Geobiology. 2018;16:279–296. wileyonlinelibrary.com/journal/gbi | 279

Received:30August2017  |  Accepted:11January2018DOI: 10.1111/gbi.12278

O R I G I N A L A R T I C L E

A morphogram for silica- witherite biomorphs and its application to microfossil identification in the early earth rock record

J. Rouillard1  | J.-M. García-Ruiz2 | J. Gong1 | M. A. van Zuilen1

1EquipeGéomicrobiologie,InstitutdePhysiqueduGlobedeParis,SorbonneParisCité,UniversitéParisDiderot,UMR7154,CNRS,Paris,France2LaboratoriodeEstudiosCristalográficos,Instituto Andaluz de Ciencias de la Tierra,ConsejoSuperiordeInvestígacìonesCientificas–UniversidaddeGranada,Granada,Spain

CorrespondenceJ.Rouillard,EquipeGéomicrobiologie,InstitutdePhysiqueduGlobedeParis,SorbonneParisCité,UniversitéParisDiderot,UMR7154CNRS,Paris,France.Email:rouillard@ipgp.fr

Funding informationtheEuropeanResearchCouncil(ERC)undertheEuropeanUnion’sHorizon2020researchandinnovationprogramme,Grant/AwardNumber:646894;theERCSeventhFrameworkProgrammeFP7/2007-2013,Grant/AwardNumber:340863

AbstractArcheanhydrothermalenvironmentsformedalikelysitefortheoriginandearlyevo-lutionoflife.Thesearealsothesettings,however,werecomplexabiologicstructurescanform.Low-temperatureserpentinizationofultramaficcrustcangeneratealka-line,silica-saturatedfluidsinwhichcarbonate–silicacrystallineaggregateswithlife-like morphologies can self-assemble. These “biomorphs” could have adsorbedhydrocarbonsfromFischer–Tropschtypesynthesisprocesses,leadingtometamor-phosedstructuresthatresemblecarbonaceousmicrofossils.Althoughthisabiogenicprocesshasbeenextensivelycitedintheliteratureandhasgeneratedimportantcon-troversy,sofaronlyonespecificbiomorphtypewithafilamentousshapehasbeendiscussedfortheinterpretationofArcheanmicrofossils.Itisthereforecriticaltopre-ciselydeterminethefulldistributioninmorphologyandsizeofthesebiomorphs,andto study the rangeof plausible geochemical conditions underwhich thesemicro-structurescanform.Here,asetofwitherite-silicabiomorphsynthesisexperimentsinsilica-saturatedsolutionsispresented,forarangeofpHvalues(from9to11.5)andbariumionconcentrations(from0.6to40mmol/LBaCl2).Underthesevaryingcondi-tions,awiderangeoflife-likestructuresisfound,fromfractaldendritestocomplexshapeswithcontinuouscurvature.Thesize,spatialconcentration,andmorphologyofthebiomorphsarestronglycontrolledbyenvironmentalparameters,amongwhichpH is themost important. This potentially limits the diversity of environments inwhichthegrowthofbiomorphscouldhaveoccurredonEarlyEarth.Giventhevari-etyoftheobservedbiomorphmorphologies,ourresultsshowthatthemorphologyof an individual microstructure is a poor criterion for biogenicity. However, bio-morphsmaybedistinguishedfromactualpopulationsofcellularmicrofossilsbytheirwide,unimodalsizedistribution.Biomorphsgrownbydiffusion insilicagelcanbedifferentiatedbytheircontinuousgradientinsize,spatialdensity,andmorphologyalong the direction of diffusion.

ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium,providedtheoriginalworkisproperlycited.© 2018 The Authors. GeobiologyPublishedbyJohnWiley&SonsLtd.

280  |     ROUILLARD et AL.

1  | INTRODUC TION

The search for biosignatures in the Early Archean geological record ischallenging,asmostsedimentaryformationsfromthistimeinter-valhavebeenexposedtohydrothermalfluids,andhaveundergoneat least prehnite–pumpellyite facies to lower greenschist faciesmetamorphism.

Under these high temperature and high pressure conditions,rocks have recrystallized and organic matter has undergone car-bonization. Therefore, any original organic remnant of microbiallifepreservedwithinthemhasbeenchemicallyalteredandhaspo-tentiallybeendeformed.Moreover, therearemanynon-biologicalprecipitation/crystallization phenomena leading to morphologiesresembling primitive fossilizedmicro-organisms. Examples are ex-foliated micas (Wacey, Saunders, Kong, Brasier, & Brasier, 2015),dispersedhematitecrystals(Marshall,Emry,&Marshall,2011;Pinti,Mineau,&Clement,2009),andmigratedhydrocarbons(eitherbio-genic or abiogenic) that filled the pore space between botryoidalquartz grains (Brasier etal., 2005). Most intriguing, however, areself-organized life-like silica–carbonate precipitates (Garcia-Ruizetal.,2003).Thesemicrostructuresformspontaneouslyinthelabo-ratorywhenalkaline-earthmetalsaremixedwithsilica-richalkalinesolutionsinthepresenceofCO2(García-Ruiz,1998).Theyhavealsobeensynthesizedbydiffusionofalkaline-earthmetalsthroughsil-icagels.Recently, it hasbeendemonstrated that these structurescanforminmodernserpentinization-derivedalkalinespringwaters(García-Ruiz,Nakouzi,Kotopoulou,Tamborrino,&Steinbock,2017).Oncethesemineral“biomorphs”areformed,theadsorptionofor-ganicmolecules(eitherbiogenicorabiogenic)ontheirsurfaceandsubsequent thermal alterationduringmetamorphismwould trans-form them into carbonaceous microstructures that truly resemble actualmicrofossils(García-Ruiz,Carnerup,Christy,Welham,&Hyde,2002;Garcia-Ruizetal.,2003;Opel,Wimmer,Kellermeier,&Cölfen,2016). For instance, it has been suggested by Garcia-Ruiz etal.(2003)thatsuchaprocessrepresentsapotentialexplanationforthecontroversialfilamentousmicrofossilsofthe3.5GaoldApexChert,Pilbara,WesternAustralia(Brasieretal.,2002;Schopf,Kudryavtsev,Agresti,Wdowiak,&Czaja,2002).

During hydrothermal serpentinization of seafloor ultramaficrocks, thepHofemanating fluids is largelycontrolledby tempera-ture.Above250°C,thesefluidsareslightlyacidic(pH<5),butwithdecreasing reaction temperature, the fluids become increasinglyalkaline, reachingapHof11at50°C (Macleod,McKeown,Hall,&Russell,1994;McCollom&Bach,2009).This increase inpH isduetolow-temperatureequilibrationwithbrucitethatformsduringtheserpentinizationprocess(McCollom&Bach,2009).Thetemperatureofhydrothermalfluidsisdependentonthedepthofcirculation,theaverage geothermal gradient, and the extent of infiltration of sea-water cooling the crust. As a consequence, although the Archeangeothermal gradient is believed to be higher than today (Arndt &Nisbet,2012),fluidsthatalteredultramaficrocksmayhavedisplayedawiderangeoftemperaturesandpHvalues.Alkalinehydrothermalserpentinization of ultramafic crust—which is observed at modern

low-temperature(<250°C)ventsitessuchas“LostCity”intheAtlanticOcean(Lang,Butterfield,Schulte,Kelley,&Lilley,2010;Proskurowskietal., 2008)—likely was an important process in aquatic environ-ments of Hadean and Early Archean age (Russell, Hall, & Martin,2010;Shibuya,Komiya,Nakamura,Takai,&Maruyama,2010),whenkomatiitescovered largeareasof theocean floor,whichcontainedmoreMgOthanbasalticrockstoday(Arndt&Nisbet,2012).

EarlyinEarthhistory,theconcentrationofsilicaintheoceanswasmuchhigherthantoday(Konhauser,Jones,Reysenbach,&Renaut,2003;Maliva,Knoll,&Simonson,2005;Siever,1992),whichisat-testedbyubiquitoussedimentarychertdepositsinArcheangreen-stonebelts(Nijman,DeBruijne,&Valkering,1998;VandenBoorn,VanBergen,Nijman,&Vroon, 2007;VanKranendonk&Pirajno,2004).Thesechertsformedeitherbydirectprecipitationofcolloi-dalsilicafromsilica-richseawater,orbyfluid-inducedsilicainfiltra-tion/replacement(silicification)ofpreexistingsediments(Stefurak,Lowe, Zentner, & Fischer, 2014; Van den Boorn etal., 2007). Insome cases, such as in the Marble Bar chert, Pilbara, WesternAustralia,softdeformationandremobilizationfeaturesshowthatthechertprecursor,atleastatsomepointofitshistory,wasahigh-viscosity gel-like material (McLoughlin, Wilson, & Brasier, 2008;VanKranendonk,2006).Debrisfrommicro-organismslivingintheupper water columnwould settle down on the ocean floor, andthusbecomeeffectivelyentombedbysilicagel.Uponburial,theseamorphoussilicadeposits(opal-A)wouldtransformtochertscon-sisting of microcrystalline quartz in which microfossil assemblages couldhavebeenpreserved.Inadditiontotheseclearsedimentarycherts, there isubiquitousevidenceforhydrothermalcherts thatformedwhen silica-rich seawater circulated through the igneouscrustalbasement.Suchhydrothermaleffluents,uponcoolingandmixingwithoverlyingseawater,wouldhaveprecipitatedsilica,di-rectlywithinthehydrothermalductornearitsvent,thuscreatingchertdykesandlocallayereddeposits(VandenBoornetal.,2007).Barite veins are often found associatedwith these hydrothermalcherts (Nijman etal., 1998; VanKranendonk,Hickman,Williams,&Nijman, 2001) indicating that significant amounts of dissolvedBa-ionswerepresent intheprecursorhydrothermalfluids.Thesehydrothermalfluidsalsocausedserpentinizationofolivinewithinthecrust,releasingH2andcausingthemetal-catalyzedsynthesisofCH4andmorecomplexhydrocarbonsbyFischer–Tropschsynthesis(McCollom&Seewald,2006;Milesietal.,2015).Thesehydrother-malenvironmentswouldalsohaveformedanimportanthabitatforthermophilicmicro-organismsthatarecapableofmetabolizingre-ducedgases,suchashydrogenormethane(chemolithoautotrophs;Ueno,Yamada,Yoshida,Maruyama,&Isozaki,2006).Organicrem-nantsofchemolithoautotrophscouldthusbeeffectivelyentombedinhydrothermalsilica,andcouldhavebeenpreservedascarbona-ceousmicrofossils inArchean chert veins. Finally, chert depositsalso form in subaerial hot spring environments.When silica-rich,andoftenalkalinethermalfluidsreachthesurface,amorphoussil-ica isprecipitatedduetoevaporationand/orasignificantdropintemperature, leadingtosilicasinters inwhichremnantsofmicro-biallifearesuperblypreserved(Campbelletal.,2015;Munoz-Saez,

