Paleoenvironmental reconstruction of the Ediacaran Araras platform (Western Brazil) from the sedimentary and trace metals record

Precambrian Research 241 (2014) 185– 202

Contents lists available at ScienceDirect

Precambrian Research

jou rn al h om epa ge: www.elsev ier .com/ locate /precamres

Paleoenvironmental reconstruction of the Ediacaran Araras platform(Western Brazil) from the sedimentary and trace metals record

Pierre Sansjofre a,b,∗, Ricardo I.F. Trindadeb, Magali Ader a, Joelson Lima Soares c,Afonso C.R. Nogueira c, Nicolas Tribovillardd

a Équipe de géochimie des isotopes stables, Institut de Physique du Globe de Paris, Sorbonne Paris Cité, Univ Paris Diderot, UMR 7154 CNRS,F-75005 Paris, Franceb Departamento de Geofisica, Instituto de Astronomia, Geofísica e Ciências Atmosféricas, Universidade de São Paulo, Rua do Matão 1226,05508-900 São Paulo, Brazilc Faculdade de Geologia, Instituto de Geociências, Universidade Federal do Pará, CEP 66.075-110 Belém, Brazild Université Lille 1 et CNRS, laboratoire Géosystèmes UMR 3298, 59655 Villeneuve d’Ascq Cedex, France

a r t i c l e i n f o

Article history:Received 12 February 2013Received in revised form19 November 2013Accepted 23 November 2013Available online 7 December 2013

Keywords:Carbonate platformRedoxNeoproterozoicSnowball EarthTrace metal geochemistry

a b s t r a c t

Ocean and atmosphere oxidation in the Ediacaran Period paved the way for the dawn of animals. In anattempt to better document the record of seawater redox state at the onset of the Ediacaran, we performeda paleoenvironmental study of post-Marinoan (635 Ma) carbonate strata from the southeastern marginof the Amazon craton, western Brazil. Five sections were sampled along the Araras carbonate platform.Outcrop-based facies analysis, complemented by petrographic description of representative samples, wasperformed on these sections. Seven facies associations (FA) were recognized. Four FA are encountered inthe inner-shelf, from the basal glacio-marine deposits of the Puga Formation to shallow and moderatelydeep platform facies, which are systematically covered by deeper CaCO3 over-saturated facies. Five FAoccur on the outer-shelf, including storm-wave influenced facies, below storm-wave base facies andslope intraformational breccia. These facies associations indicate a transgressive systems tract over thecarbonate platform. Trace metals (U, Mo, Zn, Pb, Cd, Cu, Ni, V) and Al concentrations, pyrite abundance andtotal organic carbon (TOC) contents in these sections are generally low. They nonetheless present, frombase to top of the sections, significant stratigraphic variations which can be traced along the platform, withsuccessive enrichments of (i) Pb and Zn, (ii) U, and (iii) both U and Mo in the thin marl levels containing thehighest amount of organic carbon and pyrite (0.4% and 1.9%, respectively). We interpret this succession torecord the progressive evolution of sediment pore-waters from oxic to more reducing conditions, drivenby an increase in sedimentary organic matter accumulation. The lack of evidence for persistent sulphidemineralization and associated enrichments in Mo suggests that sulfidic conditions were only achievedin sediment pore waters. In the aftermath of the Marinoan glaciation the water column must then havebeen essentially oxic on the Araras platform, with anoxia only sporadically reaching the sediment–watercolumn interface in the deepest parts of the platform.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

A Neoproterozoic Oxidation Event (NOE) is widely believed tomark the transition from anoxic to oxic conditions in the deep oceanat about the same time as the radiation of multi-cellular eukaryotes(Johnston et al., 2010, 2012; Lyons, 2007; Planavsky et al., 2011;Scott et al., 2008; Och and Shields-Zhou, 2012). More specifically,

∗ Corresponding author. Present address: Laboratoire Domaines Océaniques, Uni-versité de Bretagne Occidentale, UMR 6538, 29820 Plouzané, France.Tel.: +33 622828674.

E-mail addresses: [email protected], [email protected],[email protected] (P. Sansjofre).

most authors have argued that this oxygenation event occurredduring the Ediacarian, i.e., after the 635 Ma Marinoan glaciation.However, several authors proposed a redox-stratified ocean dur-ing this interval of time (Canfield et al., 2007; Chang et al., 2012;McFadden et al., 2007; Wang et al., 2012). For example, basedon paired carbon isotopic analyses on Ediacarian carbonates fromSouth China (the Doushantuo Formation), Ader et al. (2009) pro-posed a redox stratified water column with three different redoxzones: oxic, sulfidic and methanogenic. Using iron speciation datafrom the same succession, Li et al. (2010) proposed a redox stratifiedocean with a sulfidic zone in the inner-shelf and a ferruginous zonein the basin. Although the details of these models differ, they bothagreed on the presence of an anoxic deep-water mass during thedeposition of the thick carbonate horizon overlying the Marinoan

0301-9268/$ – see front matter © 2013 Elsevier B.V. All rights reserved.

http://dx.doi.org/10.1016/j.precamres.2013.11.004

dx.doi.org/10.1016/j.precamres.2013.11.004

http://www.sciencedirect.com/science/journal/03019268

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

http://crossmark.crossref.org/dialog/?doi=10.1016/j.precamres.2013.11.004&domain=pdf

mailto:[email protected]



dx.doi.org/10.1016/j.precamres.2013.11.004

186 P. Sansjofre et al. / Precambrian Research 241 (2014) 185– 202

diamictite in the Yangtze platform. This interpretation was rein-forced by the high !34S of sulfate during the Neoproterozoic erasuggesting high bacterial sulfate-reduction activity (Hurtgen et al.,2002). However, the model of global deep ocean anoxia during theEdiacaran Period has been challenged by new U/TOC and Mo/TOCdata (Sahoo et al., 2012), which suggests instead that the post-Marinoan ocean was dominantly oxygenated, with anoxic watermasses possibly being restricted to intracratonic basins and oxygenminimum zones. To date, most redox data for the Ediacarian Periodare derived from the Yangtze platform (Ader et al., 2009; Feng et al.,2010; Huang et al., 2011; Jiang et al., 2007; Li et al., 2010). It is there-fore essential to investigate the redox state on other platforms ofthe same age in order to better document the spatial repartition ofanoxic water-masses in the Ediacaran ocean.

Here, we present a redox paleo-reconstruction of the EdiacaranAraras platform, southeastern Amazonian craton (Western Brazil),coupled with a stratigraphic and sedimentologic study of five sed-imentary sections in the carbonates directly overlying the Pugadiamictite (of Marinoan age). Two of these sections are from deeperparts of the platform and described here for the first time, allow-ing us to update a prior paleo-environmental interpretation of theAraras platform (Nogueira et al., 2007). The depositional redox statewas constrained by trace metal and major element concentrations,known to be useful proxies for redox depositional-condition recon-struction (Algeo and Rowe, 2012; Brumsack, 1980, 2006; Hetzelet al., 2009; Tribovillard et al., 2006). The carbonate-dominatedlithology of the rocks studied here demanded specific normaliza-tion of trace-metal contents as explained below.

2. Geological setting of the Araras carbonate platform

The study area is located in Mato-Grosso State, in centralBrazil (Fig. 1a). In this region, Neoproterozoic carbonates of theAraras Group discontinuously outcrop within the Paraguay Belt(600–520 Ma; Trompette, 1994, 1997), along the southeastern mar-gin of the Amazonian craton. The Araras Group was depositeddirectly over the glacial sediments of the Puga Formation (Nogueiraet al., 2003). It is divided into four units from base to top: theMirassol d’Oeste Formation, composed of dolostones with wellpreserved sedimentary features (stromatolite, tube-like structuresand megaripples) characteristic of post-Marinoan cap dolostones(Hoffman, 2011; Nogueira et al., 2007); the Guia Formation, domi-nantly composed of gray limestones with subordinate sandstones;the Serra do Quilombo Formation, composed of dolostones, and theNobres Formation, comprising limestones and cherts. This carbon-ate succession is disconformably overlain by siliciclastic sedimentsof the Alto Paraguai Group (Fig. 1b).

Based on sedimentological and chemostratigraphic (carbonand strontium isotopes) data, and two Pb–Pb isochron ages of627 ± 32 Ma and 622 ± 33 Ma obtained at the base of the Guia For-mation, the Puga Formation has been correlated to the Marinoanglaciation (Babinski et al., 2006; Nogueira et al., 2007; Romero et al.,2013; Sansjofre et al., 2011).

Five sections containing the first two formations of the Ararasgroup (Mirassol d’Oeste and Guia) were sampled for this study(Fig. 1a and b). Quarries names, geographic coordinates, beddinginclination and stratigraphic thickness are given in Table 1.

Terconi and Tangará sections (Nogueira et al., 2003, 2007;Romero et al., 2013; Soares and Nogueira, 2008) are located in theexternal part of the Paraguay Belt, far away from the metamorphiczone (Fig. 1a; Alvarenga et al., 2007) and represent the inner-shelf ofthe paleoplatform. Camil, Carmelo and Copacel sections are locatedin the internal part of the Paraguay Belt (Fig. 1a) and were depositedin the outer-shelf. Sedimentologic and petrographic description of

Table 1

Name, coordinates, inclination and thickness of outcrops for each section.

Quarry Latitude Longitude Inclination (◦) Thickness (m)

TERCONI 15◦40′42 S 58◦04′32 W ≈0 33TANGARÁ 14◦39′25 S 57◦31′22 W ≈0 40CAMIL 16◦12′33 S 57◦34′24 W ≈90 111CARMELOa 16◦11′22 S 57◦21′10 W ≈60 NW 207COPACELa 14◦40′42 S 56◦17′51 W ≈55 SE 357

a Newly described sections.

Camil section can be found in Nogueira et al. (2007), whereas theCarmelo and Copacel sections are described here for the first time.

In the Terconi quarry, the Mirassol d’Oeste Formation is rep-resented by a pink to gray thinly laminated dolomicrite. Its baseshows stromatolitic laminations with large decimeter-scale dome-shaped structures (Fig. 2a). These are covered by stromatoliticlaminations punctuated by meter-scale vertical tubes of 3 to 6 cmin diameter (tube-like structures), filled with secondary carbonates(Fig. 2b). Its upper part presents megaripples (Fig. 2c) and microand macropeloids. In the Tangará da Serra section, only the upperpart of the Mirassol d’Oeste Formation crops out. It contains tube-like structures, megaripples, and micro and macropeloids (Fig. 2d).In both Terconi and Tangará sections, the contact between Miras-sol d’Oeste and Guia formations is marked by a 5–15 cm thicksiliciclastic-rich interval (Fig. 2e). Above this boundary, the GuiaFormation comprises alternating beds of aragonite pseudomorphcrystal fans with a maximum length of 8 cm in a micritic matrix andcross-stratified grainstone (Fig. 2f).

The Carmelo quarry outcrops near the BR-270 road (between thecities of Caceres and Cuiabá). The section begins with 20 m of white,thinly laminated dolomicrite overlying the purple diamictites ofthe Puga Formation (Fig. 3a and b). Because of its stratigraphicposition, and stromatolitic dolomicrite lithology, we tentativelycorrelate this white dolomite to the Mirassol d’Oeste Fm., despitethe fact that the typical anomalous structures found in cap dolo-stones (namely, tube-like structures and megaripples) have notbeen observed there. The section continues with 40 m of thinly lam-inated red fine-grained sandstone composed mainly of quartz andiron–oxide with carbonate cement (Fig. 3c and d), alternating withdecimetre-thick, gray carbonate beds (Fig. 3c) that we interpretto belong to the base of the Guia Formation. The Guia Forma-tion continues with 110 m of gray carbonates presenting undulatedlaminations, followed upward by planar laminated micrite (Fig. 3e)with no evidence of crystal fans. The remaining 40 m of Carmelosection comprises alternating marl and limestone, with brecciatedlevels containing decimetre-scale tabular and angular clasts (Fig. 3fand g) in the last 5 m.

