AUL OFFMAN ERIES - Princeton University · 1971; Deynoux and Trompette 1976; Hambrey and Harland...

PAUL F. HOFFMAN SERIES

The End-CryogenianGlaciation of SouthAustralia

Catherine V. Rose a,b, Adam C. Maloofa, Blair Schoenea, Ryan C.Ewingc, Ulf Linnemannd, MandyHofmannd, and John M. Cottlee

aDepartment of GeosciencesPrinceton UniversityGuyot Hall, Washington RoadPrinceton, NJ, USA 08544

bDepartment of Earth and Planetary SciencesWashington University in St. Louis1 Brookings Drive, St. Louis, MO, USA63130E-mail: [email protected]

cDepartment of Geology and GeophysicsTexas A&M University, MS 3115,College Station, Texas, TX, USA 77843

dSenckenberg Naturhistorische SammlungenDresdenMuseum für Mineralogie und GeologieKönigsbrücker Landstrasse 159, D-01109Dresden, Germany

eDepartment of Earth ScienceUniversity of CaliforniaSanta Barbara, CA, USA 93106

SUMMARYThe Elatina Fm. records the youngerCryogenian ice age in the Adelaide RiftComplex (ARC) of South Australia,which has long-held the position as thetype region for this low-latitude glacia-tion. Building upon a legacy of work,we document the pre- and syn-glacialsedimentary rocks to characterize thedynamics of the glaciation across theARC. The Elatina Fm. records an arrayof well-preserved glacial facies at manydifferent water depths across the basin,including ice contact tillites, flu-vioglacial sandstones, dropstone inter-vals, tidal rhythmites with combined-flow ripples, and turbidites. The under-lying Yaltipena Fm. records the pro-glacial influx of sediment fromencroaching land-based ice sheets. Theonset of the glaciation is heralded bythe major element ratios (ChemicalIndex of Alteration) of the pre-glacialfacies across the platform that show areduction in chemical weathering and adeterioration in climate towards thebase of the Elatina Fm. The advancingice sheets caused soft-sediment defor-mation of the beds below the glacialdiamictite, including sub-glacial pushstructures, as well as sub-glacial erosionof the carbonate unit beneath. Meas-ured stratigraphic sections across thebasin show glacial erosion up to 130 minto the carbonate platform. However,δ13C measurements of carbonate clastswithin the glacial diamictite units wereused to assess provenance and relativetiming of δ13C acquisition, and suggestthat at least 500 m of erosion occurredsomewhere in the basin. Detrital zircon

provenance data from the Elatina Fm.suggest that glacial sediment may havebeen partially sourced from the cratonsof Western Australia and that theWhyalla Sandstone, even if stratigraph-ically correlative, was not a sedimentsource. The remainder of the ElatinaFm. stratigraphy mostly records thedeglaciation and can be divided intothree facies: a slumped sandstone,dropstone diamictite, and current-reworked diamictite. The relative sealevel fall within the upper Elatina Fm.requires that regional deglaciationoccurred on the timescale of ice sheet– ocean gravitational interactions(instant) and/or isostatic rebound(~104 years). Structures previouslyinterpreted as soft-sediment folds with-in the rhythmite facies that were usedto constrain the low-latitude positionof South Australia at the time of theElatina glaciation are re-interpreted asstoss-depositional transverse rippleswith superimposed oscillatory waveripples. These combined-flow ripplesacross the ARC attest to open seaswith significant fetch during the initialretreat of local glaciers. In addition,this interpretation no longer requiresthat the magnetization be syn-deposi-tional, although we have no reason tobelieve that the low-latitude direction isa result of remagnetization, and posi-tive reversal tests and tectonic foldtests are at least consistent with syn-depositional magnetization. Together,these paired sedimentological andchemostratigraphic observations revealthe onset of the glaciation and advanceof the ice sheet from land to create aheavily glaciated terrain that wasincised down to at least the base of thepre-glacial Trezona Fm.

Geoscience Canada, v. 40, http://dx.doi.org/10.12789/geocanj.2013.40.019 © 2013 GAC/AGC®

256

GEOSCIENCE CANADA Volume 40 2013 257

SOMMAIRELa Formation d’Elatina représente laphase précoce de l’âge glaciaire duCryogénien de l’Adelaide Rift Complex(ARC) dans le sud de l’Australie, régionqui a longtemps été la région type decette glaciation de basse latitude. Àpartir d’un legs de travaux, nous noussommes appuyés sur l’étude des rochessédimentaires préglaciaires et syn-glaciaires pour caractériser ladynamique de la glaciation à traversl’ARC. La Formation d’Elatina ren-ferme une gamme de faciès glaciairesbien préservés correspondant à dif-férentes profondeurs d’eau à travers lebassin, dont des tillites de contactglaciaire, des grès fluvioglaciaires, desintervalles à galets de délestage, desrythmites tidales avec des combi-naisons de rides d’écoulement, et desturbidites. La Formation sous-jacentede Yaltipena est constituée de sédi-ments proglaciaires provenant delentilles de glace en progression. Ledébut de la glaciation est reflété dansles ratios des éléments majeurs (indiced’altération chimique) des facièspréglaciaires de la plateforme qui mon-tre une réduction de l’altération chim-ique et une détérioration du climat àl’approche de la base de la Formationd’Elatina. La progression des nappesde glace a entraîné une déformationdes lits de sédiments meubles sous ladiamictite glaciaire, montrant entreautres des structures de poussée sous-glaciaires ainsi que de l’érosion sous-glaciaire de l’unité de carbonate sous-jacente. Les mesures de coupes strati-graphiques à travers le bassin montrentque l’érosion glaciaire a enlevé jusqu’à130 m du carbonate de la plateforme.Toutefois, les signatures isotopiquesδ13C de fragments de carbonate dansles unités de diamictites glaciaires util-isées pour établir la provenance et lachronologie d’acquisition relative de lasignature δ13C des fragments, permetde penser qu’il y a eu au moins 500 md'érosion quelque part dans le bassin.Les données de provenance sur zirconsdétritiques de la Formation d’Elatinapermettent de penser que les sédimentsglaciaires provenaient partiellement descratons de l'Australie occidentale et quele grès de la Formation de Whyalla,bien que stratigraphiquement corrélé,n'a pas été une source de sédiments.Ce qui reste de la stratigraphie de la

Formation d’Elatina représente princi-palement la déglaciation, laquelle peutêtre divisée en trois faciès : un grèsplissé, une diamictite à galets dedélestage, et une diamictite remaniéepar des courants. La baisse du niveaurelatif de la mer dans la partiesupérieure de la Formation d’Elatinasuppose une déglaciation régionale surune échelle de temps de l’ordre de cellede la nappe de glace – interactionsgravitationnelles de l’océan (instanta-nées) et/ou rebond isostatique (~ 104

ans). Des structures décritesprécédemment comme des plis de sédi-ments mous dans des faciès de rhyth-mites qui impliquait une position debasse latitude pour l'Australie du Sud àl'époque de la glaciation Elatina, sontréinterprétées comme des rides sédi-mentaires transverses asymétriquesavec des rides de vagues oscillatoiressuperposées. Ces combinaisons derides d’écoulement à travers l’ARCconfirment l’existence d’un milieumarin ouvert d’une ampleur certaine aumoment de la retraite initiale des gla-ciers locaux. En outre, cette interpré-tation ne nécessite plus que la magnéti-sation soit synsédimentaire, bien quenous n'ayons aucune raison de penserque l’orientation magnétique de basselatitude soit le résultat d’une ré-aiman-tation, et que les tests de réversibilitépositifs et les tests de plissement tec-tonique sont au minimum conformes àune magnétisation synsédimentaire.Ensemble, ces observations sédimen-tologiques et chimiostratigraphiquesmettent en lumière le début de laglaciation et l'avancée du couvert deglace continental menant à une régionfortement englacée qui a été inciséejusqu'à à la base de la Formationpréglaciaire de Trezona.

INTRODUCTIONAt least two Neoproterozoic glacigenic,poorly sorted conglomeratic units arepresent on all continents exceptAntarctica, commonly interrupting car-bonate platform sequences, and insome cases found near the paleomag-netic equator. Therefore, at least twiceduring this era, continental glaciersreached sea level in the low-latitudes(Embleton and Williams 1986; Schmidtand Williams 1995; Sohl et al. 1999;Evans 2000; Macdonald et al. 2010).These two Cryogenian glaciations are

colloquially referred to as the older‘Sturtian’ and younger ‘Marinoan’ low-latitude glaciations (Hoffman andSchrag 2002; Hoffman 2005, 2011).Banded iron formation within the thickSturtian diamictite units, and sedimen-tologically and geochemically distinc-tive ‘cap’ carbonate sequences thatconsistently drape both glacial deposits(Williams 1979; Kennedy 1996;Kennedy et al. 1998; Hoffman et al.2007; Rose and Maloof 2010), mayindicate long-term isolation of theocean from the atmosphere. Together,these observations led to the contro-versial ‘snowball Earth’ hypothesis,wherein an ice-albedo runaway feed-back caused ice to advance rapidly tothe equator and seal the entire ocean inice for millions of years (Kirschvink1992; Hoffman et al. 1998, 2002).

Early work on these enigmaticlow-latitude glaciations focused onidentifying and describing the glacialdeposits around the world (Spencer1971; Deynoux and Trompette 1976;Hambrey and Harland 1981; Deynoux1985; Lemon and Gostin 1990). How-ever, this focus on the glacial depositsled to different interpretations andsome workers proposed a debris floworigin for the conglomeratic beds(Schermerhorn 1974; Eyles 1993;Arnaud and Eyles 2006). Eyles andJanuszczak (2004) argued that both theglacigenic deposits and associated car-bonate rocks could be explained bycontinental rifting, with carbonatedeposited in restricted basins starvedof clastic input. Although not wide-spread, high-latitude carbonate andevaporite rocks coexist with glacigenicsediments today (Walter and Bauld1983). A close examination of the pre-and post-glacial Neoproterozoic rocksin the North Atlantic region, however,determined that these carbonate unitswere probably deposited in warmwater and thus climatic fluctuationsbetween ‘balmy’ and ‘icy’ conditionsoccurred quite rapidly (Fairchild 1993).

The snowball Earth hypothe-sis (Kirschvink 1992; Hoffman et al.1998; Hoffman and Schrag 2002) wasproposed by Kirschvink (1992) in anattempt to explain the presence ofbanded iron formation within the thickolder Cryogenian diamictite units,which were deposited near the equator.Further work also identified the dis-

258

tinctive ‘cap’ carbonate sequences thatconsistently drape both glacial deposits(Williams 1979; Kennedy 1996;Kennedy et al. 1998; Hoffman et al.1998, 2007), which were not explainedby previous models. Both of theseobservations imply perturbations toseawater chemistry that are consistentwith long term isolation of the oceanfrom the atmosphere. Hence, since itsmost recent formulation in 1998, testsof the snowball Earth hypothesisturned their attention to the geochem-istry of the cap carbonates (Williams1979; Fairchild 1993; Hoffman et al.1998; Grotzinger et al. 2000; James etal. 2001; Kennedy et al. 2001; Shields2005; Hoffman and Schrag 2002; Hig-gins and Schrag 2003; Hoffman et al.2007; Rose and Maloof 2010; Hoff-man 2011). Despite advances towardunderstanding this non-uniformitarianclimate state, relatively little recentwork has been done on the glacial sedi-ments themselves. This inattentionarises in part because these depositsare spatially heterogeneous and diffi-cult to interpret. Recently, some limitedobservations of thick glacial depositsintercalated with wave-rippled andhummocky cross-stratified interglacialsandstones (Allen and Etienne 2008;Le Heron et al. 2011a,b) indicated ice-front mobility or advance–retreatcycles (Christie-Blick et al. 1999; Con-don et al. 2002; Leather et al. 2002;Rieu et al. 2007b; Allen and Etienne2008). These studies present theirobservations as a challenge to a ‘hard’snowball scenario, which completelyencompasses the world with ice, ques-tioning the ideas that sea ice would beglobally present and that moisturefrom sea-ice sublimation alone wouldhave been sufficient to drive thedynamics of polythermal ice withactive subglacial hydrology. However,glacial diamictite successions may bedeposited entirely during deglaciation,and such arguments for advance–retreat cycles may not be relevant topeak snowball conditions. Yet, impor-tantly, these few studies highlight theneed for a re-evaluation of low-latitudeglacial deposits themselves. Our workre-focusses attention back to the glacialdeposits, investigating the Marinoan-age Elatina Fm. in the Adelaide RiftComplex (ARC), South Australia.However, rather than studying the

diamictites in isolation, we also build acomprehensive multidisciplinarydataset from the adjacent strata to eval-uate both the transition into and out ofthe ice-house event. This study demon-strates that an approach that integratesbasin-scale analysis with detailed sedi-mentology and chemostratigraphy,when set in the context of the pre- andpost-glacial sediments, can provide newinsights into the dynamics of extensiveglaciations of the Cryogenian.

The Elatina Fm. is of globalimportance because: 1) its sedimentol-ogy is diverse and well preserved,recording transitions in glacial facies atdifferent water depths across the basin;2) it represents the type region for theMarinoan glaciation (Williams et al.2008); 3) it has yielded the most robustpaleomagnetic data for any late Cryo-genian glacigenic succession (Evansand Raub 2011), and 4) the recentlyestablished Ediacaran System and Peri-od (Knoll et al. 2006) has its GlobalStratotype Section and Point (GSSP) atthe base of the Nuccaleena Fm. over-lying the Elatina Fm. in the centralFlinders Ranges. The glacial origin ofthe Elatina Fm. was first recognized byMawson (1949) following his discoveryof diamictite containing faceted andstriated clasts in Elatina Creek in thecentral Flinders Ranges, and he pro-posed the term ‘Elatina glaciation’(Williams et al. 2008; Fig. 1 [17]). Sincethese initial observations, several stud-ies have been published on the ElatinaFm. Additional mapping extendedobservations of the Elatina Fm. lateral-ly across the ARC (Dalgarno and John-son 1964; Leeson 1970). This carefulwork led to Elatina Creek being nomi-nated as the type section (Dalgarnoand Johnson 1964) and to correlationsto other Elatina Fm. sections in thesouthern (Binks 1968; Miller 1975;Jablonski 1975) and northern (Coats etal. 1973; Preiss and Forbes 1981)Flinders Ranges. Lemon and Gostin(1990) presented a detailed sedimento-logical study of the Elatina Fm. withinthe central Flinders Ranges, whichestablished three correlative facies forthe Elatina Fm. that were interpretedas recording the advance of the ice andsubsequent deglaciation. Most workersdocument the advance of grounded iceor scouring by icebergs, which attest tothe glacial nature of the Elatina Fm.

(Lemon and Gostin 1990; Lemon andReid 1998; Williams et al. 2008). How-ever, it has been suggested that theupper Elatina Fm. consists of glaciallyinfluenced debris flows resulting fromlocal slope collapse (Le Heron et al.2011a; Le Heron 2012), and someworkers have even proposed that theentire succession records tectonicallytriggered mass flow deposits (Scher-merhorn 1974; Eyles et al. 2007).

Williams (1989, 1991, 1998,2000) and Williams et al. (2008)described rhythmic laminations withinsiltstone of the upper Elatina Fm. atWarren Gorge, Pichi Richi Pass in thesouthern Flinders Ranges, and MarinoRocks south of Adelaide (Fig. 1 [2, 4]).The rhythmites consist of 1–2 cm-thick bundles of mm-scale coupletsthat were originally interpreted asvarves deposited in a periglacial lake(Williams 1981), where each clasticlamina represents annual deposition ofsediment by glacial meltwater (Williams1981, 1985; Williams and Sonett 1985).Williams (1985) argued that the period-ic deposition of cyclic laminae corre-lated with sunspot cycles, though themechanism by which Marinoan climatewas controlled by sunspots was notclearly articulated. Furthermore, astro-nomical models suggest that thesunspot cycle during the Neoprotero-zoic was 3-10% shorter than today(Noyes et al. 1984), whereas the Elati-na Fm. data suggest that the meancycle was ~8% longer at that time(Williams 1988). By comparing theElatina rhythmites to two other puta-tively correlative rhythmite deposits inSouthern Australia, the Reynella Silt-stone Member and Chambers BluffTillite, Williams determined that thepackages of laminae represent fort-nightly cycles of spring-neap lunar tidedeposits as a distal part of an ebb-flood tidal delta (Zahnle and Walker1987; Williams 1988). The literaturesince has focused on correlating rhyth-mite periodicities with tidal periodici-ties and elucidating the implications forthe history of the Earth-Moon orbit(Williams 1997, 1998, 2000). Underthis model, one couplet represents asemi-diurnal or diurnal depositionalcycle, strictly constraining the rate ofbedform migration and aggradation ofthe rhythmite sequences.

The rhythmites also are the


centre-piece of a key paleomagneticconstraint for Cryogenian low-latitudeglaciation and the snowball Earthhypothesis. Results of a fold test onputative soft-sediment folds withinthese rhythmites indicated that thecharacteristic remnant magnetism

(ChRM) was syn-depositional in ageand constrained the Elatina glaciationto an equatorial paleolatitude (<10º)(Sumner et al. 1987). A stratigraphicallyconsistent polarity reversal test con-firmed the primary component ofChRM in the Elatina Fm. and suggest-

ed a paleolatitude of 2.7 ± 3.7º(Schmidt et al. 2009). Raub and Evans(2008) and Schmidt et al. (2009) founda steeper mean inclination for the Nuc-caleena Fm. of 27º, possibly due to theless-compacted carbonate lithology,which results in a paleolatitude of 14 ±

Flin

der

s R

ang

es

Gam

mon

Ran

ges

ADELAIDE

COMPLEXRIFT

StuartShelf

31o

32o

30o138o

139o 140o

9° paleonorth

Blinman

30

130120110 140

0 200 400 600 800 1000km

20

10120 130 140 10

20

30

160

150

150

AdelaideRiftComplex

N

volcanic rocks

undifferentiated sedimentary rocks

sandstone, siltstone, shale, limestone, conglomerate, tuff

WILPENA GROUP: sandstone, siltstone, shale, dolomite

PERMIAN-QUATERNARY

CAMBRIAN-ORDOVICIAN

CAMBRIAN

NEOPROTEROZOIC

UMBERATANA GROUP: diamictite, sandstone. siltstone, limestone, dolomite, conglomerate, shale

BURRA GROUP: siltstone, sandstone, dolomite

CALLANNA GROUP: siltstone, sandstone, carbonate, evaporite, basalt, diapiric breccia

PALEO-MESOPROTEROZOIC

undifferentiated Paleo-Mesoproterozoic rocks

major fault

syncline axisanticline axis

sections

LEGEND

N635 Ma

N0 Ma

Wilpena Group

Umberatana Group

Burra Gp

ADELAIDE RIFTSTUARTSHELF

SE

Callanna Gp

PALEOPROTEROZOIC BASEMENT OF THE GAWLER CRATON

Ediacara biota

0 5 10-5-10

0 5 10-5-10

BitterSpringsStage

SturtianGlacial

MarinoanGlacial

GaskiersGlacial

Neo

proteroz

oic

777 Ma

635 Ma

542 Ma

D R

I F

T

490 MaDelamerian orogeny

R I F

T659.7 MaSturtian

Cry

oge

nia

nEdia

cara

n

Tonian

Cambrian

NW

3036

26

28

27

313233

35

38

3940

4142

4344

45 46

47

29

37

34

12

11

876

543

2

9 10

1

16

18 192022

23

24

17

21

1314

15

III

IV

V

I

II

Fig. 8

Fig. 6

Fig. 7

δ13Ccarb (‰ VPDB)

a

b

Figure 1. (a) Simplified geological map of the study area within the Adelaide Rift Complex (ARC) adapted from Preiss andRobertson (2002). Locations of measured stratigraphic sections are denoted by red circles and labeled with numbered squares.Fold axes within the Adelaide Rift Complex are denoted by dashed lines and labeled with Roman numerals: (I) Mount Morrisanticline; (II) Mount Jeffery syncline; (III) Arkaroola syncline; (IV) Umberatana syncline; and (V) Mount Fitton anticline. Thelarge grey boxes represent the areas of the detailed maps presented in Figures 6, 7 and 8. (b) Schematic NW-SE stratigraphiccross-section of the Adelaide Rift Complex, highlighting the rift-to-drift transition and major sequence boundaries (adaptedfrom Lemon and Gostin 1990). δ13C profile adapted from Halverson et al. (2005) time-aligned with the right-hand edge of thestratigraphic cross-section. A SHRIMP U–Pb zircon age of 659.7 ± 5.3 Ma from a tuffaceous horizon in the Wilyerpa Fm., justabove the Appila (Sturtian) diamictite, provides a maximum age for the base of the interglacial sedimentary units (Fanning2006). The overlying Nuccaleena Fm. is dated by correlation to the uppermost Marinoan glacial deposits and the associated capdolostone in Oman (Bowring et al. 2007; Rieu et al. 2007a), Namibia (Hoffmann et al. 2004) and South China (Condon et al.2005), which contain ID–TIMS U–Pb zircon ages of ~635 Ma.

260

2º (Evans and Raub 2011). The syn-sedimentary fold test represents themost reliable paleomagnetically derivedlow-latitude constraint for a late Cryo-genian glacial deposit. The ‘fold’ struc-tures used are spaced ~50 cm apartwithin the rhythmite facies, have beendocumented in Warren Gorge andPichi Richi Pass, and have been inter-preted as gravity slides that were trig-gered by storm waves (Williams 1996).

The existence of multiplereversals suggests that the Elatina gla-cial epoch lasted for several 105 to afew 106 years (Schmidt and Williams1995; Sohl et al. 1999). Similarly, multi-ple magnetic reversals recorded in theNuccaleena Fm. cap dolostone suggestthat it took >105 yrs for its deposition(Trindade et al. 2003; Kilner et al.2005; Raub and Evans 2006; Schmidtet al. 2009). However, the giant waveripples present in cap dolostone unitsaround the world indicate extremelyfast aggradation (Hoffman et al. 2007;Raub and Evans 2008; Lamb et al.2012). This sedimentological constraintimplies that the multiple polaritychrons in the Nuccaleena Fm. are per-haps not true reversals but are a rarerecording of geomagnetic excursionsthat can occur in less than ~2 ky(Gubbins 1999; Hoffman et al. 2007).Thus the timescale for the duration ofthe Elatina glacial epoch remains con-troversial.

Although carbonate carbonand bulk organic carbon records haveplayed key roles in understanding thepre-glacial (Rothman et al. 2003; Fikeet al. 2006; Swanson-Hysell et al. 2010;Rose et al. 2012) and post-glacial car-bonate succession (Hoffman et al.1998; Grotzinger et al. 2000; Kennedyet al. 2001; Hoffman and Schrag 2002;Higgins and Schrag 2003; Hoffman etal. 2007; Rose and Maloof 2010), fewstudies have looked beyond the sedi-mentology with regard to the glacialdiamictite units. The Chemical Indexof Alteration (CIA) is a weatheringproxy that uses ratios of major ele-ments in siliciclastic rocks (Nesbitt andYoung 1982, 1989; Fedo et al. 1995;McLennan et al. 1993; Nesbitt andYoung 1996; Colin et al. 1998; Corco-ran and Mueller 2002; Sheldon et al.2002; Scheffler et al. 2003). High CIAvalues reflect the removal of mobilecations (Ca2+, Na+, K+) relative to stable

residual constituents (Al3+, Ti4+) duringchemical weathering, which isenhanced during humid and warm cli-mate conditions. In contrast, low CIAvalues indicate the near absence ofchemical weathering, and consequentlymight reflect cool and/or arid condi-tions. Such compositional variation ofsiliciclastic rocks has been used to eval-uate Paleoproterozoic, Cambro-Ordovician, Neogene and Quaternaryglaciations (Krissek and Kyle 1998;Young 2002; Dobrzinski et al. 2004;Bahlburg and Dobrzinski 2011). Sever-al Cryogenian glacial successionsrecord low CIA values within the gla-cial diamictites and relatively high CIAvalues in intercalated siltstone, includ-ing the Port Askaig Fm., Scotland(Panahi and Young 1997), the NantuoFm., South China (Dobrzinski et al.2004), the Huqf Fm., Oman (Rieu etal. 2007c) the Smalfjord and Morten-snes Fms., northern Norway, andGhaub Fm., Namibia (Bahlburg andDobrzinski 2011). However, this geo-chemical proxy has not been per-formed on the siliciclastic-dominatedglacial Elatina Fm., or assessed in mul-tiple sections across a basin to deter-mine the variable role of provenance,diagenesis, and grain size on CIA val-ues.

The ARC provides a uniqueopportunity to examine the three-dimensional paleo-landscape and theevolution of the glaciation that mayallow us to test specific predictions ofthe snowball Earth hypothesis. We aimto establish the extent of sub-glacialerosion and advance–retreat cycles totest the predictions of a cold-based icesheet, an attenuated hydrological cycle,and rapid onset and deglaciationassumed to be required by a snowballEarth. We aim to determine the prove-nance of the glacially transported sedi-ments to test the snowball Earth pre-diction that the onset of the glaciationin the tropics is represented by theadvance of sea-ice onto the continents(Hoffman et al. 2002). Furthermore,we aim to establish if open water exist-ed on shallow platforms during theMarinoan glaciation, which has beencited in opposition to the ‘hard’ snow-ball Earth model. Such an open waterscenario would be compatible with aJormungand climate state where a thinstrip of the tropical ocean remains

exposed (Abbot et al. 2011).In this paper, we present

detailed sedimentological observationspaired with high resolution δ13C andmajor element data from the pre-glacialcarbonate platform and syn-glacial sed-iments of the Marinoan glaciationacross the ARC, South Australia. Wequantify the degree of erosion, charac-terize the provenance, and establish thestyle of the glaciation across the basin.A total of 47 stratigraphic sectionsdocument the syn-glacial facies thatcorroborate the seminal work ofLemon and Gostin (1990), and extendanalysis to the entirety of the basin.We document a regression within theupper Elatina Fm. that suggests thatthe regional deglaciation occurred onthe timescale of gravitational with-drawal (instant) and/or isostaticrebound (~104 years). We propose thatthe ‘soft-sediment folds’ in the rhyth-mite facies across the ARC are com-bined-flow ripples, which attest toopen seas with significant fetch duringthe initial retreat of local glaciers. Thecurrent and wave ripples also castdoubt on the veracity of the syn-depo-sitional paleomagnetic constraint,requiring that the low-latitude directiononly need be pre-Late Cambrian fold-ing in age. We quantify the amount ofglacial truncation across the platformfrom 29 δ13C chemostratigraphic sec-tions through the pre-glacial carbonateplatform. This work allows correlationof formations across facies transitionsto unify and simplify the pre-existingstratigraphic classifications across thebasin. We show that δ13C–δ18O valuesfor 269 carbonate clasts within thediamictite were acquired before glacialerosion, and that deep glacial incisioninto the carbonate platform likely wasresponsible for exhumation of claststhat record the full stratigraphic rangein δ13C values. We present detrital zir-con data from the ARC and StuartShelf to establish possible temporaland spatial changes in the provenanceof the glacial sediments. Finally, wepresent new bulk compositional data(major elements) from 13 sections totest whether CIA records are coherentbasin-wide, and may record paleo-weathering intensity, or whether theCIA proxy is controlled by the second-ary processes of sorting and/or diage-nesis.


