congo Craton.Eur.J.Mineral.17,655–664. Armstrong R.and Wilson A.H.(2000)A SHRIMP U–Pb study of zircons from the layered sequence of the Great Dyke,Zimba-bwe,and a granitoid anatectic dyke.Earth Planet.Sci.Lett.

180,1–12.

Armstrong R.,Master S.and Robb L.J.(2005)Geochronology of the Nchanga Granite,and constraints on the maximum age of the Katanga Supergroup,Zambian Copperbelt.J.Afr.Earth Sci.42,32–40.

Arndt N.T.and Goldstein S.L.(1987)Use and abuse of crust-formation ages.Geology15,893–895.

Attoh K.,Dallmeyer R.D.and A?aton P.(1997)Chronology of nappe assembly in the Pan-African Dahomeyide orogeny,West

Africa:evidence from40Ar/39Ar mineral ages.Precambrian Res.

82,153–171.

Baertschi P.(1976)Absolute18O content of standard mean ocean water.Earth Planet.Sci.Lett.31,341–344.

Barry T.L.,Pearce J.A.,Leat P.T.,Millar I.L.and Le Roex A.P.

(2006)Hf isotope evidence for selective mobility of high-?eld-strength elements in a subduction setting:South Sandwich Islands.Earth Planet.Sci.Lett.252,223–244.

Batumike J.M.,Gri?n W.L.,O’Reilly S.Y.,Belousova E.A.and Pawlitschek M.(2009)Crustal evolution in the central Congo-Kasai Craton,Luebo,D.R.Congo:insights from zircon U–Pb ages,Hf-isotope and trace element data.Precambrian Res.170,107–115. Becker T.,Garoeb H.,Ledru P.and Milesi J.P.(2005)The Mesoproterozoic event within the Rehoboth Basement Inlier of Namibia:review and new aspects of stratigraphy,geochemistry, structure and plate tectonic setting.South Afr.J.Geol.108,465–492.

Be′dard J.(2006)A catalytic delamination-driven model for coupled genesis of Archaean crust and sub-continental litho-spheric mantle.Geochim.Cosmochim.Acta70,1188–1214. Begg G.C.,Gri?n W.L.,Natapov L.M.,O’Reilly S.Y.,Grand S.

P.,O’Neill C.J.,Hronsky J.M.A.,Poudjom Djomani.Y., Swain C.J.,Deen T.and Bowden P.(2009)The lithospheric architecture of Africa:seismic tomography,mantle petrology, and tectonic evolution.Geosphere5,23–50.

Bekker A.,Holland H.D.,Wang P.L.,Rumble D.,Stein H.J., Hannah J.L.,Coetzee L.L.and Beukes N.J.(2004)Dating the rise of atmospheric oxygen.Nature427,117–120.

Belousova E.A.,Kostitsyn Y.A.,Gri?n W.L.,Begg G.C., O’Reilly S.Y.and Pearson N.J.(2010)The growth of the continental crust:constraints from zircon Hf-isotope data.

Lithos119,457–466.

Bindeman I.N.and Valley J.W.(2001)Low d18O rhyolites from Yellowstone:magmatic evolution based on analysis of zircons and individual phenocrysts.J.Petrol.42,1491–1517.

Black R.,Latouche L.,Lie′geois J.P.,Caby R.and Bertrand J.M.

(1994)Pan-African displaced terranes in the Tuareg shield (central Sahara).Geology22,641–644.

Black L.P.,Kamo S.L.,Allen C.M.,Davis D.W.,Aleiniko?J.

N.,Valley J.W.,Mundil R.,Campbell I.H.,Korsch R.J., Williams I.S.and Foudoulis C.(2004)Improved206Pb/238U microprobe geochronology by the monitoring of a trace-element-related matrix e?ect;SHRIMP,ID-TIMS,ELA-ICP-MS and oxygen isotope documentation for a series of zircon standards.Chem.Geol.205,115–140.

Bleeker W.(2003)The late Archean record:a puzzle in ca.35 pieces.Lithos71,99–134.