     |  281ROUILLARD et AL.

Saltiel, Manga, Nguyen, & Gonnermann, 2016; Ruff & Farmer,2016).Recently,itwasshownthatsuchsilicasinterswithremnantmicrobial biosignatures (stromatolites, microbial palisade struc-tures,gasbubbles)occurredalreadyasearlyas3.5Gaago(Djokic,VanKranendonk,Campbell,Walter,&Ward,2017).

Overall, alkaline and silica-rich conditions (predominantlysubaqueous but also subaerial) were probably frequent inserpentinization-dominated environments during the Hadean andEarlyArchean,forminglikelysettingswherelifeappearedandevolved(Damer,2016;Djokicetal.,2017;Holm,1992;Holm&Andersson,2005;Martin & Russell, 2007; Russell etal., 2010; Schulte, Blake,Hoehler,&McCollom,2006),butalsowheresynthesisofsilica–car-bonatebiomorphscouldhavetakenplace(Garcia-Ruizetal.,2003).Onaglobalscale,theCO2partialpressureoftheEarlyArcheanat-mospherehasbeen interpretedtobehigher thantoday’s (Kasting,2014;Walker,1985;Wolf&Toon,2014),leadingtosignificantcon-centrationsofdissolvedcarbonatespeciesinArcheanseawaterandmild acidification of the seawater. Three main mechanisms can then be invoked for the formationofbiomorphs in theArchean: (i)Thediffusionofcalcium-richseawaterintogelifiedsilicadepositswouldresultintheformationofmonohydrocalciteoraragonitebiomorphs,(ii) the diffusion of a barium-rich hydrothermal fluid in gelified or-thochemicalsilicadepositswouldresultintheformationofwitheritebiomorphs,and(iii)themixingbetweenabarium-richhydrothermalfluidandasilica-richhydrothermalfluidwouldresult intheforma-tionofwitheritebiomorphs.Thepotentialco-occurrenceofearlymi-crobiallifeandmineral-based“biomorphs”inArcheanenvironmentsmustthereforebetakenintoaccountinearlyliferesearch,andthemorphologicalidentificationofcarbonaceousmicrostructuresasan-cient microfossils should be carefully assessed.

García-Ruiz,Melero-García, andHyde (2009) haveproposed amorphogenetic model describing the growth of silica–carbonatebiomorphs.Thesestructurescanbeformedwithbariumcarbonate(witherite,which is the best studied variety), strontium carbonate(strontianite), or calcium carbonate (aragonite andmonohydrocal-cite;Bittarello,RobertoMassaro,&Aquilano,2010;Voinescuetal.,2007;Zhang,2015;Zhang,Morales,&García-Ruiz,2017).Inanal-kalinesolutionenriched inSiO2,CO3

2− (derivedfromthediffusionofatmosphericCO2),andametalcation (Ba,Ca,orSr), thenucle-ationofcarbonatecrystalswillproceedspontaneously.Theirgrowthcausesa localdecrease inpHthat triggerssilicapolymerization inthedirectvicinityofthecarbonatenucleii.Growthofthecarbonatecrystals is therefore impededbyoligomeric silica impurities, caus-ingthemaingrowthfrontofthecrystalstosplitrepeatedlyatnon-crystallographicangles.Duringthisprocess,singlecrystalsymmetryisbroken,leadingtotheformationofdendritesthatexpandintotheentireavailablevolume.Thisepisodeinbiomorphformationiscalledthe“fractalgrowthregime.”

Upon sustained fractal growth, the amount of silica coatingthe witherite crystals becomes eventually high enough to inter-rupt the crystallization of dendrites. Precipitating silica increasesthelocalsaturationindexofwitherite,untilnewnucleationeventsoccuronthesilicacoating.Thedendriticmaterial isthusreplaced

byasheet-likepolycrystallinematerialconsistingofcarbonatena-norods (40nmwide, 400nm long) interspersed in an amorphoussilicamatrix,aso-calledfibrillationevent(García-Ruizetal.,2009).Theorientationofthelongaxisofthenanorodsvariescontinuouslyandsmoothlywiththepositionofthecrystalinthematerial,ageo-metrical feature which has been referred to as orientation order-ing(García-Ruizetal.,2002).Thisgeneratesanoverallcontinuouscurvature in thematerial, characteristic of the secondepisode inbiomorphformation,calledthe“curvilineargrowthregime”(García-Ruizetal.,2009).

Inthisstudy,wesynthesizedsilica-witheritebiomorphsboth inalkaline silica-rich solutions and in alkaline silica gels, representingpotential precursors for Archean cherts. For experiments realizedinalkalinesilica-enrichedstartingsolutions,a systematicoverviewis presented on the variation in morphology—defined as a “mor-phogram”—of witherite-silica biomorphs that form at the surface(water–air interface), over a rangeof pHvalues andbariumcationconcentrations.Toquantify thesizedistributionofbiomorphs,au-tomatedmeasurements in optical imageswere also performed. Insolution,biomorphstendtoaggregate,loweringthequalityofthesemeasurements.Toavoidthisproblem,thesizedistributionwaspref-erentially studied for biomorphs that were grown in silica gel. Inthismedium,biomorphs aredispersed throughout the gelwithouttouchingeachother.solution-basedexperimentsandgel-basedex-perimentscanbecomparedbecausethemorphogeneticmechanismunderlyingbiomorphgrowthinthesetwomediaisexactlythesame(Kellermeier,Cölfen,&García-Ruiz,2012).AsitisplausiblethatsilicagelscouldhaveformedincertainHadean/Archeanaquaticenviron-ments,thespecificdiffusion-relatedpatternsobservedingelexper-iments are also discussed. The results of both solution-based andgel-basedexperimentsaresubsequentlyusedtodefinecriteria fordistinctionbetweenbiomorphsandmicrofossils,andtomakeanas-sessmentofthelikelihoodthatassemblagesofbiomorphsareassoci-atedwith,ormistakenfor,truemicrofossilsinancientchertdeposits.

2  | METHODS

2.1 | Computer simulations of witherite nucleation

Numericsimulationswereperformedtocharacterizetheeffectofourexperimentalparametersonthecrystallizationbehaviorofwith-erite.Accordingtoclassicalnucleationtheory (DeYoreo&Vekilov,2003;Nývlt,Söhnel,Matuchová,&Broul,1985),thecharacteristicwaiting time for nucleation T(s)decreaseswiththesaturationindexσof themineral (withoutdimension)asshownbyEquation1.Theconstants A (s)andB (withoutdimension)arepositive,andforthecaseofwitherite,σ is defined as (Ba2+)(CO3

2−)/Ks,whereKs is the solubility constant of witherite:

The saturation index of witherite for each individual start-ing solutionwas computed usingVisualMinteQ (version 3.1) sim-ulations. The partial pressure ofCO2was fixed at 3.8×10

−4bar,

(1)T=A∗exp (B∕σ2)

282  |     ROUILLARD et AL.

representingambientatmosphericlevel.Separatesimulationsweremadeforeach individual initialbariumconcentration ([Ba]) inourexperiments. TheCl− concentrationwas fixed at twice the initial[Ba].Tomatchtheexperimentalvaluesmeasured,thepHwasvar-iedbetween9.3and11.3,withastepof0.1pHunitsbetweentwosimulations.Onlytheformationofwitheriteandamorphoussilica(gel)wasallowed.TheactivitycorrectionmethodchosenwasbasedonthetheoryofDebye–Hückelfornon-idealelectrolytesolutions.