GEOLOGICAL BACKGROUNDThe Adelaide Rift Complex (ARC) ispart of a continental margin formed tothe present-day east of the Stuart Shelf(Preiss 2000; Fig. 1a). The basementbeneath the Gawler Craton to the westof the ARC is composed of lateArchean–early Mesoproterozoic mag-matic lithologies and metasedimentaryrocks (Fig. 2). Similarly, the Curna-mona Province to the east, and the Mt.Painter Inlier to the south of the ARCconsist of Paleoproterozoic–Mesopro-terozoic granite, gneiss and metasedi-mentary rocks (Willis et al. 1983; Preiss2000; Compston et al. 1966; Coats andBlisset 1971; Teale 1993; Elburg et al.

2001). The basement beneath the ARChas only limited exposure but may be adistinctly younger geological provincethan most of the Gawler Craton(Preiss 2000). It is likely that there wasa late Paleoproterozoic precursor basinoccupying much the same area as theNeoproterozoic ARC, with sedimenta-tion and volcanism between ~1.75–1.65 Ga, and orogeny at ~1.6 Ga(Preiss 2000).

Paleoproterozoic to Mesopro-terozoic cratonic basement of theARC is overlain by a 7-12 km thickNeoproterozoic to Cambrian sedimen-tary package that is subdivided intofour units: the Callana, Burra, Umber-

atana and Wilpena Groups (Preiss1987, 2000). The ARC was a zone ofdeep subsidence, punctuated byepisodes of syn-sedimentary faultingand diapiric mobilization of CallannaGroup evaporites (Preiss 1987; Fig.1b). Neoproterozoic sediment accumu-lation is attributed to a succession ofrift and thermal subsidence phases,with the main rifting commencing at~827–802 Ma (Fanning et al. 1986;Jenkins 1990; Wingate et al. 1998). It isnot known exactly when rifting termi-nated, but large-scale crustal normalfaulting is thought to have diminishedby the Cryogenian Period (Preiss2000). Deposition ceased and the sedi-

PB

YB

GAB

PP

HCI

PCI

MII

GB

Am

AI

AFP

GV

GI

CI

MDS

NHI

BH

MP

MB

GDS

Ge

Bi

AV

M

RJI

TASMAN

FOLD

BELT~500 M

a

Ng

AFP: Albany-Fraser Province Nd:1.85-2.33 Ga, U-Pb: 1.0-1.3 GaAd: Adelaidian Rift Complex 800-550 MaAm: Amadeus Basin NeoproterozoicA: Arkara Gneiss U-Pb: 1.58 Ma AI: Arunta Inlier Nd: 2.12-2.19 GaBa: Bangemall Basin 1.45-1.3 GaBi: Birrindudu Basin 1.56 GaBH: Broken Hill Block Nd: 2.2-2.26 GaCI: Coen Inlier Nd: 1.94-2.13 GaFR: Flinders RangeGAB: Gascovne Block Nd: 2.20-3.52 GaGDS: Gairdner Dyke Swarm and Willouran Volc. 827.9 Ma GB: Gawler Block Nd: 2.54-2.69 Ga, U-Pb 2.55 and 2.31 GaGV: Gawler Volcanics U-Pb: 1.58-1.60 GaGI: Georgetown Inlier Nd: 2.17-2.29 Ga, U-Pb: 1.7-1.55 GaGe: Georgina Basin, NeoproterozoicHCI: Halls Creek Inlier Nd: 2.13-2.29 GaH: Hammersley Basin 2.77-2.20 GaK: Kimberley Basin <1.87-1.40 GaLB: Leeuwin Block U-Pb: 1200-1050 Ma, 800-650 Ma, 580-500 MaM: McArthur Basin 1.80-1.43 GaMII: Mount Isa Inlier Nd: 2.14-2.30 Ga, U-Pb: 1.62-1.52 GaMP: Mount Painter Block Nd: 2.01 GaMDS: Mundine Well Dyke Swarm 755.3 MaMuDS: Mugamurra Dyke Swarm 748.2 Ma (K-Ar)MB: Musgrave Block Nd: 1.77-1.89 Ga, U-Pb: 1.0-1.3 GaNg: Ngalia Basin, NeoproterozoicNHI: Northampton Inlier Nd: 2.02-2.08 Ga, U-Pb ~1.08 Ga, detrital zirc.<2.04 GaOf: Officer Basin NeoproterozoicPP: Paterson Province ~500 Ma w/ 690 Ma Crofton GranitePB: Pilbara Block felsic volc. & gneiss: Nd: 3.17-3.44 Ga, U-Pb: 3.45 Ga granite: Nd: 3.12-3.20 Ga, U-Pb 2.76 GaPCI: Pine Creek Inlier gneiss: Nd: 2.23-2.62 Ga, U-Pb 2.47 and 1.86 GaRJI: Rum Jungle Inlier gneiss & granite: Nd: 2.71-3.3 GaSa: Savory Basin, NeoproterozoicSS: Stuart ShelfTCI: Tennant Creek Inlier Nd: 2.27-2.50 Ga, U-Pb 1.87 GaV: Victoria River Basin 1.2 GaYB: Yilgarn Block granite: Nd: 2.60-2.75 Ga, U-Pb: 2.68 and 3.21 Ga xenocrystic zircon: U-Pb 3.4 Ga

Of

Ad

H

SSFR

Ba

K

MuDS

Mafic Dike Swarm

Outline of the Centralian Super Basin,representing the maximum flooding ofthe continents between 800-500 Ma

Tasman Line, representing maximumeastward extent of pre-Neoproterozoiccrust. Unlike Laurentia, the 87Sr/86Sr= 0.706 isopleth (measured inPhanerozoic granites) does not followa smooth line in Australia

Sa

Igneous/Metamorphic Ages

TCI

Archean-Early Proterozoic

Mesoproterozoic

Neoproterozoic-Cambrian

SEDIMENTARY BASINSMajor Lineament

2.6 Ga

2.0 Ga

1.9-1.8 Ga

1.8-1.7 Ga

1.7-1.6 Ga

1.6-1.5 Ga

1.3-1.0 Ga

0.7-0.5 Ga

LB

BASEMENT MAP OF AUSTRALIA

AGES

LEGEND

Map of Adelaide Rift Complex(Fig. 1)

Figure 2. Map of Australia denoting the ages of basement complexes and outlining the Precambrian sedimentary basins. TheU–Pb ages of the Gawler Craton and Adelaide Rift Complex (ARC), as well as the Musgrave Block, Albany-Fraser Province,Paterson Province, and Leeuwin Block, are highlighted in blue in the key (adapted from Pell et al. 1997).

262

mentary rocks were folded during theCambro-Ordovician Delamerianorogeny (ca. 514–490 Ma; Drexel andPreiss 1995; Foden et al. 2006) to cre-ate a region of elevated topographyforming the Flinders and GammonRanges (Thomson et al. 1964).

Stratigraphy of the Adelaide RiftComplexThe Burra Group consists of basalcarbonate rocks with evaporite andclastic units (Preiss 1987), and theUmberatana Group is characterized bya ~4.5 km thick interglacial successionbounded by the older Sturtian-equiva-lent and younger Marinoan-equivalentglacimarine deposits (Fig. 1b). Theseglacial deposits within Australia arereferred to as the ‘Sturt’ and ‘Elatina’glaciations respectively, because theseterms are consistent with local stratig-raphy (Williams et al. 2008). TheWilpena Group records the end-Cryo-genian post-glacial sequence, whichthen shallows upwards in two parase-quences from deep marine siltstone toshallow marine sandstone. Thesesequences are followed by transgressiveEarly Cambrian shallow-marine sand-stone and deeper water carbonate andshale (Preiss 1987).

The interglacial successionbetween the Sturtian and Marinoanglacials of the Umberatana Groupconsists of the Tapley Hill Fm., EtinaFm., Enorama Fm., and Trezona Fm.(Fig. 3). The Tapley Hill Fm. consistsof dark grey, laminated siltstonedeposited during the post-Sturt glacialsea level rise. This unit shoals up intothe Etina Fm., which consists of shal-low marine sandstone, grey cross-bed-ded oolitic grainstone and sandy lime-stone, and microbial reefs interbeddedwith green dolomitic siltstone andshale (Fig. 4a). The base of the Enora-ma Fm. marks a major flooding sur-face and consists of finely laminatedgrey-green and minor red shale withminor fine sandstone beds. To thenorth, the nomenclature for the TapleyHill–Enorama Fm. stratigraphy con-sists of the Balcanoona, Yankaninna,and Amberoona Fms. (Preiss et al.1998; Fromhold and Wallace 2011; Fig.3). The extent to which these forma-tions are laterally correlative to oneanother and to the stratigraphy withinthe central Flinders Ranges is debated.

The Enorama Fm. is followed by agradual coarsening and shoaling-upward sequence that culminates inintraclastic limestone breccia, stroma-tolite bioherms, oolitic grainstone andsiltstone of the Trezona Fm. The Tre-zona Fm. contains putative fossildebris in packstones that onlap anddrape the stromatolite bioherms andthat have been interpreted as sponge-grade metazoan body fossils (Maloofet al. 2010). Detailed mapping and 26measured stratigraphic sections showthat the Trezona carbonate faciesrecord a progressive deepening

towards the north of the ARC (Rose etal. 2012; Fig. 1). In the south, the tem-poral equivalent of the pre-glacial Tre-zona Fm. is a thick, dark red, mud-cracked sandstone and siltstone depositwith medium-coarse grit lenses(Yaltipena Fm.; Fig. 4b–d). These sedi-mentary units inter-finger with thenearshore channelized limestone grain-stones, stromatolites and associatedfacies of the Trezona Fm. in the cen-tral region (Fig. 4e, f), and transition tostormy outer shelf carbonate ribboniteand grey-green calcareous shale in thenorth (Fig. 4g). The overlying syn-gla-

STUARTSHELF

SOUTHFLINDERS

CENTRAL FLINDERS

NORTHFLINDERS

THIS STUDY

Nuccaleena Fm

Reynella Siltstone Mbr

Whyalla Sandstone

Wilmington Fmequivalent

Angepena Fm equivalent

Brighton Lst

Tapley Hill Fm

Appila Tillite Appila Tillite

Nuccaleena Fm Nuccaleena Fm Nuccaleena Fm

Reynella Siltstone Mbr

Elatina Fm Elatina Fm

Balparana Sandstone

Mt. Curtis Tillite

Fortress Hill Fm

Yaltipena Fm

Trezona Fm

Enorama Shale

Etina Fm

Tapley Hill Fm

Sunderland Fm

Wilmington Fm

Angepena Fm

Amberoona Fm

Yankaninna Fm

Tapley Hill Fm

Lyndhurst FmWilyerpa Fm

Tapley Hill Fm Tapley Hill Fm Tapley Hill Fm

Elatina Fm

Yaltipena Fm

Trezona Fm

Enorama Shale

Etina Fm

Tapley Hill Fm

Um

bera

tana

Gro

up

Nep

ouie

Su

bgro

upU

palin

na S

ubgr

oup

Yere

lina

Subg

roup

Wilp

ena

Gro

upYu

dnam

utan

aSu

bgro

up

Stur

tian

Mar

inoa

nCr

yoge

nian

Edia

cara

n

Wilyerpa Tillite

~635 Ma

643+/-2.4 Ma

659.7+/-5.3 Ma

U-Pb date: ash at top Wilyerpa Fm. (Fanning 2006)

Re-Os date: basal Tindelpina Shale Mbr. (Kendall et al. 2006)

U-Pb dates: ash within glacial Ghaub Fm., Namibia (Hoffmann et al. 2004) ash within Doushantuo cap carbonate, China (Condon et al. 2005)

STRATIGRAPHY OF ADELAIDE RIFT COMPLEX AND STUART SHELF

Balcanoona Fm

Amberoona Fm

Figure 3. Cryogenian stratigraphy of the Adelaide Rift Complex (ARC) and neigh-bouring Stuart Shelf (modified from Preiss et al. 1998). The two Cryogenianglacially related series represent lower and upper deposits of the UmberatanaGroup and are assigned to Sturtian and Marinoan time divisions (Preiss et al. 1998).Note the name changes between each area in South Australia, which historically hasmade detailed correlations of the interglacial stratigraphy difficult. The final col-umn outlines the formation names that are used in this study for the end-Cryogen-ian interglacial–glacial stratigraphy across the entire ARC in an attempt to simplifyand unify current nomenclature.


cial Elatina Fm. exhibitsimpressive facies variability,from marine sandstones in thesouth, to ice-contact tillites,fluvioglacial and shallowmarine sandstones in the cen-tral region, and debris flowsand turbidites in the north(Coats 1981; Preiss 1987;Eyles et al. 2007; Fig. 5). Tothe west, the Elatina Fm. tran-sitions to the periglacial-eolianWhyalla Sandstone on the Stu-art Shelf (Williams and Tonkin1985; Williams 1986, 1994).The Elatina–Nuccaleena Fm.contact marks the onset of thepost-glacial transgression, thebase of the Wilpena Group(Williams 1977; Plummer1979; Dyson 1992; Kennedy1996) and the beginning ofthe Ediacaran Period (Knoll etal. 2006). The Nuccaleena Fm.consists of buff-coloreddolomite grainstone that isoverlain by red laminated silt-stone and fine-grained sand-stone of the Brachina Fm.

Many irregular brecciabodies mapped throughout theARC have been interpreted assyn-sedimentary diapirs (Webb1960; Coats 1965; Dalgarnoand Johnson 1968). Thesebodies formed by the intru-sion of evaporite from theCallana Beds with subsequentdissolution near the surface,leaving a cap of insolubleinterbeds and other rocksdragged to the surface byhalite. Active salt diapirism isthought to have occurred atleast throughout the deposi-tion of the Etina Fm. andlower two thirds of the Enora-ma Fm. (Lemon 2000). Thick-ening of these formationstowards the diapir in the cen-tral Flinders Ranges, which isparticularly evident at thenorthern fold axis of the cen-tral anticline, shows that thisinterval was a time of activesalt withdrawal and diapirgrowth (Lemon 2000; Fig. 1a).

Fig

ure

4.S

edim

ento

logy

of

the

pre-

glac

ial T

rezo

na a

nd Y

altip

ena

Fms.

(a) C

ross

-bed

ded

coar

se g

rain

ston

e w

ith q

uartz

gra

ins w

ithin

the

inte

rglac

ial se

dim

enta

ryro

cks o

f th

e E

tina

Fm.,

Moo

lool

oo (F

ig. 1

[24]

). (b

-d) M

ud-c

rack

ed si

ltsto

ne, m

ud-c

hip

brec

cia a

nd sy

mm

etric

micr

o-w

ave

rippl

es w

ithin

the

Yalti

pena

Fm

. at T

re-

zona

Bor

e, in

dica

ting

inte

rmitt

ent s

ubae

rial e

xpos

ure

and

very

shall

ow w

ater

, res

pect

ively

. (e)

Stro

mat

olite

s with

in th

e pr

e-gl

acial

Tre

zona

Fm

., Tr

ezon

a Ra

nges

(Fig.

1 [1

8]).

(f) O

utcr

op o

f th

e Tr

ezon

a Fm

. sho

win

g pu

tativ

e sk

eleta

l mor

phol

ogies

in fo

ssil

debr

is on

lappi

ng a

nd d

rapi

ng a

stro

mat

olite

bio

herm

(und

er th

e 30

cm

long

ham

mer

) (M

aloof

et a

l. 20

10; F

ig. 1

[18]

). A

3D

reco

nstr

uctio

n of

one

such

foss

il is

show

n in

the

inse

t im

age.

(g) I

nter

bedd

ed c

arbo

nate

ribb

onite

and

gre

y-gr

een

calca

reou

s sha

le of

the

Trez

ona

Fm. i

n th

e no

rther

n Fl

inde

rs R

ange

s, U

mbe

rata

na (F

ig. 1

[40]

).


Radiometric Age ConstraintsThere are no direct age constraints forthe onset or duration of the ElatinaFm. A SHRIMP U–Pb zircon age of659.7 ± 5.3 Ma from a tuffaceous hori-zon in the Wilyerpa Fm., just abovethe Appila (Sturtian) diamictite, pro-vides a maximum age for the base ofthe interglacial sedimentary rocks (Fan-ning 2006). A black shale in the Tin-delpina Shale Member of the lower-most Tapley Hill Fm. has given a Re-Os age of 643.0 ± 2.4 Ma (Kendall etal. 2006). The post-glacial NuccaleenaFm. ‘cap dolostone’ may be ~635 Ma.This age is determined by correlatingthe formation, based upon sedimentol-ogy and chemostratigraphy, to the lateCryogenian glacial deposit and base-Ediacaran cap dolostone in Namibia(635.5 ± 0.6 Ma; Hoffmann et al.2004) and South China (635.2 ± 0.6Ma; Condon et al. 2005), respectively,where zircons collected from ashdeposits intercalated within the forma-tions have been dated using the U–Pbsystem.

METHODS

Chemical Index of Alteration MethodsThe chemical index of alteration (CIA)is a proxy for evaluating paleoclimaticconditions using ratios of major ele-ments in siliciclastic rocks (Nesbitt andYoung 1982, 1989; Fedo et al. 1995;McLennan et al. 1993; Nesbitt andYoung 1996; Colin et al. 1998; Corco-

ran and Mueller 2002; Scheffler et al.2003), and is expressed as:

(Eq.1 ),

using molar proportions. CaO* repre-sents CaO present in silicate minerals,as opposed to carbonate or phosphateminerals (Nesbitt and Young 1982).CIA is expressed as a dimensionlessnumber between 0 and 100.

X-ray fluorescence (XRF)determination of the major elementcompositions of 63 samples (Li2B4O7-fused glass pellets) was carried out atMichigan State University using aBruker AXS S4 Pioneer instrument.Inorganic carbon content of each sam-ple was determined at NorthwesternUniversity using a UIC CM5012Coulometer. The results were used tocalculate the carbonate-derived calciumcontent (wt %). XRF analyses for theremaining 70 samples were made atActlabs using a Panalytical AxiosAdvanced wavelength dispersive XRFspectrometer. Inorganic carbon con-tent of each sample was determined atActlabs using an Eltra CW-800 analyz-er and the coulometric technique. Theresults were used to calculate the car-bonate-derived calcium content (wt %).For all samples, the dolomite and cal-cite molar ratio was determined usingX-Ray Diffraction (XRD) analysis by aRigaku MiniFlex XRD at PrincetonUniversity.

δ13C MethodsCarbonate rocks were sampled at ~1.0

m resolution from 23 measured strati-graphic sections from across the ARC.Clean dolostone and limestone withoutsiliciclastic components, secondaryveining or cleavage were targeted. Atotal of 2439 samples were slabbedand polished perpendicular to beddingand 5 mg of powder were micro-drilled from individual laminations forisotopic analysis. Note that data from1042 of these samples have been pre-viously published in Rose et al. (2012).At the University of Michigan StableIsotope Laboratory, all powders wereheated to 200ºC to remove volatilecontaminants and water. Samples werethen placed in individual borosilicatereaction vessels and reacted at 76ºCwith 3 drops of H3PO4 in a FinniganMAT Kiel I preparation device cou-pled directly to the inlet of a FinniganMAT 251 triple collector isotope ratiomass spectrometer. δ13C and δ18O datawere acquired simultaneously and arereported in the standard delta notationas the ‰ difference from the VPDBstandard. Measured precision wasmaintained at better than 0.10‰ (1σ)for both δ13C and δ18O. At PrincetonUniversity, all carbonate powders wereheated to 110ºC to remove water. Sam-ples were then placed in individualborosilicate reaction vials and reactedat 72ºC with 5 drops of H3PO4 in aGasBench II preparation device cou-pled directly to the inlet of a ThermoDeltaPlus continuous flow isotoperatio mass spectrometer. δ13C and δ18Odata were acquired simultaneously andare reported in the standard delta nota-tion as the ‰ difference from the

CIA = x 100Al2O3

Al2O3 + K2O + Na2O + CaO*

Figure 5. (previous page) Sedimentology of the syn-glacial Elatina Fm. (a) Coarse grit lenses and trains within a pink slumpedsandstone of the Elatina Fm., Elatina Creek (Fig. 1 [17]). (b) Poorly sorted diamictite with angular and facetted clasts of a rangeof basement lithologies within a red silt matrix, Trezona Bore (Fig. 1 [18]). (c) Ripple cross-laminated sandstone derived fromthe reworking of underlying diamictite, Bunyeroo Gorge (Fig. 1 [15]). (d) Glacially striated green microgabbro clast within redsilt matrix of diamictite, Trezona Bore (Fig. 1 [18]). (e) Plan view of green basalt clasts forming lag deposits within red siltmatrix of diamictite, Bulls Gap (Fig. 1 [22]). (f) Plan view of lag within coarse, poorly sorted, pink slumped sandstone, TrezonaBore (Fig. 1 [18]). (g) Elatina Fm. diamictite resting unconformably on tidal flat sandstone of the Yaltipena Fm. as a result ofice-contact deposition, Trezona Bore (Fig. 1 [18]). (h) Soft-sediment deformation in the upper sandstone of the Elatina Fm.,Trezona Bore. (i) Sub-glacial push structure lying unconformably on the Yaltipena Fm. (right of image) at Trezona Bore. Notehammer for scale (circled). The scoured basal contact and contorted diamictite beds indicate local ice-contact deposition. (j)Wave ripples within the upper Elatina Fm., Warren Gorge (Fig. 1 [4]). (k) Plan view of geometric bifurcations of secondary rip-ples, Warren Gorge (Fig. 1 [4]). (l) Cross-section through large-scale ripple showing lamination bundles and couplets in thesouthern Flinders Ranges, Warren Gorge (Fig. 1 [4]). (m) Plan view of grainflow originating from crest of a ladder ripple, War-ren Gorge (Fig. 1 [4]). (n) Cross-section of lamination couplets in the northern Flinders Ranges, Oodnaminta Hut (Fig. 1 [33]).(o) Granitoid clast within microbialite bioherm of the Trezona Fm. at Punches Rest (Fig. 1 [36]). (p) Trezona Fm. fossiliferouspackstone clast within the glacial diamictite of the Elatina Fm., near Oodnapanicken Bore (Fig. 1 [32]). (q) Diamictite of theCurtis Tillite with quartzite, granite, basalt, and rare dolomite clasts, near Mt. Curtis (Fig. 1 [44]). (r) Diamictite reworked bygraded debris flows within the Elatina Fm., Billy Springs (Fig. 1 [47]).

266

VPDB standard. Precision and accura-cy of data are monitored throughanalysis of 21 standards that are runfor every 59 samples. Measured preci-sion is maintained at better than0.10‰ (1σ) for both δ13C and δ18O.

Geochronology MethodsTwelve zircon concentrates were sepa-rated from 2 to 4 kg of sample materi-al at the Senckenberg NaturhistorischeSammlungen Dresden using standardmethods. Zircon grains of all grainsizes and morphological types werehand-picked, mounted and analyzedfor U, Th, and Pb isotopes byLA–ICP–MS techniques at the Muse-um für Mineralogie und Geologie(GeoPlasma Lab, SenckenbergNaturhistorische Sammlungen Dres-den), using a Thermo-Scientific Ele-ment 2 XR sector field ICP–MS cou-pled to a New Wave UP–193 ExcimerLaser System. U–Th–Pb analyses wereconducted for 15 s background acqui-sition followed by 30 s data acquisition,using a laser spot-size of 25 and 35µm, respectively. A common-Pb cor-rection based on the interference- andbackground-corrected 204Pb signal anda model Pb composition (Stacey andKramers 1975) was carried out if nec-essary. The necessity of the correctionwas judged on whether the corrected207Pb/206Pb lay outside of the internalerrors of the raw, measured ratios. Rawdata were corrected for backgroundsignal, common Pb, laser-induced ele-mental fractionation, instrumentalmass discrimination, and time-depend-ent elemental fractionation of Pb/Thand Pb/U using an Excel spreadsheetprogram developed by Axel Gerdes(Institute of Geosciences, JohannWolfgang Goethe-University Frankfurt,Frankfurt am Main, Germany). Report-ed uncertainties were propagated byquadratic addition of the externalreproducibility obtained from the ref-erence zircon ‘GJ-1’ (~0.6% and 0.5-1% for the 207Pb/206Pb and 206Pb/238U,respectively) during individual analyti-cal sessions and the within-run preci-sion of each analysis. All uncertaintiesare quoted at the 95% confidence or2σ level. For further details on analyti-cal protocol and data processing, seeGerdes and Zeh (2006).

Eleven zircon concentrateswere separated from 2 to 4 kg of sam-

ple material at Princeton Universityusing standard methods. These con-centrates were annealed in an oven at900ºC for 3 days before being mount-ed in resin blocks and polished to halftheir thickness. The zircon grains wereanalyzed for U, Th and Pb isotopesusing a Laser Ablation Multi-CollectorInductively Coupled Plasma MassSpectrometer (LA–MC–ICP–MS) sys-tem housed at the University of Cali-fornia, Santa Barbara. Instrumentationconsists of a Nu Plasma MC–ICP–MS(Nu Instruments, Wrexham, UK) and a193 nm ArF laser ablation system(Photon Machines, San Diego, USA).Analytical protocol is similar to thatdescribed by Cottle et al. (2009a,b,c).U-Th-Pb analyses were conducted for15 s each using a spot diameter of 24µm, a frequency of 4 Hz and 1.2 J/cm2

fluence (equating to crater depths ofapproximately 4 µm). U–Th–Pb datafrom 4 samples were collected over 4days of continuous instrument opera-tion. A primary reference material,‘91500’ zircon (1065.4 ± 0.3 Ma207Pb/206Pb ID–TIMS age and 1062.4 ±0.4 Ma 206Pb/238U ID–TIMS age(Wiedenbeck et al. 1995)) wasemployed to monitor and correct formass bias as well as Pb/U and frac-tionation. To monitor data accuracy, asecondary reference zircon ‘GJ-1’(608.5 ± 0.4 Ma 207Pb/206Pb ID–TIMSage (Jackson et al. 2004) and 601.7 ±1.3 Ma 206Pb/238U ID–TIMS age) wasanalyzed concurrently (once for every5–7 unknown samples) and correctedfor mass bias and fractionation basedon measured isotopic ratios of the pri-mary reference material. Analyses ofthe GJ-1 secondary reference zirconduring the analytical period yielded aweighted mean 206Pb/238U age of 603.9± 0.6, MSWD = 0.9. Data reductionwas carried out using Iolite version2.1.2 (Paton et al. 2010). All uncertain-ties are quoted at the 95% confidenceor 2σ level and include contributionsfrom the external reproducibility ofthe primary reference material for the207Pb/206Pb and 206Pb/238U ratios.

RESULTS

Sedimentology

Pre-glacial Trezona and YaltipenaFormationsIn the central Flinders Ranges, the

Elatina Fm. unconformably overliesthe Trezona Fm. and the Yaltipena Fm.of the Upalinna Subgroup (Lemon andGostin 1990; Lemon and Reid 1998).The Trezona Fm., as defined by Dal-garno and Johnson (1964), is mainlyrestricted to, and attains its maximumthickness, around the central anticlineof the Flinders Ranges (Fig. 6). Palered and grey thin interbedded silt,microbialite, and intraclastic brecciacharacterize the lower part of the Tre-zona Fm. These decimetre-thickparasequences typically contain guttercasts and rip-up clasts, indicating astorm-dominated shelf environment(Myrow 1992). Up-section, microbialitetransitions into stromatolitic facieswith stromatolite flake breccia and bio-clast packstone filling the spacebetween stromatolite heads with up toone metre of synoptic relief (Rose etal. 2012). The upper Trezona is domi-nated by large stromatolitic moundsand oolitic and intraclastic breccialimestone. Stromatolites within theTrezona Fm. commonly are large,round mounds up to several metres inheight, but can be elongate lobatemounds with cuspate valleys betweenthem (Fig. 4e). To the north, the Tre-zona Fm. consists of cm-thick lime-stone laminite deposited below stormwave base that alternate with greenish-grey laminated calcareous shale andsiltstone (Fig. 4g). The amount of silt-stone between these ‘ribbonite’ pack-ages increases to the north, until atTower Gap the Trezona Fm. reaches~600 m in thickness with barely anylimestone beds (Fig. 7 [38]). To thesouth, the Trezona Fm. interfingerswith and transitions into the partiallycorrelative Yaltipena Fm (Fig. 8).