Blichert-Toft J.(2008)The Hf isotopic composition of zircon reference material91500.Chem.Geol.253,252–257.

Blichert-Toft J.and Puchtel I.S.(2010)Depleted mantle sources through time:evidence from Lu–Hf and Sm–Nd isotope systematics of Archean komatiites.Earth Planet.Sci.Lett.

297,598–606.

Bodet F.and Scha¨rer U.(2000)Evolution of the SE–Asian continent from U–Pb and Hf isotopes in single grains of zircon and baddeleyite from large rivers.Geochim.Cosmochim.Acta 64,2067–2091.

Boher M.,Abouchami W.,Michard A.,Albare`de F.and Arndt N.

T.(1992)Crustal growth in West Africa at2.1Ga.J.Geophys.

Res.97,345–369.

Bolhar R.,Weaver S. D.,Whitehouse M.J.,Palin J.M., Woodhead J.D.and Cole J.W.(2008)Sources and evolution of arc magmas inferred from coupled O and Hf isotope systematics of plutonic zircons from the Cretaceous Separation Point Suite(New Zealand).Earth Planet.Sci.Lett.268,312–324.

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120113

Borg G.and Krogh T.(1999)Isotopic age data of single zircons from the Archaean Sukumaland Greenstone belt,Tanzania.J.

Afr.Earth Sci.29,301–312.

Borg G.and Shackleton R.M.(1997)The Tanzania and NE-Zaire Cratons.In Greenstone Belts(eds.M.de Wit and L. D.

Ashwal).Clarendon Press,London,pp.608–619.

Bouvier A.,Vervoort J.D.and Patchett P.J.(2008)The Lu–Hf and Sm–Nd isotopic composition of CHUR:constraints from unequilibrated chondrites and implications for the bulk com-position of terrestrial planets.Earth Planet.Sci.Lett.273,48–

57.

Bradley D.C.(2011)Secular trends in the geologic record and the supercontinent cycle.Earth-Sci.Rev.108,16–33.

Brandl G.and de Wit M.J.(1997)The Kaapvaal Craton.In Greenstone Belts(eds.M.de Wit and L.D.Ashwal).Clarendon Press,London,pp.581–607.

Brown M.(2007)Metamorphism,plate tectonics,and the super-continent cycle.Earth Sci.Frnot.14,1–18.

Bruguier O.,Dada S.and Lancelot J.R.(1994)Early Archaean component(>3.5Ga)within a 3.05Ga orthogneiss from northern Nigeria:U–Pb zircon evidence.Earth Planet.Sci.

Lett.125,89–103.

Caby R.(2003)Terrane assembly and geodynamic evolution of central-western Hoggar:a synthesis.J.Afr.Earth Sci.37,133–159.

Cahen L.,Snelling N.J.,Delhail J.and Vail J.R.(1984)The Geochronology and Evolution of Africa.Clarendon Press, London.

Campbell I.H.and Allen C.M.(2008)Formation of supercon-tinents linked to increases in atmospheric oxygen.Nat.Geosci.

1,554–558.

Campbell I.H.,Reiners P.W.,Allen C.M.,Nicolescu S.and Upadhyay R.(2005)He–Pb double dating of detrital zircons from the Ganges and Indus Rivers:implication for quantifying sediment recycling and provenance studies.Earth Planet.Sci.

Lett.237,402–432.

Can?eld D.E.(2005)The early history of atmospheric oxygen: homage to Robert M.Garrels.Annu.Rev.Earth Planet.Sci.33, 1–36.

Carlson R.W.and Pearson D.G.(2005)Physical,chemical,and geochronological characteristics of continental mantle.Rev.

Geophys.43,2004RG000156.

Carter A.and Moss S.J.(1999)Combined detrital-zircon?ssion-track and U–Pb dating:a new approach to understanding hinterland evolution.Geology27,235–238.

Cavosie A.J.,Valley J.W.and Wilde S.A.(2005)Magmatic d18O in4400–3900Ma detrital zircons:a record of the alteration and recycling of crust in the Early Archean.Earth Planet.Sci.Lett.

