Did the Breakout of Laurentia Turn Gondwanaland Inside-Out? Pau F. HOFFMAN Comparative geology suggests that the ‘continents adjacent to northern, western, southern, and eastern Laurentia in the Late Proterozoic were Siberia, Austra Antarctica, southern Africa, and Amazonia-Baltica, respectively. Late Proterozoic fragmentation of the supercontinent centered on Laurentia would then have been followed by rapid fan-like collapse of the (present) southern continents and eventual consolidation of East and West Gondwanaland. In this scenario, a pole of rotation near ‘the Weddell Sea would explain the observed dominance of wrench tectonics in (present) east-west trending Pan-African mobile belts and subduction-accretion tec- tonics in north-south trending belts. In the process of fragmentation, rifts originating. in the interior of the Late Proterozoic supercontinent became the external margins of Paleozoic Gondwanaland; exterior margins of the Late Proterozoic supercontinent became landlocked within the interior of Gondwanaland. HIE INTERVAL ENCOMPASSING THE | Vendian to Cambrian transition (0.7 © 05 billion years ago (Ga)] is perhaps the most enduring enigma in his- torical geology. The advent and subsequent evolutionary explosion of metazoa were ac- companied by extreme fluctuations in sea level, global climate, and the isotopic and chemical compositions of seawater (1-3) ‘That these changes are related to the frag mentation of a Tate Proterozoic supercanti- rent (4) isan idea that stemmed from the observation that rifting and continental breakup occurred contemporancously around the margins of Laurentia (North America and Greenland), Baltica (Batic shield and Russian platform), Siberia, and pars of Gondwanaland (5-7). However, the relation of the hypothetical supercontinent to the continents forming Gondwanaland (South America, Aftica, Arabia, India, Ant arctica, and Australia) is unresolved because reliable paleomagnetic data for the interval are sparse. In most reconstructions of the Late Proterozoie supercontinent, Gond= ‘wanaland is weated as a coherent entity (1, 6, 8-11), despite evidence that its consolida- tion was broadly contemporaneous with fragmentation of the northern continents (12, 13). Not surprisingly, some question the realty of a supercontinent chat may have ‘acre Geaience Dikion, Geog! Saray of aoa Orso, Ones, KIA GER 7IUNE 1991 dlsintegrated before it had formed. In the absence of sea-loor magnetic anomalies, re-Mesozoie continental reconstructions are necessarily speculative. However, corre= lation of Precambrian orogenic belts that ‘were once continuous but are now truncated at modem or ancient continental margins provides 2 means of establishing former linkages between separated continents. On the basis of such evidence, T presenta qual- itative but testable model for the breakup of a Late Proterozoic supercontinent centered. on Laurentia and the subsequent assembly of Paleozoic Gondwanaland. Northeastern Siberia (all directions refer tw present-day coordinates) has been pro- posed a the conjugate margin to southwest fm Laurentia in the Late Proterozoic on geological (14) and paleomagnetic (15) ‘grounds. The evidence is permissive but not conclusive. U-Pb zircon geochzonology in- dlicates that the Anabar