2.2 | Synthesis of biomorphs

2.2.1 | Synthesis of biomorphs in solution

A15mmol/Lsilicasolutionwaspreparedbydiluting1mlofcom-mercial water glass (Sigma-Aldrich, 12.5 wt % of Si) in 350ml ofmillipore-filtratedwater.ToobtainarangeofdifferentinitialpHval-ues,thismothersolutionwasaliquotedinsmallvolumes,anddiffer-entvolumesofNaOHorHClsolutions(0.1mol/L)wereadded,from1mlHClto1mlNaOHper10mlofthesodiumsilicatesolution.Intotal,21solutionswithdifferentinitialpHwereproducedthisway.Several solutions of different BaCl2 concentration were prepared(givingarangein[Ba]from1.25to40mmol/L)upondissolutionofsolidbariumchloridedihydrateinmillipore-filtratedwater.Thesolu-tionswerethencombinedin24-wellultra-cleardishestoobtain147synthesis media with distinct initial conditions. The dishes were cov-erednon-hermeticallybyalid,toallowthediffusionofatmosphericCO2withinthesolutionwhileavoidingdustdepositionthatwoulddecreasetherateofevaporation.Thisprotocolwasusedtwice.Thefirsttime,themorphologicalevolutionofbiomorphswasspecificallytracked.Thesecondtime,thepHevolutionofasubsetof22differ-entinitialsolutionswasmeasuredfor24hrbyaThermoScientificROSSUltra™semimicropHelectrode.

2.2.2 | Synthesis of biomorphs in a silica gel

Silica gels were prepared by neutralizing a 10ml silica solution(~0.5mol/L)using3.25mlofanHClsolution(1mol/L).Oncethegel-lingprocesswascomplete,asolutionofBaCl2(0.25,0.5or1mol/L)wasaddedonthetopofthegelandallowedtodiffuseinside.Theprotocolisdescribedinmoredetail inMelero-García,Santisteban-Bailón, and García-Ruiz (2009). Under these conditions, largebiomorphsappear inafewdaysinthegel.Thisprotocolwasusedspecifically for the studies of the size distribution, as it produceswell-dispersedbiomorphs,sothatpicturescanbeeasilytreateddur-ing image analysis.

2.3 | Biomorph characterization

Biomorphcrystallization(firstappearance,growth,andmorphology)attheliquid–atmosphereinterfacewastrackedbyobtainingopticalmicroscopypicturesofeachofthe147solutionsatregulartimeinter-valsfor2days.Apetrographicmicroscope(LeicaDM2500P;InstitutdePhysiqueduGlobedeParis)wasusedintransmissionmodewith

a10×objective.Twomultifocus long-rangemicroscopeswerealsoused for separate picture acquisition at random intervals (NikonAZ100;InstitutoAndaluzdeCienciasdelaTierra,Granada,Spain).

OpticalmicroscopypictureswereprocessedusingiMageJsoftware(Abramoff,Magalhaes,&Ram,2004).Afterconversiontograylevel,athresholdandabinarizationwereappliedtothepicturestoseparatethecrystallizedparticles fromthebackground.The typeof thresh-oldwascriticalinthisaspect,assometypestendedtoincludepartsofthebackgroundtothebiomorphs,whereasothertypestendedtodiscardsomepartsofthebiomorphstothebackground.A“minimum”thresholdwaschosen,whichwasconsistentamongallpicturesandcorrectlyseparatedthebiomorphsfromtheirbackground.Theparti-cleswerethendetectedandanalyzedusingImageJparticleanalyzer,giving informationabout theareaor the shapeof theparticles.Alldataobtainedthisway,ineverypicture,foreveryinitialconditionandtimestep,werestored inafive-dimensionmatrix (usingMatlab ver-sion R2009b—MATLAB and Statistics Toolbox Release 2012b; TheMathWorks,Inc.,Natick,MA,USA)forfurtheruse.

Afterthefirstresultshadbeenobtained,specificconditionsofsynthesis representing different regimes were studied again withthesameprotocolin10mlpetridishes.Atdifferenttimeintervals,a fraction of the structures observed floating at the surface was re-trieved,rinsed(successivelyin0.01mol/LNaOHandmilli-Qwater),placed on conductive tape, and gold-coated to be observedwitha Scanning ElectronMicroscope (Auriga FEG, IPGP). Imagesweretakeninsecondaryelectronmode,eitherwithachamberdetectororwithaninlensdetector.Theworkingdistancewas7mm,exceptforafewpicturestakenwith10or18mm,andthevoltagewasbetween3and10kV.

2.4 | Size distribution studies

Pictures of biomorphs grown in gelwere acquired 2weeks aftertheonsetoftheexperimentusingadirectpetrographicmicroscope(LeicaDM2500 P, IPGP) in transmissionmodewith a 4× objec-tive.Foreveryfieldofview,~30picturesweretakenatregularlyincreasingfocaldepthsthroughthegel.Thesepictureswereusedforanextendeddepth-of-fieldreconstructionusingiMageJ(EDFpl-ugin,EPFL),whichbroughtallstructuresintofocusandimprovedulterior picture treatment. For comparison with biological data,a culture of cyanobacteria(Synechocystis sp.) from the PasteurCyanobacteria Collection (PCC6803) was grown with standardBG-11mediatostationarygrowthphase.Itwassubsequentlyim-agedwithanOlympusBX-51microscopeunder100× (oil) objec-tive with contrast enhanced using differential interference contrast (DIC)optics.Forstudiesonsizes,picturesweretreatedinthesameway as presented before with iMageJ.When particles tended toconnect, theywereseparatedautomaticallywithawatershedal-gorithm (Beucher & Lantuejoul, 1979) which detects boundariesbetween areas in grayscale images. The size is given as an equiva-lent diameter:

(2)Equivalent diameter=2

√

(

Area

π

)

     |  283ROUILLARD et AL.

To quantitatively compare size distributions, their widths arecharacterizedusing(average/standarddeviation)ratios.Theseratioswerecomputedforbiomorphsinourexperiments,abiologicaltestsample(apictureofacultureofSynechocystisspp.),andsizedistribu-tiondatafoundinArcheanmicropaleontologyliterature(Butterfield&Chandler,1992;Wacey,Kilburn,Saunders,Cliff,&Brasier,2011,SupportingInformation).

3  | RESULTS

3.1 | Biomorph nucleation

3.1.1 | Numerical simulations

Using numerical simulations described in paragraph 3.1, for eachstartingsolution,thesaturationindexσ of witherite was calculated.

The resultsareshown inFigure1a.Each initial condition is repre-sentedinasinglecellasafunctionofbothinitial[Ba]andinitialpHshown,respectively,ontheverticalandthehorizontalaxis.Asthe[CO3

2−]increaseswiththepH,theaxesofthisfigurecanalsobereadas initial [Ba]andinitial [CO3

2−].Therefore,thesaturation indexofwitherite,representedbyacolorchart,increasesfromlefttorightandfromtoptobottom.

3.1.2 | Solution experiments

Experimentsofnucleation insolutionwereperformedforeachofthe individual starting conditions shown inFigure1.Within a fewminutes to a few hours (depending on the initial conditions—seeTable1)afterreagentshadbeenmixed,biomorphswereobservedatthewater–airinterface.InaccordancewithEquation1,Figure1aand Table1, particles tend to appear earlier at higher initial [Ba]

F IGURE  1  (a)Colormapshowingthesaturationindexofwitheriteforthedifferentinitialsettingsofoursolutionexperiments.(b)EvolutionofthepHinthesynthesismediumduringthereactionfordifferentinitialpHvalues.(c)Relativeareacoveredbybiomorphsinpicturesafter26hrdependingontheinitialpHandfordifferentinitial[Ba][Colourfigurecanbeviewedatwileyonlinelibrary.com]

284  |     ROUILLARD et AL.

and higher initial pH (higher saturation indexes). The precipitatesalsodevelopedontheplasticwalls,atdepthsincreasingwithtime,and eventually reached the bottom of the wells. This is due to the progressive diffusion of atmospheric CO2(g) from top to bottom,causinganincreaseinthesaturationindexofwitherite.Duetothecontinuous diffusion of CO2(g) into the reactionmedium, the pHvaluedrops consistently after the start of the experiment, under-goinganexponentialdecreaseandreachingneutralvaluesinafewtensofhours(Figure1b).Ourresultsalsoillustratethatowingtotheincreased rate of CO2uptakeathigherpHvalues,thepHdecreasesfaster for solutions starting atmore alkaline conditions, anobser-vationalreadymadebyEiblmeier,Kellermeier,Rengstl,García-Ruiz,andKunz(2013).Thisresultsinequalequilibrationtimesforallthesolutions(around45hr).AlthoughthegeneralexponentialdecreaseofthepHinourexperimentsisinaccordancewithpreviousresults(Eiblmeieretal.,2013;Kellermeier,Melero-Garcíaetal.,2012),themeasured rate is two to three times higher. This can be ascribed to differences between the experimental protocols. Our experimen-talprotocolforpHmeasurementinvolvedfrequentopeningofthewells,thusacceleratingtherateofCO2uptake.Attheendoftheex-periments,theprecipitatecoverageatthewater–airinterface(givenin relative area,%), as a function of initial pH and starting [Ba] isshownonFigure1c.Overall, the area coveredby theprecipitatesincreaseswithincreasinginitialpH,whiletheinfluenceoftheinitial[Ba]islessconclusive.AsthepHvalueiscontrollingthecarbonateconcentrationinsolution,theextentofbiomorphgrowthismostlycontrolled by carbonate availability and not by barium availability.

3.1.3 | Gel experiments

Within a carbonate-rich silica gel, the precipitation of biomorphicwitherite is caused by the diffusion of barium through the medium.