In Camil and Copacel quarries only the Guia Formation isexposed. They both comprise gray carbonate interbedded withdarker, marl-rich, strata (Fig. 4a). In Copacel, the Guia Formationis separated from the Puga diamictite by ∼30 m interval of non-exposure (Fig. 1b). From 30 to 90 m, common cross-stratificationand wavy laminations (Fig. 4b) pass upward into planar laminatedcarbonates, then alternating marl-limestone (Fig. 4c). This succes-sion is similar to that observed in the Carmelo section. The upperpart of the section (from 240 to 350 m) comprises massive lime-stone beds and breccia with decimetre-scale tabular and angularclasts (Fig. 4d and e). The Camil section closely resembles the upperpart of Copacel and Carmelo sections (Nogueira et al., 2007), albeitwith less abundant marl.

3. Facies associations and paleo-environmental

reconstruction

Outcrop-based facies analysis, complemented by petrographicdescription of representative samples, were performed on the five

P. Sansjofre et al. / Precambrian Research 241 (2014) 185– 202 187

Fig. 1. (a) Geological location map and (b) general stratigraphy of the study area. The letters A, B, C, D and E represent the following localities: (A) Terconi, (B) Tangará daSerra, (C) Carmelo, (D) Camil and (E) Copacel quarries.

sections. The different facies observed can be grouped into sevenfacies associations (FA): glaciomarine (FA0), shallow to moder-ately deep platform (FA1), wave influenced shallow platform (FA2),wave-influenced mixed platform (FA3), storm wave-influencedplatform (upper offshore; FA4), CaCO3 over-saturated platform(FA5), deep platform below storm-wave base (lower offshore; FA6)and slope (FA7). These facies associations are represented in Fig. 5and detailed below. A summary is given in Table 2.

3.1. FA0 – glaciomarine

In the region of Mirassol d’Oeste the subhorizontal beds of thePuga glaciogenic sediments are up to 5 m thick and overly theProterozoic metamorphic rocks of the Amazonian Craton. In theNorthern Paraguay Belt, this unit is folded and foliated with thebest exposures found in the region of Cáceres, Jangada and Nobres,mostly overlying metasedimentary rocks of the Cuiaba Group. Thisassociation consists of massive diamictites, siltstone and sandstone.The diamictites are greenish gray to black and, when weath-ered, exhibit purplish to reddish color. The sandy-clayey matrixof diamictites (fine to medium sand) is often carbonatic and mica-ceous. Angular to well-rounded, matrix-supported clasts of largely

varying size (pebbles, blocks and boulders up to a meter) show vari-able composition (sandstone, granite, gneiss, volcanic rock, shale,quartz, pelitic, among others) and are often striated and faceted. Themain structures of the diamictites are massive bedding, even par-allel bedding and less frequently convolute bedding. The siltstonesexhibit scattered pebbles (maximum clast size, 5 cm) and frequentintercalations of 20 cm-thick fine sandstone with even parallel andrare cross lamination, mostly interrupted by levels of disseminatedpebbles and granules (dump structure). The granules and pebbleshave similar composition to those found in diamictites. The lami-nation of siltstones and sandstones are deformed at the base of theclasts, often forming overload structures (dropstone structure).

The description of this association is similar to that made byAlvarenga (1988) and Alvarenga and Trompette (1992), whichinterpreted it as glaciomarine continental shelf deposits with influ-ence of debris flow in the more distal domain. We concur withthat interpretation. The most proximal glacial environment canbe suggested by the lithological clast diversity of diamictites,which suggests reworking and abrasion of basement rocks. Thesame abrading process was responsible for the generation of thesandy-clayey matrix. Liquefaction processes occurred in wateroverpressured sediments. The retreat of the ice sheets would have


Fig. 2. Macroscopic observations in the Terconi and Tangará quarries. (a) Stromatolitic lamination observed at the base of the Mirassol d’Oeste formation in the Terconiquarry, the scale is given by the hammer. (b) Tube-like structures observed in the Terconi quarry. (c) Megaripples observed on an erratic block in the Terconi quarry, thehammer on the right gives the scale. (d) Macropeloids observed in the Tangará da Serra quarry. (e) General view of the Tangará da Serra quarry. (f) Crystal fans observed inthe limestones from Tangará quarry.

induced the development of restricted environments in the coastalzone where weak currents allowed siltstones and sandstones toaccumulate. Granules and pebbles disseminated in these depositsare interpreted as ice-rafted debris dropped by melting of icebergs.

3.2. FA1 – shallow to moderately deep platform

Laminated peloidal dolomudstone/dolowackestone overlapdirectly FA0 developing a laterally irregular and undulated sharpcontact. Within the first meter above the contact, the dolo-mudstone displays massive bedding, and locally incipient planarand convolute laminations. Planar laminations generally show anonlap configuration, mainly filling-in depressions (>0.7 m) in thediamictite. Peloidal doloboundstone is thinly laminated, laterallycontinuous for hundreds of meters, and is characterized by pla-nar stromatolites, rarely domal (Fig. 2a). The domal stromatolites

exhibit a preferential orientation toward 150◦ Az (gutter stroma-tolites) and undulations 70 cm in length and 10–20 cm in width;the undulation wavelength decreases upward. Fenestraes with upto 2 mm in length occur nearly parallel to the microbial and planarlaminations, and are sporadically filled by spar calcite and euhedralrhombs of dolomite cement. Synsedimentary normal faults oftendeform the microbial lamination, forming disharmonic folds. Verti-cal tube structures up to 6 cm in diameter and metric length disruptand deform the microbial lamination, with straight and sharpedges, and are filled by dolomicrite and dolomicrospar (Fig. 2b).

The sharp contact between the diamictite and the cap carbon-ate has been described by some authors as erosive, indicative ofdiastem or of a significant hiatus (Von der Borch et al., 1989;Christie-Blick et al., 1990; Fairchild, 1993; Dyson and von derBorch, 1994). Others consider these contacts as concordant, gra-dational with no evidence of erosion (Plummer, 1978; Williams,


Fig. 3. Macro and microscopic observations in the Carmelo quarry. (a) Stromatolitic lamination observed in the white dolostones. (b) Microscopic observation of the whitedolosparite with the presence of fractures. (c) Alternated carbonated-red sandstone levels observed at the base of the Guia Formation, the decameter is 15 cm in diameter.(d) Microphotographic (reflected light) of the red-sandstone bed, with fine horizontal laminations. (e) Fine laminations observed in the Guia Formation. (f) Carbonate-marlalternations observed in the Carmelo quarry, Nellore cows are used for scale. (g) Brecciated level in the upper part of the quarry, the compass is used for scale and is 7 cm indiameter.

1979; Alvarenga and Trompette, 1992; Kennedy, 1996; Hoffmanet al., 1998; Myrow and Kaufman, 1999). However, in most casestheses contacts are interpreted as transgressive surfaces gener-ated by post-glacial sea level rise. In the Araras cap carbonate, thiscontact surface is very irregular and undulated, being interpretedby Nogueira et al. (2003) as a result of liquefaction-induced soft-deformation soon after dolomudstone deposition. This suggests ashort time interval between the two depositions and admits a sud-den warming, responsible for the fast melting concomitant withpost-glacial sea level rising. On the other hand, the onlap con-figuration of the dolomudstone lamination indicates the infill ofan irregular surface. At any rate, this surface is interpreted as theproduct of post-glacial transgression related to the last Cryogenianglaciation event.

Microbial activity that induced the carbonate precipitation isindicated by the abundance of stromatolites and peloidal fabric.The stratiform stromatolites suggest growth in the seafloor with-out currents action, probably in the context of a marine platformpartially protected or below storm wave base, within the photiczone. Fenestrae cavities are formed by confined and dehydration

gas due to degradation of organic matter (Hardie, 1977; Tucker,1991). Although fenestral porosity is classically attributed to peri-tidal environments (Kendall and Warren, 1987; Pratt, 1994; Shinn,1983), there were no structures typical of shallow water envi-ronments such as tidal facies and desiccation cracks to prove theepisodic emergence of the dolostones facies.

3.3. FA2 – wave influenced shallow platform

This facies association consists predominantly of pinkish tograyish peloidal dolograinstone up to 6 m thick and laterally con-tinuous for tens of meters. It exhibits thin planar laminations withlow-angle truncations and laterally grades to megaripples bed-ding (Fig. 2c). In the sedimentary sequence it begins with cyclicinterbedding of dolograinstones with layers of fibrous crystals andeven-parallel and inclined laminations. The megaripple beddingis complex, formed by internal climbing wave-ripple-laminationthat shows an oscillatory pattern of the crests. The megaripplesare 10–20 cm wide with a wavelength of 0.15 to 0.7 m. Theircrests are elongated in the NNW direction. The peloids consist of


Fig. 4. Macro and microscopic observations in the Copacel and Camil quarries. (a) General view of the Copacel quarry with black marl levels alternating with carbonatedlevels, the truck on the right can be used for scale. (b) Undulated lamination observed at the base of the Copacel section. (c) Alternated carbonated to marl levels observed inthe Copacel quarry. (d) Brecciated nodular clast in the upper part of the Copacel quarry, scale is indicated by the pen. (e) Brecciated level with darker tabular clasts, scale isindicated by the pen. (f) Partially alizarin-red stained thin section showing the transition between carbonate and marl horizons.

microcrystalline dolomite and range in diameter from 3 mm(micropeloids) to 5 mm (macropeloids) (Fig. 2d). They generallyexhibit a grumose texture, and the interpeloidal cement is charac-terized by xenotopic crystals of dolomite. Fine terrigenous grains,like quartz, feldspar and heavy minerals occur disseminated withinthe cement. The millimetric variation in thickness of these lamina-tions reflect different peloid sizes. Micropeloids generally form thethinnest lamination (1–2 mm thick) but can also occur as isolatedgrains within the thicker lamination. These thin laminations have areduced interpeloidal space with little or no cement. Macropeloidsconsist of aggregates of micropeloids cemented by dolomite. Theyoccur discontinuously only in thicker laminations (up to 5 mmthick). The interpeloidal porosity of up to 2 mm in diameter is filledby euhedral crystals of dolomite.

FA2 was generated by oscillatory flow in moderately shal-low water environments. The peloids (micro- and macropeloids)were formed by smooth movements on the seafloor and their

preservation implies reduced abrasion during transport or fastcementation, because these are delicate structures that are easilydisaggregated (James and Narbonne, 2001; Halverson et al., 2004).Sporadic oscillatory movements are represented by inclined lami-nations while the presence of megaripple bedding showing crestsswing suggests dominance of oscillatory flow in the generation ofthis structure. Allen and Hoffman (2005) suggest extreme windsat the origin of these megaripples, however, the complex inter-nal arrangement of the laminations suggests combined flow understorm waves action (Dumas and Arnott, 2006). Similar geometri-cal arrangements are found on Svalbard, where the laminations areinterpreted as products of sediment stabilized by microbial activityand bedforms are consistent with transport by traction (Halversonet al., 2004).

Megaripples with muddrapes and wave laminations withfibrous crystals were produced by the interaction of currentsin deep water, suspension and precipitation of aragonite at the


Fig. 5. Stratigraphic correlation of studied section based on facies associations.

seafloor. The thickness of the laminated layers decreases upwardwhile those with crystals remain constant, indicating predom-inance of seafloor precipitation. When crystals are scarce theprecipitation rate was probably reduced due to the action ofcurrents related to a decrease in water depth and increased sed-imentation rate. The preservation of macropeloids in truncatedlaminations suggests reduced abrasion and fast cementation. Theabundant presence of peloids in this association is interpreted as aresult of intense microbial activity that induces carbonate nuclea-tion. It is noted in this association a tendency to shallow upward,which is visible through the scarcity of fibrous crystals at thetop and the widespread occurrence of current and wave-influencestructures.