and western Aldan, shields (Eig. 1A) of Siberia consist of crust that formed before 3.0 Ga and underwent high-grade metamorphism and anatexis at 2.8 to 2.6 Ga and again at 2.0 to 1.9 Ga (16). In Laurentia, only the Thelon-Talsson ‘magmatic zone (Fig. 1A) in northwest Can- ada has such a history (17), but this zone is truncated beneath the plains of east-central Alberta and does not appear to reach the Cordillera (18). However, the zone does cexend northward where i is truncated by the Paleozoie Franklin orogen in the Arctic (17) Thas, if the Anabar and Thelon Tale son beits were originally connected, Siberia was most likely conjugate to the Arctic (19) ‘margin of Laurentia (Fig. 1A). Jefferson (20) proposed that westera Law renta was flanked inthe Late Proterozoic by the eastern margin of the Australian craton together with the margin of the East Ant arctic craton corresponding tothe Transant- arctic Mountains (Fig. 1A). This proposal stemmed from comparative Late Protero- oie stratigraphy of the respective Australian and Laurentian margins (20). Recently, ‘Moores (21) and Dalziel (22) have bolstered the restoration and fixed its position by pointing our ehat a tectonic boundary equiv= alent tothe Grenville orogenic front of west ‘Tomas reappears in Antarctica near the ease ‘const ofthe Weddell Sea (Fig, 1A). Moores (21) suggested that the Grenvillian (1.3 to 110 Ga) belt loops back around the East Antarctic craton (by way of the Easter Ghats of India) into the Albany-Fraser belt (Fig, 1A) of southwest Australia (23). The reconstruction implies that the northwest comer of Laurenta lay adjacent t0 north central Australia in the Late Proterozoic. ‘There are indeed remarkable parallels in Early Proterozoic geology and U-Pb ‘geochronology between the Wopmay oro- zen in northwest Canada and the Mount Isa, Pine Creck, Tennant Creek, and Halls Creek inlicrs of northern Australia (Fig. 1A). The main orogenic eveat in each area ‘occurred fiom 1.90 to 1.88 Ga—the Calde- tian orogeny in Canada (24) and the Barra- ‘mundi orogeny in Australia (25). Each area ‘experienced intense, mainly felsic volcanism and platonism from 1.88 to 1.84 Ga (26, 27), and the felsic magmas were derived from crustal sources having Nd model ages of 2.4 t0 2.0 Ga (28, 29). The comparison ‘not only supports the proposed reconstruc- tion (20-22) but also implies that northwest TLaurentia, east Antarctia, and the Austra- lian eraton were fellow travelers from 1.9 until 0.6 Ga (30). If East Gondwanaland (India- Australi Anarctica) separated from western Lanren- tia then which continent oF continents lay adjacent ro eastern and southera Laurentia before the opening of the Iapetus Ocean? ier works have advocated northwest AF- rica (6), western South America (6), Arabia (7, 15), and Bakica (31), Pre-Grenvilian tectonic zones provide piercing points that pin Baltica adjacent to northwest Britain GN), which was then a part of Laurentia of southeast Greenland (Fig. 1A). To the south, the Appalachian and Ouachita mar- ‘gins of Laurentia are confined t0 the Gren- ville orogen, Therefore, cratons that are bordered by Grenvllian bets are more like'y 1m have been conjugate ro southeastern and. RevoRts 1409)| \ Fig. 1. (A) Reconstruction of the proposed Tate Proteramse super