Thebariumconcentration,andthereforealsothesaturation indexσ ofwitherite, decreases spatially away from thediffusion source(Melero-García etal., 2009). Pictures taken at equal times and atdifferent distances from the initial diffusion source show that the

TABLE  1 Tablesummarizingthetimeofappearanceofbiomorphs(measuredbyopticalmicroscopy)fordifferentinitialpH(columns)and[Ba](lines).Thedifferenttimeintervalsareasfollows:0.5:biomorphshaveappearedinopticalmicroscopybefore0.5hr,1.5:biomorphshaveappearedbetween0.5and1.5hr,2:betw.1.5and2hr,3:betw.2and3hr,4:betw.3and4hr,5:betw.4and5hr,6:betw.5and6hr,8:betw.6and8hr,23:betw.8and23hr,Nf:notfoundafter23hr

9.33 9.43 9.53 9.63 9.73 9.83 9.93 10.03 10.13 10.23

1.25mmol/L Nf Nf Nf Nf Nf Nf Nf 23 23 23

2.5mmol/L Nf Nf Nf Nf Nf 6 6 8 8 23

5mmol/L Nf Nf 6 4 4 2 3 5 4 4

10 mmol/L Nf Nf 4 2 2 2 2 2 3 3

20 mmol/L Nf Nf 3 1.5 1.5 1.5 2 1.5 2 0.5

40mmol/L Nf Nf 1.5 1.5 1.5 0.5 0.5 1.5 1.5 1.5

10.33 10.43 10.53 10.63 10.73 10.83 10.93 11.03 11.13 11.23 11.33

1.25mmol/L 23 23 23 3 3 3 3 3 0.5 0.5 0.5

2.5mmol/L 4 4 3 1.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5

5mmol/L 3 3 3 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5

10 mmol/L 3 3 3 3 1.5 0.5 0.5 0.5 0.5 0.5 0.5

20 mmol/L 1.5 3 3 2 2 0.5 0.5 0.5 0.5 0.5 0.5

40mmol/L 3 2 3 1.5 1.5 0.5 0.5 0.5 0.5 0.5

FIGURE  2 Crystallizationgradientofbiomorphsinagel.S1,S2,andS3:pictures(opticalmicroscopy)takenatdifferentdistancesfromthediffusionsource(L1:3mm,L2:8mm,L3:12mm).Scalebars:600 μm[Colourfigurecanbeviewedatwileyonlinelibrary.com]

     |  285ROUILLARD et AL.

spatialconcentrationofbiomorphsdecreaseswiththedistancetothediffusionsource(Figure2),whichisinaccordancewiththisde-crease of σawayfromthesource,andwithobservationinpreviousstudies(Melero-Garcíaetal.,2009).Whileinsolutionexperiments,the biomorphs tend to grow on interfaces (air–water interface,solid–water interface), they form throughout the gel under condi-tionswhichvarywithtimeandspace,thusdisplayingacontinuousvariationinmorphologies.

3.2 | Biomorph morphology

3.2.1 | Biomorphs in solution experiments

Awidediversityofcarbonate-silicabiomorphshapeswereobservedinthesolutionexperiments(Figures3–5).Theseshapescanbeclas-sifiedintoseveralmainmorphologicaltypes,followingthemorpho-geneticmodelpresentedinearlierstudies(García-Ruizetal.,2009;seetheIntroduction).

The first growth regime represents fractal growth. The varia-tioninthenumberofnon-crystallographicbranchingeventsalongacertain lengthofcrystal,orbranchingdensity,appearstobere-sponsibleforthevarietyoffractalgrowthshapesobserved.Atlow

branchingdensity,theindividualbranchesarephysicallyseparatedfromeachotherwithintheoverallstructure,leadingtodendrite-likeshapes(Figure3a,b).Athigherbranchingdensity,spaceconstraintsforcethebranchestodecreaseindiameterateverysplittingevent(Figure3c), and finally connect to form a smooth overall fern-likedendritic structure (Figure3d). Eventually, this splitting leads to acontinuousgrowthfrontofwitheritecrystals,leadingtoframboidal(Figure3e–h)andultimatelysphericalshapes(Figure3i).

Thesecondgrowthregimerepresentscurvilineargrowth.Oneof theprominentaspectsof this regime is thepotential curlingofsheet-likematerialonitsouteredge(García-Ruizetal.,2009).Manyoftheshapesformedduringthisregimecanbeunderstoodintermsofgrowthfrontspeedandcurlingspeed,leadingtosimplecircularsheets (no curl, growthoccurs at a similar rate in everydirection,Figure4a),leaf-shapedbiomorphs(duetoanencounteroftwocurlsofoppositehandedness, Figure4b,c), andhelicoids (due to anen-counteroftwocurlsofthesamehandedness,Figure4d,e).Ifthetwocurlshavethesamepropagationvelocityastheradialvelocityofthesheet,aperfecthelicoidisformed,butifthecurlingvelocityexceedstheradialvelocity,thetworimscoiloneachother,creatingabraid(Figure4d).If,inaddition,oneofthecurlspropagatesfasterthantheother,awormlikestructureiscreated(Figure4e).

F IGURE  3 Maintypesofbiomorphsobtainedthroughfractalgrowth.(a),(b),(c)Dendrites.(d)Fern-likebiomorph.(e)to(h)Framboidal-typebiomorphs.(i)Spheroidalbiomorph.PicturesaretakenusingScanningElectronMicroscopy(secondaryelectronmode).Scalebars:(a),(b)10μm;(c),(e),(f),(h),(i)5μm;(d),(g)20μm

286  |     ROUILLARD et AL.

The transition from fractal growth to curvilinear growth can be easily observed in some structures,with a core of fractal growthaggregates, surrounded by radially expanding sheets which wereformedbycurvilineargrowth(Figure5a,b).Insteadoftwisting,thesheets can also fold onto themselves, forming gutter shapes (5),moth-likeshapes(5d),snail-likeshapes(5e),orcombinationsofthese(5f,g).Inothercases,ring-likestructuresarisethroughthecurvilin-earregimeandgrowwithdiversethicknessandtodiverseheightstoformdonut(5h),coral,orurn-likeshapes(5i).Lastly,mushroom-likeshapes (Figure5j,k) canemerge from the fractal aggregates.Theirformationprocessisnotwellunderstoodyet.

Lookingat a singleexperiment, at a given time,onlyone typeofgrowthprocessoccurs(fractalorcurvilinear).However,themor-phologiesofindividualbiomorphsvary.Forexample,circularsheetscangrownexttoleaf-shapedbiomorphs,orspheroidscangrownextto framboids.

3.2.2 | Biomorphs in silica gel experiments

Thegrowthofwitheriteingelsoccursbyacounter-diffusionprocess(Melero-Garcíaetal.,2009).A firstprecipitateofmetal carbonateiscreatedatthegel–solutioninterfaceatapHcloseto8.5duetomixingoftheacidicmetalsolutionandthealkalinegel.Astheacidicmetalsolutiondiffusesacrossthegel,itbecomeslessconcentrated.Therefore,withinasinglegelexperiment, there isagradualvaria-tioninconditionsacrossthegel,fromrelativelyacidicandmetal-richat the interface tomorealkalineandcarbonate-richat theendofthegel.Asaconsequence, in thegelspreparedfor thisstudy, theshapeofbiomorphschangesalongthediffusiondirection(Figure2).

Close to the diffusion source, the low pH values lead to the ap-pearance of biomorphs related to the fractal growth regime. ThiswasobservedbeforeinexperimentsofMelero-Garcíaetal.(2009).Furtherawayfromthediffusionsource,pHvaluesrangeupto10.5,andbiomorphsdisplaysheets,trumpet-shapesandhelicoidscharac-teristicofthecurvilineargrowthregime,withmorenumerousandlonger helicoids observable at increasing distances from the diffu-sionsource(seealsofigure9inMelero-Garcíaetal.,2009).

3.3 | Influence of pH and [Ba] on biomorph growth in solution experiments

Tostudytheeffectofsolutionpropertiesonthemorphologicalvari-ationinbiomorphs,asetofexperimentswasconductedinwhichthe[Ba]wasvariedfrom0.5to40mmol/LandthestartingpHwasvar-iedfrom9.3to11.3.Theexperimentsdemonstratedastrongcontroloftheseparametersonthemorphologiesobtained.InFigure6,theconditionsfornormalcrystallographicgrowthofwitheritecrystals(black), fractal growth regime of biomorphs (blue), and curvilineargrowth regimeof biomorphs (green) are shown. In case of a timeevolution,anoverlayofasmallerrectangleindicatesthesecondaryregime.At [Ba]=0.625mmol/L, at any startingpH,nobiomorphsareformed.Instead,needleorprism-likepseudohexagonalwitheritecrystalsdevelopslowlyatthewater–airinterface.At[Ba]valuesof1.25mmol/Landabove,biomorphsareseentodevelop,butonlyinthehighpHrange.InthishighpHrange,[CO3

2−]issufficienttocausethedevelopmentofsilica-witheritebiomorphs.When[Ba]risesfrom1.25 to 5mmol/L, the minimum pH at which biomorphic growthoccurs is shifted to lower values (from10 to 9.5), thus expanding

F IGURE  4 ExamplesofbiomorphtypesobtainedthroughcurvilineargrowthandexplainedbythemorphogeneticmodelproposedinGarcía-Ruizetal.,2009.(a)Discoidalsheet,withcharacteristicwave-likepatterns.(b,c)Leaf-likebiomorphs.(d)Helicalbraid.(e)Wormlikebraids.PicturesaretakenusingScanningElectronMicroscopy(secondaryelectronmodeexceptEtakeninbackscatteredmode).Notethegrowthoflargewitheritecrystalsonthesurfaceofsomebiomorphs,relatedtopost-removaldrying.Scalebars:(a)10μm;(b),(c),(d),(e)20 μm

     |  287ROUILLARD et AL.

therangeofpHconditionsforwhichthesynthesisofbiomorphsispossible in theseexperiments. Inmanybiomorphsynthesisproto-cols,a[Ba]of5mmol/Lisused(Kellermeier,Glaab,Melero-García,&García-Ruiz,2013;Voinescuetal.,2007).Wethereforefirstde-scribe the biomorphs that form at [Ba]=5mmol/L and variablestartingpHvalues.