3.4. FA3 – wave influenced mixed platform

The FA3 is characterized by red to purple laminated silts andcarbonate sandstones observed mainly in Tangará da Serra andCarmelo quarry. In Tangará da Serra the FA3 displays centimet-ric to metric thicknesses and fills in depressions along synclinesdeveloped into the FA2 as a result of synsedimentary fault dis-placement (Fig. 2e). The marls and carbonate sandstones withmegaripples bedding have lateral continuity of a few dozens ofmeters and a thickness of about 6 m. Often the megaripple bed-ding is associated with ripple marks, which suggests intervals

generated by oscillatory and unidirectional flows. Toward the top,thin layers of siltstones interbedded cyclically with mudstone areevidence of a higher terrigenous input on the platform. The pres-ence of terrigenous material throughout the FA3 is one of itsmain characteristics and suggests a relative proximity to the con-tinental sources. The characteristics of this facies indicate shallowwaters influenced by currents and waves with cyclical interrup-tions associated with mudstone deposition and a large terrigenousinput.

3.5. FA4 – storm wave influenced platform

This facies association encompasses marl, limestone with wavebedding and limestone with hummocky cross-stratification. Themain exposures of this unit reach up to 10 m in thickness and occurin Carmelo and Copacel quarries. The marl presents parallel lamina-tion and cross-bedding (Fig. 3e) but are usually massive. Limestoneswith wave bedding are usually interbedded with laminated marlforming a rhythmite that reaches up to 50 m in thickness. Graylimestones with abundant terrigenous grains exhibit centimetrichummocky cross-stratification and are interbedded with the lime-stone/marl rhythmite. Terrigenous grains, usually in the size of siltto very fine sand, are disseminated in the rock and consist of quartz,feldspar and mica.


Table 2

Facies associations identified along the Araras carbonate platform. Facies, structures and deposition processes associated.

N Facies association Facies Structures Processes

FA0 Glaciomarine Diamictite with decimetric striatedblocks disseminated in a pelitic matrix

Massive bedding withdropstones, striated andfaceted clasts

Glacial abrasion of basement rocksand liberation of dropstones byicebergs (ice rafting)

FA1 Shallow to moderatelydeep platform

Dolomudstone, stromatolitedoloboundstone and peloidaldolowackestone

Stratiform stromatolitewith fenestrae porosity,Tube-like structures. Planarand convolute laminations

Microbial activity-inducedcarbonate precipitation in a quietand possibly protectedenvironment. Fluid escape relatedto degradation of stromatoliticmats

FA2 Wave influenced shallowplatform

Dolograinstone with micro and macropeloids

Parallel and tabularcross-stratification, micro-and macropeloids andmegaripples bedding

Oscillatory flow, chemicalprecipitation of carbonate andearly cementation, biologically-influenced precipitation indicatedby the presence of peloids

FA3 Wave-influenced mixedplatform

Interbedded limestone andfine-grained red sandstones

Megaripples,cross-stratification andundulated lamination

Greater detrital input; oscillatoryand current-inducedsedimentation

FA4 Storm wave influencedplatform

Mudstone, marls and limestone Undulated lamination andhummocky crossstratification

Sedimentation by decantation withoccasional storm-influenceddeposits

FA5 CaCO3 over- saturatedplatform

Thin alternation of limestone beds andmarls levels

Aragonite-pseudomorphscrystal fans, parallellaminations, neptuniandykes

Deposition by sedimentation andrapid crystallization at the seafloorin an over-saturated environment.The presence of neptunian dikessuggest syn-sedimentary tectonics

FA6 Deep platform belowstorm wave base

Fine limestone and black marls Planar lamination Sedimentation by decantation andcarbonate precipitation in quietand anoxic environments belowstorm wave base

FA7 Slope Fine limestone and black marls Planar lamination, slumpstructures andintraformational breccias

Carbonate brecciation anddeposition due to gravitationalinstability in the slope

The precipitation of carbonate mud was interrupted by spo-radic terrigenous inflows induced by low velocity currents. Theaction of oscillatory flow led to the migration of small bedformsthat alternated with periods of low energy. The platform wasperiodically reworked by storm waves indicated by hummockycross-stratification. Wave laminations were generated by lessintense oscillatory flow during the decline of the storms (Dumasand Arnott, 2006). The presence of opaque minerals, mainly pyrite,suggests a reducing marine environment.

3.6. FA5 – CaCO3 over-saturated platform

This association is characterized by a 6 m thick and laterallycontinuous sequence of tabular beds of mudstone with undu-lated top, interbedded with dark shales. It is found only in themore proximal Terconi and Tangará da Serra quarries. The mud-stones beds are composed of centimetric crystals pseudomorphsof aragonite (Fig. 2f) and dolomitized micrite. Isolated crystalsare most common at the base of the association while at the topthey become more abundant and connected. Even-parallel andwave laminations commonly occur among and over the crystals,resulting respectively from sedimentation in quiet environmentsor influenced by oscillatory flow. Mudstone and terrigenous grainsare interbedded forming convex laminations between the crystalson the top of sequence. This configuration suggests that depositionwas simultaneous or predated the crystals formation, the convex-ity of lamination resulting from laminae deformation during crystalgrowth.

The abundance of crystals is the result of high alkalinity, absenceof carbonate inhibitors in seawater, such as Fe2+ and Mn2+, and lowsedimentation rates (Sumner, 2002). The way the crystals occurmay reflect the strength of currents or oscillatory flow during sed-imentation. The isolated crystals are associated with low or no

sedimentation during crystals precipitation while connected crys-tals are related with the increase of sedimentation concomitantto crystal growth. At the top of the association, thicker layers ofdark shales are cyclically interbedded with mudstones. The pres-ence of ripple marks on top of the layers suggests the influence ofoscillatory flows that are common in shallow waters, but the occur-rence of crystals in these layers is a strong hint of deep water. Thethicker layers of dark shale are the result of increased sedimen-tation rate in calm or partially stagnant waters. The concomitantoccurrence of common shallow-water structures and deep waterprecipitates leads us to conclude that this association representsa transitional environment on a deep platform over-saturated inCaCO3. The occurrence of laminar fouling (crusts) and aragonitecrystal fans may suggest an abiotic precipitation. The intercalationof these crusts with micrite in a deep water environment is con-sistent with a sea over-saturated in aragonite and higher crystalgrowth rates than sediment accumulation rates. These conditionsof sedimentation are found in most cap carbonates around theworld (Grotzinger and Knoll, 1995; Sumner and Grotzinger, 1996;Sumner, 2002).

3.7. FA6 – deep platform below storm wave base

This facies association is the most frequent in the Guia Forma-tion, being found hundreds of kilometers across the study area. Thedominant facies are laminated limestones and calcareous shales.The limestones are fine and gray, arranged in tabular layers up to3 m thick, alternating with millimeter laminations of terrigenousgrains and/or black shales (Figs. 3f and 4c), the later with maximumthickness of 1 cm. Wave laminations and silica nodules are foundlocally while dissolution seams, stylolites and pyrite are abundant.

The generally fine-grained limestones, with dissociated struc-tures produced by wave and tidal currents, occur in continuous


Fig. 6. Paleo-environment depositional model proposed for the Araras carbonate platform. See text for description.

layers for hundreds of kilometers and are compatible with sedi-mentation in deep water below storm wave base (Pfeil and Read,1980; Stow, 1986). The normal gradation found in limestone can begenerated by turbidity currents of low density which is consistentwith the inference of a deep marine paleoenvironment, and withthe retrogradational nature of the succession studied (see Tucker,1986; Fairchild and Hambrey, 1984; James and Narbonne, 2001).The water depth facilitated the preservation of organic matter. Thesignificant influx of terrigenous grains to the carbonate platformmay have a continental origin.

3.8. FA7 – slope

This association characterizes the upper portion of the Guia For-mation and is usually associated with FA6. Limestones with slumpstructures and intraformational breccias characterize this associa-tion (Figs. 4d, e and 3g). Features such as convolute laminations,synsedimentary normal faults, fractures, folds, deformed layers,neptunian dikes and slump structures are commonly found. Theneptunian dykes are filled with intraformational breccias of lime-stone and tabular shale clasts.

The occurrence of deformed intervals between undeformed lay-ers is caused by different degrees of liquefaction in the deposit(Winterer and Sarti, 1994). The intense brecciation of layersand concomitant cementation and intumescence of the seafloorappears to be a kind of expansive crystallization. Normal faultsand slump structures indicate vertical movements associated withgravitational instability on a ramp/slope. Breccias with tabularclasts are probably related to gravitational flows of high viscos-ity, such as slipping, landslides, debris flows, liquefaction flowsand turbidity currents. Likewise, the gravitational instability is also

indicated by fault associations, slumps structures, mass flow, nep-tunian dikes and breccias. This gravitational instability and ruptureof partially lithified sediment is consistent with sedimentationrelated to deep-water slope environment.

3.9. Paleoenvironmental interpretation

Fig. 6 shows the reconstructed paleoenvironmental scenario forthe Araras carbonate platform, divided into 4 depositional phasesnoted “a” to “d”. The main scenario showed in phase “a” suggeststhe final stage of deglaciation with the presence of icebergs in theshallow to deep platform, and is similar to most paleogeograph-ical reconstitution of post-Marinoan cap carbonates and glacialdeposits (Hoffman, 2011). Diamictons with dropstones (Puga For-mation) were the result of basement rock abrasion by glaciersand were delivered to the platform during deglaciation mostly byice-rafting (Fig. 6a). The post-glacial warm climate allowed the pre-cipitation of dolomitic muds partially induced by biologic activity(Mirassol d’Oeste Formation) in the euphotic zone (e.g., Hoffmanet al., 2007; Rose and Maloof, 2010). Low-angle lamination in thefirst meters of dolostone, observed in the Terconi quarry, suggestsreworking by wave action. On the other hand, peloids and macro-peloids in the dolostone suggest early dolomitic cementation ina calm environment (Nogueira et al., 2007), probably protectedby barriers, maybe linked to the irregular morphology of diamic-ton deposits and the eroded substrate (Fig. 6b1). Deposition inwell-mixed shallow waters suggests an oxic environment for theMirassol d’Oeste dolostones, in agreement with the presence ofprimary detrital hematite in these levels (Font et al., 2005). Justabove the dolostones, a 10 cm to 20 m thick progradational succes-sion is recorded by laminated iron-rich siltstones and marls and


terrigenous fine limestone with megaripple bedding (Fig. 6b2 andc1). Terrigenous sediments represent a brief progradation of thecoastline preceding the long-term transgression responsible for thedeposition of hundreds of meters of fine limestone with increasingfrequency of shales upward (Fig. 6c1 and 6d1). Over-saturation inCaCO3 in the moderately deep platform favored the precipitationof aragonite crystal fans alternated with deposition of lime mudsand shales (Fig. 6d1). The precipitation of crystal fans occurred onlyin the cratonic area.

The stages shown in Figs. 6d1 and d2 can be reconstructed fromCarmelo, Camil, and Copacel sections deposited in the distal deepplatform. We assume here that layers with carbonate breccia atthe slope were deposited broadly contemporaneously to the stratabearing crystal fans in the inner-platform.

Together these facies associations form a transgressive systemstract over the carbonate platform. The shallow water deposits ofthe inner-shelf contain iron oxides precipitated in an oxic water-column. Deeper inner-shelf sediments and outer-shelf sedimentswith darker marl levels suggest more reducing conditions at leastin the sediment. More information is needed to estimate if thesedeposits were anoxic and if so to estimate whether anoxia wasrestricted to the sediment or if it reached the water column, asit seems to have been the case for the Doushantuo Formation (Aderet al., 2009; Li et al., 2010).