235,663–681.

Cawood P. A.and Buchan C.(2007)Linking accretionary orogenesis with supercontinent assembly.Earth-Sci.Rev.82, 217–256.

Cawood P.A.,Nemchin A.A.,Freeman M.and Sircombe K.

(2003)Linking source and sedimentary basin:detrital zircon record of sediment?ux along a modern river system and implications for provenance studies.Earth Planet.Sci.Lett.

210,259–268.

Chauvel C.,Lewin E.,Carpentier M.,Arndt N.T.and Marini J.-C.

(2008)Role of recycled oceanic basalt and sediment in generating the Hf–Nd mantle array.Nat.Geosci.1,64–67. Chen D.,Deloule E.,Cheng H.,Xia Q.and Wu Y.(2003) Preliminary study of micro-scale zircon oxygen isotopes for Dabie-Sulu metamorphic rocks:ion probe in situ analyses.

Chin.Sci.Bull.48,1670–1678.

Chen Y.X.,Zheng Y.F.,Chen R.X.,Zhang S.B.,Li Q.,Dai M.

and Chen L.(2011)Metamorphic growth and recrystallization

of zircons in extremely18O-depleted rocks during eclogite-facies metamorphism:evidence from U–Pb ages,trace elements,and O–Hf isotopes.Geochim.Cosmochim.Acta75,4877–4898. Christensen N.I.and Mooney W. D.(1995)Seismic velocity structure and composition of the continental crust:a global review.J.Geophys.Res.100,9761–9788.

Chu N.C.,Taylor R.N.,Chavagnac V.,Nesbitt R.W.,Boella M.,Milton J.A.,German C.R.,Bayon G.and Burton K.

(2002)Hf isotope ratio analysis using multi-collector induc-tively coupled plasma mass spectrometry:an evaluation of isobaric interference corrections.J.Anal.Atom.Spectrom.17, 1567–1574.

Clauer N.,Caby R.,Jeannette D.and Trompette R.(1982) Geochronology of sedimentary and metasedimentary Precam-brian rocks of the West African craton.Precambrian Res.18, 53–71.

Clemens J.D.and Vielzeuf D.(1987)Constraints on melting and magma production in the crust.Earth Planet.Sci.Lett.86,287–306.

Cloutier J.,Stevenson R.K.and Bardoux B.(2005)Nd isotopic, petrologic and geochemical investigation of the Tulawaka East gold deposit,Tanzanian Craton.Precambrian Res.139, 147–163.

Collins A.S.and Pisarevsky S.A.(2005)Amalgamating eastern Gondwana:the evolution of the Circum-Indian Orogens.

Earth-Sci.Rev.71,229–270.

Collins A.S.,Reddy S.M.,Buchan C.and Mruma A.(2004) Temporal constraints on Palaeoproterozoic eclogite formation and exhumation(Usagaran Orogen Tanzania).Earth Planet.

Sci.Lett.224,175–192.

Collins W.J.,Belousova E.A.,Kemp A.I.S.and Murphy J.B.

(2011)Two contrasting Phanerozoic orogenic systems revealed by hafunium isotope data.Nat.Geosci.4,333–337. Compston W.,Williams I.S.and Meyer C.(1984)U–Pb geochronology of zircons from lunar breccia73217using a high mass resolution ion microprobe.J.Geophys.Res.89, B525–B534.

Condie K.C.(1998)Episodic continental growth and supercon-tinents:a mantle avalanche connection?Earth Planet.Sci.Lett.

163,97–108.

Condie K.C.,Bickford M.E.,Aster R.C.,Belousova E.and Scholl D.W.(2011)Episodic zircon ages,Hf isotopic compo-sition,and the preservation rate of continental crust.Geol.Soc.

Am.Bull.123,951–957.

Cornell D.H.,Armstrong R. A.and Walraven F.(1998) Geochronology of the Proterozoic Hartley Basalt Formation, South Africa:constraints on the Kheis tectogenesis and the Kaapvaal Craton’s earliest Wilson cycle.J.Afr.Earth Sci.26, 5–27.