At[Ba]=5mmol/LandstartingpH=9.5,fractalgrowth-relatedbiomorphs (dendritic, fern-like, and framboidal shapes) grow forabout1day in themedium, reaching sizesup to severalhundredsof μm (Figure6). When the starting pH is increased, betweenpH=10.0–10.7, this fractalgrowthshiftsveryquickly (aftera few

tens ofminutes) to a curvilinear growth regime. Intricate agglom-eratescreatedby involutesheetsappear.Aftera fewhours,opensheets emerge from these aggregates, leading to diverse leaf-likeandhelical structures. InFigure7, thedetailedpH-[Ba] conditionsareshownatwhichthedifferentcurvilinearstructuresareformed,includinginvolutesheets(green),leaf-likesheets(orange),coral-likestructures(gray),andhelicalstructures(brown).Incaseofasucces-sionofmorphologies,anoverlayofasmallerrectangleindicatesthelaterformedphase.AsisseeninFigure6,atpHabove10.7,thesamefractalgrowth-relatedstructuresappearasthosethatformedatpH9.5–10.0.However,theyappearmuchfaster(withinafewminutes

F IGURE  5  (a)and(b)transitionbetweenfractalgrowthandcurvilineargrowth.Thesquaresnoted1and2indicaterespectivelyzonesoffractalgrowthandofcurvilineargrowthinthesamestructure.(c–k)Othertypesofbiomorphsobtainedthroughcurvilineargrowth.(c–g)biomorphsobtainedthroughtheinvolutionoffoldingsheets.(c)Gutter-likebiomorph.(d)Moth-likebiomorph.(e)Snail-likebiomorphs.(f,g)involutedaggregates.(h,i)biomorphsobtainedthroughthedevelopmentofclosedsheets.(h)donut-shapedbiomorph.(i)Coralorurn-shapedbiomorphs.(j,k)mushroom-likebiomorphs.PicturesaretakenusingScanningElectronMicroscopy(secondaryelectronmode).Scalebars:(a),(d),(e),(h),(j),(k)10μm;(b),(c)5μm;(f),(g),(i)20μm[Colourfigurecanbeviewedatwileyonlinelibrary.com]

288  |     ROUILLARD et AL.

after mixing of the reagents), and are much more numerous andsmaller,around10–15μminsize.This is inaccordancewithprevi-ousresultspresentedby(Eiblmeieretal.,2013).Afterafewhours,attheseconditions,newbiomorphsappear,withamorphogeneticpathverysimilartowhatwasdescribedpreviouslyforpH10–10.7.Basedontheseobserveddifferences,wecandefinethreegeneraldomainsforbiomorphgrowth,separatedbyredlinesandnumberedasfollowsintheFigure6:1)alowpHregime,definedbyslowfractalgrowth of a small number of large biomorphs, 2)An intermediatepH regime, defined by fast fractal growth, followed by curvilin-eargrowthof involutecomplexaggregates,flatsheets,andhelicalstructures,3)AhighpHregime,definedbyshortfractalgrowthofa largenumberofsmallbiomorphs.Althoughall themorphologiescouldbeinterpretedasbeingapartofthesametheoreticalmorpho-logicalcontinuum(Melero-Garcíaetal.,2009),seealsosection4.3),thiscompletecontinuumisnotobservedunderallexperimentalcon-ditions.Indomain1,thedevelopmentofbiomorphsstopsatfractalgrowthandisnotfollowedbyacurvilineargrowth.Indomain2,thefractalgrowthisveryshort,andoftennotvisibleinthefinalstruc-turesbecausecoveredbythesubsequentfoldingofsheets.ThepHrangeofthesedomainswasmodifiedwhenthe[Ba]waschanged.Anincreasein[Ba]expandsthepHrangeofdomain2(seeFigure6).TherangeinpHoverwhichhelicalstructuresaresynthesizedissig-nificantlyincreasedaswellwhen[Ba]increasesfrom5to40mmol/L(see thegrayarea inFigure7).Theyare found forpH=10.1–11.3

(upper limit of our study) for [Ba] above 20mmol/L. Overall, thehighertheconcentrationofbarium,thelargertheextensionofthecurvilinear growth regime.

After a few days, in most experiments, prism- and needle-likewitheritecrystalsstartcoveringthesurfacesofpreviouslyformedbio-morphshapes(e.g.,Figure5d;Kellermeier,Melero-Garcíaetal.,2012).

3.4 | Size distribution of biomorphs

Inthecaseofsolutiongrowth,duetothephysicaleffectoftheme-niscusinthewell,biomorphstendtoaggregateinthecenterofthewell,making the accurate size distribution analysis a difficult task.As a consequence, the size distribution analyses were realized ingelgrowthexperiments.Intheseexperiments,theslowdiffusionofthebariumsolution leads to the formationof largerbiomorphs (upto3–4mm) than theonesobtained in solution (up to~200μm for the same curvilinear growth regime).As discussedpreviously (sec-tion 4.1), the pH, [Ba2+] and nucleation index change both in timeand along the direction of diffusion in the silica gel. A wide range of growthconditionsthusexistsinasinglegel,enablingtheformationofdifferentmorphologies(section4.2).Moreover,onamm-scale,thenumberofbiomorphsdecreaseswiththedistancefromthediffusionpoint,whereastheirindividualsizeincreases.ThisspatialgradientisillustratedinFigure2(S1andS3areseparatedby9mm).Atagivendistancefromthediffusionpoint,thenucleationrateisassumedtobe

F IGURE  6 Qualitativemorphogramrepresentingthedominantgrowthregimes observed for each unique initial settingofpHandBaconcentration.Theblackshaderepresentstheoccurrenceofnon-biomorphicwitheritecrystals.Theblueshaderepresentsthefractalgrowthregime.Thegreenshaderepresentsthecurvilineargrowthregime.Smallerrectangles indicate the transition to the other regime during time. The three domainsoutlinedinredcorrespondtothethreedomainsdefinedinthetext.Scalebars:20μm[Colourfigurecanbeviewedatwileyonlinelibrary.com]

     |  289ROUILLARD et AL.

constant.Therefore,thesizedistributionofbiomorphswasmeasuredinagroupofpicturesalignedalongaplaneparalleltothediffusionfrontandseparatedfromthediffusionsourceby3mm.Inthisarea,thegrowthconditionsareassumed tobesimilar,meaning that thedifferencesinsaturationindex,pHorBa/CO3 are not large enough toaffectthegrowth.Theobtainedsizedistribution(Figure8a)isuni-modal,whichisexpectedforauniformgrowthprocess,andisclosetoanormaldistribution,withanaveragesizeof21μm and a stand-ard deviation of 8 μm.Theaverage,standarddeviation,andaverage/standard deviation ratios of this size distribution are summarized in theSupportingInformation(Table1)andcomparedtothesameval-uesinothersizedistributions(seesection3.4).

4  | DISCUSSION

4.1 | Influence of [CO32−] diffusion on biomorph

growth in solution experiments

Duringtheexperiments,pHmeasurementsofthebulksolutionshowthat CO2diffusionfromtheatmosphereprogressivelyacidifiesthegrowthmedium(Figure1b).Wehypothesizethatduringthisgradualchange,thesolutioncrossesthepreviouslydefinedpH-dependentcrystallizationdomains,andgeneratesthecorrespondingbiomorph

shapes. It is thusexpectedthatbiomorphscorrespondingtomoreacidic domains will start growing later than those correspondingto more alkaline domains. This coexistence of different domain-corresponding biomorphs is indeed observed in our experiments(Figure9a,b). For the lowest initial pHvaluesof thedomain2, in-volutedsheetsappearfirst(arrowsnoted1inFigure9),associatedafter3–4hrtoflatsheets (arrownoted2 inFigure9).After1day,large framboidal biomorphs (arrownoted3), corresponding to thedomain1,appearinthemedium.Itthereforeappearsthatthesyn-thesismediumhasreachedpHvaluescharacteristicofdomain1,andthegrowthregimehasswitchedfromacurvilineartoafractal-typegrowthprocess.Similarly, inthewholerangeofdomain3, the ini-tialdevelopmentofsmallframboidalbiomorphs isfollowedbythegrowth of involuted aggregates and other curvilinear structures cor-respondingtodomain2.ThesynthesismediumhasreachedpHval-uescharacteristicofthedomain2.ThefinalpHafterequilibrationwiththeatmosphereisthesameforallthesolutions(approximatelyatpH=7—seeFigure1b).Therefore,solutionsstartingindomain3(thehighestpHrelateddomain)willcrosstheconditionsrelatedtodomains2and1duringtheirevolution.However,thecoexistenceofmorphologiesseemslimited.Nomorphologiescorrespondingtodo-main 1 can be observed in the solutions starting in the domain 3 or inthehigherpHpartofthedomain2.Aplausibleexplanationisthat

F IGURE  7 QualitativemorphogramrepresentingthemainmorphologiesappearingthroughcurvilineargrowthforeachuniqueinitialsettingofpHandBaconcentration.Thegreenshaderepresentsaggregatesofinvolutedsheets.Theorangeshaderepresentsflatsheetsandleaf-likebiomorphs.Thebrownshaderepresentsthehelicalbiomorphs(helicoids,braids,andwormlikebraids).Thegrayshaderepresentscoral-likebiomorphs.Thesmallertherectangle,thelaterthecorrespondingmorphologyappears.Scalebars:(I),(II)10μm;(III),(IV)20μm [Colourfigurecanbeviewedatwileyonlinelibrary.com]

290  |     ROUILLARD et AL.

whenthesolutionsreachtheconditionsofdomain1,thesaturationindexofwitheriteisnotsufficientanymoretoallowfurthernuclea-tionorgrowthofbiomorphs,andthemorphologiescorrespondingtothedomain1willnotbeformed.Aswasdiscussedinsection3.1,therateofpHdecreaseishigherforhigherinitialpH.Forthewholerange in initialpH’sofdomain3, theconditionsrelatedtodomain2will be reachedwhile the saturation indexofwitherite is abovethelimittonucleatebiomorphscorrespondingtodomain2.Forso-lutions starting in domain 2, however, the pHwill decreasemuchslower,andthesolutionsstartingathighpHofthisdomainwillonlyreachtheconditionsofdomain1whenthesaturationindexisunderthelimitallowinggrowthofdomain1-relatedbiomorphs.Theshapesoftheobservedbiomorphs,theirevolutionthroughoutthedurationoftheexperiment,andtheirassociationwithotherbiomorphshapescan all be related to crossing by the solution of various growth-controllingpHand[Ba]domains.

4.2 | Morphology as a criterion to detect biomorphs in the fossil record

4.2.1 | Morphology of solution- grown biomorphs

Manyofthediversebiomorphshapesthatweobtainedinourso-lutionexperimentsresembleshapesofmodernmicro-organisms.

The fractal growth process gives rise to biomorphs resemblingcommonmorphogroupsofbacteria.Purelyspheroidalbiomorphs(Figure10a)canbecomparedtotheshapesofmonococcibacte-ria,whereas framboidal biomorphs (Figure10c) resemble for in-stance bacterial clusters obtained through the random divisional patternofstreptococci(e.g.,Microcystis flosaquae inFigure10d).Whentheinternalgeometryofbiomorphsisaccessible,theymaybedistinguishedbytheir internalcontinuity,whileindividualmi-crobial cells in clusters are separated bymembranes andwalls.IntheArcheanrockrecord,spheroidalandclusteredmicrostruc-tures have been identified in the Strelley Pool formation, theFarrellQuartzite, and theMoodiesGroup (Figure10b,e; Javaux,Marshall, & Bekker, 2010; Sugitani, Grey, Nagaoka, Mimura, &Walter,2009;Sugitanietal.,2010).Thecurvilineargrowthpro-cess generates biomorphs with helicoidal structures such asbraided (Figure10h)orcylindrical,segmented (wormlike)shapes(Figure10f),which can be compared to, for example, the stalksproduced byMariprofundus ferroxydans (Figure10i, Singer etal.,2011), or awide range of filamentous bacteria (e.g.,Oscillatoria princeps). Inthefirstcase,there isanobviousdifference insize.Inthelattercase,biomorphscanusuallybedistinguishedbytheirshorter length and coiled internal geometry, although cyano-bacteria such as Spirulinasp.displaythesameinternalgeometry(Figure10g).Helicoidal,andmorepreciselywormlikebiomorphs,

F IGURE  8 Sizedistributionofgel-grownbiomorphs(a)at3mmfromthediffusionsource(see4.6)comparedtothesizedistributionofSynechocystissp.(b)[Colourfigurecanbeviewedatwileyonlinelibrary.com]

F IGURE  9  (a,b)Opticalmicrographs(transmissionmode,scalebars200μm)oftheparticlesobservedatthesurfaceofthesolutionforaninitialBaconcentrationof1.25mmol/LandaninitialpHof10.64.(a)picturetakenafter4hr.(b)picturetakenafter26hr.Arrow1:Involutedsheets.Arrow2:Flatsheets.Arrow3:Framboidalbiomorphs[Colourfigurecanbeviewedatwileyonlinelibrary.com]

     |  291ROUILLARD et AL.

havebeencomparedbeforetothecontroversialApexChertmi-crostructures(Garcia-Ruizetal.,2003;Figure10j).ItisimportanttonotethattheflangedmicrostructuresfoundintheStrelleyPoolFormationandtheFarrellQuartzite(Sugitani,Mimura,Nagaoka,Lepot, & Takeuchi, 2013; Sugitani etal., 2009, 2010) have nomorphological equivalents among the biomorphs grown in thisstudy. This confirms the often discussed difficulty of distinguish-ing actual microbial remnants from abiotic counterparts basedonmorphologicalcriteriaalone(Brasieretal.,2002;Buick,1990;Garcia-Ruizetal.,2003;Schopfetal.,2002),especiallyinancientmetamorphosedrocks.

4.2.2 | Variation of morphologies in populations of solution- grown biomorphs

Incontrasttoindividualmorphologies,thevariationinshapewithinapopulationofmicrostructurescouldpotentiallyserveasacriterion

todiscriminatebetweenanabioticandbiologicorigin.Asreportedinourresults (endofsection4.2.1)andas isseeninFigure9a,fora single constant synthesis condition, biomorphs with a range ofsmoothlyvaryingmorphologiescanbeobserved.Forexample,atagiventimeinasinglestartingsolution,biomorphsthatwereformedthrough fractal growth can range from dendritic shapes tomulti-globular, biglobular, or even spheroidal shapes. This demonstratestheinfluenceof localheterogeneities(suchassupersaturationval-uesorimpurities)insolutions.Micro-organismsfromasinglestrain,however,exhibitverysimilarshapes,andamicrobialassemblageofseveralstrainsdisplaysadiscontinuousspectrumofmorphologies.ThiscriterioncanbeappliedtoArcheanmicrofossils, forexample,carbonaceousmicrostructuresdescribedinthe3,4GaStrelleyPoolChert(Sugitanietal.,2013).Thestructuresdescribedthereinhavespheroidal shapes and sizes (~10μm) similar to some biomorphs.However, their shape is also remarkably homogeneous across thepopulation.Basedonourexperimentalresults,itcanbeconcluded

F IGURE  10 Morphologicalcomparisonofspecificbiomorphshapes,bacteria(orbacteria-relatedstructures)andArcheanmicrostructures.(a)spheroidalbiomorph,(b)populationofspheroidsfromtheStrelleyPoolFormation,(c)framboidalbiomorphs,(d)clusterofMicrocystis flosaquae,(e)clusterofspheroidsfromtheStrelleyPoolformation,(f)wormlikebiomorphs,(g)Spirulinasp.(h)helicalbraid-likebiomorph,(i)helicalstalksproducedbyMariprofundus ferrooxydans,(j)filamentousstructuresfromApexChert.Sourcesofthepictures:(a),(c),(f),(h)thisstudy.(b),(e)Sugitanietal.(2015).(d),(g)PicturesbyPr.Tsukii,ProtistInformationServer.(i)Singeretal.(2011).(j)Waceyetal.(2015).Scalebars:(a),(i)1μm;(b),(d),(e),(f),(g),(h)50μm;(c),(j)10μm[Colourfigurecanbeviewedatwileyonlinelibrary.com]

292  |     ROUILLARD et AL.

thatitishighlyunlikelythatfractalbiomorphsgrownfromsolutioncouldhaveproducedsuchahomogeneouspopulationofspheroids,whichdisprovesthebiomorphhypothesisinthiscase.

4.3 | Size as a criterion to detect biomorphs in the fossil record

4.3.1 | Size of solution- grown biomorphs

The size range (from a fewμm to a few tens of μm) displayed bybiomorphsgrowninsolutionsintheupper-pHdomains,noted2and3inFigure6,lieswithintherangeofprokaryoticsizes.AtlowerpHvalues (domain noted 1 in Figure6), fractal-type biomorphs oftengrowtoseveralhundredsofmicrometersindiameter,whichisoneortwoordersofmagnitudelargerthanindividualmodernprokary-otes. Only very few strains of modern bacteria reach these sizes. However,manybacteria tend toassociate incolonies, forming forexampleclustersorfilaments,withsizescomparabletothelargestbiomorphsobservedindomain1.Overall,theabsolutesizerangesofbiomorphsandbacteria(individualorcolonial)aresimilar.Therefore,it is concluded that the absolute size of individual microstructures doesnotrepresentarobustcriterionforbiogenicityofmicrostruc-turesintheArcheanrockrecord.