The Yaltipena Fm. was firstrecognized and distinguished from theoverlying Elatina Fm. by Dalgarno andJohnson (1964), leading to Bulls Gapbeing nominated as its type section(Lemon and Reid 1998). Previously,Lemon and Reid (1998) determinedthat the Yaltipena Fm. commenceswith a transgressive, red intraclasticconglomerate, comprised of flat dis-coidal pebbles of lithified micrite in amatrix of sand-sized intraclasts that isin disconformable contact with under-lying stromatolitic, oolitic and intraclas-tic limestones of the Trezona Fm.However, this nominated bed is indis-


δ1

3C

carb

‰

01

0-1

05

-5

50

75

10

0C

IA

δ1

3C

carb

‰

01

0-1

05

-5

50

75

10

0C

IA

δ1

3C

carb

‰

01

0-1

05

-5

50

75

10

0C

IA

10

0

20

0

50

0

60

0

70

0

80

0

90

0

0

10

0

20

0

30

0

10

0

20

0

30

0

40

00

80

0

90

0

10

00

0

10

0

20

0

11

00

12

00

13

00

14

00

30

0

80

0

90

0

10

00

11

00

12

00

13

00

14

00

TREZONA FM

ETINA FM

ELATINA FM

NU

CCA

LEEN

A F

M

ENO

RAM

AFM

YALTIPENA FM

δ13

Cca

rb (‰

VP

DB

)

carb

onat

e un

itssi

licic

last

ic u

nits

mic

rob

ialit

e /

intr

acla

st b

recc

iap

acks

ton

eg

rain

sto

ne

oo

lite

stro

mat

olit

eri

bb

on

ite

mar

l

san

dst

on

e

LITHOFACIES

limes

ton

e

silt

sto

ne

shal

e

Nu

ccal

een

a ca

p d

olo

sto

ne

gla

cial

dia

mic

tite

rst

app

eara

nce

of T

rezo

na

foss

ilsC

hem

ical

Ind

ex o

f Alt

erat

ion

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

δ1

3C

carb

‰

01

0-1

05

-5

50

75

10

0C

IA

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

δ1

3C

carb

‰

01

0-1

05

-5

50

75

10

0C

IA

Emu

Gap

Elat

ina

Cre

ek

Trez

on

a B

ore

Mo

olo

olo

o

0

10

0

20

0

30

0

40

0

50

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

40

0

Bu

lls G

ap

meter level

STR

ATI

GR

AP

HIC

SEC

TIO

NS

IN T

HE

CEN

TRA

L FL

IND

ERS

RA

NG

ES

NO

RTH

SOU

TH

0

100

0

100

200

0

100

200

300

Ben

net

t Sp

rin

gs

Do

nke

y G

ully

Wel

l

An

go

rich

ina

War

cow

ie

13

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

0

10

0

20

0

19

20

22

23

24

inec

tion

po

int

Trez

ona

δ13C

an

omal

y

BRACHINA FM

BRACHINA FM

16

17

18

31°4

0’

!"!

#$

138°

30’

138°

40’

138°

50’

139°

0’

31°1

0’

31°2

0’

138°

30’

138°

40’

138°

50’

139°

0’

16

18

19

2022

23

24

17

21

13

1415

N

Eno

ram

a Fm

Etin

a Fm

un

di

eren

tiat

ed u

nit

s

Bra

chin

a Fm

Elat

ina

FmTr

ezo

na

Fm

Nu

ccal

een

a Fm

Fig

ure

6.S

elect

ion

of st

ratig

raph

ic se

ctio

ns fr

om th

e ce

ntra

l ant

iclin

e of

the

Ade

laide

Rift

Com

plex

(ARC

) num

bere

d to

cor

resp

ond

with

Fig

ure

1. P

re-g

lacial

δ13C

data

with

in th

e E

tina

and

Trez

ona

Fms.

are

deno

ted

by so

lid b

lue

circle

s and

δ13C

data

with

in th

e po

st-g

lacial

Nuc

calee

na F

m. c

ap d

olos

tone

are

den

oted

by

open

blue

circ

les (R

ose

and

Malo

of 2

010)

. All

sect

ions

are

hun

g fr

om a

dat

um c

hose

n at

the

infle

ctio

n po

int a

t the

bas

e of

the

reco

very

from

the

Trez

ona

nega

tive

δ13C

anom

aly. T

he Y

altip

ena

Fm. t

hins

unt

il it

is m

issin

g en

tirely

at t

he W

arco

wie

[13]

, Elat

ina

Cree

k [1

7], A

ngor

ichin

a [2

3], a

nd M

oolo

oloo

sect

ions

[24]

. Diff

eren

tial

glac

ial e

rosio

n tr

unca

ted

the

Yalti

pena

Fm

. dur

ing

the

ice a

dvan

ce. F

urth

erm

ore,

the

δ13C

data

indi

cate

trun

catio

n of

the

repr

oduc

ible

carb

on is

otop

ic tre

nd in

the

Trez

ona

Fm.,

ther

eby

quan

tifyin

g th

e de

gree

and

late

ral v

ariab

ility

of

glac

ial e

rosio

n. N

ote

that

the

Eno

ram

a Fm

. con

sists

of

shale

with

no

asso

ciate

d δ13

C da

ta a

ndth

eref

ore

is no

t sho

wn

in th

e st

ratig

raph

ic se

ctio

ns fo

r clar

ity. T

he a

vera

ge th

ickne

ss o

f th

e E

nora

ma

Fm. i

s ~30

0 m

. The

loca

tion

of th

e m

appe

d ar

ea w

ithin

the

ARC

is o

utlin

ed b

y a

grey

box

in F

igur

e 1.

268

Gammon Ranges

31o

30o

139o

140o

30

36

26

28

27

31

32

33

35

38

39

40

41

42

43

44

45

46

47

29

37

34

N

25 k

m

TREZONAFM

ELAT

INA

FM ETINA

FMmeter level

Tayl

or

Cre

ek

Oo

dn

apan

icke

n

Bo

re

Pu

nch

es R

est

0

40

0

20

0

10

0

01

0-1

05

-5

δ1

3C

carb

‰

Nan

nip

inn

a C

reek

0

40

0

30

0

20

0

10

0

50

0

60

0

70

0

80

0

0

40

0

30

0

20

0

10

0

50

0

60

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

0

40

0

30

0

20

0

10

0

50

0

60

0

70

0

80

0

90

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

0

30

0

20

0

10

0

Loo

kou

t H

ill

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

Um

ber

atan

a

0

20

0

10

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

Wee

too

tla

29

0

40

0

30

0

20

0

10

0

50

0

60

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IAN

UCC

ALE

ENA

FM

Win

yag

un

na

Idin

ha

Sp

rin

g

Tow

er G

ap

ST

RA

TIG

RA

PH

IC S

ECT

ION

S IN

TH

E N

OR

TH

ERN

FLI

ND

ERS

RA

NG

ES

NO

RT

H

SO

UT

H

0

30

0

20

0

10

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

0

40

0

30

0

20

0

10

0

50

0

60

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

0

40

0

30

0

20

0

10

0

50

0

60

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

30

0

50

0

60

0

70

0

80

0

40

0

50

0

60

0

70

0

80

0

90

0

10

00

11

00

ENORAMA FM?

δ1

3C

carb

(‰ V

PD

B)

carb

onat

e un

itssi

licic

last

ic u

nits

pac

ksto

ne

gra

inst

on

eo

olit

est

rom

ato

lite

rib

bo

nit

em

arl

san

dst

on

e

LITHOFACIES

limes

ton

e

silt

sto

ne

shal

e

Nu

ccal

een

a ca

p d

olo

sto

ne

gla

cial

dia

mic

tite

rst

app

eara

nce

of T

rezo

na

foss

ils

cove

r

Ch

emic

al In

dex

of A

lter

atio

n

inec

tion

po

int

Trez

ona

δ13C

an

omal

y

31

32

34

35

36

37

38

39

40

50

75

10

0C

IA

TAPLEY HILL FM

mic

rob

ialit

e /

intr

acla

st b

recc

ia

/ te

pee

bre

ccia

Fig

ure

7.S

tratig

raph

ic se

ctio

ns o

f th

e no

rther

n Fl

inde

rs R

ange

s num

bere

d to

cor

resp

ond

with

Fig

ure

1. P

re-g

lacial

δ13C

data

with

in th

e E

tina

and

Trez

ona

Fms.

are

deno

ted

by so

lid b

lue

circle

s and

δ13C

data

with

in th

e po

st-g

lacial

Nuc

calee

na F

m. c

ap d

olos

tone

are

den

oted

by

open

blu

e cir

cles (

Rose

and

Malo

of 2

010)

. All

sect

ions

are

hun

g fr

om a

dat

um c

hose

n at

the

infle

ctio

n po

int a

t the

bas

e of

the

reco

very

from

the

Trez

ona

nega

tive

δ13C

anom

aly. F

ewer

δ13C

data

poi

nts i

n th

isre

gion

mak

e co

rrela

tions

tent

ativ

e w

hen

usin

g th

is δ13

C in

flect

ion

poin

t. Th

e E

nora

ma

Shale

is o

f un

know

n th

ickne

ss w

ith c

onta

cts d

eter

min

ed b

y th

e δ13

C va

lues

of th

e un

derly

ing

Etin

a Fm

. and

ove

rlyin

g Tr

ezon

a Fm

. The

loca

tion

of th

e m

appe

d ar

ea w

ithin

the

Ade

laide

Rift

Com

plex

is o

utlin

ed b

y a

grey

box

in F

igur

e 1.


tinguishable from many of the intra-clastic beds within the Trezona Fm.and the disconformity is not laterallycontinuous. We propose the base ofthe red siltstone to represent the begin-ning of the Yaltipena Fm.

The Yaltipena Fm. (<100 m)is a coarsening upwards sequence, withred siltstone grading upwards intovery-fine sandstones that transition tomedium-grained sandstones upsection(Lemon and Reid 1998). The base ofthe thick siltstone interval is character-ized by low-amplitude starved ripplesand lenticular bedding. Further up sec-tion, these beds become generally well-laminated and show many indicationsof shallow water, including small wave-length (~20 mm) symmetrical and lin-guoid ripples, planed-off ripples, ladderripples, and interference ripples (Fig.4d). Abundant desiccation cracks andpossible raindrop impressions indicateintermittent subaerial exposure (Fig.4b). Thin channels of clean, whitequartz sandstone with mud rip-upclasts at the bases are the first indica-tion of the coarser interval above thesiltstone (Fig. 4c). These coarser sand-stones are both planar and troughcross-bedded with foresets between0.05-0.2 m thick. Glacial clasts up to 1m across pierce the top of theYaltipena Fm. and the upper contacttypically is contorted and scoured (Fig.5g, i).

In the southern FlindersRanges, the Yaltipena Fm. (referred tolocally as the Wilmington Fm.) consistsof thinly bedded red siltstone, withmudchips, desiccation cracks, and shortwavelength interference ripples, but theformation reaches a thickness of >500m (Fig. 8). Near the base of the forma-tion, rare discontinuous metre-scalestromatolite bioherms are present thatlikely are correlative to the Etina Fm.Towards the top of the formation, thesiltstone layers coarsen up into grey-green and red-brown, fine to medium-grained well-sorted, planar cross-bed-ded quartzite, sandstone, sandy silt-stone, and minor arkose. The YaltipenaFm. progressively thins towards thecentral anticline and at Bulls Gap thebasal contact is transitional as the redsiltstone interfingers with the intraclastlimestone of the upper few metres ofthe Trezona Fm. (Fig. 6 [22]). There-fore, the Yaltipena Fm. was being

5 k

m

N

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

TREZONA FM

ETINA FM

ELATINA FM

NU

CCA

LEEN

A F

M

ENORAMA FM

War

ren

Go

rge

4C

ock

roac

h V

alle

y

6

Stev

e’s

Go

rge

7

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

Par

taco

on

a

meter level

STRA

TIG

RAPH

IC S

ECTI

ON

S IN

TH

E SO

UTH

ERN

FLI

ND

ERS

RAN

GES

NO

RTH

SOU

TH

Bu

ckar

ing

a G

org

e

9P

etta

na

Cre

ekM

idd

le G

org

e1

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

0

10

0

20

0

30

0

0

10

0

20

0

30

0

40

0

50

0

60

0

0

10

0

20

0

0

10

0

20

0

30

0

40

0

50

0

60

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

0

10

0

20

0

30

0

40

0

0

10

0

20

0

30

0

40

0

50

0

60

0

0

10

0

20

0

01

0-1

05

-5

δ1

3C

carb

‰

50

75

10

0C

IA

δ1

3C

carb

(‰ V

PD

B)

carb

onat

e un

itssi

licic

last

ic u

nits

mic

rob

ialit

e /

intr

acla

st b

recc

iap

acks

ton

eg

rain

sto

ne

oo

lite

stro

mat

olit

eri

bb

on

ite

mar

l

san

dst

on

eLITHOFACIES

rhyt

hm

ite

limes

ton

esi

lt

Nu

ccal

een

a ca

p d

olo

sto

ne

gla

cial

dia

mic

tite

first

ap

pea

ran

ce o

f Tr

ezo

na

foss

ils

gri

tg

rain

tra

ins

pla

nar

lam

inat

ion

scr

oss

lam

inat

ion

str

ou

gh

cro

ss-b

edd

ing

Ch

emic

al In

dex

of A

lter

atio

n

YALTIPENA FM

TAPLEY HILL FM

11

12

32°4

0’

32°2

0’

32°0

’

010

km

138°

0’N

12

11

87

6

54

3

2

91

0 1En

ora

ma

Fm

Etin

a Fm

un

diff

eren

tiat

ed u

nit

s

Bra

chin

a Fm

Elat

ina

FmTr

ezo

na

Fm

Nu

ccal

een

a Fm

Fig

ure

8.S

elect

ion

of st

ratig

raph

ic se

ctio

ns fr

om th

e so

uthe

rn F

linde

rs R

ange

s num

bere

d to

cor

resp

ond

with

Fig

ure

1. P

re-g

lacial

δ13C

data

with

in th

e E

tina

and

Trez

ona

Fms.

are

deno

ted

by so

lid b

lue

circle

s and

δ13C

data

with

in th

e po

st-g

lacial

Nuc

calee

na F

m. c

ap d

olos

tone

are

den

oted

by

open

blu

e cir

cles (

Rose

and

Mal-

oof

2010

). A

ll se

ctio

ns a

re h

ung

from

a d

atum

at t

he b

ase

of th

e N

ucca

leena

Fm

. The

loca

tion

of th

e m

appe

d ar

ea w

ithin

the

Ade

laide

Rift

Com

plex

is o

utlin

edby

a g

rey

box

in F

igur

e 1.

270

deposited in the south during deposi-tion of the Enorama and TrezonaFms. in the central anticline (Fig. 8).

Syn-glacial Elatina Formation

Central Flinders RangesA thin basal conglomerate and overly-ing massive boulder diamictite up to 5m thick are present locally in the cen-tral Flinders Ranges. These discontinu-ous diamictite beds at the base of theElatina Fm. are contorted, have ascoured basal contact, and containlarge (~1 m) extrabasinal boulders ofgranitic gneiss that pierce the top ofthe underlying Yaltipena Fm. (Lemonand Gostin 1990; Fig. 5g-i; Fig. 9 [18,19, 22]). Some of the clasts in thediamictite beds are faceted and striated,attesting to their glacially influenceddeposition, with the striations typicallyoriented parallel to the long-axes ofthe clasts (Mawson 1949; Dalgarno andJohnson 1964; Preiss 1987; Lemon andGostin 1990; Fig. 5b, d). This clastsuite consists predominantly of basalt,well-rounded quartzite, and graniticgneiss.

The remainder of the glacialfacies overlying these discontinuousdiamictite beds can be traced acrossthe central anticline of the FlindersRanges (Fig. 1), and can be describedby three distinct facies (Lemon andGostin 1990): 1) a lower slumped sand-stone; 2) a middle interval of drop-stone diamictites; and 3) an upperinterval of current reworked diamictite.The first facies consists of a suite ofcoarse, slumped sandstone beds thatmay directly overlie the basal uncon-formity at the base of the Elatina Fm.Approximately 5 m of channeled,cross-bedded, coarse-grained sand-stone are present at the base of theunit (Lemon and Gostin 1990). Thissandstone grades upwards into flaser-bedded, muddy sandstone, and is over-lain by several 1 m beds with large ball-and-pillow structures (Lemon andGostin 1990; Fig. 5h). Above this softsediment deformation, lies a ~50 mthick interval of white, pink- to red-dish-brown, poorly sorted feldspathicsandstone (Facies 1). These feldspathicsandstone beds contain sub-angular tosub-rounded quartz and feldspargrains, plus lithic fragments and heavymineral grains (Preiss 1987), and gener-

Buny

eroo

Lo

okou

tD

raft

ing

Yard

s

STRA

TIG

RAPH

IC S

ECTI

ON

S TH

ROU

GH

ELA

TIN

A F

M, C

ENTR

AL

FLIN

DER

S

40200

106

m

Brac

hina

>1 k

m

Trez

ona

- Etin

a6040200

73 m

Tr

ezon

a

160

140

120

100806040200

90 m

Ya

ltipe

na

120

100806040200

110

mYa

ltipe

na

6040200

50 m

Yalti

pena

100806040200

430

mYa

ltipe

na- T

rezo

na

6040200

35 m

Yalti

pena

- Tre

zona

Ark

aba

40200

100

mYa

ltipe

na

Emu

Gap

Elat

ina

Cree

kTr

ezon

a Bo

reM

oolo

oloo

Bulls

Gap

Benn

ett S

prin

gsD

onke

y G

ully

Wel

lA

ngor

ichi

naW

arco

wie

13

EAST

ERN

FL

IND

ERS

Pana

ram

atee

carb

onat

e un

its

silic

icla

stic

uni

ts

mic

robi

alite

/ in

trac

last

bre

ccia

pack

ston

egr

ains

tone

oolit

est

rom

atol

iterib

boni

tem

arl

sand

ston

e

LITH

OFA

CIES

rhyt

hmite

silts

tonegl

acia

l dia

mic

tite

grit

grai

n tr

ains

plan

ar la

min

atio

nscr

oss

lam

inat

ions

trou

gh c

ross

-bed

ding

granite

basa

lt

quartzite

sandstone

cher

t

limestone

dolostone

silt

gnei

ssdi

amic

tite

mea

n cl

ast

orie

ntat

ion

clast lithology

slum

ped

sand

ston

e

Elat

ina

Fm fa

cies

drop

ston

e di

amic

tite

rew

orke

d di

amic

tite

turb

idite

s

bedd

ed d

iam

ictit

e

Faci

es 1

Faci

es 2

Faci

es 3

120

100806040200

150

mYa

ltipe

na- T

rezo

na

806040200

400

mYa

ltipe

na- T

rezo

na

ELATINA FM

NUCC

ALEE

NA F

M

806040200

200

mYa

ltipe

na- T

rezo

na

1 km

Br

achi

na

320

280

240

200

160

12080400

360

1415

1617

1819

2021

2223

2425

100806040200

75 m

Trez

ona

31°4

0’

010

km

138°

30’

138°

40’

138°

50’

139°

0’

31°1

0’

31°2

0’

138°

30’

138°

40’

138°

50’

139°

0’

16

18

19

2022

23

24

17

21

13

1415

N

Enor

ama

FmEt

ina

Fmun

diffe

rent

iate

d un

its

Brac

hina

Fm

Elat

ina

FmTr

ezon

a Fm

Nuc

cale

ena

Fm

Fig

ure

9.D

etail

ed st

ratig

raph

ic se

ctio

ns o

f th

e E

latin

a Fm

. acr

oss t

he c

entra

l Flin

ders

Ran

ges n

umbe

red

to c

orre

spon

d w

ith F

igur

e 1.

The

Elat

ina

Fm. c

an b

ede

scrib

ed b

y th

ree

dist

inct

facie

s (Le

mon

and

Gos

tin 1

990)

: 1) a

low

er, c

oars

e, slu

mpe

d sa

ndst

one

with

con

torte

d gr

anul

e lay

ers a

nd tr

ough

cro

ss-b

eddi

ng; 2

) am

iddl

e in

terv

al of

well

-lam

inat

ed re

d sil

tsto

ne d

rops

tone

diam

ictite

; and

3) a

n up

per i

nter

val o

f cu

rren

t-rew

orke

d di

amict

ite w

ith d

istin

ctiv

e gr

avel

lags a

nd ri

pple

cros

s-lam

inat

ions

. The

ove

rlyin

g N

ucca

leena

Fm

. cap

dol

osto

ne v

aries

from

buf

f cr

oss-

stra

tified

dol

omite

gra

inst

one

with

m-s

cale

gian

t wav

e rip

ples

to re

d sil

tydo

lom

ite ri

bbon

ite w

ith lo

w-a

ngle

cros

s-st

ratif

ied la

min

atio

ns (R

ose

and

Malo

of 2

010)

. The

‘rib

bon’

facie

s con

sists

of

swale

y do

losil

tite.

All

sect

ions

are

hun

g fr

oma

datu

m a

t the

bas

e of

the

Nuc

calee

na F

m. T

he lo

catio

n of

the

map

ped

area

with

in th

e A

delai

de R

ift C

ompl

ex is

out

lined

by

a gr

ey b

ox in

Fig

ure

1.


ally are bimodal, marked by a domi-nance of coarse silt to fine sand andvery coarse sand to granule fractions.The sandstones lack distinct beddingbut have thin, discontinuous, and com-monly contorted granule layers (Fig.5a). The less deformed intervals con-tain slumped, trough cross-bedded setsup to 1 m thick with normal grading.Rare isolated pebbles to large over-sized clasts also may be present inthese sandstone layers, typically associ-ated with dolomite, quartz and granitefragments.

The slumped sandstone bedsgrade upward into a massive to well-laminated red siltstone with a fewdropstones (Facies 2). These diamictitebeds are dominated by carbonate claststhat have been presumed to originatefrom the underlying Trezona Fm., andbasalt clasts (Lemon and Gostin 1990).Finally, current-reworked diamictitebeds dominate the top third of theElatina Fm. (Facies 3). These diamic-tites have undergone significantreworking, creating prominent beds ofdistinctive gravel lags and lonestonescapped by mm- to cm-scale ripplecross-laminated fine- to very fine-grained sandstone within an otherwisemassive unit (Fig. 5c, f). Sections meas-ured at the northern rim of the centralanticline and between Elatina Creekand Trezona Bore record anotherdiamictite at the top of the formationwithin the current-reworked interval(e.g. Moolooloo; Fig. 9 [24]). Thesebeds contain basalt clasts and, in con-trast to previous observations byLemon and Gostin (1990) andWilliams et al. (2008), we did not noteclasts of ooid and algal limestone fromthe Trezona Fm. (Fig. 5e).

North Flinders RangesDespite variations in thickness of eachfacies, the total thickness of the Elati-na Fm. remains relatively uniform(~70–100 m) in the central anticline.However, to the east and north, theElatina Fm. thickens to >300 m andeventually the diamictites transitioninto stratified debris flows and tur-bidites (Fig. 5q, r; Fig. 10). At PunchesRest, isolated rounded ~0.2–1.0 mdiameter granite and gabbro clasts areincorporated within the upper twometres of a Trezona Fm. stromatolitebioherm that records the highest δ13C

Idin

ha S

prin

gO

odna

pani

cken

Bo

reN

anni

pinn

a Cr

eek

Punc

hes

Rest

Mt.

Lynd

hurs

tTa

ylor

Cre

ekW

adm

ore

Mt.

Curt

is

200

40200

160

140

120

100806040200

160

140

120

100806040200

100806040200

6040200

160

140

120

100806040200

160

140

120

100806040200

180

160

140

120

100806040200

180

Bend

Wel

l

160

140

120

100806040200

180

260

220

280

240

200

Mt.

Broo

kBi

lly S

prin

gsU

mbe

rata

naFo

rtre

ss H

illM

ayna

rds W

ell

Nob

bler

s Wel

lO

odna

min

taH

ut

120

100806040200

806040200

6040200

6040200

Cham

bers

G

orge

Mt.

McT

agga

rt

200

200

806040200

40200

190

m

Trez

ona

45 m

Br

achi

na

11 m

Tr

ezon

a

47 m

Br

achi

na

30 m

Tr

ezon

a

60 m

Tr

ezon

a

6 m

Tr

ezon

a

800

m

Trez

ona

- Et

ina

815

m

Trez

ona

- Et

ina

50 m

Br

achi

na12

0 m

Br

achi

na

135

m

Trez

ona

270

m

Fort

ress

30 m

Br

achi

na

300

m

Fort

ress

60 m

Fo

rtre

ss

35 m

Fo

rtre

ss

18 m

Br

achi

na

100

m

Brac

hina

carb

onat

e un

its

silic

icla

stic

uni

ts

mic

robi

alite

/ in

trac

last

bre

ccia

pack

ston

egr

ains

tone

oolit

est

rom

atol

iterib

boni

tem

arl

sand

ston

e

LITH

OFA

CIES

rhyt

hmite

silts

tonegl

acia

l dia

mic

tite

grit

grai

n tr

ains

plan

ar la

min

atio

nscr

oss

lam

inat

ions

trou

gh c

ross

-bed

ding

granite

basa

lt

quartzite

sandstone

cher

t

limestone

dolostone

silt

gnei

ssdi

amic

tite

mea

n cl

ast

orie

ntat

ion

clast lithology

Elat

ina

Fm fa

cies

turb

idite

s

bedd

ed

diam

ictit

ela

min

ated

silt

sand

ston

e

STRA

TIG

RAPH

IC S

ECTI

ON

S TH

ROU

GH

ELA

TIN

A F

M, N

ORT

HER

N F

LIN

DER

S30

36

ELATINA FM

NUCC

ALEE

NA F

M

6040200

Tow

er G

ap26

2827

3132

3335

3839

4041

4243

44

Gammon Ranges

31o

30o

139o

140o

4546

47

30

36

26

28 27

3132

3335

38

39

4041

42

4344

4546

47

29

3734

186

m

Trez

ona

N

25 k

m

Fig

ure

10.

Det

ailed

stra

tigra

phic

sect

ions

of

the

Elat

ina

Fm. a

cros

s the

nor

ther

n Fl

inde

rs R

ange

s num

bere

d to

cor

resp

ond

with

Fig

ure

1. A

ll se

ctio

ns a

re h

ung

from

a d

atum

at t

he b

ase

of th

e N

ucca

leena

Fm

. The

loca

tion

of th

e m

appe

d ar

ea w

ithin

the

Ade

laide

Rift

Com

plex

is o

utlin

ed b

y a

grey

box

in F

igur

e 1.

272

following the Trezona anomaly (Roseet al. 2012; Fig. 5o; Fig. 7 [36]). Thislayer is laterally discontinuous on thekm-scale and has not been document-ed elsewhere.