Cumming G.L.and Richards J.R.(1975)Ore lead isotope ratios in a continuously changing earth.Earth Planet.Sci.Lett.28, 155–171.

Dada S.S.(1998)Crust-forming ages and proterozoic crustal evolultion in Nigeria:a reappraisal of current interpretations.

Precambrian Res.87,65–74.

De Carvalho H.,Tassinari C.,Alves P.H.,Guimaraes F.and Simoes M.C.(2000)Geochronological review of the Precam-brian western Angola:links with Brazil.J.Afr.Earth Sci.31, 383–402.

De Waele B.and Fitzsimons I.C.W.(2007)The nature and timing of Palaeoproterozoic sedimentation at the southeastern margin of the Congo Craton;zircon U–Pb geochronology of plutonic, volcanic and clastic units in northern Zambia.Precambrian Res.

159,95–116.

De Waele B.,Johnson S.P.and Pisarevsky S. A.(2008) Paleoproterozoic to Neoproterozoic growth and evolution of

114T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120

the eastern Congo Craton:its role in the Rodinia puzzle.

Precambrian Res.160,127–141.

de Wit M.J.,Roering C.,Hart R.J.,Armstrong R.A.,de Ronde

C.E.,Green R.W.E.,Tredoux M.,Peberdy E.and Hart R.A.

(1992)Formation of an Archaean continent.Nature,553–562. Deniel C.,Vidal P.,Fernandez A.,Le Fort A.and Peucat J.-J.

(1987)Isotopic study of the Manasulu granite(Himalaya Nepal):inferences on the age and source of Himalayan leucogranites.Contrib.Mineral.Petrol.96,78–92.

Dhuime B.,Hawkesworth C.and Cawood P.(2011a)When continents formed.Science331,154–155.

Dhuime B.,Hawkesworth C.J.,Storey C. D.and Cawood P.(2011b)From sediments to their source rocks:Hf and Nd isotopes in recent river sediments.Geology39, 407–410.

Dickinson W.R.(2008)Impact of di?erential zircon fertility of granitoid basement rocks in North America on age populations of detrital zircons and implications for granite petrogenesis.

Earth Planet.Sci.Lett.275,80–92.

Dodson M.H.,Compston W.,Williams I.S.and Wilson J.F.

(1988)A search for ancient detrital zircons in Zimbabwean sediments.J.Geol.Soc.Lond.145,977–983.

Drummond M.S.and Defant M.J.(1990)A model for trondhjemite-tonalite-dacite genesis and crustal growth via slab melting–Archean to modern comparisons.J.Geophys.Res.95, 21503–21521.

Eggins S.M.,Kinsley L.P.J.and Shelley J.M.G.(1998) Deposition and elemental fractionation processes during atmo-spheric pressure laser sampling for analysis by ICP-MS.Appl.

Surf.Sci.127–129,278–286.

Eglinton B.M.(2006)Evolution of the Namaqua-Natal Belt, southern Africa–a geochronological and isotope geochemical review.J.Afr.Earth Sci.46,93–111.

Eiler J.M.(2001)Oxygen isotope variations of basaltic lavas and upper mantle rocks.Rev.Mineral.Geochem.43,319–364. Ekwueme B.N.and Kro¨ner A.(2006)Single zircon ages of migmatitic gneisses and granulites in the Obudu Plateau:timing of granulite-facies metamorphism in southeastern Nigeria.J.

Afr.Earth Sci.44,459–469.

Ferre′ E.,Gleizes G.and Caby R.(2002)Obliquely convergent tectonics and granite emplacement in the Trans-Saharan belt of Eastern Nigeria:a synthesis.Precambrian Res.114,199–219.

Foley S.,Tiepolo M.and Vannucci R.(2002)Growth of early continental crust controlled by melting of amphibolite in subduction zones.Nature417,837–840.

France-Lanord C.and Le Fort P.(1988)Crustal melting and granite genesis during the Himalayan collision orogenesis.

Trans.Roy.Soc.Edinburg:Earth Sci.79,183–195.