4.3.2 | Distribution of sizes in populations of gel- grown biomorphs

In contrast to sizes of individual biomorphs ormicro-organisms,the frequency distribution of sizes of biomorph/micro-organisminasample,athinsectionorapicture,isapowerfuldescriptorofan assemblage. As automated sizemeasurement protocols havebeendeveloped(usingElasticLightScatteringorEpifluorescenceMicroscopy data), the size distribution of single strain unicellu-lar bacterial populations could be studiedmore effectively thanbefore (Harvey & Marr, 1966; Katz etal., 2003; Uysal, 2001).According to these studies, bacteria as diverse as Escherichia coli,Synechococcus spp.,Pseudomonas aeruginosa,Staphylococcus aureus or Bacillus subtilis all display unimodal, slightly positivelyskewed distributions. Using pictures of Synechocystis spp. (PCC6803) cultures taken in opticalmicroscopy (bright field orDIC),we also found this kind of distribution (Figure8b). This is in ac-cordancewithearlymathematicalmodelswhichtookintoaccountabinarydivision,aGaussiandistributionofsizesofcellsenteringdivision, aGaussiandistributionof sizesofdaughter-cells, andadependencyof thegrowthrateonthecellvolume (Koch,1966).Therefore,mostmonospecificunicellularpopulationscanbeex-pectedtodisplaythiskindofdistribution.Naturalpopulationsareusually composed of several strains, and their size distributionsarethusexpectedtobeplurimodal.However,inthePrecambrianmicropaleontological record, unimodal distributions are also re-ported(Butterfield&Chandler,1992;Sugitanietal.,2013;Waceyetal.,2011).Thiscanbeascribedtothreepotentialfactors:(i)mi-cropaleontologists often make size measurements on groups of

microstructuresthatwereseparatedbefore,basedontheirmor-phology;(ii)somegroupsoforganismscanbeselectivelylostdur-ingdiagenesisbecausetheyaremorefragilethanothers,reducingtheresultingdiversityinthefossilassemblage;(iii)theplurimodalnature of a distribution can be difficult to detect when the size of thepopulationissmallandthedistributionofsinglestrainsover-lapeachother.

Thewidthofsizedistributionsisquantitativelycompared(seesection 3.4) using the Average/Standard deviation ratios (here-after noted A/SD). This ratio was found to be lower for silica-witherite biomorphs and for abiotic spherulites from the Gwnagroup (Waceyetal.,2011) than forbacteria (0.8 for thespheru-lites, between 1.7 and 2.7 for biomorphs, 4.9 for bacteria). InSchopf, Kudryavtsev, Sugitani, andWalter (2010) and inWaceyetal. (2011), the authors already mentioned the large width ofthesizedistributionofpseudofossils(abioticspherulitesfromtheGwnagroup,butalsohematitepseudofossils fromtheLakhandaformationandkerogenpseudofossilsfromtheMarra–Mambafor-mation).Thewidthofthesizedistributions indicatesdifferencesin the genetic process.While daughter-cells arise through a celldivisionprocessfromamother-cell,biomorphs,singlecrystalsandcrystalaggregatesgrowthroughanaccretionaryprocess.SeveralofthePrecambrianmicrofossilpopulationsstudiedhere(TableS1;Butterfield & Chandler, 1992; Wacey etal., 2011) display highA/SDvaluessimilartothevalueoftheSynechocystispopulation(between 4 and 5), which indicates that these populations mayrepresent single strain microbial cells. In contrast, a populationof spheroidal cells from the Gunflint formation (Wacey etal.,2011)andapopulationofspheroidsfromAguBay(Butterfield&Chandler,1992)displaylowerA/SDvalues(between2and3).Astheirbiogenicity isnotquestioned, thesewidedistributionsmaybeinterpretedascrypticplurimodaldistributions,producedbytheassemblageofseveralstrainsofbacteria.Alternatively,thesewidedistributionscouldbeexplainedbydegradationalgradients(Knoll&Golubic,1979).

4.4 | Spatial gradients displayed by gel- grown biomorphs

In silica gel experiments, the diffusion process results in a con-tinuouschange in size, indensityandmorphologyofbiomorphs(seepicture1andsection3.4)observableonashortspatialscale,a typeofspatialgradientwhich isnotobserved inbiologicalas-semblages. In thegeological record, thepresenceof thiskindofgradientwithinchertswouldprovideastrong indication that: (i)thestructuresareabiotic,and(ii)thattheprecursorofthedeposithadaviscosityhighenoughtoallowslowdiffusionprocesses.

4.5 | Biomorph growth conditions and Archean paleoenvironments

Inthecurrentstudywehavechosentoperformbiomorphsynthe-sisbyclassicalatmosphericCO2diffusionintoanalkaline,silica-rich,

     |  293ROUILLARD et AL.

andBa-richsolution,andtostudythefullrangeofmorphologiesthatareformedatvariablestartingpH.DuringtheArchean,thissetofconditionscouldhavebeenpresentinshallow–marinehydrothermalvent systemsor subaerial geothermal springs. Inourexperiments,fractalgrowthcanoccuroverawiderangeofconditions,atpHval-uesbetween9.5 and11.3 and [Ba] between1.25 and40mmol/L(seeFigure6),whilethecurvilineargrowthofhelicoidalbiomorphsis restrictedbetweenpH10.1and11.3,and [Ba]between10and40mmol/L.Thesebariumconcentrationsaremuchhigherthanthatof modern seawater ([Ba]=0.15μmol/kg, Herzig & Hannington,2006), and modern hydrothermal fluids ([Ba]=1.3–54μmol/kg,(Charlou etal., 2005; Ludwig, Kelley, Butterfield, Nelson, & Früh-Green,2006;VonDamm,1995).However,inArcheanhydrothermalfluidsthatleachedpredominantlyultramaficcrust,thebariumcon-centrationmayhavebeenhigher.Forinstance,komatiitesfromtheBarbertonGreenstoneBelt,SouthAfricacontain2–93ppmbarium(Parman, Shimizu,Grove,&Dann, 2003),while typicalMORBba-saltscontain0.5ppmbarium(Workman&Hart,2005).ThefactthatmanyArcheanhydrothermaldepositscontainbarite (Nijmanetal.,1998;Philippotetal.,2007;VanKranendonk,2006)suggests thatsignificant amounts of barium were available in these hydrothermal fluidsbutalsoshowsthat,atthetimethesedepositsformed,bariumsulfateratherthanbariumcarbonatewouldprecipitate.Thesolubil-ityproductofbariumsulfate(Ksp=1.0810

−10 at T=25°C)isslightlysmallerthanthatofbariumcarbonate(Ksp=2.5810

−9 at T=25°C;Lide,1947).Thisdifference insolubilityproductbetween the twophasesissosmallthattheformationofeitherbariteorwitheritewilllargelydependonthepartialpressureofCO2(intheocean,hydro-thermalfluids,orporefluids),andonpH,whichcontrolsthespecia-tion of carbonic acid. An overall high CO2pressureandlocalhighpHwillcreatehighersupersaturationvaluesforwitheritethanforbarite.Furthermore,underthereducingconditionsoftheHadeanandearlyArchean,inmanyenvironments,theconcentrationofdissolvedsul-fatewouldbemuchlowerthantoday,andwitheriteinsteadofbaritewould form when barium was available. Note that for strontium and calcium,theotheralkaline-earthelementsthatformbiomorphs,thesolubilityproductofstrontianite(SrCO3,Ksp=5.6010

−10at25°C)isalmostthreeordersofmagnitudelowerthanthatofcelestite(SrSO4,Ksp=3.4410

−7at25°C)whilethesolubilityproductofaragonite(theCaCO3phaseformingbiomorphs,Ksp=6.010

−9at25°C)isfiveorderofmagnitudelowerthantheoneofgypsum(Ksp=3.1410

−5at25°C;Lide, 1947). Therefore, in silica-rich alkaline solutions, strontiumshouldprecipitate in the formofstrontianite-silicabiomorphsandcalciumshouldprecipitateintheformofaragonite-silicabiomorphs.Our experiments allow for the first time to limit the geochemi-cal conditionsunderwhichcarbonate-silicabiomorphs couldhaveformed in surface environments of the early Earth.

5  | CONCLUSION

The experimental studyof the silica-witherite biomorph systematambientconditionsdemonstratestheextensiverangeofbiomimetic

morphologies that can be formed in this simple and abiogenic chemical system with only subtle changes in environmental condi-tions.Alltheobtainedmorphologiescanbeinterpretedasdifferentstepsofamorphogeneticcontinuumaccordingtoamodelpreviouslydescribed(García-Ruizetal.,2009).Thevaluesofbariumconcentra-tionandpH,which influences thespeciationofdissolvedCO2 and SiO2andthereforethemineralsaturationindexes,haveastrongef-fect on the finalmorphologies obtained. They control the numberandthesizeofbiomorphs,controlatransitionfromthefractaltothecurvilineargrowthregime,andaffectthetwistingduringcurvilineargrowth.

SeveralpaleoenvironmentalinferencespointtoplausiblescenariosfortheformationofbiomorphsduringtheArcheanEon,eitherinsolu-tionoringel.However,thisstudydemonstratesthatthehighbariumconcentrationsandhighpHvaluesnecessaryforbiomorphgrowthtooccurinsurfaceenvironments(influencedbyCO2diffusion),stronglyrestrict the number of settings where this growth could have occurred duringtheArchean.Itremainstobedeterminedwhatlimitsareimposedonbiomorphgrowthunderanextendedrangeofconditions,especiallywhenfluidsarewarmer(thesolubilityproductofwitheriteislowered,butthatofsilicaisincreased)orcontainmoreCO3

2−(thepHofthesys-temisbuffered,enablingmoretimeforslownucleationprocesses).