a continent. Restoration of Baie s fiom Gower (31), Amazonia fom (6), aod Austal: Arercica om i) and 22) The lone feof te Congo, Kalahas, Wet Alcs, and ‘Amazouia cators ree changes isthe size and shape a3 venof {Sonic shorening during the be eens of Cond intl (PateAicnn orogeny) (8) Late Cantvian plogeogrpiy af te the breshou of Parent Gi), Abandonment of Bact (5), and ‘=ouiag amalgamation of Com swanaland (56) Dated os are Sul ils around «pole of rom on inthe Wesel Se southem Laurent than those lke the West ‘Aftican craton that lack Grenvilian belts, Grenvillian bekss marginal to cratons in Gondwanaland include the Rondonia-Sua- sas belt (32) of the Amazonis craton, the Itumide and Kibaride belts (13) of the Con: {go craton, and the Namaqua-Natal belt (33) of the Kalahari craton (Fig. 1A). The abser- vation that Grenvillian rocks in both the ‘Namaqua-Natal belt and southern Laurentia have Nd model ages that cluster around 1.4 Ga (4, 35) lends further eredence t0 the proposed reconstruction (Fig. 1A) in which they are in close proximity. Other than Baltica, none of the Eurasian cratons is known to have a marginal Grenvllian bet, Late Proterozoic juxtaposition of Arabia and Laucentia is unlikely: Arabia is com posed of Late Proteraznic orogenic belts containing much juvenile crust (36). The ‘only equivalent belts in Laurentia (the Ava- Jonian terranes of the Appalachians) did not arrive there until 0.4 to 0.3 Ga (37). There fone, the cratons most likely to have been conjugate to eastern and southern Laurentia in the Late Proterozoic ate Baltica, Amazo nia, Congo, and Kalahari (Fig. 1A). ‘The separation of Austaliz-Antarctica and Amazonia-Baltica ftom western and ‘eastern Laurentia, respectively, implies that 1819 Gondwanaland was assembled through a fan-lke collapse of its constituent cratons Collapse could have been achieved by ‘counter-clockwise rotation of Antarctics: Australia and dextral translation (with or ‘without clockwise rotation) of Amazonia, relative to Laurentia (Batic is left behind) (Fig. 1B), Available paleomagnetic data are ‘compatible with the implied rotation (10, 11, 22) and with large-scale convergence between Hast and West Gondwanaland, Possibly lasting until che Late Cambrian (12), Rotation about a pote near the Wed dell Sea would account for the dominant ‘wanscurrent motion observed in east 10 east- northeast trending belts of Pan-African (0.8 to 05 Ga) age. Such belts would have approximated small circles of rotation. In the Borborema-Cameroun belt (Fig. 1B), there is geological (38, 39) and paleomag: netic (40) evidence for dextral shear. For the Damara-Zambezi belt (Fig. 1B), there is _geological evidence for dominanely trans rent shear (41) and paleomagneric evidence for a 35° clockwise rotation of the Keapvaal craton with respect to che Congo cratoa (33, 42). For orthogonal Pan-Affiean belts, lke the Mozambique belt (Fig. 1B) that welds East and West Gondwanaland, a pole of rotation near the Weddell Sea would ac: ‘count for the observed dominance of thst: ing (38, 41, 43), the northward inevease in the width of crust produced by subduction: accretion tectonics (36, 44), and the south ‘ward increase in metamorphic grade and basement reactivation (44, 45). The south- SCHENCE, VOL, 252crn limit