Overall, there are three broadfacies to the Elatina Fm. in the northFlinders Ranges: a laminated siltstone,a diamictite, and a coarse feldspathicsandstone. These three facies are later-ally widespread but are not always allpresent at every locality. The first faciesconsists of grey-green (weathering red-dish-brown), finely laminated siltstonewith rare scattered pebbles, cobbles,and gritty lenses. Known as theFortress Hill Fm., this siltstone is over-lain sharply by sandstone and a diamic-tite layer. This diamictite is the secondfacies, referred to locally as the MountCurtis Tillite, and consists of sparselonstones of pebble- to boulder-size ina grey-green, massive and laminated,sandy siltstone matrix. Diamictite clastlithologies are mostly quartzite, lime-stone and dolostone, and less com-monly granite and gneiss, and the longaxes of clasts retain a general east-westorientation (Fig. 10 [31]). Some of theclasts are faceted and striated, and raregranite boulders may reach up to 3 m x8 m (Williams et al. 2008). Sub-angularlimestone packstone clasts of the Tre-zona Fm. (~20 cm), which contain dis-tinctive skeletal fossil debris (Maloof etal. 2010), are present as clasts withinthe Elatina Fm. diamictite near Punch-es Rest and Oodnapanicken Bore (Fig.5f). The third facies consists of palegrey and brownish grey, medium-grained, feldspathic sandstone withinterbeds and lenses of calcareous silt-stone and pebble conglomerate andoverlies the Mt. Curtis Tillite (referredto locally as the Balparana Sandstone).The clasts are mostly of vein quartz,with some quartzite, siltstone and gran-ite clasts. At Billy Springs, which is themost northern locality within the ARC,the Elatina Fm. consists of decimetre-to metre-thick reverse-graded turbiditesthat coarsen upwards from planar lami-nated fine sand to coarse sand withquartzite and sandstone clasts (Fig. 5q,r; Fig. 10 [47]).

South Flinders RangesIn the southern Flinders Ranges, thebase of the Elatina Fm. is not alwaysclear, but was taken to coincide with

the first influx of lithicgranules (Fig. 11). The for-mation can be split into alower member of grey, finesandstone with granuletrains, a middle member ofpurple siltstone and finesandstone with scatteredcoarse grains, and an uppermember of pale grey, finesandstone with troughcross-bedding (Jablonski1975; Miller 1975). Themiddle purple siltstonemember commonly gradesinto a planar laminated unitwithin the upper 50 m ofthe Elatina Fm. This faciesoutcrops almost continu-ously between Saltia Creekto Buckaringa Gorge,although the best exposureis at Warren Gorge wherethe unit is ~30 m thick.The following descriptionsare based on observationsmade at Warren Gorge(Fig. 1 [4]).

The planar lamina-tions consist of ~1–2 cm-thick bundles of thickeningand thinning, normallygraded very fine sandstoneto siltstone couplets (Fig.5l, n). The number of cou-plets within each bundlevaries, however, fifteen isthe maximum number perbundle. These rhythmitesdocument a nested hierar-chy of periodicities consis-tent with the number andrelative thickness of cou-plets that are interpreted tobe tidal in origin (Williams1989, 1991, 1998, 2000).Superimposed on theselaminae throughout thestratigraphy are three cate-gories of bedforms, whichare herein referred to asprimary, secondary, and ter-tiary bedforms. The pri-mary bedforms have thelargest wavelength, typicallybetween 20 and 40 cm, anda mean amplitude of 1.3cm (n=44) (Fig. 12a). Thebedforms are very slightlyasymmetrical with a pale-

Gra

ham

’s Cr

eek

200

mBr

achi

na

silic

icla

stic

uni

ts

sand

ston

e

LITH

OFA

CIES

rhyt

hmite

silts

tone

glac

ial d

iam

ictit

e

grit

grai

n tr

ains

plan

ar la

min

atio

nscr

oss

lam

inat

ions

trou

gh c

ross

-bed

ding

cove

r

slum

ped

sand

ston

e

Elat

ina

Fm fa

cies

rew

orke

d sa

ndst

one

rhyt

hmite

s

rhyt

hmite

sam

ples

carb

onat

e un

itsgr

ains

tone

ribbo

nite

War

ren

Gor

ge1

Cock

roac

h Va

lley

2St

eve’

s G

orge

3Pa

rtac

oona

7Bu

ckar

inga

G

orge

4Pe

ttan

a Cr

eek

6M

iddl

e G

orge

5

100806040200

STRA

TIG

RAPH

IC S

ECTI

ON

S TH

ROU

GH

ELA

TIN

A F

M, S

OU

THER

N F

LIN

DER

S

ELATINA FM

NUCC

ALEE

NA F

M

Sout

h G

orge

Blue

Rid

geG

orge

Pich

i Ric

hi

100806040200

6040200

100806040200

120

100806040200

100806040200

806040200

140

mYa

ltipe

na

100806040200

250

mYa

ltipe

na

- Etin

a

100806040200

6040200

120

100806040200

140

140

mEt

ina

600

mTr

ezon

a-E

tina

90 m

Yalti

pena

110

mYa

ltipe

na26

0 m

Yalti

pena

- Etin

a

200

50

75

10

0C

IA5

07

51

00

CIA

Han

cock

Lo

okou

t

89

1011

12

turb

idite

sbe

dded

di

amic

tite

5 km

N

32°4

0’

32°2

0’

32°0

’

010

km

138°

0’N

12

11

87

6

54

3

2

910

1En

oram

a Fm

Etin

a Fm

undi

ffere

ntia

ted

units

Brac

hina

Fm

Elat

ina

FmTr

ezon

a Fm

Nuc

cale

ena

Fm

Fig

ure

11.

Det

ailed

stra

tigra

phic

sect

ions

of

the

Elat

ina

Fm. a

cros

s the

sout

hern

Flin

ders

Ran

ges n

umbe

red

to c

orre

spon

d w

ith F

igur

e 1.

All

sect

ions

are

hun

gfr

om a

dat

um a

t the

bas

e of

the

Nuc

calee

na F

m. T

he lo

catio

n of

the

map

ped

area

with

in th

e A

delai

de R

ift C

ompl

ex is

out

lined

by

a gr

ey b

ox in

Fig

ure

1.


oflow towards 325º, and have gentlysinuous, bifurcating crestlines (Fig. 5j;Fig. 12b). The secondary bedformshave a wavelength between 4 and 12cm and a mean amplitude of 0.8 cm(Fig. 12a). These smaller symmetricalbedforms occur within the troughs ofthe larger ripples and have linear crest-lines perpendicular to the crests of theprimary bedforms (Fig. 5j), creating aladder ripple morphology (Fig. 5j). Inplan view, the crestlines terminate withdistinct geometric bifurcations (Fig.5k). In cross-section, the primary andsecondary ripples show vertical aggra-dation of crestlines for many metreswith only very slight sinuous laterallymigrating crestlines (Fig. 5l; Fig. 12b).There is evidence of rare micro-fault-ing showing mm-scale normal offseton the limbs of the secondary bed-forms. The tertiary bedforms have par-allel to slightly oblique crestlines tothose of the secondary bedforms andhave a wavelength of 4–8 cm. Thesecrestlines are typically discontinuousand decrease in amplitude towards thecentre of the secondary troughs (Fig.5k, m). Grainflows from the crests ofthe secondary bedforms are commonand can extend up to 27 cm and typi-cally trend to 030º (Fig. 5m). Furtherup-section, the distinction between pri-mary and secondary ripples is blurred

as the bedding surface becomes cov-ered with interference ripples of equalwavelength that creates a polygonaldistribution of crests. The number ofcouplets recorded on either side of thecrest of an individual bedform doesnot vary systematically, suggesting thatthe bedforms do not influence therhythmic laminations. Above the rhyth-mite unit is a reappearance of a finesandstone with some granule layersand scattered coarse grit throughout(Jablonski 1975; Miller 1975).

An isolated outcrop of similarcyclically laminated facies also can befound in the northern Flinders at Ood-naminta Hut (Fig. 1 [33]; Fig. 5n). Atthis locality, the rhythmic laminationsare most pronounced at the base ofthe section and typically display siltdrapes. At the base of the rhythmites,there are isolated rounded granules <1cm, suggesting overlying ice was pres-ent at least at the onset of deposition.Further up-section, the primary bed-forms are well developed with fewexamples of secondary ripples. Theaverage wavelength and amplitude ofthe primary bedforms are ~30 cm and0.9 cm, respectively. In contrast to thebedforms in the south, these havestrongly sinuous laterally migratingcrestlines, leading to truncation of theplanar laminae (Fig. 12b) and record a

strong asymmetry to the NW and weakasymmetry to the SE. The rhythmitesin the Flinders Ranges have been cor-related to the Reynella Siltstone Mem-ber of the Marino Arkose, near Ade-laide, suggesting a wide area of deposi-tion across the basin (Preiss 1987).

Two observations of theupper Elatina Fm. contact with theoverlying Nuccaleena Fm. suggest thatthere is not a low-angle unconformityon outcrop scale (Forbes and Preiss1987; Lemon and Gostin 1990; Preiss2000; Williams et al. 2008). Firstly, thebasal Nuccaleena Fm. contact does notdisplay an angular cross-cutting rela-tionship in any of our 47 measuredsections (Rose and Maloof 2010). Sec-ondly, the contact may be winnowed,knife sharp, or transitional with silt andice-rafted debris. This variably sharpcontact may suggest the presence of alocal disconformity between the twoformations. However, although it isuncertain as to how much time is miss-ing in each section, it is known that thebasal cap carbonate was depositedwhen glaciers were present (Rose andMaloof 2010).

Geochemistry

Chemical Index of AlterationA total of 133 samples from 13 strati-

Warren Gorgeprimarysecondary

primarysecondary

Oodnaminta

0

10

20

30

40

50

60

70

0

5

10

15

20

25

30

35

40

Wavelength [cm]

War

ren

Gor

ge s

trat

igra

phic

hei

ght [

m]

Ood

nam

inta

str

atig

raph

ic h

eigh

t [m

]

50 600 10 20 30 40

a b

Freq

uenc

y

Blue Ridge Gorge

vy/vx ratio

Warren Gorge

Oodnaminta Hut

-6 -4 -2 0 2 4 6 8 10 12

10

5

0

15

10

5

0

10

5

0

15

20 1 2

3

Figure 12. (a) Graph plotting the wavelengths for the primary and secondary ripples at Warren Gorge and Oodnaminta Hutversus stratigraphic height. (b) Histogram of bedform climbing vectors at Blue Ridge Gorge, Warren Gorge, and OodnamintaHut, where vy = rate of accumulation and vx = rate of migration determine the angle of climb of the bedforms (Rubin andHunter 1982). Mean values are denoted with solid circles and 1σ errors are shown with a shaded box behind the histogram.Note that the ripples at Oodnaminta Hut show the greatest lateral migration in two directions. The inset shows cross-sectionsof the primary (1) and secondary (2) ripples at Warren Gorge, and the primary ripples at Oodnaminta Hut (3).

274

graphic sections across the ARC recorda decline in CIA values through theYaltipena Fm. from the underlyinginterglacial stratigraphy to the lowestvalues in the glacial diamictite, beforerecovering to pre-glacial values in theoverlying Brachina Fm. (Fig. 13; Fig.14; S1 in supplementary online materi-al). The transition in the Yaltipena Fm.declines from ~67 to ~58 at the basalElatina Fm. contact. There is signifi-cant scatter in CIA values within theElatina Fm. However, this scatter has abroad geographic distribution with sec-tions in the central Flinders Rangestypically recording values below 60,whilst sections to the northern regionsrecord values greater than 60.

To the north, the PunchesRest section records CIA valuesbetween 56 and 72 (Fig. 7 [36]). TheEnorama records a mean of 70 thatgradually declines through the TrezonaFm. to values between 64 and 68 with-in the glacial facies. Similarly, the CIAvalues at Nannipinna Creek show agradual decline in the CIA valuestowards the glacial unit but recover toa relatively high index (72) within thediamictites (Fig. 7 [31]). The CIAdataset in the southern Flinders Rangesis limited, which in part is due to thelimited outcrop and arenaceous faciesthat reduces the sampling selection.The Yaltipena and Elatina Fms. recordsimilar CIA values of 65–66. Fourrhythmite samples from Blue RidgeGorge and Buckaringa Gorge haveconsistently low CIA values between59 and 61.

Carbon isotopesThe Etina Fm. is characterized by sus-tained high δ13C values with a mean of~+8‰. An increase in scatter and adecrease in mean δ13C values occuralong parasequence boundary surfacesbetween limestone and siltstone, likelyrepresenting local secondary fluid alter-ation. The Enorama Fm. consists of~300 m of shale with no associatedδ13C data. There is no evidence to sug-gest unconformities at either the top ofthe Etina Fm. or the base of the Tre-zona Fm. Thus, all stratigraphic sec-tions within the central anticline of theARC show an inferred dramatic shiftin δ13C from ~+9‰ to ~–9‰ withinthe Enorama Fm. The overlying pre-glacial Trezona Fm. remains at –9‰

0

100

200

300

400

CHEMICAL INDEX OF ALTERATION COMPOSITE SECTION carbonate units siliciclastic units

packstonegrainstone

stromatolite

sandstone

LITHOFACIES

siltstone

ELAT

INA

FM

met

er le

vel

NUCCALEENA FM

ENO

RAM

A FM

TREZ

ONA

FM

YALT

IPEN

A FM

16 Emu Gap

BRAC

HINA

FM

50 60 70 80CIA

50 60 70 80CIA

500

600

700

Blue Ridge Gorge

Buckaringa GorgeSteve’s Gorge

SOUTH

Emu GapElatina CreekTrezona BoreBennett Springs

Moolooloo

Donkey Gully WellBulls Gap

Angorichina

CENTRAL

Nannipinna CreekPunches Rest

NORTH

3 pt moving average

Figure 13. Stratigraphic variations in the Chemical Index of Alteration (CIA)weathering proxy through the Umberatana Group. Each formation with CIA datafrom stratigraphic sections across the Adelaide Rift Complex was adjusted to fit thethickness of the correlative formations within the nominated Emu Gap referencesection (Fig. 6 [16]). Low CIA values are associated with glacial conditions inferredfrom sedimentological facies, and relatively high CIA values are associated withpre- and post-glacial deposits. The upper Yaltipena Fm. and Elatina Fm. recordlower minimum CIA values compared to those of the other pre-glacial and post-glacial formations. Note that the stratigraphic section is simplified and all the sam-ples analyzed for major elements were siltstones or the siltstone matrix of thediamictites.


for up to 150 m before graduallyrecovering towards 0‰. The upper~15 m of the Trezona Fm. typicallyrecord a ~1‰ decline in δ13C, with anincrease in scatter and a decrease inmean δ18O values (Rose et al. 2012).However, these characteristics are trun-cated in the Warcowie, Elatina Creek,Angorichina and Moolooloo sections,where the Yaltipena Fm. also is entirelymissing (Fig. 6 [13, 17, 23-24]).

Despite the increase in silt-stone towards the north of the ARC,sections record a similarly dramatic~18‰ shift in δ13C across pre-glacialEtina- and Trezona-equivalent forma-tions. The Nannipinna Creek section(Fig. 7 [31]) records the most completerecord of the Etina-equivalent any-where within the ARC. The δ13C valuesgradually increase from 0‰ towards~+10‰. All other northern sectionsrecord a portion of this Etina-equiva-lent enriched δ13C signature. The Tre-zona-equivalent Fm. records a similarδ13C trajectory as the central sections.The complete Trezona δ13C trends arepresent in the most northerly sectionswithin the ARC basin, for example,Taylor Creek shows a gradual rise from–9‰ to –2‰ (Fig. 7 [39]). In contrast,the Weetootla, Nannipinna Creek andOodnapanicken Bore δ13C trajectories

to the south are severely truncated,with the upper value of the TrezonaFm. reaching only ~–8‰ (Fig. 7 [29,31-32]). The entire Trezona δ13C trendis missing from both the Winyagunnaand Idinha Spring sections (Fig. 7 [34-35]).

To the south of the FlindersRanges, the Etina Fm. laterally pinchesout to intermittent, thin stromatoliticbioherms that record ~+9‰ values(e.g., Cockroach Valley; Fig. 8 [6]). Sim-ilarly, the Trezona Fm. thins and atPartacoona the formation spans lessthan 15 m, recording δ13C values of~–2‰, before laterally transitioninginto the siliciclastic Yaltipena Fm. fur-ther south.

In paleomagnetic studies, aconglomerate test is used to determinewhether clasts in a conglomerate weremagnetized prior to transport and dep-osition (preserving random magneticdirections) or after deposition (preserv-ing uniform magnetic directions,despite random clast orientations).Analogously, we performed an isotopeconglomerate test (DeCelles et al.2007; Husson et al. 2012) on the Elati-na Fm. diamictites at 8 localities acrossthe ARC to assess the provenance andrelative timing of acquisition of δ13Cand δ18O by measuring the isotopic val-

ues of 269 carbonate clasts of theElatina Fm. (Fig. 15). Collectively, thediamictites record δ13C variabilitybetween –9‰ and +10‰. However,the distribution of δ13C values for indi-vidual diamictite localities can vary dra-matically, even on a short spatial scale.For example, the Trezona Bore andEnorama Creek localities are less than1 km apart but the clast δ13C data varyfrom –5‰ to +10‰ and 0‰ to+2.5‰, respectively. At Oodnapanick-en Bore, clasts of the distinct Trezonafossiliferous packstone facies were ana-lyzed and record a mean δ13C value of–7.3‰ compared to a mean of –5.4‰for δ13C of generic carbonate clastsfrom the same locality. Note that thereare numerous clasts with δ13C valuesbetween 0‰ to +4‰, which fall in ‘noman’s land’ because these carbon iso-tope values are rarely recorded withineither the pre-glacial stratigraphy ofthe Etina or Trezona Fms. (Fig. 15).

Detrital zirconsU–Pb LA–ICP–MS ages have beendetermined for detrital zircons from 22samples across the ARC within thepre-glacial Trezona and Yaltipena Fms.,and the syn-glacial Elatina and WhyallaFms. (Fig. 16; S2 in supplementaryonline material). The zircon-age spec-tra for the Trezona Fm. show a single~1.2 Ga peak. Two samples were ana-lyzed for detrital zircon from theYaltipena Fm. at Trezona Bore (Fig. 1[18]). The spectra show peaks at ~690Ma, ~1.1 Ga and ~1.7 Ga producing amore varied signal than the underlyingTrezona Fm. to the north.

Eleven detrital zircon spectrafrom the syn-glacial Elatina Fm. allrecord dominant peaks at ~1.1 Ga and~1.2 Ga. In addition, localities to thenorth and south show young peaks atHalletts Cove (~665 Ma), Lame HorseGully (~760 Ma), and ChambersGorge and Walters Well (~730 Ma).The Stuart Shelf records a prominent~1.7 Ga peak throughout the stratigra-phy from the Mesoproterozoic Pandur-ra Fm. to the directly overlying Elatina-equivalent Whyalla Fm. In addition, theWhyalla Fm. shows ~1.6 Ga, ~1.1 Gaand ~1.2 Ga peaks from five samplesacross the shelf.

DISCUSSIONThe eastern edge of the Gawler Cra-

Emu Gap

Blue Ridge Gorge

Buckaringa GorgeSteve’s Gorge

Nannipinna CreekPunches Rest

Elatina CreekTrezona BoreBennett Springs

Moolooloo

Donkey Gully WellBulls GapAngorichina

Al2O3

K2OCaO* + Na2O

weath

erin

gprovenance

sorting

CIA

90

80

70

60

50

tona

lite

gran

ite

gran

odio

rite

CaO* + Na2O

Al2O3SECTIONSFORMATIONS

Elatina FmBrachina Fm

Trezona FmYaltipena Fm

Enorama Fm

a b

Figure 14. Compositional variations of siltstone samples of the pre-, syn-, andpost-glacial stratigraphy illustrated in A–CN–K (Al2O3–CaO+Na2O–K2O) compo-sitional space, which are colour-coded by formation (a) and stratigraphic section(b). The trend defined by the data with low Chemical Index of Alteration (CIA)values roughly follows that expected from variable extents of chemical weatheringof a granodioritic source (solid black star). However, this trajectory diverges withhigher CIA values, indicating a change in source and/or diagenesis. Based on theweathering trajectory being parallel to the A-CN boundary and minimal field evi-dence for a diagenetic origin, we calculate that 45% of the trend can be explainedby weathering and 55% is a result of a change in provenance.

276

-20 -15 -10 -5 0 5-15

-10

-5

0

5

10

δ13C

(‰)

δ18O (‰)N

O M

AN

’S

LAN

D

Tower Gap

0

0.2

0.3

0.4

0.1

0 10-10

+ δ13Cμ = 2.7σ = 2.0n = 10

- δ13Cμ = -1.4σ = 1.0n = 16

Anzac Bore

0

0.2

0.3

0.4

0.1

0 10-10

- δ13Cμ = -4.1σ = 1.2n = 23

Oodnapanicken 2

0

0.2

0.3

0.4

0.1

0 10-10

- δ13Cμ = -5.6σ = 1.0n = 44

Trez pckst δ13C μ = -7.3 σ = 0.5 n = 11

Oodnapanicken 1

0

0.2

0.3

0.4

0.1

0 10-10

- δ13Cμ = -5.2σ = 0.9n = 43

Trez pckst δ13C μ = -7.4 σ = 0.5 n = 18

Trezona Bore+ δ13Cμ = 4.6 σ = 3.0n = 24

0

0.2

0.3

0.4

0.1

0 10-10

- δ13Cμ = -3.6σ = 1.9n = 17

Enorama Creek

0

0.2

0.3

0.4

0.1

0 10-10

+ δ13Cμ = 1.4σ = 0.6n = 14

0

0.2

0.3

0.4

0.1

0 10-10

Oodnapanicken 3- δ13Cμ = -5.5σ = 0.8n = 32

Trez pckst δ13C μ = -7.1 σ = 0.3 n = 6

CLAST δ13C HISTOGRAMS North Bore

+ δ13Cμ = 3.6 σ = 2.5n = 11

0

0.2

0.3

0.4

0.1

0 10-10

δ13C (‰)δ13C (‰)

b

-10 -5

50

100

150

200

250

Tow

er G

apA

nzac

Bor

eO

odna

pani

cken

1

Ood

napa

nick

en 2

Ood

napa

nick

en 3

Trez

ona

Bore

TREZONA FM

δ13C

interpolated datareal data

Emu Gap

c

a

North Bore

Anzac Bore

Enorama Creek

Tower Gap

Trezona Bore

Oodnapanicken 1Oodnapanicken 2Oodnapanicken 3

Etina and Trezona FmsTrezona fossiliferous packstone

Stratigraphic sections

Clasts within Elatina Fm

FINGERPRINTING

δ18O vs. δ13C CROSSPLOT

-20 -15 -10 -5 0 5-15

-10

-5

0

5

10

δ13C

(‰)

δ18O (‰)

NO

MA

N’S

LA

ND

d

C HIST13AST δCLa AMSOGRT b

aper GwoTTo0 4

C HIST13AST δCLeorth BNor

0.1

0.4

0.3

0.2

0-10 0

a

0 4

0.1

0.4

0.3

0.2

0-10

AMS OGRT

n = 11σ = 2.5μ = 3.6

C13+ δ

10Oodnap

Oodna

0

n = 18σ = 0.5 μ = -7.4

ez pckst δrT

en 2panick

en 1panick

10

C13δ

n = 43σ = 0.9μ = -5.2

C13- δ

b

c

0.1

0.4

0.3

0.2

0

n = 16σ = 1.0μ = -1.4

C13- δ

-10 0eornzac BA0.4

0.3 σ = 1.2μ = -4.1

C13- δ

n = 10σ = 2.0μ = 2.7

C13+ δ

10

0.1

0.4

0.3

0.2

0-10

0.4

0.3

n = 11σ = 0.5 μ = -7.3

ez pckst δrT

0

σ = 0.3μ = -7.1

ez pckst δrT

Oodnap

150

C13δ

n = 44σ = 1.0μ = -5.6

C13- δ

10

C13δ

σ = 0.8μ = -5.5

C13- δ

en 3panick

T

200

250

ONATREZFINGERc

MA FPRINTING

0.1

0.2

0

n = 23

-10 0

eorona BezrT0.4

0.3

0.2n = 17σ = 1.9μ = -3.6

C13- δ

10

nn = 24σσ = 3.0μμ = 4.6+ C13+ δ

En0.4

0.3

0.2

0.1

0.2

0-10

eekrama Cor

n = 6

0

50

n = 14σ = 0.6μ = 1.4

C13+ δ

n = 32

10100

150

eorap

G en 3

en 2

en 1 e

ore

or

0.1

0-10 0

C (‰13δ

ous packserossilifona fezrTmsona FezrTtina andE

tionsaphic secrtigatrS

OSSC CR13. δsO v18δd

10

0.1

0-10

‰)

onet

TOSPL

1

10

5 C (‰

)3

δ

0C (‰)13δ

S’A

NO

M

-10

1C

(‰)

3δ

10

5

10real dainterp

δ

Emu Gap

LAN

DS

AN

OM

-5

AN

DS’

AN

M

nzac

BA

er G

woTTo

odna

pani

cke

Ood

napa

nick

eO

odna

pani

cke

O

atapolated data

C13δ

ona

BezrT

p

ona

BezrT

eorth BNor

eornzac BA

eekrama CEnor

odnapanicken 3Oodnapanicken 2Oodnapanicken 1O

eorona BezrT

aper GwoTTo

mtina Flasts within ElaC

-20 -15-15

-10

-5

0 NO

N

-10 -5 0O (‰18δ

-20 -15 -1-15

-10

-5

0

L

5‰)

NO

0 -5 0 5O (‰)18δ

LLAON

OM

5

Figure 15. (a) Histograms of δ13C for carbonate clasts within the Elatina Fm. The clast count locations are labeled on Figure 1.Note that the purple and red histogram bars at Oodnapanicken correspond to data from carbonate clasts and Trezona fossilifer-ous packstone (pckst) clasts, respectively. Hollow histogram bars indicate where the sample size for clasts with positive or nega-tive δ13C values was too small to be included in statistical ‘fingerprinting’ analysis. The positive and negative δ13C values are treat-ed as different clast populations that are putatively sourced from the Etina Fm. and Trezona Fm., respectively. Dots denote themean δ13C for positive and negative δ13C values and/or Trezona fossiliferous packstone clasts at each location, and the paleshaded areas mark the range of δ13C covered by one standard deviation from either side of the mean for each clast group. (b)Fossiliferous packstone clast from the Trezona Fm. and generic grey limestone clasts within the glacial diamictite of the ElatinaFm., near Oodnapanicken Bore. (c) Trezona Fm. ‘fingerprinting’ solutions for carbonate clasts within the Elatina Fm. Both theinterpolated and real δ13C data sets for the Trezona Fm. are depicted. Coloured bars mark the stratigraphic range of possiblesourcing for the clasts. Hollow bars indicate unreliable ‘fingerprinting’ solutions due to a non-normal distribution and/or smallsample size (n<20) (Chen 2012). (d) Crossplots showing δ18O vs δ13C data from the Etina and Trezona Fms. from the Emu Gap,Elatina Creek, Trezona Bore, Bulls Gap and Moolooloo stratigraphic sections (Fig. 1; Fig. 6 [16-18, 22, 24]), and data from car-bonate clasts within the Elatina Fm. The crossplot on the right shows the mean and standard deviation (1σ) of δ13C and δ18Ofor each dataset. The stratigraphic Etina and Trezona Fms. ellipses are opaque and highlighted with a dashed border. Note thatmany of the δ13C values of the carbonate clasts predominantly fall in ‘no man’s land’ between –2.5‰ to +5‰, where the strati-graphic sections of the Trezona and Etina Fms. rarely have δ13C values within this zone.