4. Rationale for the use of trace and minor elements as

paleoredox indicators

Redox sensitive trace elements (e.g., U, Mo, Zn, Pb) are fre-quently used to infer paleoredox conditions in sediments (e.g.,Algeo et al., 2012; Algeo and Rowe, 2012; Brumsack, 2006; Lyonset al., 2003; McManus et al., 2006; Piper et al., 2007; Scott and Lyons,2012; Tribovillard et al., 2008). U and Mo are of particular interestbecause of their following shared chemical properties: (1) they arepresent in low concentration in the continental crust (U ∼2.7 ppmand Mo ∼3.7 ppm; (McLennan, 1989)) and are mainly derived fromoxidative weathering and subsequent riverine input, (2) they arenot concentrated in marine phytoplankton, (3) they have residencetimes of ∼450,000 years (U) and ∼780,000 years (Mo), that ensuressteady state equilibrium and uniform concentration in the mod-ern ocean (Chaillou et al., 2002; Colodner et al., 1995), (4) they areconservative under oxic conditions and trapped into the sedimentunder anoxic conditions (Brumsack, 1986). As a consequence, sed-iment enrichments in U and Mo are generally related to anoxicand/or euxinic environments, where these elements can be easilyscavenged or diffusively trapped across the sediment–water inter-face from the oceanic reservoir into the sediment.

Although U and Mo share many properties in seawater, the pro-cess by which each are incorporated into anoxic sediments aredifferent (Algeo and Maynard, 2004; Algeo and Tribovillard, 2009).In modern settings, part of dissolved U(VI) diffuses into the sed-iment. When it reaches the ferruginous redoxcline (where Fe3+

reduces to Fe2+), the U(VI) is reduced to U(IV) and fixed into thesediment by adsorption or precipitation (Klinkhammer and Palmer,1991). It has been suggested that Mo trapping proceeds throughthe formation of thiomolybdate (MoOxS4

−x2), which is then fixedin Fe–sulfide and/or organic matter molecules (Tribovillard et al.,2004; Zheng et al., 2000). Since thiomolybdate requires free H2Sto form, Mo accumulation in the sediment requires more reduc-ing conditions than U accumulation, and occurs dominantly undereuxinic conditions (Helz et al., 1996; see also Helz et al., 2011;Tribovillard et al., 2012). The Mo/TOC ratio in euxinic settings isa good proxy for assessing the global dissolved-Mo content of theocean (Algeo and Lyons, 2006; Algeo and Rowe, 2012). In present-day oceans, where euxinic conditions are rare, Mo is dominantly

delivered to the ocean by riverine inputs and accumulates con-servatively. The present oceanic Mo concentration is 105 nmol/kg,which makes of it one of the most abundant trace metals inthe ocean. When euxinia occurs only locally, Mo sequestration isnot limited by Mo availability (because Mo is available from theremainder of the oxic ocean), leading to high Mo/TOC values (Algeoand Rowe, 2012). In pre-Marinoan rocks, Mo/TOC ratios are low(<30), suggesting low dissolved-Mo contents in the oceans probablyresulting from the efficient trapping of Mo into the sediments due towidespread euxinia (Scott et al., 2008). In contrast, post-Marinoanrocks from the Doushantuo Formation have shown high Mo/TOCratios, commonly exceeding 70, suggesting a dramatic increase inthe ocean redox state in the Marinoan glacial aftermath (Sahooet al., 2012).

Although zinc (Zn), lead (Pb), copper (Cu), cadmium (Cd),nickel (Ni) and vanadium (V) accumulation in sediments dependson redox conditions, their geochemical cycles are also highlyinfluenced by biological activity, detrital input or hydrothermalactivity, complicating their use in redox environment reconstruct-ions (Tribovillard et al., 2006; Wangersky, 1986). Several of theseelements are important components of enzymes (e.g., Cu, Zn andV) and can be concentrated in modern phytoplankton. In somecases, enrichment of these elements can result from an increase inorganic matter accumulation (Beveridge and Murray, 1976, 1980;Bruland, 1980). During sedimentation and early diagenesis, a sig-nificant amount of the organic matter is remineralized. Under oxicconditions, the associated trace metals are solubilized and releasedback into the water column, which may prevent their accumula-tion in sediments (Beck et al., 2008; Hatch and Leventhal, 1992;Ripley et al., 1990; Wangersky, 1986). Some trace metals, such asZn, Pb, Ni and Cu, can be fixed on oxides and are known to sub-stitute for manganese and iron in the most common oxides andhydroxides (Tribovillard et al., 2006). The solubility of these oxy-hydroxides, and thus their capacity to transport and accumulatetrace metals in the sediment, is strongly redox-dependent (Calvertand Pedersen, 1993; Caplan and Bustin, 1999; Morford et al., 2001;Algeo and Maynard, 2004). Trace metals accumulated in oxic sed-iments by oxy-hydroxides will be released in the porewater onlyif redox conditions in the sediment reach the manganous and fer-ruginous zones where oxy-hydroxides are reduced (Canfield andThamdrup, 2009). Whether trace metals have been released in theporewater by organic matter mineralization or by oxy-hydroxidesreductive dissolution, if the porewater redox conditions reach thesulfidic zone, some elements, such as Zn and Pb, will be precipi-tated and trapped into the sediment as sulphides (e.g., Tribovillardet al., 2006).

More thorough reviews of trace metal sequestration in sedi-ments are presented in Brumsack (2006), Tribovillard et al. (2006),Sageman et al. (2003) and Lyons et al. (2009). In the current study,because most of the lithology are carbonates that have elementalconcentrations different from black shale samples, a careful separa-tion of authigenic from bulk elemental concentrations was required(see Section 5.2).

5. Methods

5.1. Analyses

A total of 95 bulk-rock samples were analyzed: 37 from theTerconi section, which includes 10 samples from Font et al.(2006), 19 from the Tangará section, 8 from the Camil section,21 from the Carmelo section and 20 from the Copacel section.Rock fragments free of apparent fractures or recrystallization werecarefully selected and crushed in an agate mortar. Chemical anal-ysis were performed by ICP-AES (major and minor elements)


and ICP-MS (trace elements) at the spectrochemical laboratory(SARM) of the Centre de Recherches en Pétrographie et Géochimiein Vandœuvre-les-Nancy. Samples were prepared by fusion withLiBO2 (980 ◦C during 60 min), followed by dissolution in a HNO3(1 mol L−1)–H2O2 (∼0.5% v/v)–glycerol (∼10% v/v) mixture in orderto obtain a diluted solution (Carignan et al., 2001). Precisionand accuracy were both better than 1% of the measured value(mean 0.5%) for major–minor elements and 8% for trace metals, aschecked by international standards and analysis of replicate sam-ples (Carignan et al., 2001).

Pyrite content was quantified gravimetrically after precipitationof Ag2S following the chromium reduction method of Canfield et al.(1986). Based on replicate analysis, the external reproducibility fol-lowing this technique is better than 0.2%.

For total organic carbon (TOC) quantification, samples weredecarbonated in 6 N HCl overnight at room temperature, followedby 2 h at 80 ◦C for dolomite samples. Residues were washed withdistilled water until a neutral pH was achieved, centrifuged anddried at 50 ◦C. For each sample, 10–60 mg of decarbonated powderwas loaded into a quartz tube together with cupper oxide wires.Tubes were connected to a vacuum line and sealed under sec-ondary vacuum (<10−5 mbar). They were then heated at 950 ◦C for6 h to oxidize the organic matter. The resulting CO2 was cryogeni-cally separated in a vacuum line and manometrically quantifiedusing a Toepler pump. TOC contents were calculated from the CO2quantification with an approximate precision of ±10%.

5.2. Quantification of trace metal authigenic contents

It is important to stress here that most rocks in the studiedsections are dominated by carbonate; this mineral fraction of thesediment acts as a sterile, diluting phase that does not incorpo-rate the redox-sensitive trace metals in significant proportions. Inorder to minimize dilution effects by carbonate, elemental concen-trations are commonly normalized to aluminum (see discussion inVan der Weijden, 2002; Tribovillard et al., 2006). However, it hasbeen shown that for samples with low aluminum content (<1%), Al-normalization can be biased because of an aluminum-enrichmentin the oxide phases (Kryc et al., 2003; Van der Weijden, 2002). Forour sample set, which contains more than 95% carbonate (with [Al]<1%), Al-normalization is thus inappropriate. More importantly,Van der Weijden (2002) showed that spurious correlations areobserved when one element (Velement) shows much higher vari-ability (i.e., the standard deviation divided by the mean) than theother ones. In our case, aluminum content varies from 0.02% to sev-eral percent, which represents a Val = 0.98 for the Terconi sampleset, which is higher than that of other elements (e.g., VU = 0.69). Forinstance, Font et al. (2006) reported trace metal data for the sameMirassol d’Oeste pink dolostones in the Terconi Quarry. Most of thetrace metals showed high enrichment factors (expressed using theconventional Al-normalization) and were interpreted as indicatinganoxic pore-water influenced by bacterial sulfate-reduction. How-ever, these rocks also contain detrital hematite (Font et al., 2005;Trindade et al., 2003), which cannot be reconciled with pore waterH2S production, known to promote hematite reductive dissolution(Canfield et al., 1992; Raiswell and Canfield, 1996). A more straight-forward interpretation is that these high enrichment factors resultsimply from the calculation artifacts described above.

An alternative way to normalize the geochemical data thatcircumvents the Al-normalization pitfalls discussed above is toestimate the authigenic fraction of an element concentration bysubtracting the expected detrital fraction from the total concen-tration (Tribovillard et al., 2006). The expected detrital fractionis obtained by multiplying the Al content of the sample by theelement-to-Al ratio of the standard used. The common standarduse for Phanerozoic and Proterozoic sediments is the Post Archean

Australian Shale (PAAS) (McLennan, 1989). Hence, instead of nor-malizing the whole rock composition by the PAAS standard, theresults are given here as the authigenic fraction of the element,noted Xauth, and calculated as follows:

Xauth = Xtotal − Xdetrital (1)

with

Xdetrital =

(

X

Al

)

PAAS× Alsample (2)

The pitfall of this calculation is that it implies that the X/Alratio of the terrigenous fraction is close to that of the PAAS.This assumption is legitimate when the depositional setting col-lects well-mixed terrigenous particles originating from large-scaledrainage systems, which is probably the case for the regional car-bonated platform studied here. In what follow, we link temporaland spatial variations in Xauth values to relative redox changes.

6. Results of trace element geochemistry

Authigenic concentrations of Cd, Cu, Mo, U, Ni, V, Co, Cr, Zn, Pb,as well as Al, FeS2, MnO and TOC contents in the Mirassol d’Oesteand Guia formations for the 5 sections are given in SupplementaryMaterial. The most relevant elements for redox interpretations arepresented in Fig. 7. For the Terconi section, authigenic concentra-tions recalculated from Font et al. (2006) are also reported in theSupplementary material table of contents and Fig. 7. Trace metals(U, Mo, Zn, Pb, Cd, Cu, Ni, V) authigenic contents and Al concen-trations, pyrite abundance and total organic carbon (TOC) contentsin these sections are generally low. Nonetheless, they present sig-nificant stratigraphic variations, which can be traced along theplatform. Four zones can be identified in the platform, from base totop:

6.1. Zone A

The first one (zone A, Fig. 7) occurs in the dolomicrite at thebase of the Mirassol d’Oeste Formation, where sedimentary fea-tures indicate shallow water deposition (FA1). It is characterizedby very low Al and TOC contents (mean of 0.10% and 0.09%, respec-tively), and no pyrite. This zone shows the lowest Uauth contents,comprised between 0.33 and 1.30 ppm, with an average value of0.8 ppm. The abundances of Moauth and Crauth are too low to bedetected (<0.35 and <5 ppm, respectively). Abundances of Znauthand Pbauth average 20.6 ± 9.7 and 9.5 ± 3.5 ppm, respectively. Coauthand Cuauth show low contents except for the first sample of the Ter-coni quarry (with values of 3.4 and 18.9 ppm, respectively), whichcorrespond to the diamictite–dolomite transition. Zone A is thuscharacterized by low trace metal contents, except for Niauth andMnO. Niauth contents are higher than those found in the other zones,with values between 7.5 and 14.4 ppm and an average of 9.1 ppm.MnO content is also relatively high compared to other zones, andvaries between 0.11% and 0.44%, with an average of 0.14%, the high-est value being recorded by the basal sample of the Terconi Quarry.For the Carmelo section, the only value of MnO content measuredin this zone is 1.2% (Fig. 8).