of the Mozambique belt is uncee tain, Some workers have envisaged that the [Late Proterozoic collision zone extends be- tween the Kalahari and East Antarctic cra tons (#4). Others have. interpreed. the southernmost Mozambique bel as essential ly Grenvilian (Lurian) in age (46): this relation implies that the Kalahari ceatoa belongs to Fast Gondwanaland and thatthe collion zone between East and West Gonwanaland is transformed from the Meocambique belt tothe Gariep bel by way of the Damara-Zamberi belt (Fig. IB). Ace cordingly, the Damaca-Zambe bet should hhaveexpevienced net dexral (33) rather than sinistral (41) tanscurens shear “The proposed reasembly armed the con- ssituent eratons of Gondovanaland inside. cout: the external Paleozoic margins of Gondwanaland (he Tasman, Transantare- tic, Cape, and Andean margins of Fig. 1B) originated 2s rift zones inthe interior of che Late Proterozoic supetcontinent. Converse- ly, the external margins ofthe Late Protero oie superconsinent, some of which had been consuming plate boundaries since be- fore 0.7 Ga, became landlocked within the interior of Gondvwanaland. Such margins border the Mozambigqe (44), Arabia-Nubia (36, 47), Benin-Worss (38, 48), Gariep-Fe ano-beia (49), and Paraguay-Aragual- Rockelide (8, 50, 51) mobile belts (Fig 1B) Oibvionsy, i s impossible for Aust Antarctica fo have been connected eo west em Laurentia and Amazonia wo have been Simultaneously connected 10 eastern Laue rentia if Gondwanaland was alecady acoher- cencentity: Thus, he final assembly of Gond- wanaland mast poste the breakour of Laurentia for the scenario proposed heze to tbe viable, This requirement constiates. a clear test ofthe model Rifting had begun in westem Laorentia by 0.78 Ga (17) and in eastern Lauentia by 0.60 Ga (52). A mii rum age for continental breakup of 0.62 £0 (0.56 Ga is inferred from the onsee of long: lived thermal subsidence along both margins (6) and swatigraphie evidence suggests an Enrly Cambrian nifero-drit transition (3) However, the Cordillera and Appalachians contain ao equivalents of the Late Protero- oie Beardmore (54) and Pampean (50) frogenics that affected the Transantarctic and Andean margins, respectively. Ths ob- servation implies, iF these orogens are cor rectly dated, that Fase and West Gondwana- land had separated from Laurentia before the end of the Proterozoic. The age of consolidation of East and West Gondwana Jundis also uncertain. Subdueron of oceanic lehosphere in the Arabian-Nubian. shield ended at 0.64 10 0.62 Ga (36, 47), To the southeast, however, a terminal collision of FONE 1991 7 latest Cambrian age (0.51 10 0.50 Ga) has been postulated between an active margin in northeast Somalia and Ease Gondwanaland (40), Ie may be significant thatthe searater Syfsr curve reaches a peak of 0.7096 ia the Middle Cambrian that isthe highest ‘alu in the lat blion years (3), Evidently, the Middle Cambrian was atime of extcine- Iy rapid erosion of old radiogenic crust. The crosion must be a consequence of tizonic vplif, as ie cannot be accounted for by lication (0 continental gacations of Cambrian age) or exstary (global sea level rose daring Camb tin). Tn the absence ‘of major conemporancous orogeny outside fof Gonduanaland, the seawater ©'Sqf°St carve suggests that