Flinders Ranges

Gammon Ranges

AD

EL

AID

E

CO

MP

LE

XR

IFT

4 378

9

10

31

o

32

o

30

o1

39

o

14

0o

Bli

nm

an

11v

olc

an

ic r

ock

su

nd

iffe

ren

tia

ted

se

dim

en

tary

ro

cks

san

dst

on

e, s

ilts

ton

e, s

ha

le,

lim

est

on

e, c

on

glo

me

rate

, tu

ff

WIL

PE

NA

GR

OU

P

PER

MIA

N-Q

UA

TER

NA

RY

CA

MB

RIA

N-O

RD

IVIC

IAN

CA

MB

RIA

N

NEO

PR

OTE

RO

ZO

IC

UM

BE

RA

TAN

A G

RO

UP

BU

RR

A G

RO

UP

CA

LL

AN

NA

GR

OU

P

PALE

O-M

ESO

PR

OTE

RO

ZO

ICu

nd

iffe

ren

tia

ted

Pa

leo

-

Me

sop

rote

rozo

ic r

ock

s

ma

jor

fau

lt

syn

clin

e a

xis

an

ticl

ine

ax

is

PEA

K A

GES

WIT

HIN

ELA

TIN

A F

M (G

a)

~1.2

Ga

~0.7

Ga

sam

ple

loca

liti

es

*Hal

let C

ove

n=47

SOU

THER

N F

LIN

DER

S RA

NG

ES

1000

2000

3000

Age

(Ma)

ELAT

INA

Fm

635

1000

2000

3000

Age

(Ma)

*†Bi

lly S

prin

gsn=

28

Lam

e H

orse

n=45

Nan

nipi

nna

Bore

n=11

79

10

11

NO

RTH

ERN

FLI

ND

ERS

RAN

GES

ELAT

INA

Fm

635

†Em

u G

ap

†Moo

lool

oo

n=39

n=43

n=48

1000

2000

3000

Age

(Ma)

*Wal

ters

Wel

ln=

34

Cham

bers

Gor

gen=

34

*†El

atin

a Cr

eek

n=18

6

Relative probability

*Bla

ck H

ill W

ell

n=11

9

8

7 6

5

3

2

1

CENTRAL FLINDERS RANGES

†Ben

nett

Spr

ings

ELAT

INA

Fm

635

STU

ART

SH

ELF

1000

2000

3000

Age

(Ma)

WH

YALL

AFm

n=91

n=10

6

n=10

6

PAN

DU

RRA

Fm

635

WH

YALL

AFm

*n=1

29n=

107

*n=1

20*n

=103

N1

38

o1

37

o

N6

35

Ma

N0

Ma Stu

art

Sh

elf

9° p

aleo

no

rth

2

6

Ga

wle

r C

rato

n

12

12

13

14

13

14

TREZ

ON

AFm

Nan

nipi

nna

Bore

n=11

99

1000

2000

3000

Age

(Ma)

28 m

, n=4

7YA

LTIP

ENA

Fm

*Tre

zona

Bor

e

0 m

, n=2

4

4

PRE-

ELAT

INA

STR

ATIG

RAPH

Y

*Tre

zona

Bor

e4

635

1

5

12

Fig

ure

16.

Map

of

sam

ple

loca

lities

and

indi

vidu

al pr

obab

ility

dist

ribut

ion

func

tions

for U

–Pb

detri

tal z

ircon

age

s col

lecte

d fr

om th

e Pa

ndur

ra a

nd W

hyall

a Fm

s.on

the

Stua

rt Sh

elf, a

nd fr

om th

e Tr

ezon

a, Ya

ltipe

na a

nd E

latin

a Fm

s. fr

om a

cros

s the

Ade

laide

Rift

Com

plex

(ARC

). N

ote

that

the

prob

abili

ty d

istrib

utio

n fu

nc-

tion

for E

latin

a Cr

eek

inclu

des z

ircon

gra

ins c

ollec

ted

from

two

sam

ples

. The

sam

ples

yiel

ding

the

youn

gest

zirc

on g

rain

s are

den

oted

by

an a

ster

isk (*

). Se

ctio

nsw

ith sa

mpl

es ta

ken

from

mor

e th

an 5

0 m

abo

ve th

e ba

se o

f th

e fo

rmat

ion

are

deno

ted

with

a c

ross

(†).

The

~1.

7 G

a zi

rcon

gra

ins w

ithin

the

Why

alla

Fm o

n th

eSt

uart

Shelf

are

sour

ced

from

the

unde

rlyin

g Pa

ndur

ra F

m. S

ampl

es fr

om lo

calit

y 15

are

from

the

eolia

nite

facie

s of

the

Why

alla

Fm. N

ote

that

~1.

7 G

a zi

rcon

grain

s are

not

foun

d w

ithin

the

pre-

glac

ial o

r Elat

ina

Fm. s

tratig

raph

y w

ithin

the

ARC

, ind

icatin

g th

at th

e W

hyall

a eo

lian

sand

shee

t did

not

act

as t

he m

ain so

urce

of se

dim

ent d

urin

g th

e gl

aciat

ion.

Lik

ely so

urce

s for

the

~1.

2 G

a zi

rcon

s with

in th

e A

RC a

re th

e M

usgr

ave

Bloc

k an

d A

lban

y-Fr

azier

Pro

vinc

e w

ithin

Aus

tralia

,an

d/or

the

Wilk

es P

rovi

nce

of E

ast A

ntar

ctica

. To

the

north

of

the

ARC

, the

Elat

ina

Fm. r

ecor

ds a

n ~

0.7

Ga

peak

for w

hich

the

zirco

n co

uld

be so

urce

d fr

omth

e M

t. Cr

ofto

n G

rani

te in

Pat

erso

n Pr

ovin

ce a

nd/o

r the

Lee

uwin

Com

plex

of

Wes

tern

Aus

tralia

. Not

e th

at 8

pot

entia

lly y

oung

zirc

on g

rain

s (<

690

Ma)

from

the

Elat

ina

and

Why

alla

Fms.

have

bee

n da

ted

usin

g ID

–TIM

S (T

able

S3 in

supp

lemen

tary

onl

ine

mat

erial

).

278

ton runs approximately north-southand it is this paleo-coastline and thedeepening of the ARC basin to thenorth that controls the variability indepositional environments and faciesof the Etina–Elatina succession. Thecross-bedded oolitic and sandy faciesof the Etina Fm. was deposited underpredominantly high-energy conditions,possibly as migrating shoals that wereintercalated with small stromatolitebioherms and calcareous siltstone.Preiss (1987) proposed that the pres-ence of quartz throughout the Etinafacies is evidence for an influx of sandfrom the Gawler Craton. Towards thesouth, the Etina Fm. is restricted to afew small stromatolite bioherms (<5m) surrounded by siliciclastic sand.These stromatolites may have been col-onizing the edges of tidal channels thatwere prograding out towards the shelfalong the Gawler Craton coastline(Preiss 1987). The basin deepenstowards the north, resulting in the stro-matolitic and oolitic facies of the cen-tral region transitioning to siltier facies.There is a prograding reef complex tothe northeast at Oodnaminta Hut thatis eroded, creating a submarine escarp-ment with large olistoliths that markthe shelf–slope transition (Giddings etal. 2009). Farther north, the faciesdeepen to silt and carbonate ribboniteinterbeds that record δ13C values of upto +10‰. Thus, we provide geochemi-cal evidence that the Balcanoona,Yankaninna and Amberoona Fms. arethe northern equivalents of the EtinaFm. Farther to the northwest, theEtina Fm. consists of carbonate tepeebreccia, which indicates subaerial expo-sure. This tepee breccia is restricted inlateral extent and surrounded by silt.Such a juxtaposition of facies suggeststhat a paleo-high or promontory exist-ed in the north of the basin. Suchpaleotopography, which occurs over ashort spatial scale, may have been gen-erated by salt diapirs that caused afringing reef to become periodicallyexposed. The closest diapir to thispaleohigh crosscuts the lower 50 m ofthe Trezona Fm. in map view to thenorthwest of Nannipinna Creek, sug-gesting that diapirism was activethroughout the deposition of the EtinaFm. (Fig. 1 [31]).

In addition to abrupt lateralfacies changes over short spatial scales,

there is a distinct relationship betweenlateral facies variability in the pre- andsyn-glacial sedimentary rocks and theaxes of three 50-km scale south verg-ing-open folds (Rose and Maloof 2010;Fig. 1I–V). Generally, across each foldthe pre-glacial Etina Fm. records grain-stone and stromatolite facies south ofthe fold axis and siltstone, shale andolistostromes to the north (Coats et al.1973; Preiss 1987; Giddings and Wal-lace 2009). Similarly, the Trezona Fm.transitions from siltstone and lime-stone ribbons on the southern foldlimb to turbidites on the northern limb(Coats et al. 1973). The Elatina Fm.facies change from laminated siltstoneand rarer boulder tillite on the south-ern fold limbs to conglomeratic debrisflows, diamictite, and massive grittysiltstone on the northern limbs (Roseand Maloof 2010). We propose thatthe sedimentary facies and the pres-ence of basin-bounding normal faultsinfluenced the subsequent Delameriantectonic deformation to create the dis-tinct relationship between lateral faciesvariability in the pre- and syn-glacialstratigraphy and the axes of the folds(Rose and Maloof 2010). We agreewith the interpretation of Giddingsand Wallace (2009) that the pre-glacialEtina Fm. facies variation marks a reefmargin to slope setting across theArkaroola syncline (Fig. 1 III), and fur-ther suggest that the abrupt lateralfacies changes across the folds to thenorth represent the transition from theouter ramp or upper slope to the lowerslope (Rose and Maloof 2010). Abovethe Etina Fm., the Enorama Fm. ispresent both north and south of thefold axes and represents a marinetransgression that caused inundation ofthe central Flinders Ranges carbonateplatform and silt deposition. This for-mation of finely laminated green silt-stone and claystone lacks current orwave ripples, cross-bedding or coarserclastic interbeds, until near the contactwith the Trezona Fm., indicating thatdeposition occurred in a quiet, marinesetting.

The Advance of Land IceThe onset of the glaciation is notrecorded at the top of the TrezonaFm. by a basin-wide disconformityand/or karsted surface (Lemon andGostin 1990). The Yaltipena Fm.

records the encroaching glaciation. Adrop in sea level and/or an increase insiliciclastic sediment supply caused thetidal flat and beach facies to progradetowards the east and north, until mud-cracked siltstone interfingered with thecarbonate reef of the Trezona Fm.(Fig. 17). The progradation of the tidalflat is linked to the advancing icesheets as glacial clasts up to 1 m acrosspierce the top of the Yaltipena Fm.and the upper contact typically is con-torted and scoured. Associated soft-sediment deformation of the Yaltipenasilt layers below the diamictite suggeststhe silt was unlithified and was ruckedup during glaciation by overriding ice(Rose et al. 2012). Furthermore, in thenorthern ARC, isolated rounded gran-ite and gabbro clasts within the uppertwo metres of a Trezona Fm. stroma-tolite bioherm, which records the high-est δ13C following the Trezona anom-aly, represent the onset of glaciation(Rose et al. 2012). This ice-rafteddebris requires floating, debris-ladenicebergs in water possibly no deeperthan the photic zone during terminalcarbonate deposition on the outer shelf(Fig. 17). In addition to the sedimen-tology, the geochemistry of the pre-and syn-glacial sedimentary rocks canprovide information about theencroaching ice sheet and the Cryogen-ian climate state.

The major-element geochem-istry of siliciclastic rocks depends onthe intensity of chemical weatheringand, thus, should preserve a record ofsevere climatic changes (Nesbitt andYoung 1982, 1996; Johnsson 1993;Fedo et al. 1997b; Corcoran andMueller 2002; Rieu et al. 2007c). TheCIA values for the Elatina Fm. have amean value of 63 ± 5 (1σ error). If theCIA values accurately reflect the inten-sity of chemical paleoweathering, thenthis low value corroborates theglacigenic interpretation of the ElatinaFm. based on independent sedimento-logical evidence (Fig. 13). This CIAvalue is typical of other Marinoan gla-cial successions around the world(averages: Ghaub Fm., 63 ± 5; Morten-snes and Smalfjord Fms., 66 ± 1; PortAskaig Fm., 70 ± 5; all 1σ errors;Bahlburg and Dobrzinski 2011).Although the Enorama and TrezonaFms. record values of 68 ± 4 and 67 ±4 (1σ error), respectively, which are


lower than expected for a pre-glacialwarm, tropical carbonate platform,these CIA values still show a relativedecline across the Yaltipena Fm. Thetight downturn in CIA from 68 to ~55recorded in three sections across theYaltipena Fm. (average 62 ± 5 (1σ

error)) attests to the onset of theglaciation with an overall lowering ofsea level and progradation of the tidalflat facies across the basin. In addition,the upper Yaltipena Fm. and ElatinaFm. record lower minimum CIA valuescompared to those of the other pre-

glacial and post-glacial formations. Themean CIA value for the overlyingBrachina Fm. is 67 ± 5 (1σ error), indi-cating a recovery to pre-glacial values.However, important factors other thanclimate need to be considered whenevaluating the major element composi-

F1F2

Trezona Fm stromatolites

S N

2

3

4

push moraine

glacial erosion and ice-rafted debris

reef slopeproglacial tidal flat

1

prograding Yaltipena Fm

ice sheet

ice sheet

ice sheet

GLA

CIAT

ION

DEG

LACIAT

ION

6

7

8

5

sea level

local ice sheet retreat

isostatic rebound and/or loss of ice sheet gravitational attraction

global deglaciation

cap dolostone deposition

F3

LEGEND

dropstone diamictite

Nuccaleena Fm cap dolostone

Brachina Fm

slumped sandstone

S N

ice sheet

F1F2

Yaltipena Fm

siltstone

debris laden icebergs Facies 1 Facies 2 Facies 3

Elatina Fm

F1F2F3

F1F2F3

current-reworked diamictite

rhythmite

sandstone

diamictite

siltstone

Figure 17. Schematic depositional model for the Elatina glaciation and deglaciation across the southern, central, and northernregions of the Adelaide Rift Complex. During the onset of the glaciation, ice-contact deformation of the Yaltipena Fm. andglacial truncation of the Trezona Fm. occured in the central fold [3]. To the north, synchronous deposition of dropstones with-in stromatolites is recorded in the uppermost Trezona Fm. [4]. During the deglaciation, slumped sandstone (F1) and dropstonediamictite (F2) were deposited in the central anticline [5]. As the local ice sheet continued to retreat, loss of ice sheet gravita-tional attraction and/or isostatic rebound led to a regression and current reworking of the underlying diamictite (F3) [6]. Theglobal deglaciation is recorded by a final diamictite within the upper Elatina Fm. and the precipitation of the Nuccaleena Fm.cap dolostone and overlying Brachina Fm. [7-8].

280

tion of siliciclastic sediments and sedi-mentary rocks, including the influenceof source rock composition (McLen-nan et al. 1993; Fralick and Kronberg1997), sediment recycling (McLennanet al. 1993; Cox et al. 1995), tectonicsetting (McLennan et al. 1993; Corco-ran and Mueller 2002), relief(Grantham and Velbel 1988; Johnsson1993), sediment transport (Malmon etal. 2003), and diagenesis (Nesbitt andYoung 1989; Fedo et al. 1995, 1997a).

The geochemical compositionof sediments is strongly controlled bythe composition of the rocks andregolith from which the sediment isderived (Nesbitt and Young 1989;McLennan et al. 1993; Fedo et al. 1996;Fralick and Kronberg 1997). AnA–CN–K (Al2O3–CaO+Na2O–K2O)ternary plot shows that the major ele-ment composition of samples from theTrezona, Yaltipena and Elatina Fms.that have been little affected by chemi-cal weathering were derived from rockswith an average granodioritic composi-tion (Nesbitt and Young 1984, 1989;Fig. 14a). However, data from theElatina Fm. are shifted towards theK2O apex, suggesting that the glacia-tion introduced rocks derived from amore granitic sediment source. In addi-tion, the trend of the complete data setin the A–CN–K plot does not perfect-ly parallel the A–CN boundary (Fig.14b). This drift mostly is representedby samples from northern sections,suggesting that there may be somevariation in the composition of thesediment source rocks specific to thisregion. This observation is corroborat-ed by the detrital zircon suites fromthe Elatina Fm. that record <760 Mayoung zircon peaks in the northernFlinders Ranges (Fig. 16 [7-8]), andinterestingly, both the Leeuwin Com-plex and Paterson Province of the Yil-garn Craton, which are proposedsources for these young zircon grains,are granitic.

The observed weatheringtrend also reflects the degree to whichrecycled sedimentary rocks have beenmixed and incorporated into theglacigenic deposits (Bahlburg andDobrzinski 2011). Glacial depositsrecycle large quantities of sedimentthat may retain CIA values reflective ofprogressive weathering in previous cli-mates, causing scatter in the CIA val-

ues (Cox et al. 1995). We interpret thevariability in the CIA values for theElatina Fm. to reflect in part sedimentderived from the underlying YaltipenaFm., which may record progressiveweathering of sedimentary materialunder an equatorial climate over aseries of recycling events. TheYaltipena Fm. sediment may have orig-inated from different basins that expe-rienced a different number and/orintensity of weathering regimes. Basedon the weathering trajectory being par-allel to the A-CN boundary and mini-mal field evidence for a diagenetic ori-gin, we calculate that ~45% of thetrend can be explained by weatheringand ~55% is a result of a change inprovenance.

An alternative hypothesis forwhy the Elatina Fm. samples plottowards the K-apex in A–CN–K com-positional space is that the formationexperienced burial diagenesis, withpotassium metasomatism changing thebulk composition, and consequentlythe CIA, of sedimentary rocks (Fedoet al. 1995). Potassium metasomatismis suspected when the compositionalvariations in a sedimentary successiondeviate from the ideal weatheringtrend, showing enrichment in potassi-um and thus diagenetically loweredCIA values. However, the underlyingYaltipena Fm. records a tight decliningCIA trend that reaches a nadir of 55,and neither formation shows any phys-ical evidence for pervasive diagenesis,such as significant liesegang commonlyassociated with migrating alterationfronts. Thus, the drop in mean CIAvalues from pre-glacial to glacial faciesis more likely a result of climate deteri-oration and/or different weatheringsources.

CIA values strongly dependon the abundance and composition ofclay minerals, and therefore may beinfluenced by the effects of hydrody-namic sorting during deposition.Although this study is limited to silt-stone rocks in order to reduce theeffect of differential sorting on com-position, subtle differences in grainsize may exist between the samples.However, the magnitude of variabilityin CIA within the Elatina Fm. is notlikely to be a consequence of suchminor facies differences between sam-ples.

Overall, the CIA results indi-cate that compositional variations inmajor element geochemistry within theARC provide limited information whenevaluating the paleoclimatic signifi-cance of the Marinoan siliciclastic suc-cession because it is hard to accuratelydetermine the degree to which a vari-able source rock and climate controlledthe CIA values. Despite the lack of anunambiguous interpretation for thesyn-glacial CIA values, we interpret thelarge scale CIA record to represent realclimatic shifts during the pre- andpost-glacial transitions to and from theElatina glaciation. Therefore, the evi-dence for the Yaltipena representingencroaching land-based ice is a) thearchitecture of the Yaltipena Fm. as aprograding tidal flat complex, b) thepresence of glacial dropstones in thestromatolites beyond the tidal flat edge,c) the sub-glacial deformation of theunlithified Yaltipena sediments, and d)the tight declining CIA trend thatreaches a nadir of 55 towards the basalcontact of the Elatina Fm.

Subglacial ErosionLocal glacial erosion variably truncatesthe Yaltipena Fm., such that this for-mation is entirely missing in parts ofthe central Flinders anticline (Fig. 6[17, 23]). The minimum amount ofglacial incision into the underlying car-bonate platform may be quantified bycorrelating the stratigraphic sections,using the inflection point at the base ofthe recovery from the Trezona negativeδ13C anomaly, and assuming that themaximum thickness of the TrezonaFm. in the central Flinders Ranges isrecorded in the Emu Gap section, towhich all sections within the centralanticline are referenced (Halverson etal. 2002; Fig. 1 [16]). Thus, the minimumamount of glacial incision is calculatedfrom the stratigraphic thickness of theTrezona δ13C signature recorded abovethe inflection point. A total of 9 meas-ured sections through the Trezona-Elatina Fms. across central FlindersRanges show that up to ~130 m havebeen truncated by glacial erosion. Tothe north, although the stratigraphiccorrelations are more tentative and theTrezona Fm. could consist entirely ofsiltstone, it appears that the completeTrezona Fm. and overlying stratigra-phy, up to 500 m thick, have been


removed and the Elatina Fm. is in con-tact with the Enorama Shale (Brench-ley-Gaal 1985; Fig. 7 [34]). Progressivetruncations of beds within the TrezonaFm. indicate that the unconformity ismore pronounced to the north, westand south of Trezona Bore. The sec-tions north of Punches Rest do notrecord glacial truncation, and we inter-pret these sections that record hun-dreds of metres of stratigraphy withδ13C values remaining at ~–2‰ in theupper Trezona Fm. (Fig. 7 [38-40]) torecord a deeper environment and anincrease in accommodation space.

Further information on theextent of glacial incision comes fromthe carbonate clasts within the diamic-tites. Previous work determined that allcarbonate clasts in the central anticlinein the Elatina Fm. are thought toderive from the underlying TrezonaFm., due to distinctive ooid clasts iden-tified near Bulls Gap (Lemon andGostin 1990; Fig. 15). Although wefind clasts of fossiliferous packstonenear Oodnapanicken Bore (Fig. 15b),the remaining carbonate clasts aregeneric cream to grey limestone with-out unique sedimentary features todetermine their origin. If the clastsacquired their δ13C values in situ on thecarbonate platform, then they shouldexhibit a random collection of valuesrepresenting the full isotopic rangepresent in the eroded carbonate plat-form δ13C profile at the time ofdiamictite deposition (DeCelles et al.2007; Husson et al. 2012). In contrast,if the extremely negative δ13C values(down to –9‰) in the carbonate plat-form are a result of post-depositionaldiagenesis, then clasts from individualdiamictite beds should either (a) reflectthe original pre-diagenetic isotopic val-ues in the platform (i.e., not extremelydepleted in δ13C), or (b) have consistentdiagenetic values that are roughlyhomogeneous within diamictite units.The wide variability of the clastsshows that they record the full isotopicrange from –9‰ to +10‰ present inthe carbonate platform δ13C profile,and thus, rules out late-stage burial dia-genesis (Fig. 15a). This isotope conglomer-ate test (DeCelles et al. 2007; Husson etal. 2012) does not preclude that theclasts reflect early meteoric diageneticalteration of the carbonate platformδ13C profile prior to local glacier

advance. However, while a meteoricdiagenesis hypothesis for the TrezonaFm. can be made consistent with thetiming of exposure of the platform, itis inconsistent with top-down modifi-cation of the platform by meteoric flu-ids (Swart and Kennedy 2012) andwith the lack of permeability-depend-ent δ13C spatial variability between dif-ferent lithofacies across the platform(Rose et al. 2012).

The negative δ13C clasts withvalues < –2‰ can be attributed to theupper ~180 m of the Trezona Fm.(Fig. 15c). The fossiliferous packstoneclasts have consistently lower δ13C val-ues in comparison to generic limestoneclasts at Oodnapanicken Bore. Thisδ13C value is characteristic of thesepackstone units, as they are most com-monly associated with the initial prolif-eration of stromatolites within thelower half of the Trezona Fm. stratig-raphy in the central anticline of theARC. Only the Weetootla, Nannipinna,and Winyagunna sections in the north-ern Flinders Ranges document trunca-tion of the Trezona δ13C anomaly tothe inflection point of –9‰ (Fig. 7[29, 31, 34]). This observation suggeststhat either the clasts were not trans-ported a long distance from Nannipin-na Creek or the site of erosion is notexposed within the ARC. At Oodna-panicken Bore, the δ13C values of theclasts ranges from –3‰ to –9‰, andalmost record the full isotopic range ofthe Trezona Fm.; the δ18O values, how-ever, are ~5‰ more negative thanthose of the Trezona Fm. stratigraphy(Fig. 15d). This observation suggeststhat diagenesis has locally altered theδ18O values. The timing of this diagen-esis is unknown, but because only theclasts at Oodnapanicken Bore showmodified δ18O compared to the othernorthern clasts (Fig. 15a), it likelyreflects late stage diagenesis local toOodnapanicken Bore, rather than pan-basin early alteration during transporta-tion of broken clasts in δ18O-depletedglacial meltwater.

The clasts with isotopically-enriched δ13C values between +5‰and +10‰ can be attributed to theEtina Fm. and northern equivalents(Fig. 15a). However, nowhere withinthe ARC does glacial erosion removeall of the Trezona and Enorama Fms.,juxtaposing Elatina Fm. on top of the

Etina Fm. Active salt diapirismthroughout the deposition of the EtinaFm. and much of the Enorama Fm.are thought to be responsible forentraining brecciated blocks of theunderlying stratigraphy and bringingthem to the surface (Lemon 1985;Lemon and Gostin 1990; Lemon2000). In map view, many of thediapirs cross-cut the Etina Fm. and thelower two-thirds of the Enorama Fm.in the central region (Lemon 2000),whilst in the north they impinge on thelowermost ~50 metres of the TrezonaFm. Despite the small Oratunga diapirin the central region that cross-cuts theUmberatana Group stratigraphy, forthe most part diapirs were emergentand most active before deposition ofthe Trezona Fm. and Elatina Fm.(Webb 1960; Coats 1965; Lemon2000).

One possibility is that the iso-topically-heavy clasts originated from adistal diapir that sampled the EtinaFm., and was then eroded and trans-ported by the ice sheet. Many of theclasts fall between –2‰ and +5‰,which are not values recorded in eitherthe Etina or Trezona Fm. stratigraphy.These values may represent diagenesisof the Etina clasts but this seemsunlikely given that the values are highlyvariable within a 5 m x 5 m cross-sec-tional area of diamictite. The onlyplace in the interglacial stratigraphythat records these values is the transi-tion from the Tapley Hills Fm. foundbelow the Etina Fm., suggesting thateither erosion of a diapir or a nearly 1km deep glacial incision at an unob-served locality sourced these deeplyburied formations. Alternatively, theshale of the Enorama Fm. might tran-sition to carbonate facies, presentlyunobserved, and record the full iso-topic down-turn of the Trezona δ13Canomaly from +5‰ at the top of theEtina Fm. to the –9‰ nadir at thebase of the Trezona Fm. Thus, the car-bonate clasts might reflect at least~500 m of glacial incision, possiblyoutside of the ARC. The JakobshavnIsbræ ice stream in Greenland creates afjord ~700 m deep (Holland et al.2008) and paleofjords during the Ceno-zoic are thought to have incised up to1 km below sea level in Antarctica(Young et al. 2011). Thus, althougherosion of a diapir could source the

282

clasts, given a sufficiently large ice capin Australia it seems likely an icestream would be capable of erodingthe entire Etina Fm.