6.2. Zone B

The second zone (zone B, Fig. 7) corresponds to the upper part ofthe Mirassol d’Oeste Formation, where the occurrence of micro andmacro peloids suggests a shallow water depositional setting (FA2).Al and TOC contents are slightly higher than in zone A and increaseupward from 0.10% to 0.27% and from 0.09% to 0.24%, respec-tively. Pyrite is absent except in TeS 22 with a content of 0.33%.Uauth content is low (1 ppm on average) and increases slightly in


Fig. 7. Authigenic Zn, Pb, Mo, U enrichments and pyrite contents along the different section studied. Znauth , Pbauth , Moauth and Uauth are given in ppm, pyrite content in %.

Fig. 8. MnO contents (%), along the Terconi, Tangará and Carmelo sections.

the upper part of zone B. Moauth shows the highest concentra-tions in the whole data set (maximum value of 16.4 ppm in theCarmelo quarry). At Terconi, Moauth increases progressively from0.6 to a maximum of 1.4 ppm near the dolomite-limestone con-tact (Fig. 7). At the same interface, Znauth and Pbauth present thehighest contents (291.8 and 175.7 ppm, respectively) while Cdauth,Cuauth and Vauth increase moderately with averages of 3.2, 18.8and 3.3 ppm, respectively. This zone also presents the highest MnOcontent (Fig. 8), with an average value of 0.20% in the Terconi andTangará sections, and only one value of 1.6% at the Carmelo quarry.

Niauth is higher than in zone C and D with an average of 7.3 ppm.Trace metal contents are generally higher in the outer-shelf thanin the inner-shelf. This is particularly true for the Znauth and Pbauthcontents that reach extremely high values in the Carmelo quarry(2231 and 2195 ppm, respectively).

6.3. Zone C

The third zone (zone C, Fig. 7) is observed in the Guia Formationand ranges from the inner-shelf and CaCO3 over-saturated facies


Fig. 9. TOC (%) vs. Uauth (ppm) and Total Sulfur (TS) content (%) vs. Moauth (ppm) for all the sampled analyzed. TS is calculated from the pyrite extraction plus the sulfurcontained in Pb and Zn assumed to be as sphalerite and galena.

(FA5), down to the outer-shelf with storm wave-influenced facies(FA4). Al contents vary between 0.2 and 6.6% with an average of1.5%, and TOC are slightly higher than zone B with an averageof 0.45%. This zone is characterized by a Uauth enrichment thatis decoupled from the Al content (R2

AlvsU = 0.03). Uauth contentsincrease from 0.8 to 3.0 ppm on average, while Moauth contentsremain below 0.35 ppm. Znauth and Pbauth have low average con-centrations (11.9 and 5.7 ppm) compared to zones A and B. In thiszone, MnO content and Niauth decreases down to a value of ∼0.05%and 2.5 ppm respectively. Pyrite is present only at the base of theGuia Formation in the Terconi quarry, where its content reaches0.8% in TeS 30, the sample with the lowest carbonate content inthis section.

6.4. Zone D

The fourth zone (zone D, Fig. 7) also occurs in the Guia For-mation and ranges from the outer-shelf to the slope with themarl-limestone facies association (FA6) and brecciated carbonate(FA7). It corresponds to the deepest part of the platform and gen-erally shows the lowest average values for most of the elementsstudied here, except for a few marl beds presenting slight enrich-ments in Uauth, Moauth and pyrite contents (Fig. 7). Uauth and Moauthenrichments are not very high and stay below 2.3 and 1.1 ppmrespectively. In the marl levels, pyrite contents show the highestvalues reaching a maximum value of 3.1%. TOC and Al contents aresimilar to the one in zone C except in the marl levels, which showmoderate enrichment with value of TOC and Al contents reaching0.38 and 3.48%, respectively.

7. Discussion

7.1. Reliability of trace metals as paleoredox indicators in theAraras platform

Before the trace metal authigenic contents can be interpretedin terms of redox variations, three conditions must be met: (i)trace metals must be (abundantly) present in seawater; (ii) theirvariations should not be primarily controlled by changes in sedi-ment accumulation rates; (iii) post-depositional migration of tracemetals must be limited.

In a Snowball Earth scenario (Hoffman et al., 1998), isolation ofthe ocean by a thick ice blanket would act as a barrier to light, stop-ping photosynthetic production. Under these conditions the oceanswould have been quickly stripped of available dissolved oxygen,

leading to a massive transfer of redox-sensitive and/or sulphide-forming elements to the sediment, which would in turn deplete theoceans in trace metals. In this scenario, authigenic Mo and U levelswould have approached zero despite benthic anoxia, as observedfor example for North-American late-Devonian black shales (Algeo,2004). For the Araras Group, one way to check whether U andMo were present in the ocean is to test if there are U and Moenrichments in the sediment at anoxic or even euxinic intervals.The observation that TOC roughly correlates with U content sug-gests that U was present in the ocean at a detectable concentration(Fig. 9a). Mo enrichments in zone B and D correspond to inter-vals where the sulfur content increases indicating conditions moreprone to sulphate reduction reactions (Figs. 9b and 7). We thusassume that Mo and U were present in the water column duringdeposition of most of the Araras carbonates and that the lack ofenrichment in these trace metals did not result from any “reservoireffect” (cf. Algeo, 2004).

Variations in the deposition rate of sediments can either diluteor enrich trace metal contents independently of the redox condi-tions at or below the sediment–water interface. This is hard to testdirectly in the studied sections because of the low detrital con-tent and the lack of control on carbonate sedimentation rate. Yet,if sedimentation rate was the major process controlling variationsin authigenic trace metal contents, they should all vary synchro-neously and proportionally to sedimentation rate. This is not thecase in the studied sections.

Finally, since carbonate rocks are prone to secondary recrystal-lization and migration of diagenetic fluids, the trace metals signalcan be disturbed (Banner and Hanson, 1990; Brand and Veizer,1980). In our sections, the general preservation of a carbonatemicritic matrix suggests that recrystallization was insignificant,but possible vertical fluid migration may have occurred locally inthe Mirassol d’Oeste Formation through its tube-like structures(Fig. 2b). Such a vertical fluid migration could explain the Mo, Znand Pb enrichments observed in zone B, which directly overliesthese tube-like structures. However, because the tube-like struc-tures softly deform the laminations their cementation must haveoccurred prior to lithification and compaction, limiting the tracemetals migration through the tube-like structure and their trappingin the zone B. This possibility is explored further in the discussion.

The presence of local bitumen accumulations has been reportedin the zone B of the Terconi section (Elie et al., 2007). Bitumenmigration could have modified the primary trace metal content(and in particular the Ni and V content which are often associ-ated with organic matter). However, the lack of source rocks apart


Fig. 10. Composite stratigraphic log of the Araras carbonate platform and redox paleo-environmental interpretation for the different zones observed. Zone A: Trace metalcontent is low, most of trace metals are trapped and adsorbed on the organic matter but liberated in the water column because of a high organic matter oxidative degradationrate. Zones B and C: Detrital inputs increase together with primary productivity. Organic matter content increases in the sediment, enhancing anaerobic degradation ofthe organic matter by sulfate-reduction (SR) and dissimilatory reduction of iron (DRI). DRI induced dissolution of iron oxides. Zone D: Brief sulfidic episodes in marl levelsinducing coupled U and Mo slight enrichments in the sediment.

from the actual host-rocks and the presence of bitumen in stylolitesis more compatible with a limited bitumen migration (Elie et al.,2007). Moreover the fact that Mo, Zn and Pb authigenic enrich-ments are also observed in other sections free of bitumen (Tangaráand Carmelo), strongly support an initial, i.e., authigenic or earlydiagenetic, signal at the scale of the Araras platform with minimal,if any, post lithification vertical contamination.

7.2. Paleoredox in zones A and B: oxic to anoxic pore waters inthe shallow platform

Zones A and B were identified in three sections dominatedby shallow platform facies in the inner-shelf of the Araras

platform. This interval typically comprises stromatolites and wave-influenced sediments, which indicate deposition within the photiczone in shallow and well-mixed waters. The pink dolomite at thebase of the Mirassol d’Oeste Formation (zone A) is characterizedby low authigenic trace metal contents, the presence of Fe andMn oxides, the absence of pyrite and very low TOC contents. Thiscombination of characteristics is best interpreted as resulting fromdepositional processes under oxic conditions (above the ferrug-inous zone) both in the sediment and water column (Fig. 10Aand B). This interpretation is in agreement with the ubiquitouspresence of primary hematite in the basal part of this formation(Font et al., 2005, 2010), which also suggests that the sedimentpore-water remained oxic. We suspect that both an extremely low


organic matter delivery to the sediment and/or its extremely effi-cient remineralization in surface sediments must have preventedthe development of oxygen-limited conditions deeper in the sedi-ment during early diagenesis.

The gray dolomite making up the upper part of the Mirassold’Oeste Fm. (zone B) was probably formed in shallower watersthan zone A. These rocks show enrichments in Znauth, Pbauth firstand then in Moauth, which coincide with an increase in organicmatter content and the appearance of pyrite (Table in Supplemen-tary material). This combination suggests increasingly reducingconditions in the sediment pore-waters driven by organic matterremineralization through bacterial sulphate reduction. In sedimen-tary rocks or sediments not affected by hydrothermal fluids, assuggested by the REE pattern observed in the Terconi quarry (Fontet al., 2006), Zn and Pb are dominantly transferred to the sedimentvia organic-matter particle settling. After their release within thepore fluids through organic matter degradation, they either diffusetoward the overlying water column (as during zone A deposition)or precipitate as sulphides (mainly galena, PbS, and sphalerite, ZnS)if H2S is present in the pore fluids (Tribovillard et al., 2006). It ishence possible that Zn and Pb migrated upward from the underly-ing oxic zone A, maybe via the tube structures, and were trappedin zone B where sulfate reduction occurred, adding up to the Znand Pb scavenged from the water column. The increase in organicmatter content also account for the modest Niauth enrichments, ausual companion-trace metal for organic matter (Tribovillard et al.,2006). The increase in Moauth content coincides with the disappear-ance of Fe and Mn oxides, suggesting that the redox context reachedsulfidic conditions at least at shallow depths within the sediment,with possible temporary development of euxinia in bottom water.Given the shallow water depth, this would be compatible withbiomarker evidences for sulfate reducing and green sulfur bacteriain the zone B of Terconi section (Elie et al., 2007).

Similarity in Moauth, Pbauth, Znauth and MnO contents (Fig. 8)between the zone B in Terconi, Tangará and the white dolomiteof the Carmelo section supports the fact that the later belongs tothe Mirassol d’Oeste Formation. However, Pbauth, Znauth and Moauthcontents in the Carmelo quarry are one order of magnitude higherthan those of the inner-shelf sections (Terconi and Tangará) andare associated with high Mn contents (Fig. 8). We speculate thatpore-water redox conditions were more reducing in the outer-shelfwhen compared to the inner-shelf sections due to a stronger partic-ulate shuttle effect linked to cycles of manganese oxides and/or tohigher bioproductivity (Algeo and Tribovillard, 2009; Dellwig et al.,2010).

We thus interpret the transition from zones A to B as resultingfrom an increase in organic matter burial rate, which must haveactivated the trace-metals trapping process. Sulfate content wasprobably low compared to modern concentrations but the increasein oxic weathering of emerged land must have brought sulfate ionsand possibly trace metals such as U and Mo to the ocean (Lyonset al., 2009; Och and Shields-Zhou, 2012).