Frimmel H. E.and Frank W.(1998)Neoproterozoic tectono-thermal evolution of the Gariep Belt and its basement,Namibia and South Africa.Precambrian Res.90,1–28.

Frimmel H. E.,Zartman R. E.and Spa¨th A.(2001)The richtersveld igneous complex,South Africa:U–Pb zircon and geochemical evidence for the beginning of Neoproterozoic continental breakup.J.Geol.109,493–508.

Gabert G.(1990)Lithostratigraphic and tectonic setting of gold mineralization in the Archean cratons of Tanzania and Uganda,East Africa.Precambrian Res.46,59–69.

Gao S.,Rudnick R.L.,Yuan H.L.,Liu X.M.,Liu Y.S.,Xu W.

L.,Ling W.L.,Ayers J.,Wang X.C.and Wang Q.H.(2004) Recycling lower continental crust in the North China craton.

Nature432,892–897.

Gilliam C.E.and Valley J.W.(1997)Low d18O magma,Isle of Skye,Scotland:evidence from zircons.Geochim.Cosmochim.

Acta61,4975–4981.Goldstein S.L.,O’Nions R.K.and Hamilton P.J.(1984)A Sm–Nd isotopic study of atmospheric dusts and particulates from major river systems.Earth Planet.Sci.Lett.70,221–236. Goldstein S.L.,Arndt N.T.and Stallard R.F.(1997)The history of a continent from U–Pb ages of zircons from Orinoco River sand and Sm–Nd isotopes in Orinoco Basin River.Chem.Geol.

139,271–286.

Grantham G.H.,Maboko M.and Eglington B.M.(2003)A review of the evolution of the Mozambique Belt and implica-tions for the amalgamation and dispersal of Rodinia and Gondwana.In Proterozoic East Gondwana:Supercontinent Assembly and Breakup,vol.206(eds.M.Yoshida, B. F.

Windley and S.Dasgupta).Geol.Soc.London Spec.Publ.,pp.

401–425.

Gri?n W.L.,Pearson N.J.,Belousova E.,Jackson S.E.,van Achterbergh E.and O’Reilly S.(2000)The Hf isotope compo-sition of cratonic mantle:LAM-MC-ICPMS analysis of zircon megacrysts in kimberlites.Geochim.Cosmochim.Acta64,133–147.

Guitreau M.,Blichert-Toft J.,Martin H.,Mojzsis S.J.and Albare`de F.(2012)Hafnium isotope evidence from Archean granitic rocks for deep-mantle origin of continental crust.Earth Planet.Sci.Lett.337–338,211–223.

Ha¨lbich I.W.and Alchin D.J.(1995)The Gariep belt: stratigraphic-structural evidence for obliquely transformed grabens and back-folded thrust stacks in a combined thick-skin thin-skin structural setting.J.Afr.Earth Sci.21,9–33. Hanson R. E.(2003)Proterozoic geochronology and tectonic evolution of southern Africa.In Proterozoic East Gondwana: Supercontinent Assembly and Breakup,vol.206(eds.M.

Yoshida,B.F.Windley and S.Dasgupta).Geol.Soc.London Spec.Publ.,pp.427–463.

Hanyu T.,Tatsumi Y.,Nakai S.,Chang Q.,Miyazaki T.,Sato K., Tani K.,Shibata T.and Yoshida T.(2006)Contribution of slab melting and slab dehydration to magmatism in the NE Japan arc for the last25Myr:constraints from geochemistry.

Geochem.Geophys.Geosyst.7,Q08002.https://www.doczj.com/doc/b413032514.html,/

10.1029/2006GC001220.

Harrison T.M.,Blichert-Toft J.,Muller W.,Albare`de F.,Holden P.and Mojzsis S.J.(2005)Heterogeneous Hadean hafnium: evidence of continental crust at4.4to4.5Ga.Science310, 1947–1950.

Hawkesworth C.J.and Kemp A.I.S.(2006)Evolution of the continental crust.Nature443,811–817.