The Retreat of the Ice SheetThe basal diamictite of the Elatina Fm.has a scoured contact and contorteddiamictite beds at the base of theElatina Fm. at the Trezona Bore sec-tion, indicating local ice-contact andsub-glacial push structures formed dur-ing an ice advance (Lemon and Gostin1990; Fig. 5g-i; Fig. 17). Althoughnumerous ice sheet advances couldhave occurred, these preserved faciesrecord a singular major advance withinthe Elatina Fm. The overlying Elatinafacies consist of cross-bedded channel-ized fluvioglacial sands that grade intoshallow-marine sands with pervasiveslumping (Facies 1) and dropstonediamictite deposits (Facies 2). Thesefacies record the transgression associat-ed with the subsequent deglaciation.Despite observations of some ripplecross-laminated and cross-stratifiedsandstone intercalated with diamictitethroughout the succession (Le Heronet al. 2011a; Le Heron 2012), we docu-ment that the top third of the ElatinaFm. is dominated by current-reworkeddiamictite beds with distinctive gravellags and lonestones capped by mm- tocm-scale ripple cross-laminated fine- tovery fine-grained sandstone (Facies 3).This final facies which occurs through-out the central and southern FlindersRanges records upward shallowing andreworking of the underlying diamictitebeds. These observations and interpre-tations corroborate those of the semi-nal work of Lemon and Gostin (1990).The upper part of Facies 3 is correla-tive to the terminal shallowingsequence that transitions into grittysandstone in the southern FlindersRanges. In the central region, a thindiamictite is present in the upper partof Facies 3, just below the Nuccaleenacap carbonate, which we correlate tothe gritty sandstone just below the capdolostone at Warren Gorge and thesurrounding region (Fig. 9 [24]; Fig. 11[4]). Despite the northern FlindersRanges predominantly consisting ofdeep marine diamictite, there are threebroad facies to the Elatina Fm.: a lami-nated siltstone, a diamictite, and acoarse feldspathic sandstone. The

upper sandstone facies consists of palegrey and brown feldspathic sandstonewith lenses of calcareous siltstone andpebble conglomerate. We interpret thisBalparana Sandstone to reflect thebasin-wide shallowing within thenorthern ARC. Overall, water levelsdeepened through deposition of thefirst two units (Elatina Facies 1 and 2),shallowed during the deposition of theupper unit (Facies 3), before the post-glacial sea level rise recorded with thedeposition of the Nuccaleena andBrachina Fms. (Fig. 17).

The rhythmite facies withinthe Elatina Fm. consists of very finesandstone uniformly distributedthroughout ~30 m of stratigraphy andacross at least ~4500 km2. At WarrenGorge, the previously interpreted par-allel cuspate folds within the rhythmitefacies (Williams 1996) have a character-istic wavelength and show birfurcatingcrestlines in plan view (Fig. 5j, k). Incross section, these bedforms haveslightly asymmetric crestlines and showpreservation of the entire crest of thebedform (Fig. 5l; Fig. 12b). At Ood-naminta in the north Flinders Ranges,these bedforms have strongly sinuouslaterally migrating crestlines in cross-section, often with reversing trunca-tions of the topsets of the bedforms.In addition, there is no evidence forsoft-sediment deformation at any scalethroughout these bedforms. Together,these observations are not compatiblewith the previous interpretation as par-allel cuspate folds and slump struc-tures, particularly as the slumps likelywould migrate only in the direction ofgravity and generate a variety of soft-sediment deformation at a range ofspatial scales. Thus, we interpret thesesedimentary structures to be stoss-depositional, reversing three-dimen-sional bedforms with superimposedsymmetrical bedforms that orthogonal-ly intersect the primary crestlines(Rubin and Hunter 1982). The geome-try of these bedforms indicates com-bined flow between unidirectional andoscillatory currents. The overall east-ward migration of the primary crestsindicates an off-shore transport direc-tion with ongoing wave action generat-ing the secondary crestlines in a marinesetting above fair weather wave base.Tertiary crests appear near the top ofthe rhythmite section that are mostly

parallel to the secondary crests. Thetertiary crests may indicate a change toa shallower water depth and the gener-ation of ripples of a shorter wave-length. Thus, a relative sea level dropalso is recorded within the upperrhythmites before continued shallowingthat is marked by the gritty sandstonefurther up section. We interpret theupper rhythmites to correlate with theFacies 2–3 transition within the ElatinaFm.

The Balparana Sandstone, thecurrent-reworked diamictite, and thetertiary combined-flow ripples in therhythmite unit are time-synchronousfacies across the ARC that all record aregression. This basin-wide relative sealevel fall occurs in the upper part ofFacies 2 and Facies 3 within the overalltransgressive deglaciation sequence ofthe Elatina Fm. This regression couldbe caused by the retreat of local gla-ciers and the associated instantaneousloss of gravitational attraction that theice sheet had on the nearby ocean.During the Pleistocene-Holocene, uni-form melting across the Greenland icesheet of a volume of ice that wouldhave been equivalent to raising globalsea level by 1 m, would have generatedbetween ~2–10 m of local sea level fallaround the Greenland coast (Mitrovicaet al. 2009; Kopp et al. 2009).Although this value varies with paleo-geography and the geometry of themelting ice sheet, a similar sea levelfluctuation would be significantenough to generate the regressionrecorded in the Elatina Fm. Alterna-tively, the regression could be in partor solely caused by longer term (104

years) isostatic rebound associated withthe shrinking ice sheet. However, ifregional isostatic rebound represents asignificant part of the sea level signalthen the characteristic timescale forisostatic rebound, which is controlledprimarily by mantle viscosity, con-strains the retreat of local glaciers fromthe basin to at least 104 years beforethe global deglaciation. This scenariowould suggest that regional isostaticrebound occurred prior to the globalglacio-eustatic sea level rise that con-trolled deposition of the overlyingNuccaleena Fm. cap dolostone.Together, our sedimentological obser-vations suggest that local ice likely waspresent until the Elatina Facies 2–3


transition (Fig. 17). This persistence oflocal ice suggests that the instanta-neous loss of gravitational attraction isthe more parsimonious explanation forthe regression within the upper ElatinaFm.

The wave component of theripples within the rhythmite facies inthe southern ARC requires open tropi-cal seas with significant fetch that isprior to, or at least contemporaneouswith local ice retreat. This timing con-trasts with the model proposed for evi-dence of open water during peakglaciation, where local glaciers arestarved of moisture and ice is subli-mated away from restricted basins(Halverson et al. 2004). If the clastswithin the upper diamictite in Facies 3represent an extra-basinal source theycould be derived from the melting ofdistant debris-laden icebergs during theglobal deglaciation (Fig. 17). Thisarrival of far-traveled icebergs to thebasin would shift the onset of theglobal deglaciation from the basal con-tact of the Nuccaleena Fm. to withinthe upper Elatina Fm. (Raub andEvans 2008). However, a large propor-tion of the clasts within this diamictitein the northern region of the centralFlinders Ranges are basalt and previ-ous work attributed the source of theclasts to active diapirism. Such an intra-basinal source could represent aregional readvance that results frommoisture derived from the openingoceans feeding the continental glaciers.However, there is no evidence that gla-ciers reached the extent of those thatdeposited Facies 1 given the lack ofsubglacial deformation and proximaldeposits. Future work could test thesehypotheses by determining the clastprovenance and detrital zircon signa-ture of the upper diamictite, grittysandstone, and the Balparana Sand-stone across the ARC.

Tides During the DeglaciationThe re-interpretation of the putativesoft-sediment folds as ripples weakensthe iconic syn-sedimentary fold testthat constrained the low-latitude posi-tion of South Australia at the time ofthe Elatina glaciation. While the tec-tonic fold test still requires that thelow-latitude paleomagnetic direction ispre-Late Cambrian (Foden et al. 2006),the magnetization no longer must be

syn-depositional. That said we have noreason to believe that the low-latitudedirection is a result of remagnetization,and the positive reversal tests (Sohl etal. 1999) are at least consistent withsyn-depositional magnetization. Tauxeand Kent (1984) showed that the detri-tal remnant magnetization of hematitein the modern Soan River deposits,northern Pakistan, may record an incli-nation (e.g. 25º) that is significantlyshallower than the inclination of the insitu field (e.g. 50º). Similarly, laboratoryexperiments have shown that crystallo-graphic orientations of hematite crys-tals, which are determined by the prin-cipal susceptibility axis, are dominatedby depositional rather than magneticfield conditions (Lovlie and Torsvik1984), such that the inclination of thedetrital remanent magnetization heldby hematite is significantly shallowerthan the ambient field. These resultssuggest that sorting by oscillatory,wave-induced currents may align platyhematite grains with the rhythmitelaminations, which could account forthe positive fold test (Sumner et al.1987).

The minimal winnowing ofthe bedform crests, the limited migra-tion of crestlines, and the absence offiner sand and clays in the troughsindicate that sediment was not distrib-uted or sorted solely by wave actionacross the region. Williams (1989,1991, 2000) proposed a distal ebb-tidaldelta for the depositional setting wherefine-grained sediment is entrained byebb-tidal currents and transportedmainly in suspension by turbid currentsand jets via the main ebb channel todeeper water offshore. With increasingdistance of transport, such jets trans-form to hyperpycnal plumes and sort-ed, suspended sediment settles to formnormally graded laminae in distal set-tings. Similarly, glaciofluvial outlets atthe terminus of a glacier could gener-ate vast plumes of suspended sedimentthat would be delivered in diurnallyand seasonally controlled pulses that,in addition to the tidal signal, mighthave influenced the deposition ofrhythmite couplets and bundles.

At Oodnaminta Hut, rarequartz and feldspar granules at theonset of the rhythmite facies suggestoverlying debris-laden icebergs. Claydrapes are restricted to these lower few

metres and the very fine sand coarsensupwards throughout the section, withshorter wavelength secondary ripplesonly documented at the top of the sec-tion, indicating a shallowing. The pri-mary ripples have convex-up profilestypical of oscillatory waves. These bed-forms record a strong asymmetry, withripples climbing to the northwest and aweaker asymmetry to the southeast,suggesting that the offshore currentresponsible for the bedforms was notconstant but fluctuated in strength.The current was strongest to thenorthwest, and during periods whenthe current weakened or ceased theoscillatory flow became more domi-nant and the bedforms drifted to thesoutheast. These consistently orientedbedforms throughout the stratigraphysuggest perhaps that the waves wererefracted within a protected embay-ment.

The bedform characteristicsoutlined above, particularly the verticalaggradation and small mm-scale faultson the limbs of some bedforms atWarren Gorge, are indicative of rapidsedimentation rates. However, workersfor more than 35 years have argued fora tidal origin for the rhythmite facies inthe southern Flinders Ranges.Although a single periodic variation inthe rhythmites could suggest any num-ber of sources, time series analysisshows a nested hierarchy of bundlesthat contain ~15 couplets (Williams1989, 1991, 2000; Budnick 2012). Diur-nal and seasonal fluctuations in sedi-ment delivery by glacio-fluvial and/orkatabatic wind sources could generatethe individual couplet and annual peri-odicities. In fact, eolian delivery ofsand by daily katabatic winds mayexplain why the couplets record a diur-nal, and not the expected semi-diurnal,signal dominant in most tropicalregions today. However, the cyclicnature of the couplets strongly sug-gests a neap-spring tidal origin thatoperates on the order of 14-15 days.Thus, we agree with a tidal interpreta-tion, which can explain most of theperiodicities recorded in the rhyth-mites.

A tidal origin places a timeconstraint for the deposition of therhythmites: using the number ofspring-neap bundles and assuming thatit takes a month to deposit two bun-

284

dles, the rhythmite stratigraphy at War-ren Gorge took a minimum of 33years to accumulate. At OodnamintaHut, the rhythmite laminations are notas well preserved as those at WarrenGorge. However, a minimum of 19bundles are deposited whilst the bed-form migrates to the northwest, and aminimum of 9 bundles are counted asthe bedform migrates to the southeast.Thus, the tidal timescale implies thatthe hyperpycnal current delivering thesediment was active for ~10 months,but diminished or switched off for thefollowing ~5 months. This temporalvariability in the current strength couldrepresent a seasonally controlled sub-glacial or glacial-fluvial source.

The Whyalla Fm. on the adja-cent Stuart Shelf could have acted asthe eolian sediment source for therhythmites. Winds reworking thiseolian sand sheet may have entrainedand transported the very fine sand andsilt fraction offshore to supply therhythmite facies, and the silt fractionwould continue to be blown to moredistal localities. Winds capable oftransporting very fine sand in suspen-sion would need to be 56 km/h and toblow most days to supply sediment forthe diurnal couplets for the entireduration of rhythmite deposition(Eastwood et al. 2012). The variabilityin couplet thickness would be modulat-ed by the spring-neap tidal cycle, wherethe stronger spring tide would depositmore sand due to enhanced currentspeed and tide volume, but the weakerneap tide would not be able to entrainsand, resulting in deposition of thin siltlaminae. Katabatic winds at the steepterminus of coastal ice slopes areamong the strongest surface winds;localities in Antarctica can experienceyearly average winds in excess of ~70km/h (Turner et al. 2009). Thus, simi-lar winds could have supplied a diurnaland seasonal source of sand for therhythmites. However, this scenariorequires that wind blown sand wouldbe transported and evenly distributedover an extensive area every day for~33 years while glaciers remain in thevicinity.

Sediment Provenance and Direction of Ice TransportDetrital zircon data may shed light onthe source of sediment for the Elatina

rhythmites, as well as constrain thedirection of ice transport, the extent ofice coverage, and the maximum age ofthe glaciation. Previous work has pro-posed that the Whyalla Fm. acted as asource for the Elatina Fm. across theARC (Lemon and Gostin 1990;Williams et al. 2008). This work report-ed that the slumped sandstone unit atthe base of the Elatina Fm. is markedby a dominance of coarse silt to finesand and very coarse sand and granulefractions (Lemon and Gostin 1990;Williams et al. 2008). The Whyalla Fm.includes the missing medium to coarsesand fraction, and it is suggested thatspatial sorting by eolian reworking gen-erated the different grain size distribu-tions in the Whyalla and Elatina Fms.(Lemon and Gostin 1990; Williams etal. 2008). The fine sediment wouldhave been transported to the ElatinaFm. by wind, whereas the very coarsesand fraction would have been deliv-ered by fluvial systems. However, iffluvial systems were draining the StuartShelf or adjacent uplands, it seemsunlikely that only the coarse, and notthe medium sand fraction, would betransported from the Whyalla Fm. Fur-thermore, detrital zircon data showthat the provenance of the sand in theWhyalla and Elatina are different.Many zircon grains within the Elatina-equivalent Whyalla Fm. have similarages to the Elatina Fm. within theARC (~1.6 Ga, ~1.1 Ga and ~1.2 Gapeaks; Fig. 16). However, a pervasive~1.7 Ga peak present throughout theWhyalla Fm. is absent from all the pre-glacial and Elatina Fm. stratigraphythroughout the ARC. These zircongrains likely are derived from the Yava-pai-Mazatzal Province of Laurentiaand/or East Antarctica that was juxta-posed to Australia for the previous 300my (Karlstrom and Bowring 1988;Hoffman 1991; Goodge et al. 2008)and sourced locally from the PandurraFm. that underlies the Whyalla Fm. inmany localities across the Stuart Shelf.The majority of the zircon grainsresponsible for the 1.7 Ga peak are lessthan 80 µm and would require similarwindspeeds as the larger, fine sandfraction for them to be carried in sus-pension. Thus, it seems unlikely thatthe Whyalla sand sheet supplied sedi-ment for the glacial deposits in thecentral region, and suggests that the

source for the rhythmites must eitherbe sub-glacially derived or from anoth-er sand sheet.

Throughout the Trezona–Elatina Fm. stratigraphy in the centralARC and the Nannipinna and BillySprings localities to the north, the zir-con age spectra record a dominant~1.2 Ga peak (Fig. 16). Several grani-toid suites were intruded into the Mus-grave Block in central Australia thatyield U–Pb zircon ages of~1225–1190 Ma, associated withGrenville-age orogeny and the amalga-mation of Rodinia (Maboko et al.1991; Dalziel 1991; Hoffman 1991;Moores 1991; Camacho and Fanning1995). Similarly, the quartz monzoniteplutons and gneiss of the Albany-Fras-er Province outcrop near the southernmargin of the Yilgarn Block and aredated at ~1175 Ma (Pidgeon 1990) and~1200 Ma (Black and Shaw 1992),respectively. Thus, the ~1.2 Ga peakrecorded in the Flinders Ranges is con-sistent with an Australian intra-conti-nental source from central and/orwestern Australia. Alternatively, thispeak could originate from theGrenville-age Wilkes Province of EastAntarctica that abutted Australiathroughout the Neoproterozoic(Goodge et al. 2008).

In the northern FlindersRanges between the Moolooloo andLame Horse Gully localities, the domi-nant peaks in the detrital zircon spectraare between ~665–760 Ma (Fig. 16 [7-9]). Within the Australian continent,the Paterson Province to the east ofthe Pilbara Craton and the LeeuwinComplex to the southwest of the Yil-garn Craton in Western Australia arethe only regions hosting such youngNeoproterozoic magmatic events (Fig.2). The Paterson Province consists ofgranite, such as the Mt. Crofton Gran-ite, with ages between ~600–700 Ma(McNaughton and Goellnicht 1990;Nelson 1995). The Leeuwin Complexis dominated by granitic gneiss withages that suggest that the protoliths ofthe gneiss formed over 600 millionyears in distinct magmatic pulses at1200–1050 Ma, 800–650 Ma, and580–500 Ma (Nelson 1996, 1999, 2002;Collins and Fitzsimons 2001; Wildeand Nelson 2001). Both of thesesource areas suggest that ice likely cov-ered at least half of the Australian


continent. Zircons ranging from~600–750 Ma previously have beenreported in sedimentary rocks from theARC (Compston et al. 1987), LachlanFold Belt (Williams et al. 1988), andNew Zealand and Marie Byrd Land,Antarctica (Ireland et al. 1994), but thesource regions for the sediment stillremain a matter of conjecture. It hasbeen suggested that the sediment ini-tially was deposited in a sub-marineturbidite fan fed from the risingDelamerian-Ross Orogen of easternAntarctica and southeastern Australia(Coney et al. 1990). Thus, another pos-sibility is that the ~665–760 Ma zir-cons are derived by recycling sedimentfrom a regional eastern Antarctic andsoutheastern Australian source, or if,since the sediment being recycled has~760 Ma zircons, it too could be tap-ping the cratonic sources in WesternAustralia.

The predominance of <760Ma young zircons at several northernFlinders localities suggests that perhapsan additional or different source sup-plied sediment to this region comparedto the rest of the ARC. Furthermore,this area coincides with the localitiesthat record the deepest glacial incisioninto the carbonate platform (Fig. 7).The modern Antarctic ice cap isdrained by ice streams where regionsof rapidly moving land ice are in con-tact with slower moving ice on eitherside (Alley et al. 2004). Although theyaccount for only 10% of the volumeof the ice sheet, ice streams are size-able features, up to 50 km wide, 2000m thick, and can reach up to 700 kmlong (Joughin et al. 2001). These fastflowing zones can erode up to a kilo-metre deep into bedrock (Young et al.2011) and represent conveyor beltsthat potentially could transport sedi-ment to the ice sheet margins. Studieshave indicated that ice streams havehigh sediment fluxes between~100–1000 m3 yr−1 per metre of icefront, and sediment transport to conti-nental shelves by paleo-ice streams waseven greater (Alley et al. 2007). Theestimated sediment flux for the Nor-wegian Channel paleo-ice stream is8000 m3 yr−1 per metre width of the icestream front (Nygard et al. 2007). Thissediment is derived from large interiorbasins. The Rutford Ice Stream in WestAntarctica drains a ~45,600 km2 catch-

ment basin into the Ronne Ice Shelf(Joughin and Bamber 2005), but someAntarctic ice streams can drain basinsof >100 000 km2 (Dowdeswell et al.2006). The young age of these zircongrains suggests that such a Cryogenianice stream would have had to flow atleast 2000 km across the continentfrom Paterson Province and/orLeeuwin Complex in Western Aus-tralia. Alternatively, these young zircongrains (<760 Ma) may represent differ-ent zircon populations transported by amore local ice stream from sources insoutheastern Australia and/or easternAntarctica.

Previous geochronologicalstudies in the Adelaide Rift Complexhave placed constraints on the maxi-mum age of the Elatina glaciation. Ire-land et al. (1998) noted several youngdetrital zircons at ca. 650 Ma from thecorrelative Marino Arkose Member inthe Hallett Cove area south of Ade-laide. In particular, a detrital zirconU–Pb age of 657 ± 17 Ma wasobtained for a single grain that mayplace an upper limit on the deposition-al age of the Marino Arkose (Ireland etal. 1998; Preiss 2000). Our data cor-roborate these findings: the youngestzircon grains between ca. 650–665 Maare found within the Marino Arkose atHalletts Cove (Fig. 16), the Elatina Fm.at Elatina Creek and Walter’s Well (Fig.16 [3,7]), and in the correlative WhyallaFm. on the Stuart Shelf (n=7; Fig. 16[13,14]).

CONCLUSIONSThe basin scale architecture and widerange of sedimentary facies of the pre-and syn- late Cryogenian glacial sedi-mentary rocks of the ARC establishimportant constraints on the dynamicsof the Elatina glaciation in South Aus-tralia, providing insight into severalcontentious aspects of the snowballEarth hypothesis. The Yaltipena Fm.tidal flat sediments interfingered withand prograded over the pre-glacial car-bonate platform, which heralded theonset of the glaciation. This influx ofsediment and possible associated sealevel fall suggests that ice originatedfrom land, and thereby challenges acorollary of the snowball Earth modelthat suggests the first ice in the tropicswould arrive by the advance of sea gla-ciers (Hoffman et al. 2002). Measured

stratigraphic sections and carbon iso-topes were used to quantify ~130 m ofglacial erosion across the carbonateplatform and at least 500 m of erosionis inferred based upon the correlationof carbonate clasts within the diamic-tite to the underlying regionalchemostratigraphy. The δ13C measure-ments of carbonate clasts within theglacial diamictites give insight to therelative timing of δ13C acquisition. Thewide variability of the clasts shows thatthey record the full isotopic rangefrom –9‰ to +10‰ present in thecarbonate platform δ13C profile. Thisisotope conglomerate test supports theconclusion that the Trezona δ13Canomaly was recorded long before bur-ial diagenesis could have occurred(Rose et al. 2012).

Evidence for sub-glacial ero-sion of the carbonate platform by icestreams and ice-front instability withinan overall deglacial sequence remainscompatible with snowball Earth mod-els (Donnadieu et al. 2003; Halversonet al. 2004). Our evidence suggests thatthe local deglaciation and instanta-neous loss of gravitational attraction ofthe ice sheet on the nearby oceancaused a relative sea level fall, whichmay be comparable to Greenland dur-ing Pleistocene deglaciations. Currentand wave-generated combined-flowripples across the ARC attest to openseas with significant fetch during theinitial retreat of local glaciers. The cur-rent and wave-generated combined-flow ripples also figure prominentlyinto determining the low-latitude ofthe Elatina glaciation. We re-interpretthe folds used in the syn-sedimentarypaleomagnetic fold test, as stoss-depo-sitional transverse ripples with super-imposed oscillatory wave ripples.Although these observations weakenthe existing paleomagnetic constraintbecause the low paleolatitude is nowonly required to be pre-Late Cambrian,there is no evidence to suggest that thelow-latitude direction is a result ofremagnetization, and the positive rever-sal tests are at least consistent withsyn-depositional magnetization (Sohl etal. 1999).

Detrital zircon data providenew constraints on the provenance ofthe glacial sediment. The distributionof zircon ages indicate that at leastsome glacial sediment derived from the

286

cratons of Western Australia. Addi-tionally, zircon data from the late Cryo-genian periglacial Whyalla Fm., whichlong has been held as the stratigraphicequivalent and potential source for theElatina Fm., are sufficiently differentfrom the Elatina Fm. to infer that theWhyalla Fm. does not provide a signifi-cant source of sediment to the ElatinaFm.

Environmental and climaticconditions are challenging to interpretfrom the spatially heterogeneous glacialsedimentary rocks, especially whenstudied in isolation. However, ourwork within the Elatina Fm. demon-strates that an approach that integratesbasin-scale analysis with detailed sedi-mentology and chemostratigraphy,when set in the context of the pre- andpost-glacial sediments, can provide newinsights into the dynamics of extensiveglaciations of the Cryogenian. Newstudies pairing the sedimentology, geo-chemical and stratigraphic methodolo-gies applied to the Elatina Fm. inSouth Australia to the pre-, syn- andpost-glacial deposits on other conti-nents will be required to make progressin understanding Cryogenian glacialsediments.

ACKNOWLEDGEMENTSField, stable isotope and geochronolo-gy work was supported by NSF grantEAR-0842946 and a Sloan FoundationFellowship awarded to Maloof. JonHusson provided numerous construc-tive comments on drafts that greatlyimproved the paper. Paul Myrow pro-vided stimulating discussion in the fieldand Mauricio Perillo gave thoughtfulcomments concerning bedforms withinthe Elatina Fm. Fiona Best, BlakeDyer, Brehnin Keller, Laura Poppick,Justin Strauss, Nick Swanson-Hysell,Erica Wallstrom and Nora Xu provid-ed enthusiastic assistance in the field.Darren Crawford and ArthurCoulthard gave invaluable help access-ing the Flinders and Gammon RangesNational Parks. We are very grateful tothe landowners and pastoralists forland access. Ayami Aoyama, Claire Cal-met, Galen Gorski, Will Jacobsen,Jacquie Nesbit, Laura Poppick, JustinStrauss, and Nora Xu helped with sam-ple preparation. Some stable isotopemeasurements were performed at theUniversity of Michigan by Lora

Wingate and Kacey Lohmann and atPrinceton University by Laura Poppick.Some major element analyses for theCIA samples were run at MichiganState University by Tom Vogel, andcarbonate content of these sampleswas determined at Northwestern Uni-versity by Petra Sheaffova and BradSageman. We are thankful to GeraldPoirier and Nan Yao for help with theXRD analyses at Princeton Institutefor the Science and Technology ofMaterials. Andrew Kylander-Clark andGareth Seward are thanked for assis-tance with U–Pb geochronologic analy-ses at UCSB.

REFERENCESAbbot, D.S., Voigt, A., and Koll, D., 2011,

The Jormungand global climate stateand implications for Neoproterozoicglaciations: Journal of GeophysicalResearch, v. 116, D18103,http://dx.doi.org/10.1029/2011JD015927.

Allen, P.A., and Etienne, J.L., 2008, Sedi-mentary challenge to Snowball Earth:Nature Geoscience, v. 1, p. 817–825,http://dx.doi.org/10.1038/ngeo355.

Alley, R.B., Anandakrishnan, S., Dupont,T.K., and Parizek, B.R., 2004, Icestreams —fast, and faster?: ComptesRendus Physique, v. 5, p. 723–734,http://dx.doi.org/10.1016/j.crhy.2004.08.002.

Alley, R.B., Anandakrishnan, S., Dupont,T.K., Parizek, B.R., and Pollard, D.,2007, Effect of sedimentation on ice-sheet grounding-line stability: Science,v. 315, p. 1838–1841,http://dx.doi.org/10.1126/sci-ence.1138396.

Arnaud, E. and Eyles, C.H., 2006, Neopro-terozoic environmental change record-ed in the Port Askaig Formation,Scotland: Climatic vs tectonic controls:Sedimentary Geology, v. 183, p.99–124.

Bahlburg, H. and Dobrzinski, N., 2011, Areview of the Chemical Index ofAlteration (CIA) and its application tothe study of Neoproterozoic glacialdeposits and climate transitions, inArnaud, E., Halverson, G.P., andShields–Zhou, G., eds., The GeologicalRecord of Neoproterozoic Glacia-tions: Geological Society London,Memoirs, v. 36, p. 81–92,http://dx.doi.org/ 10.1144/M36.6.

Binks, P.J., 1968, Orroroo South Australia1:250 000 geological series map. SheetSI/54-1, Bureau of MineralResources, Australia.