7.3. Paleoredox in zones C and D: pore-water anoxic conditions inthe deep platform

Zone C corresponds to the siliciclastic-rich level and overlyinglimestones of the base of the Guia Formation deposited in a deeper-environment than zones A and B. This zone has higher Uauth andpyrite contents than zones A and B, but Mo contents are belowdetection limit. Algeo and Tribovillard (2009) have shown thatdepending on the position of the oxic–anoxic transition (hereafterreferred to as the redoxcline), the mechanisms by which Mo andU are trapped may differ drastically. When the redoxcline is toodeep in the sediment, the downward diffusion of U from seawateris too slow to allow significant enrichment. When the redoxcline

rises in the sediment, uranium starts to be trapped, initially moreefficiently than Mo, which requires the presence of free H2S to besignificantly trapped, producing Uauth enrichments without thoseof Moauth. When the redoxcline reaches the water-column, Uauthenrichments increase and are usually coupled to Moauth enrich-ments (see Fig. 5 of Algeo and Tribovillard, 2009). Uauth contents inzone C are higher than those observed in zone B, but remain mod-erate and are not accompanied by a marked Mo-content increase.Hence we propose that the chemocline shallowed up, but remaineddeep enough below the sediment–water interface, preventing anymarked Mo accumulation. This scenario suggests that the watercolumn was oxic during the deposition of zone C (Fig. 10C).

The upper part of the Guia Formation (zone D) was depositedin the deepest environment. Here, carbonate is interbedded withmarl. In carbonate beds, the trace metal contents are similar tothose of zone C, likewise suggesting an oxic water column. Howeverthe marl levels are sometimes simultaneously enriched in Uauth,Moauth and pyrite (Fig. 7). This suggests episodic upward excursionsof anoxic conditions near or above the water–sediment interface(Fig. 10D), although zone D sediments were still deposited in adominantly oxic water column.

In summary the sedimentological and geochemical featuresof the formations studied here consistently point to a carbonateramp bathed by oxygenated waters in the aftermath of Marinoanglaciation. Reducing conditions developed only in the sedimentpore-space upon organic matter decay and were restricted mostof the time to below the sediment–water interface. The deepestparts of the platform may have been exposed to sporadic upwardexcursions of anoxic conditions at or above the sediment–waterinterface but these conditions have only seldom been met.

7.4. Implication for the lower Ediacarian ocean redox state

The snowball Earth model invokes a glaciation with an anoxicocean and an abrupt deglaciation under very warm climate. Theseconditions would have led to generalized, massive and rapidprecipitation of cap-carbonates during the deglaciation and thestabilization of pre-existing bottom water anoxia (Hoffman andSchrag, 2002; Hoffman et al., 2007). The presence of Marinoanbanded iron formations interbedded with diamictites indeed sug-gests marine anoxic conditions during the glaciation that heraldedthe Ediacaran Period (Klein and Ladeira, 2004; Piacentini et al.,2007). Yet, the redox evolution of the oceans in between the glacia-tion and the important oxidation event documented by severalisotopic and geochemical indicators at the end of the Ediacarantimes (Och and Shields-Zhou, 2012) remains poorly constrained,both spatially and temporally.

To date, the two Ediacarian successions for which the mostdirect redox indicators have been obtained are the DoushantuoFormation in South China (Ader et al., 2009; Huang et al., 2011;Jiang et al., 2010; Li et al., 2010; Sahoo et al., 2012) and the East-ern European Platform (Johnston et al., 2012) and NortwesternCanada (Johnston et al., 2013). Taken together, the Mirassol d’Oesteand Guia formations are probably roughly correlated to the lowerDoushantuo Formation. The Mirassol d’Oeste Formation is a typ-ical Marinoan cap dolostone, and hence correlates with the basalDoushantuo Formation cap carbonate. The base of the Guia Forma-tion probably represents a diachronous (transgressive) surface. Inthe shelf, it rests conformably over the Mirassol d’Oeste Formation,representing the overlying levels of lower Ediacarian age (Nogueiraet al., 2007), whereas in the outer-shelf and slope depositional envi-ronment it rests directly on the Puga diamictite and its base mightbe time-equivalent to the Mirassol d’Oeste of the shallower sectorof the Araras basin (Fig. 1).

Redox reconstructions for the Doushantuo Formation suggesta redox stratified water column over the Yangtze platform (Ader


et al., 2009; Li et al., 2010). The Doushantuo cap dolostones havebeen interpreted, on the basis of iron speciation and REE patterns asdeposited in ferruginous waters influenced by hydrothermal input,which is compatible with their relatively deep depositional envi-ronment (Huang et al., 2011; Jiang et al., 2006, 2011). The redoxstate of marls and carbonates deposited above the Doushantuocap dolostones has been identified based on iron speciation, pyritemorphology and sulfur isotopes to be dominantly oxic in shal-low depositional environments of the shelf margin and dominantlyanoxic with euxinic occurrences both in the inner-shelf and in thedeep basin (Li et al., 2010). While this redox reconstruction is com-patible with the idea of a globally stratified ocean throughout thelower Ediacarian, the recent U/TOC and Mo/TOC ratios reported forMember 2 of Doushantuo Formation point to a dominantly oxy-genated global ocean following deposition of the cap dolostonedeposition (Sahoo et al., 2012). In this scenario, anoxic conditionsmust then have been localized, and possibly restricted to intra-cratonic basins and oxygen minimum zones in upwelling areas ofsome continental platforms. Presumably, the Doushantuo Forma-tion would have been deposited in such an anoxic setting (Sahooet al., 2012).

The results of the present study are compatible with a scenario ofa dominantly oxygenated Ediacaran ocean. The waters bathing theAraras platform during the early Ediacarian were oxic down to theplatform slope, where anoxia was restricted to the sediment pore-water and only episodically reached the sediment–water interface.The Mo/TOC value of 72 in the upper part of the Mirassol d’OesteFormation in the Carmelo section is comparably high to valuesreported by Sahoo et al. (2012) for the Doushantuo Formation,further attesting to widespread oxygenated waters.

8. Conclusion

The sedimentological and geochemical analyses presented hereon post-Marinoan carbonates of the Araras Group (Mato Grosso,Brazil) complement the growing record of redox conditions in theaftermath of the Marinoan glaciation and allow the identification ofthe first Early Ediacaran platform dominated by oxic conditions. Theresults shows low trace metal contents but with significant strati-graphic variations, which can be traced along the platform withsuccessive enrichments of (i) Pb and Zn, (ii) U, and (iii) both U andMo in the thin marl levels containing the highest amount of organiccarbon and pyrite (0.4% and 1.9%, respectively). We interpret thissuccession to record the progressive evolution of pore-waterstoward more reducing conditions, driven by an increase in organicmatter flux possibly due to the post-glacial recovery of bioproduc-tivity (Elie et al., 2007; Kunzmann et al., 2013). The lack of evidencefor persistent sulfide mineralization and associated enrichmentsin Mo suggest that sulfidic conditions were restricted to belowthe sediment–water interface. In the aftermath of the Marinoanglaciation, the water column must have been essentially oxic onthe Araras platform, anoxia being restricted to the sediment pore-water and only episodically reaching the sediment–water interfaceat the end of cap-dolostones deposition and in the deepest part ofthe platform. This redox reconstruction is compatible with the ideaof a dominantly oxygenated ocean in the lower Ediacaran, in whichanoxic water masses developed locally in intra-cratonic basins andoxygen minimum zones.

Acknowledgments

Research was supported by a French MRT doctoral fellowshipand a SETSI grant to P. Sansjofre and two INSU (SYSTER) grantsto M. Ader, as well as an Emergence grant from Paris council (PIM. Bonifacie). R.I.F. Trindade and A.C.R. Nogueira were supported

by the INCT-Geociam, NAP-GEOSEDEx and by FAPESP and CNPqgrants. This is IPGP contribution n◦ 3471.

Appendix A. Supplementary data

Supplementary material related to this article can befound, in the online version, at http://dx.doi.org/10.1016/j.precamres.2013.11.004.

References

Ader, M., Macouin, M., Trindade, R.I.F., Hadrien, M.H., Yang, Z., Sun, Z., Besse, J., 2009.A multilayered water column in the Ediacaran Yangtze platform? Insights fromcarbonate and organic matter paired !13C. Earth and Planetary Science Letters288 (1–2), 213–227.

Algeo, T.J., 2004. Can marine anoxic events draw down the trace element inventoryof seawater? Geology 32 (12), 1057–1060.

Algeo, T.J., Lyons, T.W., 2006. Mo-total organic carbon covariation in modern anoxicmarine environments: implications for analysis of paleoredox and paleohydro-graphic conditions. Paleoceanography 21 (1), PA1016.

Algeo, T.J., Maynard, J.B., 2004. Trace-element behavior and redox facies in coreshales of Upper Pennsylvanian Kansas-type cyclothems. Chemical Geology 206(3–4), 289–318.

Algeo, T.J., Morford, J., Cruse, A., 2012. New applications of trace metals as proxiesin marine paleoenvironments. Chemical Geology 306–307, 160–164.

Algeo, T.J., Rowe, H., 2012. Paleoceanographic applications of trace-metal concen-tration data. Chemical Geology 324–325, 6–18.

Algeo, T.J., Tribovillard, N., 2009. Environmental analysis of paleoceanographic sys-tems based on molybdenum–uranium covariation. Chemical Geology 268 (3–4),211–225.

Allen, P.A., Hoffman, P.F., 2005. Extreme winds and waves in the aftermath of aNeoproterozoic glaciation. Nature 433 (7022), 123–127.

Alvarenga, C.J.S., 1988. Turbiditos e a Glaciacao do final do Proterozoico Superior nocinturao Paraguali, Mato Grosso. Revista Brasileira de Geociências 18, 323–327.

Alvarenga, C.J.S., Trompette, R., 1992. Glacially influenced sedimentation in the LaterProterozoic of the Paraguay belt (Mato Grosso, Brazil). Palaeogeography, Palaeo-climatology, Palaeoecology 92 (1–2), 85–105.

Alvarenga, C.J.S., Dardenne, M.A., Santos, R.V., Brod, E.R., Gioia, S.M.C.L., Sial, A.N.,Dantas, E.L., Ferreira, V.P., 2007. Isotope stratigraphy of Neoproterozoic capcarbonates in the Araras group, Brazil. Gondwana Research 186, p11.

Babinski, M., Trindade, R.I.F., Alvarenga, R.V., 2006. Geochronological constraintson the Neoproterozoic glaciations in Brazil. In: Snowball Earth, Monte Verita,Tricino (Switzerland), 16–21 July, pp. 19–20.

Banner, J.L., Hanson, G.N., 1990. Calculation of simultaneous isotopic and trace ele-ment variations during water–rock interaction with applications to carbonatediagenesis. Geochimica et Cosmochimica Acta 54 (11), 3123–3137.

Beck, M., Dellwig, O., Schnetger, B., Brumsack, H.-J.R., 2008. Cycling of trace metals(Mn, Fe, Mo, U, V, Cr) in deep pore waters of intertidal flat sediments. Geochimicaet Cosmochimica Acta 72 (12), 2822–2840.

Beveridge, T.J., Murray, R.G., 1976. Uptake and retention of metals by cell walls ofBacillus subtilis. Journal of Bacteriology 127 (3), 1502–1518.

Beveridge, T.J., Murray, R.G., 1980. Sites of metal deposition in the cell wall of Bacillussubtilis. Journal of Bacteriology 141 (2), 876–887.

Brand, U., Veizer, J., 1980. Chemical diagenesis of a multicomponent carbonate sys-tem. 1: trace elements. Journal of Sedimentary Research 50 (4), 1219–1236.

Bruland, K.W., 1980. Oceanographic distributions of cadmium, zinc, nickel, and cop-per in the North Pacific. Earth and Planetary Science Letters 47 (2), 176–198.

Brumsack, H.-J., 1980. Geochemistry of Cretaceous black shales from the AtlanticPcean (DSDP Legs 11, 14, 36 and 41). Chemical Geology 31 (0), 1–25.

Brumsack, H.-J., 2006. The trace metal content of recent organic carbon-rich sed-iments: implications for Cretaceous black shale formation. Palaeogeography,Palaeoclimatology, Palaeoecology 232 (2–4), 344–361.

Brumsack, H.J., 1986. The inorganic geochemistry of Cretaceous black shales (DSDPLeg 41) in comparison to modern upwelling sediments from the Gulf ofCalifornia. Geological Society of London 21 (1), 447–462.