Hawkesworth C.J.,Dhuime B.,Pitranik A.B.,Cawood P.A.and Storey C. D.(2010)The generation and evolution of the continental crust.J.Geol.Soc.Lond.167,229–248.

Hessler A.M.and Lowe D.R.(2006)Weathering and sediment generation in the Archean:an integrated study of evolution of siliciclastic sedimentary rocks of the3.2Ga Moodies Group, Barberton Greenstone Belt,South Africa.Precambrian Res.

151,185–210.

Hiess J.,Bennett V.C.,Nutman A.P.and Williams I.S.(2009)In situ U–Pb,O and Hf isotopic compositions of zircon and olivine from Eoarchaean rocks,West Greenland:new insights into making old crust.Geochim.Cosmochim.Acta73,4489–4516.

Holland H.D.(2002)The oxygenation of the atmosphere and oceans.Phil.Trans.Roy.Soc.B361,903–915.

Holm-Denoma C.S.and Das R.(2010)Bimodal volcanism as evidence for Paleozoic extensional accretionary tectonism in the southern Appalachians.Geol.Soc.Am.Bull.122,1220–1234. Horstwood M.S.A.,Nesbitt R.W.,Noble S.R.and Wilson J.F.

(1999)U–Pb zircon evidence for an extensive early Archean craton in Zimbabwe:a reassessment of the timing of craton formation,stabilization,and growth.Geology27,707–710.

T.Iizuka et al./Geochimica et Cosmochimica Acta107(2013)96–120115

2019年7月时政热点试题及答案

2019年7月时政热点试题及答案 1. (单选题)7月1日从应急管理部消防救援局获悉,6月28日,安徽省蚌埠市消防员在营救遇险工人时身受重伤,后抢救无效不幸牺牲。7月1日,应急管理部批准同志为烈士。 A. 丁振军 B. 孙雷宇 C. 古剑辉 D. 周鹏 参考答案: B 2. (单选题)中宣部7月1日向全社会宣传发布黄文秀的先进事迹,追授她“”称号。 A. 时代楷模 B. 全国三八红旗手 C. 全国脱贫攻坚模范 D. 优秀共产党员

参考答案: A 3. (单选题)国家市场监督管理总局发布的《养老机构等级划分与评定》国家标准7月1日正式实施,这是我国发布养老机构星级评定的“”。 A. 首次、标准 B. 首次、文件 C. 首次、国标 D. 首次、等标 参考答案: C 4. (单选题)近期,中国科学院合肥智能机械研究所设计制备出一种高效的比色荧光纳米探针,用其制成荧光试纸,根据试纸颜色变化可快速识别“”。 A. 水中铅污染 B. 水中氢污染 C. 水中核污染

D. 水中硼污染 参考答案: A 5. (单选题)2019世界新能源汽车大会7月2日在召开。习致贺信,对大会的召开表示热烈祝贺。 A. 海南博鳌 B. 浙江杭州 C. 广东深圳 D. 河北石家庄 参考答案: A 6. (单选题)中央气象台7月2日发布今年,预计南海热带低压强度将有所加强,可能发展为今年第4号台风,并将于3日凌晨到上午在海南文昌到陵水一带沿海登陆。 A. 首个台风红色预警 B. 首个台风黄色预警 C. 首个台风橙色预警

D. 首个台风蓝色预警 参考答案: D 7. (单选题)7月3日,“2019互联网应用技术·万宝峰会”在娄底开幕,同时,娄底以高铁站为核心区,创建湖南省首个集聚互联网产业的特色小镇——“”。 A. 娄底南·特色小镇 B. 娄底站·联创小镇 C. 娄底南·联创小镇 D. 娄底站·网络小镇 参考答案: C 8. (单选题)从海南省气象台获悉,今年第4号台风“”(热带风暴级)已于3日零时45分前后在海南省万宁市和乐镇沿海登陆,登陆时中心附近最大风力8级(18米/秒),成为今年登陆我国的首个台风。 A. 木恩 B. 纳尔吉斯

相关主题
文本预览
相关文档 最新文档