mountain builing2ss0- Gated with the collision of East and West Gondwanaland exkminared in the Middle Cambrian. Thus, the proposed scenario may or may snot be viable, depending on which set of age cetimites is confirmed. AS some events it the erie interval are dated isotopically and others biosrargraphicaly, accurate chrono: tetrc callration of fossil zones is ll in portant. Recent evidence thatthe bas of the ‘Cambrian might bein the age range 0.54 «0 0.52 Ga (55), in contrast to-carlier estimates .f 0.59 10 0.37 Ga, highlights this problem Regardless, it scans safe to conclude that if iy scenario is valid, fragmentation of the Lite Proteogoie supercontinent and the ensuing amalgamation of Gondwanaland must bave occurred rapidly. The contempo- rancous crises in the biotic, oceanic, and limatie reams may have occurred in re- Sponse to such an abnupe tectonic eorgani- zation "REFERENCES AND NOTES 1M. A, 8. McMenamin tod DLS, MeMecamin, ‘he Emergence of intl The’ Carbon Bk vag (Cant Une. Pr Now ors 1990), 2S Caoway Mow, ms Se 78, 156 (1987) AJ Kaulimn tad ATL Kool, Ge See Am hee Drage 21,24 (198). 3, TH Donne JH. Shergold, PN. Southat, C V tains, a Paphorc Boer nd Derlpnen, AGI. Nethat sad L Tavs, Ba (Spe atl 32 Geological Sacey. Bake, One, 190), fy. ory 4.1. Vaenine ad EM. Moores Nui 28,657 (970). AG. Fiber, in Cares ond Earth “Rao, W. A Benen and 8 Van Comer, Eat (Pence Unie. Pros, Procten NJ 983), Te, Wore, Dane J Lay, Mar Gel 5, 973 (1984) Wordey, JB Moody, Rc. Neves n Te Carb Cyd wel rama Cy Neel Pars ea re eT. Sondgut and 8. Bros, ee {toma 32, Amaicen Geophysical Union, Wash Ingion DC. 1985) pp. 861572 5. Buf Sits, Gey 8,27 (1976) AV hi, fo. 38, 1231 (1990) 6.G.C.Bood PLA Nekzton M.A. Kemi Bah Plenty StL. 70,325 (1988 7. L:oscsiin, NT. Kuan M, V. Keno, ‘i, 74,108 (1988), 4, PeMori and E. Irving, J. Cl. 86,535 (1978. 9. A.G. Smit, A.M Husky, Ben, Panerai 2. 1 18 w, a, 2 2. ¥. 2 2. 20. a Peseta Walt Maye (Cambige Uae eso Carbide, 195). ReVingerVoRC. Metibe,C.R Scot in Pte ewnstcin fs Paes Palmas Te Van (ie Voo, CR. Sense, N. Bonhusnmet, Bale {Cindjmin Sor 2, nian Gephyse Unsa, ‘Wahiogron, BC, 1884), pp. 131-138, 1b Rese paper posed the 28 Kier ‘anal Geologie Conse, Washingt, DC, 22 Iu 198 hare 2 194 (1989), ‘MLO. Mein in Prnalan Pie Tero, A. yon El (Beer, Asean, 1981), 612 om E-Gaten and N. J. Sing, The Geohonsagy and Botton fr (Chae, Oxo, 1983) FW Seurand 1A. Des, Clg 6 267 (1978), [B.A Piper, Bah Plas Se Le 39,61 (1982) E,W ito, TV. Gexchera, A. Makan A N. icy OM. Koen, Tore Cage 6. 183 fast 1986), N-1, Dobe, V. Le Doak VT iJ, Souect tan Earth 4,289 (1990), B-. Hofinaa, in Te Gly of Nob Ame Az Orem, AW. Bll and A Panes, Ei (Geslogeal Sodiay of Americ, Boule” CO, 198), pp. 447-813. The Telon"Tasion ek ies sweety lave tc (2.00 © 197 Ga) and ‘Strabo clsionsl (296 ro 191 Gs) phe ‘dal ek Scope on Ashen coe GEM Rows Rh Past ME Valois, S.A, Bowring, CJ Ea Sdn pres Sera ay bate tly speed fom Laren 50127 Gai he Macken die wae and op Freon and teak in norte Canad sg ctiaeatal epraton (AN Lecheaiaat apd U IMC eane ork Per St Lan 96, 38 (1989), EW. Jueton, Geol Soe mts Pg 10 139 (107%) see abo GH. Bubacher Delerzeae Pubeslna Peles. 