Calvert, S.E., Pedersen, T.F., 1993. Geochemistry of recent oxic and anoxic marinesediments: implications for the geological record. Marine Geology 113 (1–2),67–88.

Canfield, D.E., Poulton, S.W., Narbonne, G.M., 2007. Late-neoproterozoic deep-oceanoxygenation and the rise of animal life. Science 315, 92–94.

Canfield, D.E., Raiswell, R., Bottrell, S.H., 1992. The reactivity of sedimentary ironminerals toward sulfide. American Journal of Science 292 (9), 659–683.

Canfield, D.E., Raiswell, R., Westrich, J.T., Reaves, C.M., Berner, R.A., 1986. The useof chromium reduction in the analysis of reduced inorganic sulfur in sedimentsand shales. Chemical Geology 54 (1–2), 149–155.

Canfield, D.E., Thamdrup, B., 2009. Towards a consistent classification scheme forgeochemical environments, or, why we wish the term ‘suboxic’ would go away.Geobiology 7 (4), 385–392.

Caplan, M.L., Bustin, R.M., 1999. Devonian-Carboniferous Hangenberg massextinction event, widespread organic-rich mudrock and anoxia: causes andconsequences. Palaeogeography, Palaeoclimatology, Palaeoecology 148 (4),187–207.



http://refhub.elsevier.com/S0301-9268(13)00345-8/sbref0005














































































































































































































































































































































































































































































































































































































Carignan, J., Hild, P., Mevelle, G., Morel, J., Yeghicheyan, D., 2001. Routine analyses oftrace elements in geological samples using flow injection and low pressure on-line liquid chromatography coupled to ICP-MS: a study of geochemical referencematerials BR, DR-N, UB-N, AN-G and GH. Geostandards Newsletter 25 (2–3),187–198.

Chaillou, G., Anschutz, P., Lavaux, G., Schäfer, J., Blanc, G., 2002. The distribution ofMo, U, and Cd in relation to major redox species in muddy sediments of the Bayof Biscay. Marine Chemistry 80 (1), 41–59.

Chang, H.-J., Chu, X.-L., Feng, L.-J., Huang, J., 2012. Progressive oxidation of anoxicand ferruginous deep-water during deposition of the terminal Ediacaran LaobaoFormation in South China. Palaeogeography, Palaeoclimatology, Palaeoecology321–322, 80–87.

Colodner, D., Edmond, J., Boyle, E., 1995. Rhenium in the Black Sea: comparison withmolybdenum and uranium. Earth and Planetary Science Letters 131 (1–2), 1–15.

Christie-Blick, N., von der Borch, C.C., Di Bona, P.A., 1990. Working hypotheses forthe origin of the Wonoka canyons (Neoproterozoic), South Australia. AmericanJournal of Science 290-A, 295–332.

Dellwig, O., Leipe, T., März, C., Glockzin, M., Pollehne, F., Schnetger, B., Yakushev,E.V., Böttcher, M.E., Brumsack, H.-J., 2010. A new particulate Mn–Fe–P-shuttleat the redoxcline of anoxic basins. Geochimica et Cosmochimica Acta 74 (24),7100–7115.

Dumas, S., Arnott, R.W.C., 2006. Origin of hummocky and swaley cross-stratification:the controlling influence of unidirectional current strength and aggradation rate.Geology 34 (12), 1073–1076.

Dyson, I.A., von der Borch, C.C., 1994. Sequence stratigraphy of an incised valley fill:the Neoproterozoic Seacliff Sandstone, Adelaide geosyncline, South Australia. In:Dalrymple, R.W., Boyd, R., Zaltlin, B.A. (Eds.), Incised-Valley Systems: Origin andSedimentary Sequences. SEPM (Soc. for Sediment. Geol.) Spec. Publ. 51, 209–222.

Elie, M., Nogueira, A.C.R., Nédélec, A., Trindade, R.I.F., Kenig, F., 2007. A red algalbloom in the aftermath of the Marinoan Snowball Earth. Terra Nova 19, 303–308.

Fairchild, I.J., 1993. Balmy shores and ice wastes: the paradox of carbonates asso-ciated with glacial deposits in Neoproterozoic times. Sedimentology Review 1,1–16.

Fairchild, I.J., Hambrey, M.J., 1984. The Vendian succession of northeastern Spits-berg petrogenesis of a dolomite-tillite association. Precambrian Research 26,111–167.

Feng, L.-J., Chu, X.-L., Huang, J., Zhang, Q.-R., Chang, H.-J., 2010. Reconstruction ofpaleo-redox conditions and early sulfur cycling during deposition of the Cryoge-nian Datangpo Formation in South China. Gondwana Research 18 (4), 632–637.

Font, E., Nédélec, A., Trindade, R.I.F., Macouin, M., Charrière, A., 2006. Chemostratig-raphy of the Neoproterozoic Mirassol d’Oeste cap dolostones (Mato Grosso,Brasil): an alternative model for Marinoan cap dolostone formation. Earth PlanetScience Letters 250, 89–103.

Font, E., Nédélec, A., Trindade, R.I.F., Moreau, C., 2010. Fast or slow melting of theMarinoan snowball Earth? The cap dolostone record. Palaeogeography, Palaeo-climatology, Palaeoecology 250, 215–225.

Font, E., Trindade, R.I.F., Nédélec, A., 2005. Detrital remanent magnetization inhaematite-bearing Neoproterozoic Puga cap dolostone, Amazon craton: a rockmagnetic and SEM study. Geophysical Journal International 163, 491–500.

Grotzinger, J.P.S., Knoll, A.H., 1995. Anomalous carbonate precipitates: is the pre-cambrian the key to the permian. PALAIOS 10, 578–596.

Hardie, L.A., 1977. Sedimentation on the modern carbonate tidal flats of northwestAndros Island, Bahamas, vol. 22. Johns Hopkins University, Baltimore.

Hatch, J.R., Leventhal, J.S., 1992. Relationship between inferred redox potential ofthe depositional environment and geochemistry of the Upper Pennsylvanian(Missourian) Stark Shale Member of the Dennis Limestone, Wabaunsee County,Kansas, USA. Chemical Geology 99 (1–3), 65–82.

Halverson, G.P., Maloof, A.C., Hoffman, P.F., 2004. The Marinoan glaciation (Neopro-terozoic) in northeast Svalbard. Basin Research 16 (3), 297–324.

Helz, G.R., Bura-Nakic, E., Mikac, N., Ciglenecki, I., 2011. New model for molybdenumbehavior in euxinic waters. Chemical Geology 284 (3–4), 323–332.

Helz, G.R., Miller, C.V., Charnock, J.M., Mosselmans, J.F.W., Pattrick, R.A.D., Garner,C.D., Vaughan, D.J., 1996. Mechanism of molybdenum removal from the sea andits concentration in black shales: EXAFS evidence. Geochimica et CosmochimicaActa 60 (19), 3631–3642.

Hetzel, A., Bottcher, M.E., Wortmann, U.G., Brumsack, H.-J., 2009. Paleo-redox condi-tions during OAE 2 reflected in Demerara Rise sediment geochemistry (ODP Leg207). Palaeogeography, Palaeoclimatology, Palaeoecology 273 (3–4), 302–328.

Hoffman, P.F., Kaufman, A.J., Halverson, G.P., Schrag, D.P., 1998. A neoproterozoicsnowball Earth. Science 281, 1342–1346.

Hoffman, P.F., Schrag, D.P., 2002. The snowball Earth hypothesis: testing the limitsof global change. Terra Nova 14, 129–155.

Hoffman, P.F., Halverson, G.P., Domack, E.W., Husson, J.M., Higgins, J.A., Schrag, D.P.,2007. Are basal Ediacaran (635 Ma) post-glacial “cap dolostone” diachronous?Earth and Planetary Science Letters 258, 114–131.

Hoffman, P.F., 2011. Strange bedfellows: glacial diamictite and cap carbonate fromthe Marinoan (635 Ma) glaciation in Namibia. Sedimentology 58 (1), 57–119.

Huang, J., Chu, X., Jiang, G., Feng, L., Chang, H., 2011. Hydrothermal origin of ele-vated iron, manganese and redox-sensitive trace elements in the c. 635 MaDoushantuo cap carbonate. (3), 805–816.

Hurtgen, M.T., Arthur, M.A., Suits, N.S., Kaufman, A.J., 2002. The sulfur isotopic com-position of Neoproterozoic seawater sulfate: implications for a snowball Earth?Earth and Planetary Science Letters 203 (1), 413–429.

James, N.P., Narbonne, G.M.K.K., 2001. Late Neoproterozoic cap carbonates: Macken-zie Mountains, northwestern Canada: precipitation and global glacial meltdown.Canadian Journal of Earth Sciences 38, 1229–1262.

Jiang, G., Kaufman, A.J., Christie-Blick, N., Zhang, S., Wu, H., 2007. Carbon isotopevariability across the Ediacaran Yangtze platform in South China: implicationsfor a large surface-to-deep ocean !13C gradient. Earth and Planetary ScienceLetters 261 (1–2), 303–320.

Jiang, G., Kennedy, M.J., Christie-Blick, N., Wu, H., Zhang, S., 2006. Stratigraphy, sed-imentary structures, and textures of the late neoproterozoic Doushantuo capcarbonate in South China. Journal of Sedimentary Research 76 (7), 978–995.

Jiang, G., Shi, X., Zhang, S., Wang, Y., Xiao, S., 2011. Stratigraphy and paleogeogra-phy of the Ediacaran Doushantuo formation (ca. 635–551 Ma) in South China.Gondwana Research 19 (4), 831–849.

Jiang, G., Wang, X., Shi, X., Zhang, S., Xiao, S., Dong, J., 2010. Organic carbonisotope constraints on the dissolved organic carbon (DOC) reservoir at theCryogenian–Ediacaran transition. Earth and Planetary Science Letters 299 (1–2),159–168.

Johnston, D.T., Poulton, S.W., Dehler, C., Porter, S., Husson, J., Canfield, D.E., Knoll,A.H., 2010. An emerging picture of Neoproterozoic ocean chemistry: insightsfrom the Chuar Group, Grand Canyon, USA. Earth and Planetary Science Letters290 (1–2), 64–73.

Johnston, D.T., Poulton, S.W., Goldberg, T., Sergeev, V.N., Podkovyrov, V., Vorob’eva,N.G., Bekker, A., Knoll, A.H., 2012. Late Ediacaran redox stability and metazoanevolution. Earth and Planetary Science Letters 335–336, 25–35.

Johnston, D.T., Poulton, S.W., Tosca, N.J., O’Brien, T., Halverson, G.P., Schrag, D.P.,Macdonald, F.A., 2013. Searching for an oxygenation event in the fossiliferousEdiacaran of northwestern Canada. Chemical Geology 362, 273–286.

Kendall, C.G.S.C., Warren, J., 1987. A review of the origin and setting of tepees andtheir associated fabrics. Sedimentology 34 (6), 1007–1027.

Kennedy, M.J., 1996. Stratigraphy, sedimentology, and isotopic geochemistry ofAustralian Neoproterozoic post glacial cap dolostones; deglaciation, delta 13Cexcursions, and carbonate precipitation. Journal of Sedimentary Research 66,1050–1064.

Klein, C., Ladeira, E.A., 2004. Geochemistry and mineralogy of Neoproterozoicbanded iron – formations and some selected, siliceus manganese formationsfrom the Urucum district, Mato Grosso do Sul. Brazil Economic Geology 99 (6),1233–1244.

Klinkhammer, G.P., Palmer, M.R., 1991. Uranium in the oceans: where it goes andwhy. Geochimica et Cosmochimica Acta 55 (7), 1799–1806.

Kryc, K.A., Murray, R.W., Murray, D.W., 2003. Al-to-oxide and Ti-to-organic linkagesin biogenic sediment: relationships to paleo-export production and bulk Al/Ti.Earth and Planetary Science Letters 211 (1–2), 125–141.