Black, L.P., and Shaw, R.D., 1992, U–Pb

zircon chronology of prograde Pro-terozoic events in the Central andSouthern Provinces of the AruntaBlock, central Australia: AustralianJournal of Earth Sciences, v. 39, p.153–171, http://dx.doi.org/10.1080/08120099208728012.

Bowring, S.A., Grotzinger, J.P., Condon,D.J., Ramezani, J., Newall, M.J., andAllen, P.A., 2007, Geochronologicconstraints on the chronostratigraphicframework of the NeoproterozoicHuqf Supergroup, Sultanate ofOman: American Journal of Science,v. 307, p. 1097–1145,http://dx.doi.org/10.2475/10.2007.01.

Brenchley-Gaal, A.J., 1985, The influenceof faulting on late Proterozoic sedi-mentation and Delamerian tectonicdevelopment within the Nuccaleenaarea, Central Flinders Ranges, SouthAustralia: Unpublished B.Sc. (Hon-ours) thesis, University of Adelaide,Adelaide.

Budnick, A.S., 2012, Origins of the ElatinaRhythmites and the history of theEarth–Moon orbit: Junior Paper,Princeton University, Princeton, NJ.

Camacho, A., and Fanning, C.M., 1995,Some isotopic constraints on the evo-lution of the granulite and upperamphibolite facies terranes in the east-ern Musgrave Block, central Australia:Precambrian Research, v. 71, p.155–181, http://dx.doi.org/10.1016/0301-9268(94)00060-5.

Chen, C., 2012, Insights on Earth’s deepestδ13C excursion from paleocanyonstructures at Saint Ronan, Beltana Sta-tion, South Australia: Junior Paper,Princeton University, Princeton, NJ.

Christie-Blick, N., Sohl, L.E., and Kennedy,M., 1999, Considering a Neoprotero-zoic Snowball Earth: Science, v. 284,p. 1087, http://dx.doi.org/10.1126/science.284.5417.1087a.

Coats, R., 1965, Diapirism in the AdelaideGeosyncline: Australian PetroleumExploration Association Journal, v.1965, p. 98–102.

Coats, R.P., 1981, Late Proterozoic (Ade-laidean) tillites of the Adelaide Geo-syncline, in Hambrey, M.J., and Har-land, W.B., eds., Earth’s Pre-PleistoceneGlacial Record, Cambridge UniversityPress, Cambridge, p. 537–548.

Coats, R.P., and Blisset, A.H., 1971,Regional and economic geology of theMount Painter Province: GeologicalSurvey of South Australia Bulletin, no.43, p. 66–88.

Coats, R., Callen, R., Williams, A., andThompson, B., 1973, Copley 1:250000 Geological Map Sheet: Technical


report, Geological Survey of SouthAustralia, Department of Mines Ade-laide.

Colin, C., Kissel, C., Blamart, D., andTurpin, L., 1998, Magnetic propertiesof sediments in the Bay of Bengaland the Andaman Sea: impact ofrapid North Atlantic Ocean climaticevents on the strength of the Indianmonsoon: Earth and Planetary ScienceLetters, v. 160, p. 623–635,http://dx.doi.org/10.1016/S0012-821X(98)00116-2.

Collins, A., and Fitzsimons, I., 2001, Struc-tural, isotopic and geochemical con-straints on the evolution of theLeeuwin Complex, SW Australia(abstract), in Sircombe, K., and Li, Z.,eds., From Basins to Mountains:Rodinia at the turn of the century:Geological Society of AustraliaAbstract Series 65, p. 16–19.

Compston, W., Crawford, A., and Bofinger,V., 1966, A radiometric estimate ofthe duration of sedimentation in theAdelaide Geosyncline, South Australia:Journal of the Geological Society ofAustralia, v. 13, p. 229–276,http://dx.doi.org/10.1080/00167616608728611.

Compston, W., Williams, I.S., Jenkins,R.J.F., Gostin, V.A., and Haines, P.W.,1987, Zircon age evidence for the LatePrecambrian Acraman ejecta blanket:Australian Journal of Earth Sciences,v. 34, p. 435–445, http://dx.doi.org/10.1080/08120098708729424.

Condon, D.J., Prave, A.R., and Benn, D.I.,2002, Neoproterozoic glacial-rainoutintervals: Observations and implica-tions: Geology, v. 30, p. 35–38,http://dx.doi.org/10.1130/0091-7613(2002)030<0035:NGRIOA>2.0.CO;2.

Condon, D., Zhu, Maoyan, Bowring, S.,Wang, Wei, Yang, Aihua, and Jin,Yugan, 2005, U–Pb ages from theNeoproterozoic Doushantuo Forma-tion, China: Science, v. 308, p. 95–98,http://dx.doi.org/10.1126/sci-ence.1107765.

Coney, P.J., Edwards, A., Hine, R., Morri-son, F., and Windrim, D., 1990, Theregional tectonics of the Tasman oro-genic system, eastern Australia: Jour-nal of Structural Geology, v. 12, p.519–543, http://dx.doi.org/10.1016/0191-8141(90)90071-6.

Corcoran, P.L., and Mueller, W.U., 2002,The effects of weathering, sorting andsource composition in Archaean high-relief basins: examples from the SlaveProvince, Northwest Territories, Cana-da, in Altermann, W., and Corcoran,P., eds., Precambrian Sedimentary

Environments: A Modern Approachto Ancient Depositional Systems: Spe-cial Publication 33 of the Internation-al Association of Sedimentologists, p.183–211.

Cottle, J.M, Horstwood, M.S.A., and Par-rish, R.R., 2009a, A new approach tosingle shot laser ablation analysis andits application to in situ Pb/Ugeochronology: Journal of AnalyticalAtomic Spectrometry, v. 24, p.1355–1363,http://dx.doi.org/10.1039/b821899d.

Cottle, J.M., Jessup, M.J., Newell, D.L.,Horstwood, M.S.A., Noble, S.R., Par-rish, R.R., Waters, D.J., and Searle,M.P., 2009b, Geochronology of gran-ulitized eclogite from the Ama DrimeMassif: Implications for the tectonicevolution of the South TibetanHimalaya: Tectonics, v. 28, p. TC1002,http://dx.doi.org/10.1029/2008TC002256.

Cottle, J.M., Searle, M.P., Horstwood,M.S.A., and Waters, D.J., 2009c, Tim-ing of midcrustal metamorphism,melting and deformation in the MountEverest region of southern Tibetrevealed by U(–Th)–Pb geochronolo-gy: The Journal of Geology, v. 117, p.643–664,http://dx.doi.org/10.1086/605994.

Cox, R., Lowe, D.R., and Cullers, R.L.,1995, The influence of sediment recy-cling and basement composition onevolution of mudrock chemistry inthe southwestern United States:Geochimica et Cosmochimicha Acta,v. 59, p. 2919–2940,http://dx.doi.org/10.1016/0016-7037(95)00185-9.

Dalgarno, C.R. and Johnson, J.E., 1964,Glacials of the Marinoan Series, cen-tral Flinders Ranges: Geological Sur-vey of South Australia QuarterlyGeology Notes, v. 11, p. 3–4.

Dalgarno, C.R. and Johnson, J.E., 1968,Diapiric structures and late Precambri-an–early Cambrian sedimentation inthe Flinders Ranges, South Australia:American Association of PetroleumGeologists Special Volumes, Memoirv. 8, p. 301–314.

Dalziel, I.W.D., 1991, Pacific margins ofLaurentia and East Antarctica–Aus-tralia as a conjugate rift pair: Evidenceand implications for an Eocambriansupercontinent: Geology, v. 19, p.598–601, http://dx.doi.org/10.1130/0091-7613(1991)019<0598:PMOLAE>2.3.CO;2.

DeCelles, P.G., Quade, J., Kapp, P., Fan,M., Dettman, D.L., and Ding, L.,2007, High and dry in central Tibetanduring the Late Oligocene: Earth and

Planetary Science Letters, v. 253, p.389–401, http://dx.doi.org/10.1016/j.epsl.2006.11.001.

Deynoux, M., 1985, Terrestrial or waterlainglacial diamictites? Three case studiesfrom the late Proterozoic and lateOrdovician glacial drifts in WestAfrica: Palaeogeography, Palaeoclima-tology, Palaeoecology, v. 51, p. 97–141,http://dx.doi.org/10.1016/0031-0182(85)90082-3.

Deynoux, M. and Trompette, R., 1976,Late Precambrian mixtites: Glacialand/or nonglacial? Dealing especiallywith the mixtites of West Africa (Dis-cussion): American Journal of Science,v. 276, p. 1302–1315,http://dx.doi.org/10.2475/ajs.276.10.1302.

Dobrzinski, N., Bahlburg, H., Strauss, H.,and Zhang, Qirui, 2004, Geochemicalclimate proxies applied to the Neopro-terozoic glacial succession on theYangtze Platform, South China, inJenkins, G.S., McMenamin, M.A.S.,McKay, C.P., and Sohl, L., eds., TheExtreme Proterozoic: Geology, Geo-chemistry, and Climate: AmericanGeophysical Union, GeophysicalMonograph Series v. 146, p. 13–32.

Donnadieu, Y., Fluteau, F., Ramstein, G.,Ritz, C., and Besse, J., 2003, Is there aconflict between the Neoproterozoicglacial deposits and the snowballEarth interpretation: an improvedunderstanding with numerical model-ing: Earth and Planetary Science Let-ters, v. 208, p. 101–112,http://dx.doi.org/10.1016/S0012-821X(02)01152-4.

Dowdeswell, J.A., Ottesen, D., and Rise, L.,2006, Flow switching and large-scaledeposition by ice streams draining for-mer ice sheets: Geology, v. 34, p.313–316,http://dx.doi.org/10.1130/G22253.1.

Drexel, J., and Preiss, W.V., eds., 1995, Thegeology of South Australia, Volume 2,The Phanerozoic: Geological Surveyof South Australia Bulletin, v. 54, 347p.

Dyson, I., 1992, Stratigraphic nomencla-ture and sequence stratigraphy of thelower Wilpena Group, Adelaide Geo-syncline: the Sandison Subgroup:Geological Survey of South AustraliaQuarterly Geology Notes, v. 122, p.2–13.

Eastwood, E.N., Kocurek, G., Mohrig, D.,and Swanson, T., 2012, Methodologyfor reconstructing wind direction,wind speed and duration of windevents from aeolian cross-strata: Jour-nal of Geophysical Research, v. 117,F03035, http://dx.doi.org/

288

10.1029/2012JF002368.Elburg, M.A., Bons, P.D., Dougherty-Page,

J., Janka, C.E., Neumann, N., andSchaefer, B., 2001, Age and metaso-matic alteration of the Mt Neill Gran-ite at Nooldoonooldoona Waterhole,Mt Painter Inlier, South Australia:Australian Journal of Earth Sciences,v. 48, p. 721–730, http:dx.doi.org/10.1046/j.1440-0952.2001.485890.x.

Embleton, B.J.J., and Williams, G.E., 1986,Low palaeolatitude of deposition forlate Precambrian periglacial varvites inSouth Australia: implications forpalaeoclimatology: Earth and Plane-tary Science Letters, v. 79, p. 419–430,http://dx.doi.org/10.1016/0012-821X(86)90197-4.

Evans, D.A.D., 2000, Stratigraphic,geochronological, and paleomagneticconstraints upon the Neoproterozoicclimatic paradox: American Journal ofScience, v. 300, p. 347–433,http://dx.doi.org/10.2475/ajs.300.5.347.

Evans, D.A.D. and Raub, T.D., 2011, Neo-proterozoic glacial palaeolatitudes: aglobal update, inI Arnaud, E., Halver-son, G.P., and Sheilds-Zhou, G., eds.,The Geological Record of Neopro-terozoic Glaciations: The GeologicalSociety, London, Memoirs, v. 36, p.93–112,http:dx.doi.org/10.1144/M36.7.

Eyles, C.H., Eyles, N., and Grey, K., 2007,Palaeoclimate implications from deepdrilling of Neoproterozoic strata inthe Officer Basin and Adelaide RiftComplex of Australia; a marine recordof wet-based glaciers: Palaeogeogra-phy, Palaeoclimatology, Palaeoecology,v. 248, p. 291–312, http://dx.doi.org/10.1016/j.palaeo.2006.12.008.

Eyles, N., 1993, Earth’s glacial record andits tectonic setting: Earth–ScienceReviews, v. 35, p. 1–248,http://dx.doi.org/10.1016/0012-8252(93)90002-O.

Eyles, N., and Januszczak, N., 2004, ‘Zip-per-rift’: a tectonic model for Neopro-terozoic glaciations during thebreakup of Rodinia after 750 Ma:Earth–Science Reviews, v. 65, p. 1–73,http://dx.doi.org/10.1016/S0012-8252(03)00080-1.

Fairchild, I.J., 1993, Balmy shores and icywastes: The paradox of carbonatesassociated with glacial deposits inNeoproterozoic times, in Wright, V.P.,Sedimentology Review, v. 1: BlackwellPublishing Ltd, Oxford, UK, p. 1–16,http://dx.doi.org/10.1002/9781444304534.ch1.

Fanning, C., Ludwig, K., Forbes, B., andPreiss, W., 1986, Single and multiple

grain U–Pb zircon analyses for theEarly Adelaidean Rook Tuff,Willouran Ranges, South Australia(abstract): Geological Society of Aus-tralia, Abstract Series AB 15, p. 71–72.

Fanning, M.C., 2006, Constraints on thetiming of the Sturtian glaciation fromsouthern Australia; IE for the trueSturtian (abstract): Geological Societyof America Abstracts with Programs,v. 38, p. 115.

Fedo, C.M., Nesbitt, H.W., and Young,G.M., 1995, Unraveling the effects ofpotassium metasomatism in sedimen-tary rocks and paleosols, with implica-tions for paleoweathering conditionsand provenance: Geology, v. 23, p.921–924, http://dx.doi.org/10.1130/0091-7613(1995)023<0921:UTEOPM>2.3.CO;2.

Fedo, C.M., Eriksson, K.A., and Krogstad,E.J., 1996, Geochemistry of shalesfrom the Archean (~3.0 Ga) BuhwaGreenstone Belt, Zimbabwe: Implica-tions for provenance and source-areaweathering: Geochimica et Cos-mochimicha Acta, v. 60, p. 1751–1763,http://dx.doi.org/10.1016/0016-7037(96)00058-0.

Fedo, C.M., Young, G.M., Nesbitt, H.W.,and Hanchar, J.M., 1997a, Potassic andsodic metasomatism in the SouthernProvince of the Canadian Shield: Evi-dence from the Paleoproterozoic Ser-pent Formation, Huronian Super-group, Canada: Precambrian Research,v. 84, p. 17–36, http://dx.doi.org/10.1016/S0301-9268(96)00058-7.

Fedo, C.M., Young, G.M., and Nesbitt,H.W., 1997b, Paleoclimatic control onthe composition of the Paleoprotero-zoic Serpent Formation, HuronianSupergroup, Canada: a greenhouse toicehouse transition: PrecambrianResearch, v. 86, p. 201–223,http://dx.doi.org/10.1016/S0301-9268(97)00049-1.

Fike, D.A., Grotzinger, J.P., Pratt, L.M., andSummons, R.E., 2006, Oxidation ofthe Ediacaran Ocean: Nature, v. 444,p. 744–747, http://dx.doi.org/10.1038/nature05345.

Foden, J., Elburg, M.A., Dougherty-Page,J., and Burtt, A., 2006, The timing andduration of the Delamerian Orogeny:Correlation with the Ross Orogen andimplications for Gondwana assembly:The Journal of Geology, v. 114, p.189–210,http://dx.doi.org/10.1086/499570.

Forbes, B. and Preiss, W., 1987, Stratigra-phy of the Wilpena Group, in Preiss,W.V., ed., The Adelaide Geosyncline:Late Proterozoic stratigraphy, sedi-mentation, palaeontology and tecton-

ics: Geological Survey of South Aus-tralia Bulletin, v. 53, p. 211–248.

Fralick, P.W. and Kronberg, B.I., 1997,Geochemical discrimination of clasticsedimentary rock sources: Sedimenta-ry Geology, v. 113, p. 111–124,http://dx.doi.org/10.1016/S0037-0738(97)00049-3.

Fromhold, T.A., and Wallace, M.W., 2011,Nature and significance of the Neo-proterozoic Sturtian– MarinoanBoundary, Northern Adelaide Geo-syncline, South Australia: AustralianJournal of Earth Sciences, v. 58, p.599–613, http://dx.doi.org/10.1080/08120099.2011.579624.

Gerdes, A. and Zeh, A., 2006, CombinedU–Pb and Hf isotope LA–(MC–)ICP–MS analyses of detrital zircons:Comparison with SHRIMP and newconstraints for the provenance andage of an Armorican metasediment inCentral Germany: Earth and PlanetaryScience Letters, v. 249, p. 47–61,http://dx.doi.org/10.1016/j.epsl.2006.06.039.

Giddings, J.A., and Wallace, M.W., 2009,Facies-dependent δ13C variation from aCryogenian platform margin, SouthAustralia: Evidence for stratified Neo-proterozoic oceans?: Palaeogeography,Palaeoclimatology, Palaeoecology, v.271, p. 196–214, http://dx.doi.org/10.1016/j.palaeo.2008.10.011.

Giddings, J.A., Wallace, M.W., and Woon,E.M.S., 2009, Interglacial carbonatesof the Cryogenian UmberatanaGroup, northern Flinders Ranges,South Australia: Australian Journal ofEarth Sciences, v. 56, p. 907–925,http://dx.doi.org/10.1080/08120090903005378.

Goodge, J.W., Vervoort, J.D., Fanning,C.M., Brecke, D.M., Farmer, G.L.,Williams, I.S., Myrow, P.M., andDePaola, D.J., 2008, A Positive Test ofEast Antarctica–Laurentia juxtaposi-tion within the Rodinia Superconti-nent: Science, v. 321, p. 235–240,http://dx.doi.org/10.1126/sci-ence.1159189.

Grantham, J.H., and Velbel, M.A., 1988,The influence of climate and topogra-phy on rock-fragment abundance inmodern fluvial sands of the southernBlue Ridge Mountains, North Caroli-na: Journal of Sedimentary Research,v. 58, p. 219–227, http://dx.doi.org/10.1306/212F8D5F-2B24-11D7-8648000102C1865D.

Grotzinger, J.P., Watters, W.A., and Knoll,A.H., 2000, Calcified metazoans inthrombolite-stromatolite reefs of theterminal Proterozoic Nama Group,Namibia: Paleobiology, v. 26, p.


334–359, http://dx.doi.org/10.1666/0094-8373(2000)026<0334:CMITSR>2.0.CO;2.

Gubbins, D., 1999, The distinctionbetween geomagnetic excursions andreversals: Geophysical Journal Interna-tional, v. 137, p. F1–F4,http://dx.doi.org/10.1046/j.1365-246x.1999.00810.x.

Halverson, G.P., Hoffman, P.F., Schrag,D.P., and Kaufman, A.J., 2002, Amajor perturbation of the carboncycle before the Ghaub glaciation(Neoproterozoic) in Namibia: Preludeto snowball Earth?: Geochemistry,Geophysics, Geosystems, v. 3(6), p.1–24, http://dx.doi.org/10.1029/2001GC000244.

Halverson, G.P., Maloof, A.C., and Hoff-man, P.F., 2004, The Marinoan glacia-tion (Neoproterozoic) in northeastSvalbard: Basin Research, v. 16, p.297–324, http://dx.doi.org/10.1111/j.1365-2117.2004.00234.x.

Halverson, G.P., Hoffman, P.F., Schrag,D.P., Maloof, A.C., and Rice, A.H.N.,2005, Toward a Neoproterozoic com-posite carbon-isotope record: Geologi-cal Society of America Bulletin, v. 117,p. 1181–1207,http://dx.doi.org/10.1130/B25630.1.

Hambrey, M.J., and Harland, W.B., 1981,Earth’s Pre-Pleistocene GlacialRecord: Cambridge University Press,Cambridge, 1024 p.

Higgins, J.A., and Schrag, D.P., 2003, After-math of a snowball Earth: Geochem-istry, Geophysics, Geosystems, v. 4, p.1028, http:dx.doi.org/10.1029/2002GC000403.

Hoffman, P.F., 1991, Did the breakout ofLaurentia turn Gondwanalandinside–out?: Science, v. 252, p.1409–1412, http://dx.doi.org/10.1126/science.252.5011.1409.

Hoffman, P.F., 2005, On Cryogenian (Neo-proterozoic) ice-sheet dynamics andthe limitations of the glacial sedimen-tary record: 28th DeBeers Alex. DuToit Memorial Lecture, 2004, SouthAfrican Journal of Geology, v. 108, p.557–577,http://dx.doi.org/10.2113/108.4.557.

Hoffman, P.F., 2011, Strange bedfellows:glacial diamictite and cap carbonatefrom the Marinoan (635 Ma) glacia-tion in Nambia: Sedimentology, v. 58,p. 57–119, http://dx.doi.org/10.1111/j.1365-3091.2010.01206.x.

Hoffman, P.F., and Schrag, D.P., 2002, Thesnowball Earth hypothesis: testing thelimits of global change: Terra Nova, v.14, p. 129–155, http://dx.doi.org/10.1046/j.1365-3121.2002.00408.x.

Hoffman, P.F., Kaufman, A.J., Halverson,

G.P., and Schrag, D.P., 1998, A Neo-proterozoic Snowball Earth: Science,v. 281, p. 1342–1346,http://dx.doi.org/10.1126/sci-ence.281.5381.1342.

Hoffman, P.F., Halverson, G.P., andGrotzinger, J.P., 2002, Are Proterozoiccap carbonates and isotopic excur-sions a record of gas hydrate destabi-lization following Earth’s coldest inter-vals? (Comment): Geology, v. 30, p.286–287, http://dx.doi.org/10.1130/0091-7613(2002)030<0286:APCCAI>2.0.CO;2.

Hoffman, P.F., Halverson, G.P., Domack,E.W., Husson, J.M., Higgins, J.A., andSchrag, D.P., 2007, Are basal Edi-acaran (635 Ma) post-glacial “capdolostones” diachronous?: Earth andPlanetary Science Letters, v. 258, p.114–131, http://dx.doi.org/10.1016/j.epsl.2007.03.032.

Hoffmann, K.-H., Condon, D.J., Bowring,S.A., and Crowley, J.L., 2004, A U–Pbzircon date from the NeoproterozoicGhaub Formation, Namibia: Con-straints on Marinoan glaciation: Geol-ogy, v. 32, p. 817–820,http://dx.doi.org/10.1130/G20519.1.

Holland, D.M., Thomas, R.H., de Young,B., Ribergaard, M.H., and Lyberth, B.,2008, Acceleration of JakobshavnIsbræ triggered by warm subsurfaceocean waters: Nature Geoscience, v. 1,p. 659–664,http://dx.doi.org/10.1038/ngeo316.

Husson, J.M., Maloof, A.C., and Schoene,B., 2012, A syn-depositional age forEarth’s deepest δ13C excursionrequired by isotope conglomeratetests: Terra Nova, v. 24, p. 318–325,http://dx.doi.org/10.1111/j.1365-3121.2012.01067.x.

Ireland, T.R., Bradshaw, J.D., Muir, R.,Weaver, S., and Adams, C., 1994, Zir-con age distributions in granites,greywackes, and gneisses from thesouthwest Pacific Gondwana region(abstract): 8th International Conferenceon Geochronology, Cosmochronologyand Isotope Geology, p. 151.

Ireland, T.R., Flöttmann, T., Fanning, C.M.,Gibson, G.M., and Preiss, W.V., 1998,Development of the early PaleozoicPacific margin of Gondwana fromdetrital-zircon ages across theDelamerian orogen: Geology, v. 26, p.243–246, http://dx.doi.org/10.1130/0091-7613(1998)026<0243:DOTEPP>2.3.CO;2.

Jablonski, H., 1975, Late Precambrian geol-ogy of the Warren–Buckaringa Gorgearea, Flinders Ranges, South Australia:Unpublished B.Sc. (Honours) thesis,University of Adelaide, Adelaide.

Jackson, S.E., Pearson, N.J., Griffin, W.L.,and Belousova, E.A., 2004, The appli-cation of laser ablation- inductivelycoupled plasma-mass spectrometry toin situ U–Pb zircon geochronology:Chemical Geology, v. 211, p. 47–69,http://dx.doi.org/10.1016/j.chem-geo.2004.06.017.

James, N.P., Narbonne, G.M., and Kyser,T.K., 2001, Late Neoproterozoic capcarbonates: Mackenzie Mountains,northwestern Canada: precipitationand global glaciation meltdown: Cana-dian Journal of Earth Sciences, v. 38,p. 1229–1262,http://dx.doi.org/10.1139/e01-046.

Jenkins, R.J.F., 1990, The Adelaide FoldBelt: Tectonic reappraisal, in Jago, J.B.,and Moore, P.S., eds., The Evolution ofa Late Precambrian–Early PalaeozoicRift Complex: the Adelaide Geosyn-cline: Geological Society of AustraliaSpecial Publication, v. 16, p. 396–420.

Johnsson, M.J., 1993, The system control-ling the composition of clastic sedi-ments, in Johnsson, M.J., and Basu, A.,eds., Processes Controlling the Com-position of Clastic Sediments, Geo-logical Society America, SpecialPapers, v. 284, p. 1–20,http://dx.doi.org/10.1130/SPE284-p1.

Joughin, I. and Bamber, J.L., 2005, Thick-ening of the ice stream catchmentsfeeding the Filchner–Ronne Ice Shelf,Antarctica: Geophysical Research Let-ters, v. 32, p. L17503,http://dx.doi.org/10.1029/2005GL023844.

Joughin, I., Fahnestock, M.F., MacAyeal,D., Bamber, J.L., and Gogineni, P.,2001, Observation and analysis of iceflow in the largest Greenland icestream: Journal of GeophysicalResearch, v. 106, p. 34021–34034,http://dx.doi.org/10.1029/2001JD900087.

Karlstrom, K.E., and Bowring, S.A., 1988,Early Proterozoic assembly oftectonostratigraphic terranes in south-western North America: The Journalof Geology, v. 96, p. 561–576,http://dx.doi.org/10.1086/629252.

Kendall, B., Creaser, R.A., and Selby, D.,2006, Re–Os geochronology of post-glacial black shales in Australia: Con-straints on the timing of “Sturtian”glaciation: Geology, v. 34, p. 729–732,http://dx.doi.org/10.1130/G22775.1.

Kennedy, M.J., 1996, Stratigraphy, sedimen-tology, and isotopic geochemistry ofAustralian Neoproterozoic postglacialcap dolostones; Deglaciation, δ13Cexcursions, and carbonate precipita-tion: Journal of Sedimentary Research,

290

v. 66, p. 1050–1064,http://dx.doi.org/10.2110/jsr.66.1050.

Kennedy, M.J., Runnegar, B., Prave, A.R.,Hoffmann, K.-H., and Arthur, M.A.,1998, Two or four Neoproterozoicglaciations?: Geology, v. 26, p.1059–1063, http://dx.doi.org/10.1130/0091-7613(1998)026<1059:TOFNG>2.3.CO;2.

Kennedy, M.J., Christie-Blick, N., and Sohl,L.E., 2001, Are Proterozoic cap car-bonates and isotopic excursions arecord of gas hydrate destabilizationfollowing Earth’s coldest intervals?:Geology, v. 29, p. 443–446,http://dx.doi.org/10.1130/0091-7613(2001)029<0443:APCCAI>2.0.CO;2.

Kilner, B., Mac Niocaill, C., and Brasier,M., 2005, Low-latitude glaciation inthe Neoproterozoic of Oman: Geolo-gy, v. 33, p. 413–416,http://dx.doi.org/10.1130/G21227.1.