81, 331 1988), RT, Tel ©. W. lebanon, is Pains, Pe Rie Congr 47 (Asaren Ins of Mining sd tab evil, Vos, As, 1983), E.M, Moores, Gelgy 19,425 (199), EW.D. Dalal sy. 558 MLR. Kaw, J. Gs 9, 66 (1989; B.S. Grew fd Wl Moo, Peni io. 38° 123 (1986), ‘ers CP. Del LB Hats 40/41, 1 (10a); BT. Pld, hi. 47,187 (199. PB Holman sol S.A. Bowing, Celgy 12, 68 i}. ROW! nig, Bue Res 4041, 1 88) tdi 8 Wks, dh, p21 WS Haulewng, P. Hotinan, S.A. Bower, J Vale. Gato. Res, 83,99 (19) SA Bowring spd EA Path, are Pawe Si Tat. 94,217 (198) MT. MCs, 2 Prireci Like Bos tim, A Ker Ba. (Goat. Ser 1, Amesian {GSiphyssl Union, Wetingron, BC, 198%, pp nett ‘The pit comangaincty of dhe srt aed ese ‘scan cores Brow feed couvergee be ‘teen hem crng the Perma Ranges orogeny (07 t006 Ga) aed Alc Spangseropeny (0.8 1 93 Ga) 5: Myers, Gesogy 1, 83710990) GE. Gowes, Gel ten, Siashnin Fk, 112,127 (1960) Hae x porte ro have ndergone 4 {1 deci roto, reamwe to Leen Se tucea P6-and Ll Ga, Ara rey peo Green ‘ic sons that were former ceasonous with ‘how in Laren farmed teh oppo {sore the time ofthe Greslianoregey. Nt Lith, BA Kibo, EA. O'Connor, PE J: Bt, ae 314, 385 (1985). {C Hhsmnady,PJouber, © Siowe, Epes 8, 236 (i985), BLM. Eglington, RE, Hamer, A. Ker Pr Iran oe 4b 199 (1909). Ff. Baca and. Rua, J. Gea. 97, 085 (1989), J. Palle, J.C Cale 8. Soeser,)- Quer {Gatgy 1, 85 (1900) D. i Seoesee snd VB Gang Ga: SoA Bll 96,817 85). Van der Von, Gt Sac im Bul. 100, 311 (988): Marphy snd RD. Nace, Gri, 17, 735 (1989). REPORTS MilMapes IM. Cosa, C Cast, Pream- in Abstracts, Intemational Confrence on Geocho- 8. Raby, Gol. Sue Aim. Se Pa 290, 1546. RSet, Go) enfin Beara) Pos MGos Rig Comamay od fae Gg Ce, 8. SE Hower, Gok A Cot Spe: Pep 37, Sh jan Bd 6 458 198. Kad GotogialSoty of swab, Caer, 485.0950), 47 MT Hwsetn, Bll Am Abe Peel Gel 73,1880) p 3h 40. TCO and RB. Hargis, Nawe289, 181°” 117 (iam, $6, CR ec, pn Pt ek en sap 48. KGa, Uh Renaud, ©. Pe, an.J. “ow Ch. Rep 9, | 41 MIP Goward and MC: Dat, Praia Rec 24, © Be i 26 1198 1989). ‘oj ni x Geog, One oF ee | 27 ase 428. Doras, Taonophts 87,287 (1979) 25, Asi, 1987). | 42. PR Reme, T.-C, Onwon, M.S. DiAgesfte, 50. VA Ramon pects, 168 8. 97. Bey HD, Ry A Sn B.A.T | E.G. Par W Tein, Eth iat St Let OL, SL JP. Lace, RD, Daley M, Vleoave, Harper oe 19, 7 (1981). | 589 (00), ‘GaSe ce Spc Pup. 250,151, (1989) "$8, Thrace nity conc comment on 48. RM. Siadlcon, in Colson Tela, M. P. 52. 1, Wiuns aad 8 NH, Goer 18104 manwerpe (2, 2) at Tal evened T Gonacd and AG Ris, Bas (Spee Pa 1, (1887 Sank © Ros fr permis o ier | ‘Gcogial Sey, Blt, Gur 985), pp 53. GG Bond. N. Chee Rice M.A Koning, WJ. (18) pes RT Ba JP. Grong Can 220-20 Mara and Abn, TwemrB, "Dey Nene 316,42 (988), A Thomas, Sah Ke Darah and to aaron ceenes 4385 (989), We Beak FF Bonin Ge: Ge. Sum Bul 108, $18 (1991). imadesoeesion te ped improve the Be fain, R. Scher T Tasagsy Ge Mag, $4 SG. Bang D, J, Dek BM. Smith, J. man Geolghal Srey of Carma contion Tae, 67 (1989) arp, 96, 847 (1990), Ses. | 44, V6 Ravin, Cotcnce 2,213 (1988) 85. GB" Odn eral, Natre 301,21 (1983, W. 46. SM. Babe, J. Gal Sc Lowen 147, 41,0990, *” Gop ES Witams, | Kile Zhang, 4 Janay 1991; acepted Mach 1991 cy ‘SCIENCE, VOL. 252Reprint Series et eee tee SciENCE Did the Breakout of Laurentia Turn Gondwanaland Inside-Out? Paut. F, HorrMan Copyright © 1991 by the American Association for the Advancement of Science