Kunzmann, M., Halverson, G.P., Sossi, P.A., Raub, T.D., Payne, J.L., Kirby, J., 2013.Zn isotope evidence for immediate resumption of primary productivity aftersnowball Earth. Geology 41 (1), 27–30.

Li, C., Love, G.D., Lyons, T.W., Fike, D.A., Sessions, A.L., Chu, X., 2010. A stratified redoxmodel for the Ediacaran Ocean. Science 328 (5974), 80–83.

Lyons, T.W., 2007. Oxygen’s rise reduced. Nature 448, 1005–1006.Lyons, T.W., Anbar, A.D., Severmann, S., Scott, C., Gill, B.C., 2009. Tracking Euxinia in

the ancient ocean: a multiproxy perspective and proterozoic case study. AnnualReview of Earth and Planetary Sciences 37 (1), 507–534.

Lyons, T.W., Werne, J.P., Hollander, D.J., Murray, R.W., 2003. Contrasting sulfur geo-chemistry and Fe/Al and Mo/Al ratios across the last oxic-to-anoxic transitionin the Cariaco Basin, Venezuela. Chemical Geology 195 (1–4), 131–157.

McFadden, K.A., Huang, J., Chu, X., Jiang, G., Kaufman, A.J., Zhou, C., Yuan, X., Xio,S., 2007. Pulsed oxidation and biological evolution in the Ediacaran Doushan-tuo formation. Proceedings of the National Academy of Science 105, 3197–3202.

McLennan, S.M., 1989. Rare earth elements in sedimentary rocks; influence of pro-venance and sedimentary processes. Reviews in Mineralogy and Geochemistry21, 169–200.

McManus, J., Berelson, W.M., Severmann, S., Poulson, R.L., Hammond, D.E.,Klinkhammer, G.P., Holm, C., 2006. Molybdenum and uranium geochemistryin continental margin sediments: paleoproxy potential. Geochimica et Cos-mochimica Acta 70 (18), 4643–4662.

Morford, J.L., Russell, A.D., Emerson, S., 2001. Trace metal evidence for changes inthe redox environment associated with the transition from terrigenous clay todiatomaceous sediment, Saanich Inlet, BC. Marine Geology 174 (1–4), 355–369.

Myrow, P.M., Kaufman, A.J., 1999. A newly discovered cap carbonate aboveVaranger-age glacial deposits in Newfoundland, Canada. Journal of SedimentaryResearch 69 (3), 784–793.

Nogueira, A.C.R., Riccomini, C., Sial, A.N., Moura, C.A.V., Fairchild, T.R., 2003. Soft-sediment deformation at the base of the Neoproterozoic Puga cap carbonate(southwestern Amazon craton, Brazil): confirmation of rapid icehouse to green-house transition in snowball Earth. Geology 31 (7), 613–616.

Nogueira, A.C.R., Riccomini, C., Sial, A.N., Moura, C.A.V., Trindade, R.I.F., Fairchild,T.R., 2007. Carbon and strontium isotope fluctuations and paleoceanographicchanges in the late Neoproterozoic Araras carbonate platform, southern Amazoncraton, Brazil. Chemical Geology 237, (1–2), 168–190.

Och, L.M., Shields-Zhou, G.A., 2012. The Neoproterozoic oxygenation event: environ-mental perturbations and biogeochemical cycling. Earth-Science Reviews 110,26–57.

Pfeil, R.W., Read, J.F., 1980. Cambrian carbonate platform margin facies, shadydolomite, southwestern Virginia, USA. Journal of Sedimentary Petrology 50,91–116.

Piacentini, T., Boggiani, P.C., Yamamoto, J.K., Freitas, B.T., Cruz Campanha, G.A., 2007.Formac ao ferrifera associada à sedimentac ao glaciogênica da Formac ao Puga(marinoano) na Derra da Bodoquena, MS. Revista Brasileira de Geociências 37,530–541.





















































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































Piper, D.Z., Perkins, R.B., Rowe, H.D., 2007. Rare-earth elements in the Permian Phos-phoria formation: paleo proxies of ocean geochemistry: deep sea research partII. Topical Studies in Oceanography 54 (11–13), 1396–1413.

Planavsky, N.J., McGoldrick, P., Scott, C.T., Li, C., Reinhard, C.T., Kelly, A.E., Chu, X.,Bekker, A., Love, G.D., Lyons, T.W., 2011. Widespread iron-rich conditions in themid-Proterozoic ocean. Nature 477 (7365), 448–451.

Plummer, P.S., 1978. Note on the palaeoenvironmental significance of theNuccaleena formation (upper Precambrian), central Flinders Ranges, SouthAustralia. Geological Society of Australia 25, 395–402.

Pratt, B.R., 1994. Seismites in the Mesoproterozoic Altyn formation (Belt Super-group), Montana: a test for tectonic control of peritidal carbonate cyclicity.Geology 22 (12), 1091–1094.

Raiswell, R., Canfield, D.E., 1996. Rates of reaction between silicate iron and dissolvedsulfide in Peru Margin sediments. Geochimica et Cosmochimica Acta 60 (15),2777–2787.

Ripley, E.M., Shaffer, N.R., Gilstrap, M.S., 1990. Distribution and geochemical charac-teristics of metal enrichment in the New Albany Shale (Devonian-Mississippian),Indiana. Economic Geology 85 (8), 1790–1807.

Romero, J.A.S., Lafon, J.M., Nogueira, A.C.R., Soares, J.L., 2013. Sr isotope geochemistryand Pb–Pb geochronology of the Neoproterozoic cap carbonates, Tangara daSerra, Brazil. International Geology Review 55, 185–203.

Rose, C.V., Maloof, A.C., 2010. Testing models for post-glacial ‘cap dolostone’ deposi-tion: Nuccaleena formation, South Australia. Earth and Planetary Science Letters296 (3–4), 165–180.

Sageman, B.B., Murphy, A.E., Werne, J.P., Ver Straeten, C.A., Hollander, D.J., Lyons,T.W., 2003. A tale of shales: the relative roles of production, decomposition,and dilution in the accumulation of organic-rich strata, Middle-Upper Devonian,Appalachian bas. Chemical Geology 195 (1–4), 229–273.

Sahoo, S.K., Planavsky, N.J., Kendall, B., Wang, X., Shi, X., Scott, C., Anbar, A.D., Lyons,T.W., Jiang, G., 2012. Ocean oxygenation in the wake of the Marinoan glaciation.Nature 489 (7417), 546–549.

Sansjofre, P., Ader, M., Trindade, R.I.F., Elie, M., Lyons, J., Cartigny, P., Nogueira, A.C.R.,2011. A carbon isotope challenge to the snowball Earth. Nature 478 (7367),93–96.

Scott, C., Lyons, T.W., 2012. Contrasting molybdenum cycling and isotopic propertiesin euxinic versus non-euxinic sediments and sedimentary rocks: refining thepaleoproxies. Chemical Geology 324–325, 19–27.

Scott, C., Lyons, T.W., Bekker, A., Shen, Y., Poulton, S.W., Chu, X., Anbar, A.D., 2008.Tracing the stepwise oxygenation of the Proterozoic ocean. Nature 452 (7186),456–459.

Shinn, E.A., 1983. Birdseyes, fenestrae, shrinkage pores, and loferites; a reevaluation.Journal of Sedimentary Research 53 (2), 619–628.

Soares, J.L., Nogueira, A.C.R., 2008. Depósitos carbonáticos de Tangará da Serra (MT):uma nova ocorrência de capa carbonática neoproterozóica no sul do CrátonAmazônico. Revista Brasileira de Geociências 34 (4), 715–729.

Stow, D.A.V., 1986. Deep clastic seas. In: Reading, H.G. (Ed.), Sedimentary Environ-ments: Processes, Facies and Stratigraphy. Blackwell Scientific Public., Oxford,pp. 399–444.

Sumner, D.Y., 2002. Decimeter-thick encrustations of calcite and aragonite onthe sea-floor and implications for Neoarchaean and Neoproterozoic ocean

chemistry. In: Altermann, W., Corcoran, P.L. (Eds.), Precambrian SedimentaryEnvironments: A Modern Approach to Ancient Depositional Systems. Spe-cial Publication, vol. 33. International Association of Sedimentologists,pp. 107–120.

Sumner, D.Y., Grotzinger, J.P., 1996. Were kinetics of Archean calcium carbonateprecipitation related to oxygen concentration? Geology 24 (2), 119–122.

Tribovillard, N., Algeo, T.J., Baudin, F., Riboulleau, A., 2012. Analysis of marineenvironmental conditions based on molybdenum–uranium covariation – appli-cations to Mesozoic paleoceanography. Chemical Geology 324–325, 46–58.

Tribovillard, N., Algeo, T.J., Lyons, T., Riboulleau, A., 2006. Trace metals as paleoredoxand paleoproductivity proxies: an update. Chemical Geology 232 (1–2), 12–32.

Tribovillard, N., Bout-Roumazeilles, V., Algeo, T., Lyons, T.W., Sionneau, T., Montero-Serrano, J.C., Riboulleau, A., Baudin, F., 2008. Paleodepositional conditions in theOrca Basin as inferred from organic matter and trace metal contents. MarineGeology 254 (1–2), 62–72.

Tribovillard, N., Riboulleau, A., Lyons, T., Baudin, F., 2004. Enhanced trapping ofmolybdenum by sulfurized marine organic matter of marine origin in Mesozoiclimestones and shales. Chemical Geology 213 (4), 385–401.

Trindade, R.I.F., Font, E., D’Agrella-Filho, M.S., Nogueira, A.C.R., Riccomini, C., 2003.Low-latitude and multiple geomagnetic reversals in the Neoproterozoic Pugacap carbonate, Amazon craton. Terra Nova 15, 441–446.

Trompette, R., 1994. Geology of Western Gondwana (2000–500 Ma): Pan-African–Brasiliano Aggregation of South America and Africa. Balkema,Rotterdam, NL Public, 350 p.

Trompette, R., 1997. Neoproterozoic (∼600 Ma) aggregation of Western Gondwana:a tentative scenario. Precambrian Research 82 (1–2), 101–112.

Tucker, M.E., 1986. Formerly aragonitic limestones associated with tillites in theLate Proterozoic of Death Valley, California. Journal of Sedimentary Petrology56, 818–830.

Tucker, M.E., 1991. Sedimentary petrology: an introduction to the origin of sedi-mentary rocks, second ed. Blackwell Scientific Publications, Oxford.

Van der Weijden, C.H., 2002. Pitfalls of normalization of marine geochemical datausing a common divisor. Marine Geology 184 (3–4), 167–187.

Von der Borch, C.C., Grady, A.E., Eickhoff, K.H., Di Bona, P., Christie-Blick, N., 1989. LateProterozoic Patsy Springs canyon, Adelaide Geosyncline: submarine or subaerialorigin? Sedimentology 36, 777–792.

Wang, J., Chen, D., Yan, D., Wei, H., Xiang, L., 2012. Evolution from an anoxic tooxic deep ocean during the Ediacaran–Cambrian transition and implications forbioradiation. Chemical Geology 306–307, 129–138.

Wangersky, P.J., 1986. Biological control of trace metal residence time and specia-tion: a review and synthesis. Marine Chemistry 18 (2–4), 269–297.

Williams, G.E., 1979. Sedimentology, stable-isotope geochemistry and palaeoenvi-ronment of dolostones capping late Precambrian glacial sequences in Australia.Geological Society of Australia 26, 377–386.

Winterer, E.L., Sarti, M., 1994. Neptunian dykes and associated features in southernSpain: mechanics of formation and tectonic implications. Sedimentology 41 (6),1109–1132.

Zheng, Y., Anderson, R.F., van Geen, A., Kuwabara, J., 2000. Authigenic molybdenumformation in marine sediments: a link to pore water sulfide in the Santa BarbaraBasin. Geochimica et Cosmochimica Acta 64 (24), 4165–4178.

























































































































































































































































































































































































































































































































































































































































































































































































Paleoenvironmental reconstruction of the Ediacaran Araras platform (Western Brazil) from the sedimentary and trace metals record

Documents