Kirschvink, J.L., 1992, Late Proterozoiclow-latitude global glaciation: Thesnowball Earth, in Schopf, J.W., andKlein, C., eds., The Proterozoic Bios-phere: a multidisciplinary study: Cam-bridge University Press, p. 51–52.

Knoll, A., Walter, M., Narbonne, G., andChristie-Blick, N., 2006, The Edi-acaran Period: a new addition to thegeologic time scale: Lethaia, v. 39, p.13–30, http://dx.doi.org/10.1080/00241160500409223.

Kopp, R.E., Simons, F.J., Mitrovica, J.X.,Maloof, A.C., and Oppenheimer, M.,2009, Probabilistic assessment of sealevel during the last interglacial stage:Nature, v. 462, p. 863–868,http://dx.doi.org/10.1038/nature08686.

Krissek, L.A., and Kyle, P.R., 1998, Geo-chemical indicators of weathering andCenozoic palaeoclimates in sedimentsfrom CRP-1 and CIROS-1, McMurdoSound, Antarctica: Terra Antartica, v.5, p. 673–680.

Lamb, M.P., Fischer, W.W., Raub, T.D., Per-ron, J.T., and Myrow, P.M., 2012, Ori-gin of giant wave ripples in snowballEarth cap carbonate: Geology, v. 40, p.827–830,http://dx.doi.org/10.1130/G33093.1.

Leather, J., Allen, P.A., Brasier, M.D., andCozzi, A., 2002, Neoproterozoicsnowball Earth under scrutiny: Evi-dence from the Fiq glaciation ofOman: Geology, v. 30, p. 891–894,http://dx.doi.org/10.1130/0091-7613(2002)030<0891:NSEUSE>2.0.CO;2.

Leeson, B., 1970, Geology of the Beltana1:63 360 map area: Technical report,Geological Survey of South Australia.

Le Heron, D.P., 2012, The Cryogenianrecord of glaciation and deglaciationin South Australia: Sedimentary Geol-ogy, v. 243–244, p. 57–69,http://dx.doi.org/10.1016/j.sed-geo.2011.09.013.

Le Heron, D.P., Cox, G., Trundley, A., andCollins, A.S., 2011a, Two Cryogenianglacial successions compared: Aspectsof the Sturt and Elatina sedimentrecords of South Australia: Precambri-an Research, v. 186, p. 147–168,http://dx.doi.org/10.1016/j.precam-res.2011.01.014.

Le Heron, D.P., Cox, G., Trundley, A., andCollins, A., 2011b, Sea ice-free condi-tions during the Sturtian glaciation(early Cryogenian), South Australia:Geology, v. 39, p. 31–34,http://dx.doi.org/10.1130/G31547.1.

Lemon, N., 1985, Physical modeling ofsedimentation adjacent to diapirs andcomparison with late PrecambrianOratunga breccia body in centralFlinders Ranges, South Australia:American Association of PetroleumGeologists Bulletin, v. 69, p.1327–1338.

Lemon, N.M., 2000, A Neoproterozoicfringing stromatolite reef complex,Flinders Ranges, South Australia: Pre-cambrian Research, v. 100, p. 109–120,http://dx.doi.org/10.1016/S0301-9268(99)00071-6.

Lemon, N., and Gostin, V., 1990,Glacigenic sediments of the late Pro-terozoic Elatina Forma- tion andequivalents, Adelaide Geosyncline,South Australia, in Jago, J.B., andMoore, P.S., eds., The Evolution of aLate Precambrian–Early PalaeozoicRift Complex: the Adelaide Geosyn-cline: Geological Society of AustraliaSpecial Publication Geological Societyof Australia Special Publication, v. 16,p. 149–164.

Lemon, N.M., and Reid, P.W., 1998, TheYaltipena Formation of the CentralFlinders Ranges: MESA Journal, v. 8,p. 37–39.

Løvlie, R., and Torsvik, T., 1984, Magneticremanence and fabric properties oflaboratory-deposited hematite-bearingred sandstone: Geophysical ResearchLetters, v. 11, p. 221–224,http://dx.doi.org/10.1029/GL011i003p00221.

Maboko, M.A.H., Williams, I.S., andCompston, W., 1991, Zircon U–Pbchronometry of the pressure and tem-perature history of granulites in theMusgrave Ranges, central Australia:The Journal of Geology, v. 99, p.675–697,http://dx.doi.org/10.1086/629532.

Macdonald, F.A., Schmitz, M.D., Crowley,J.L., Roots, C.F., Jones, D.S., Maloof,A.C., Strauss, J.V., Cohen, P.A., John-ston, D.T., and Schrag, D.P., 2010, Cal-ibrating the Cryogenian: Science, v.327, p. 1241–1243, http://dx.doi.org/10.1126/science.1183325.

Malmon, D.V., Dunne, T., and Reneau,S.L., 2003, Stochastic theory of parti-cle trajectories through alluvial valleyfloors: The Journal of Geology, v.111, p. 525–542, http://dx.doi.org/10.1086/376764.

Maloof, A.C., Rose, C.V., Beach, R.,Samuels, B.M., Calmet, C.C., Erwin,D.H., Poirier, G.R., Yao, Nan, andSimons, F.J., 2010, Possible animal-body fossils in pre-Marinoan lime-stones from South Australia: NatureGeoscience, v. 3, p. 653–659,http://dx.doi.org/10.1038/ngeo934.

Mawson, D., 1949, A third occurrence ofglaciation evidenced in the AdelaideSystem: Transactions of the RoyalSociety of South Australia, v. 73, p.117–121.

McLennan, S.M., Hemming, S., McDaniel,D.K., and Hanson, G.N., 1993, Geo-chemical approaches to sedimentation,provenance and tectonics, in Johnsson,M.J., and Basu, A., eds., ProcessesControlling the Composition of Clas-tic Sediments: Geological Society ofAmerica Special Papers, v. 284, p.21–40, http://dx.doi.org/10.1130/SPE284-p21.

McNaughton, N.J., and Goellnicht, N.M.,1990, The age and radiothermal prop-erties of the Mount Crofton Granite,Telfer area, Western Australia: Aus-tralian Journal of Earth Sciences, v.37, p. 103–106, http://dx.doi.org/10.1080/08120099008727909.

Miller, R.K., 1975, The late Precambriangeology of the WyaccaBluff–Buckaringa Gorge area, FlindersRanges, South Australia: UnpublishedB.Sc. (Honours) thesis, University ofAdelaide, Adelaide.

Mitrovica, J.X., Gomez, N., and Clark, P.U.,2009, The sea-level fingerprint ofWest Antarctic collapse: Science, v.323, p. 753, http://dx.doi.org/10.1126/science.1166510.

Moores, E.M., 1991, Southwest US–EastAntarctic (SWEAT) connection: Ahypothesis: Geology, v. 19, p. 425–428,http://dx.doi.org/10.1130/0091-7613(1991)019<0425:SUSEAS>2.3.CO;2.

Myrow, P.M., 1992, Pot and gutter castsfrom the Chapel Island Formation,Southeast Newfoundland: Journal ofSedimentary Research, v. 62, p.992–1007,


http://dx.doi.org/10.2110/jsr.62.992.Nelson, D.R., 1995, Compilation of

SHRIMP U–Pb zircon geochronologydata, 1994: Geological Survey ofWestern Australia Geological Survey,Record 1995/3, 244 p.

Nelson, D.R., 1996, Compilation ofSHRIMP U–Pb zircon geochronologydata, 1995: Geological Survey ofWestern Australia, Record 1996/ 5,168 p.

Nelson, D.R., 1999, Compilation ofgeochronology data, 1998: GeologicalSurvey of Western Australia, Record1999/2, 222 p.

Nelson, D., 2002, Compilation ofGeochronological Data, 2001: Geo-logical Survey of Western Australia, v.2.

Nesbitt, H.W., and Young, G.M., 1982,Early Proterozoic climates and platemotions inferred from major elementchemistry of lutites: Nature, v. 299, p.715–717,http://dx.doi.org/10.1038/299715a0.

Nesbitt, H.W., and Young, G.M., 1984,Prediction of some weathering trendsof plutonic and volcanic rocks basedon thermodynamic and kinetic consid-erations: Geochimica et Cos-mochimicha Acta, v. 48, 1523–1534,http://dx.doi.org/10.1016/0016-7037(84)90408-3.

Nesbitt, H.W., and Young, G.M., 1989,Formation and diagenesis of weather-ing profiles: The Journal of Geology,v. 97, p. 129–147,http://dx.doi.org/10.1086/629290.

Nesbitt, H.W., and Young, G.M., 1996, Pet-rogenesis of sediments in the absenceof chemical weathering: effects ofabrasion and sorting on bulk composi-tion and mineralogy: Sedimentology, v.43, p. 341–358, http://dx.doi.org/10.1046/j.1365-3091.1996.d01-12.x.

Noyes, R.W., Weiss, N.O., and Vaughan,A.H., 1984, The relation between stel-lar rotation rate and activity cycle peri-ods: Astrophysics Journal, v. 287, p.769–773,http://dx.doi.org/10.1086/162735.

Nygård, A., Sejrup, H.P., Haflidason, H.,Lekens, W.A.H., Clark, C.D., and Bigg,G.R., 2007, Extreme sediment and icedischarge from marine-based icestreams: New evidence from theNorth Sea: Geology, v. 35, p. 395–398,http://dx.doi.org/10.1130/G23364A.1.

Panahi, A. and Young, G.M., 1997, A geo-chemical investigation into the prove-nance of the Neoproterozoic PortAskaig Tillite, Dalradian Supergoup,western Scotland: PrecambrianResearch, v. 85, p. 81–96,

http://dx.doi.org/10.1016/S0301-9268(97)00033-8.

Paton, J., Woodhead, J.D., Hellstrom, J.C.,Hergt, J.M., Greig, A., and Maas, R.,2010, Improved laser ablation U–Pbzircon geochronology through robustdownhole fractionation correction:Geochemistry, Geophysics, Geosys-tems, v. 11, Q0AA06,http://dx.doi.org/10.1029/2009GC002618.

Pell, S.D., Williams, I.S., and Chivas, A.R.,1997, The use of protolith zircon-agefingerprints in determining the proto-source areas for some Australian dunesands: Sedimentary Geology, v. 109, p.233–260, http://dx.doi.org/10.1016/S0037-0738(96)00061-9.

Pidgeon, R.T., 1990, Timing of plutonismin the Proterozoic Albany Mobile Belt,southwestern Australia: PrecambrianResearch, v. 47, p. 157–167,http://dx.doi.org/10.1016/0301-9268(90)90036-P.

Plummer, P.S., 1978, Note on the palaeoen-vironmental significance of the Nuc-caleena formation (upper Precambri-an), central Flinders Ranges, SouthAustralia: Journal of the GeologicalSociety of Australia, v. 25, p. 395–402,http://dx.doi.org/10.1080/00167617808729049.

Preiss, W.V., 1987, Stratigraphic nomencla-ture and classification: in Preiss, W.V.,ed., The Adelaide Geosyncline: LateProterozoic stratigraphy, sedimenta-tion, palaeontology and tectonics:Geological Survey of South Australia,v. 53.

Preiss, W.V., 2000, The Adelaide Geosyn-cline of South Australia and its signifi-cance in Neoproterozoic continentalreconstruction: Precambrian Research,v. 100, p. 21–63, http://dx.doi.org/10.1016/S0301-9268(99)00068-6.

Preiss, W.V., and Forbes, B.G., 1981,Stratigraphy, correlation and sedimen-tary history of Adelaidean (late Pro-terozoic) basins in Australia: Precam-brian Research, v. 15, p. 255–304,http://dx.doi.org/10.1016/0301-9268(81)90054-1.

Preiss, W. and Robertson, R., 2002, SouthAustralian mineral explorers guide:Technical report, Department of Pri-mary Industries and Regions (PIRSA),Government of South Australia.

Preiss, W., Dyson, I., Reid, P., and Cowley,W., 1998, Revision of lithostratigraph-ic classification of the UmberatanaGroup: MESA Journal, v. 9, p.36–42.Raub, T.D., and Evans, D.A.,2006, Magnetic reversals in basal Edi-acaran cap carbonates: A criticalreview (abstract): Eos, Transaction-

sAmerican Geophysical Union, v. 87,p. GP41B–02.

Raub, T. and Evans, D., 2008, Paleolati-tudes of Neoproterozoic snowball gla-cial deposits: Biases and synthesis: The33rd International Geological Con-gress, Oslo.

Rieu, R., Allen, P.A., Cozzi, A., Kosler, J.,and Bussy, F., 2007a, A compositestratigraphy for the NeoproterozoicHuqf Supergroup of Oman: integrat-ing new litho-, chemo- and chronos-tratigraphic data of the Mirbat area,southern Oman: Journal of the Geo-logical Society, v. 164, p. 997–1009,http://dx.doi.org/10.1144/0016-76492006-114.

Rieu, R., Allen, P.A., Plötze, M., and Pettke,T., 2007b, Climatic cycles during aNeoproterozoic “snowball” glacialepoch: Geology, v. 35, p. 299 – 302,http://dx.doi.org/10.1130/G23400A.1.

Rieu, R., Allen, P.A., Plotze, M., and Pettke,T., 2007c, Compositional and miner-alogical variations in a Neoproterozoicglacially influenced succession, Mirbatarea, south Oman: Implications forpaleoweathering conditions: Precam-brian Research, v. 154, p. 248–265,http://dx.doi.org/10.1016/j.precam-res.2007.01.003.

Rose, C.V., and Maloof, A.C., 2010, Testingmodels for post-glacial ‘cap dolostone’deposition: Nuccaleena Formation,South Australia: Earth and PlanetaryScience Letters, v. 296, p. 165–180,http://dx.doi.org/10.1016/j.epsl.2010.03.031.

Rose, C.V., Swanson-Hysell, N.L., Husson,J.M., Poppick, L.N., Cottle, J.M.,Schoene, B., and Maloof, A.C., 2012,Constraints on the origin and relativetiming of the Trezona δ13C anomalybelow the end-Cryogenian glaciation:Earth and Planetary Science Letters, v.319–320, p. 241–250,http://dx.doi.org/10.1016/j.epsl.2011.12.027.

Rothman, D.H., Hayes, J.M., and Sum-mons, R.E., 2003, Dynamics of theNeoproterozoic carbon cycle: Pro-ceedings of the National Academy ofSciences, v. 100, p. 8124–8129,http://dx.doi.org/10.1073/pnas.0832439100.

Rubin, D.M., and Hunter, R., 1982, Bed-form climbing in theory and nature:Sedimentology, v. 29, p. 121–138,http://dx.doi.org/10.1111/j.1365-3091.1982.tb01714.x.

Scheffler, K., Hoernes, S., and Schwark, L.,2003, Global changes during Car-boniferous–Permian glaciation ofGondwana: Linking polar and equato-

292

rial climate evolution by geochemicalproxies: Geology, v. 31, p. 605–608,http://dx.doi.org/10.1130/0091-7613(2003)031<0605:GCDC-GO>2.0.CO;2.

Schermerhorn, L.J.G., 1974, Later Precam-brian mixites: Glacial and/or non-gla-cial?: American Journal of Science, v.274, p. 673–824, http://dx.doi.org/10.2475/ajs.274.7.673.

Schmidt, P.W., and Williams, G.E., 1995,The Neoproterozoic climatic paradox:Equatorial palaeolatitude for Marinoanglaciation near sea level in South Aus-tralia: Earth and Planetary ScienceLetters, v. 134, p. 107–124,http://dx.doi.org/10.1016/0012-821X(95)00106-M.

Schmidt, P.W., Williams, G.E., andMcWilliams, M.O., 2009, Palaeomag-netism and magnetic anisotropy oflate Neoproterozoic strata, South Aus-tralia: Implications for the palaeolati-tude of late Cryogenian glaciation, capcarbonate and the Ediacaran System:Precambrian Research, v. 174, p.35–52, http://dx.doi.org/10.1016/j.precamres.2009.06.002.

Sheldon, N.D., Retallack, G.T., and Tanaka,S., 2002, Geochemical Climofunctionsfrom North American Soils andApplication to Paleosols across theEocene–Oligocene Boundary in Ore-gon: The Journal of Geology, v. 110,p. 687–696,http://dx.doi.org/10.1086/342865.

Shields, G.A., 2005, Neoproterozoic capcarbonates: a critical appraisal ofexisting models and the plumeworldhypothesis: Terra Nova, v. 17, p.299–310, http://dx.doi.org/10.1111/j.1365-3121.2005.00638.x.

Sohl, L.E., Christie–Blick, N., and Kent,D.V., 1999, Paleomagnetic polarityreversals in Marinoan (ca. 600 Ma) gla-cial deposits of Australia: Implicationsfor the duration of low-latitude glacia-tion in Neoproterozoic time: Geologi-cal Society of America Bulletin, v. 111,p. 1120–1139, http://dx.doi.org/10.1130/0016-7606(1999)111<1120:PPRIMC>2.3.CO;2.

Spencer, A.M., 1971, Late Pre-cambrianglaciation in Scotland: GeologicalSociety, London, Memoirs, v. 6, p.5–100, http://dx.doi.org/10.1144/GSL.MEM.1971.006.01.02.

Stacey, J.S., and Kramers, J.D., 1975,Approximation of terrestrial lead iso-tope evolution by a two-stage model:Earth and Planetary Science Letters, v.26, p. 207–221, http://dx.doi.org/10.1016/0012-821X(75)90088-6.

Sumner, D.Y., Kirschvink, J.L., and Run-negar, B.N., 1987, Soft-sediment pale-

omagnetic field tests of late Precam-brian glaciogenic sediments (abstract):EOS Transactions of the AmericanGeophysical Union, v. 68, p. 1251.

Swanson-Hysell, N.L., Rose, C.V., Calmet,C.C., Halverson, G.P., Hurtgen, M.T.,and Maloof, A.C., 2010, Cryogenianglaciation and the onset ofcarbon–isotope decoupling: Science, v.328, p. 608– 611, http://dx.doi.org/10.1126/science.1184508.

Swart, P.K., and Kennedy, M.J., 2012, Doesthe global stratigraphic reproducibilityof δ13C in Neoproterozoic carbonatesrequire a marine origin? A Pliocene–Pleistocene comparison: Geology, v.40, p. 87–90,http://dx.doi.org/10.1130/G32538.1.

Tauxe, L., and Kent, D.V., 1984, Propertiesof a detrital remanence carried byhematite from study of modern riverdeposits and laboratory redepositionexperiments: Geophysical Journal ofthe Royal Astronomical Society, v. 76,p. 543–561, http://dx.doi.org/10.1111/j.1365-246X.1984.tb01909.x.

Teale, G., 1993, Mount Painter and MountBabbage Inliers, in Drexel, J., Preiss,W., and Parker, A., eds., The Geologyof South Australia I. The Precambri-an: Geological Survey South AustraliaBulletin, v. 54, p. 93–100.

Thomson, B., Coats, R., Mirams, R.,Forbes, B., Dalgarno, C., and Johnson,J., 1964, Precambrian rock groups inthe Adelaide Geosyncline: new subdi-vision.: Quarterly Journal of the Geo-logical Survey of South Australia, v. 9,p. 1–19.

Trindade, R.I.F., Font, E., D’Agrella-Filho,M.S., Nogueira, A.C.R., and Riccomi-ni, C., 2003, Low-latitude and multiplegeomagnetic reversals in the Neopro-terozoic Puga cap carbonate, Amazoncraton: Terra Nova, v. 15, p. 441–446,http://dx.doi.org/10.1046/j.1365-3121.2003.00510.x.

Turner, J., Chenoli, S.N., abu Samah, A.,Marshall, G., Phillips, T., and Orr, A.,2009, Strong wind events in theAntarctic: Journal of GeophysicalResearch, v. 114, D18103,http://dx.doi.org/10.1029/2008JD011642.

Walter, M.R., and Bauld, J., 1983, Theassociation of sulphate evaporites,stromatolitic carbonates and glacialsediments: Examples from the Pro-terozoic of Australia and the Caino-zoic of Antarctica: PrecambrianResearch, v. 21, p. 129–148,http://dx.doi.org/10.1016/0301-9268(83)90008-6.

Webb, B., 1960, Diapiric structures in theFlinders Ranges, South Australia: Aus-

tralian Journal of Science, v. 22, p.390–391.

Wiedenbeck, M., Alle, P., Corfu, F., Griffin,W., Meier, M., Oberli, F., Von Quadt,A., Roddick, J., and Spiegel, W., 1995,Three natural zircon standards forU–Th–Pb, Lu–Hf, trace element andREE analyses: Geostandards Newslet-ter, v. 19, p. 1–23, http://dx.doi.org/10.1111/j.1751-908X.1995.tb00147.x.

Wilde, S. and Nelson, D., 2001, Geology ofthe western Yilgarn Craton andLeeuwin Complex, Western Aus-tralia—a field guide: Geological Sur-vey of Western Australia Record 2001,v. 15.

Williams, G.E., 1977, Late Precambriandolomites, Vendian glaciation, andsynchroneity of Vendian glaciations: Adiscussion: The Journal of Geology, v.85, p. 250–251,http://dx.doi.org/10.1086/628291.

Williams, G.E., 1979, Sedimentology, sta-ble-isotope geochemistry andpalaeoenvironment of dolostones cap-ping late Precambrian glacialsequences in Australia: Journal of theGeological Society of Australia, v. 26,p. 377–386, http://dx.doi.org/10.1080/00167617908729104.

Williams, G.E., 1981, Sunspot Periods inthe late Precambrian glacial climateand solar-planetary relations: Nature,v. 291, p. 624–628,http://dx.doi.org/10.1038/291624a0.

Williams, G.E., 1985, Solar affinity of sedi-mentary cycles in the Late Precambri-an Elatina Formation: Australian Jour-nal of Physics, v. 38, p. 1027–1044,http://dx.doi.org/10.1071/PH851027.

Williams, G.E., 1986, Precambrian per-mafrost horizons as indicators ofpalaeoclimate: Precambrian Research,v. 32, p. 233–242, http://dx.doi.org/10.1016/0301-9268(86)90008-2.

Williams, G.E., 1988, Cyclicity in the latePrecambrian Elatina Formation, SouthAustralia: Solar or Tidal Signature?:Climate Change, v. 13, p. 117–128,http://dx.doi.org/10.1007/BF00140565.

Williams, G.E., 1989, Late Precambriantidal rhythmites in South Australia andthe history of the Earth’s rotation:Journal of the Geological Society, v.146, p. 97–111, http://dx.doi.org/10.1144/gsjgs.146.1.0097.

Williams, G.E., 1991, Upper Proterozoictidal rhythmites South Australia: sedi-mentary features, deposition, andimplications for the Earths paleorota-tion, in Smith, D.G., Reinson, G.E.,Zaitlin, B.A., and Rahmani, R.A., eds.,Clastic Tidal Sedimentology: CanadianSociety of Petroleum Geologists,


Memoir, v. 16, p. 161–178.Williams, G.E., 1994, The enigmatic Late

Proterozoic glacial climate: an Aus-tralian perspective, in Deynoux, M.,Miller, J.M.G., Domack, E.W., Eyles,N., Fairchild, I., and Young, G.M., eds.,Earth’s Glacial Record: CambridgeUniversity Press, p. 146–164,http://dx.doi.org/10.1017/CBO9780511628900.012.

Williams, G.E., 1996, Soft-sediment defor-mation structures from the Marinoanglacial succession, Adelaide foldbelt:implications for the palaeoaltitude oflate Neoproterozoic glaciation: Sedi-mentary Geology, v. 106, p. 165–175,http://dx.doi.org/10.1016/S0037-0738(96)00062-0.

Williams, G.E., 1997, Precambrian lengthof day and the validity of tidal rhyth-mite paleotidal values: GeophysicalResearch Letters, v. 24, p. 421–424,http://dx.doi.org/10.1029/97GL00234.

Williams, G.E., 1998, Late Neoproterozoicperiglacial aeolian sand sheet, StuartShelf, South Australia: Australian Jour-nal of Earth Sciences, v. 45, p.733–741, http://dx.doi.org/10.1080/08120099808728429.

Williams, G.E., 2000, Geological con-straints on the Precambrian history ofEarth’s rotation and the Moon’s orbit:Reviews of Geophysics, v. 38, p.37–59, http://dx.doi.org/10.1029/1999RG900016.

Williams, G.E., and Sonett, C.P., 1985,Solar signature in sedimentary cyclesfrom the late Precambrian ElatinaFormation, Australia: Nature, v. 318,p. 523–527,http://dx.doi.org/10.1038/318523a0.

Williams, G.E., and Tonkin, D.G., 1985,Periglacial structures and palaeoclimat-ic significance of a late Precambrianblock field in the Cattle Grid coppermine, Mount Gunson, South Aus-tralia: Australian Journal of Earth Sci-ences, v. 32, p. 287–300,http:dx.doi.org/10.1080/08120098508729331.

Williams, G.E., Gostin, V.A., McKirdy,D.M., and Preiss, W.V., 2008, TheElatina glaciation, late Cryogenian(Marinoan Epoch), South Australia:Sedimentary facies and palaeoenviron-ments: Precambrian Research, v. 163,p. 307–331, http://dx.doi.org/10.1016/j.precamres.2007.12.001.

Williams, I., Chen, Y., Chappell, B., and

Compston, W., 1988, Dating thesources of Bega batholith granites byion microprobe (abstract): 9th Aus-tralian Geological Convention, p. 424.

Willis, I.L., Brown, R.E., Stroud, W.J., andStevens, B.P.J., 1983, The Early Pro-terozoic Willyama supergroup: Strati-graphic subdivision and interpretationof high to low-grade metamorphicrocks in the Broken Hill Block, NewSouth Wales: Journal of the Geologi-cal Society of Australia, v. 30, p. 195–224, http://dx.doi.org/10.1080/00167618308729249.

Wingate, M.T.D., Campbell, I.H., Comp-ston, W. and Gibson, G.M., 1998, Ionmicroprobe U–Pb ages for Neopro-terozoic basaltic magmatism in south-central Australia and implications forthe breakup of Rodinia: PrecambrianResearch, v. 87, p. 135–159,http://dx.doi.org/10.1016/S0301-9268(97)00072-7.

Young, D.A., Wright, A.P., Roberts, J.L.,Warner, R.C., Young, N.W., Green-baum, J.S., Schroeder, D.M., Holt, J.W.,Sugden, D.E., Blankenship, D.D., vanOmmen, T.D., and Siegert, M.J., 2011,A dynamic early East Antarctic IceSheet suggested by ice-covered fjordlandscapes: Nature, v. 474, p. 72–75,http://dx.doi.org/10.1038/nature10114.

Young, G.M., 2002, Stratigraphic and tec-tonic settings of Proterozoic glacio-genic rocks and banded iron-forma-tions: relevance to the snowball Earthdebate: Journal of African Earth Sci-ences, v. 35, p. 451–466,http://dx.doi.org/10.1016/S0899-5362(02)00158-6.

Zahnle, K.J., and Walker, J.C.G., 1987, Cli-matic oscillations during the Precam-brian era: Climatic Change, v. 10, p.269–284, http://dx.doi.org/10.1007/BF00143906.

Received January 2013Accepted as revised August 2013

For access to Rose et al. supplementary material and large fold-out figures 6 through 11, please visit the GAC’s open source GCData Repository at: http://www.gac.ca/wp/?page_id=306.

AUL OFFMAN ERIES - Princeton University · 1971; Deynoux and Trompette 1976; Hambrey and Harland...

Documents

Transcript of AUL OFFMAN ERIES - Princeton University · 1971; Deynoux and Trompette 1976; Hambrey and Harland...