Becauseoftheircloseresemblancetobiologicalshapesandpro-cesses—forexample,celldivision—findingtherightcriteriatodistin-guishbiomorphsfromactualmicrofossilsiscrucialforlifedetectionprotocols.Basedonthisstudy,wecanlistthreecriteriatodiscardthebiomorphhypothesiswhenlookingforremnantsofprokaryoticlife:(i)Biomorphpopulationsdisplayawide,bell-likedistributionofsizeswhilesinglestrain,unicellularbacterialpopulationshaveanarrowersizedistributionandassemblagesof several strainsdisplayapluri-modal distribution, (ii) Biomorphs grown in gel-like environmentsdisplayashort-scalecontinuousvariationinsizeanddensity,thatisveryrarelyseeninbiologicalcommunities,and(iii)Biomorphsdisplayasmoothvariationofshapesunderconstant[Ba]-pHconditions.

ACKNOWLEDG MENTS

ThisprojecthasreceivedfundingfromtheEuropeanResearchCouncil(ERC)undertheEuropeanUnion’sHorizon2020researchandinno-vationprogramme(grantagreementnº646894)andundertheERCSeventh Framework Programme FP7/2007-2013 (grant agreementn°340863).JMG-RalsoacknowledgestheMinisteriodeEconomíayCompetitividadofSpainthroughtheprojectCGL2016-78971-P.Weacknowledge the analytical platform PARI and Stefan BorenstazjnforSEMimaging.Prof.Y.TsukiiandTheProtist InformationServer(http://protist.i.hosei.ac.jp/)areacknowledgedfortheuseofpicturesofcyanobacteria.ThisisIPGPcontributionn°3912.Wearegratefultotwoanonymousreviewersfortheirhelpfulcomments.

CONFLIC T OF INTERE S T

The authors declare that there is no conflict of interest regarding thepublicationofthisarticle.

294  |     ROUILLARD et AL.

ORCID

J. Rouillard http://orcid.org/0000-0002-7335-1536

R E FE R E N C E S

Abramoff,M.,Magalhaes,M.,&Ram,S. (2004). ImageprocessingwithImageJ.Biophotonics International,11,36–42.

Arndt,N.T.,&Nisbet,E.G.(2012).Processesontheyoungearthandthehabitats of early life. Annual Review of Earth and Planetary Sciences,40,521–549.https://doi.org/10.1146/annurev-earth-042711-105316

Beucher,S.,&Lantuejoul,C.(1979).Useofwatershedsincontourdetec-tion.PresentedattheInternationalworkshoponimageprocessing:real-timeEdgeandMotiondetection/estimation,Rennes.

Bittarello,E.,RobertoMassaro,F.,&Aquilano,D. (2010).Theepitaxialroleofsilicagroups inpromotingtheformationofsilica/carbonatebiomorphs:Afirsthypothesis.Journal of Crystal Growth,312,402–412.https://doi.org/10.1016/j.jcrysgro.2009.11.004

Brasier,M.D.,Green,O.R.,Jephcoat,A.P.,Kleppe,A.K.,VanKranendonk,M.J.,Lindsay,J.F.,…Grassineau,N.V.(2002).Questioningtheev-idence for Earth’s oldest fossils. Nature, 416, 76–81. https://doi.org/10.1038/416076a

Brasier,M.D.,Green,O.R.,Lindsay,J.F.,Mcloughlin,N.,Steele,A.,&Stoakes, C. (2005). Critical testing of Earth’s oldest putative fos-sil assemblage from the ~3.5 Ga Apex chert, Chinaman Creek,WesternAustralia.Precambrian Research,140, 55–102. https://doi.org/10.1016/j.precamres.2005.06.008

Buick,R.(1990).MicrofossilrecognitioninArcheanrocks:Anappraisalofspheroidsandfilamentsfroma3500M.Y.OldChert-BariteUnitat North Pole,Western Australia. Palaios,5, 441–459. https://doi.org/10.2307/3514837

Butterfield,N.J.,&Chandler,F.W.(1992).Palaeoenvironmentaldistri-butionofProterozoicmicrofossils,withanexamplefromtheAguBayformation,BaffinIsland.Palaeontology,35,943–957.

Campbell,K.A., Lynne,B.Y.,Handley,K.M., Jordan,S.,Farmer, J.D.,& Guido, D. M. (2015). Tracing biosignature preservation of geo-thermally silicified microbial textures into the geological record.Astrobiology,15,858–882.https://doi.org/10.1089/ast.2015.1307

Charlou,J.,Donval,J.P.,Fouquet,Y.,Jean-Baptiste,P.,Dehairs,F.,Holm,N.,&Godfroy,A. (2005).Organics in hydrothermal fluids from ultra-mafics on the Mid-Atlantic Ridge (MAR)-Abiogenic and/or biogenic ori-gin?PresentedattheAGUFallMeetingAbstracts.

Damer,B.(2016).AfieldtriptotheArchaeaninsearchofDarwin’swarmlittlepond.Life,6,21.https://doi.org/10.3390/life6020021

DeYoreo,J.J.,&Vekilov,P.G.(2003).Principlesofcrystalnucleationandgrowth. Reviews in Mineralogy and Geochemistry,54,57–93.https://doi.org/10.2113/0540057

Djokic,T.,VanKranendonk,M.J.,Campbell,K.A.,Walter,M.R.,&Ward,C. R. (2017). Earliest signs of life on land preserved in ca. 3.5Gahot spring deposits.Nature Communications,8, 15263. https://doi.org/10.1038/ncomms15263

Eiblmeier,J.,Kellermeier,M.,Rengstl,D.,García-Ruiz,J.M.,&Kunz,W.(2013). Effect of bulk pH and supersaturation on the growth be-haviorof silicabiomorphs in alkaline solutions.CrystEngComm,15,43–53.https://doi.org/10.1039/C2CE26132D

García-Ruiz,J.M.(1998).Carbonateprecipitationintoalkalinesilica-richenvironments. Geology,26,843–846.https://doi.org/10.1130/0091-7613(1998)026&lt;0843:CPIASR&gt;2.3.CO;2

García-Ruiz, J. M., Carnerup, A., Christy, A. G., Welham, N. J.,& Hyde, S. T. (2002). Morphology: An ambiguous indica-tor of biogenicity. Astrobiology, 2, 353–369. https://doi.org/10.1089/153110702762027925

Garcia-Ruiz, J., Hyde, S. T., Carnerup, A. M., Christy, A. G., VanKranendonk, M. J., &Welham, N. J. (2003). Self assembled silica

carbonate structures and detection of ancient microfossils. Science,302,1194–1197.https://doi.org/10.1126/science.1090163

García-Ruiz,J.M.,Melero-García,E.,&Hyde,S.T.(2009).Morphogenesisof self-assembled nanocrystalline materials of barium carbon-ate and silica. Science, 362, 362–365. https://doi.org/10.1126/science.1165349

García-Ruiz, J. M., Nakouzi, E., Kotopoulou, E., Tamborrino, L., &Steinbock, O. (2017). Biomimetic mineral self-organization fromsilica-richspringwaters.Science Advances,3,e1602285.https://doi.org/10.1126/sciadv.1602285

Harvey,R.J.,&Marr,A.G.(1966).Measurementofsizedistributionsofbacterial cells. Journal of Bacteriology,92,805–811.

Herzig,P.,&Hannington,M.(2006).Inputfromthedeep:Hotventsandcoldseeps.InH.D.Schulz&M.Zabel(Eds.),Marine geochemistry(pp.457–479).Berlin:Springer.https://doi.org/10.1007/3-540-32144-6

Holm, N. G. (1992). Chapter 1 Why are hydrothermal systems pro-posedasplausibleenvironmentsfortheoriginoflife?Origins of Life and Evolution of the Biosphere, 22, 5–14. https://doi.org/10.1007/BF01808015

Holm,N.G.,&Andersson,E. (2005).Hydrothermal simulationexperi-ments as a tool for studies of the origin of life on earth and other terrestrial planets:A review.Astrobiology,5, 444–460. https://doi.org/10.1089/ast.2005.5.444

Javaux,E.J.,Marshall,C.P.,&Bekker,A.(2010).Organic-walledmicro-fossils in 3.2-billion-year-old shallow-marine siliciclastic deposits.Nature,463,934–938.https://doi.org/10.1038/nature08793

Kasting,J.F.(2014).AtmosphericcompositionofHadean–earlyArcheanEarth:TheimportanceofCO.InGeological Society of America Special Papers. Geological Society of America,pp.19–28.

Katz, A., Alimova, A., Xu,M., Rudolph, E., Shah,M. K., Savage, H. E.,…Alfano,R. R. (2003). Bacteria size determination by elastic lightscattering. IEEE Journal of Selected topics in Quantum Electronics,9,277–287.https://doi.org/10.1109/JSTQE.2003.811284

Kellermeier, M., Cölfen, H., & García-Ruiz, J. M. (2012). Silica bio-morphs:Complexbiomimetichybridmaterialsfrom“sandandchalk”.European Journal of Inorganic Chemistry,2012, 5123–5144. https://doi.org/10.1002/ejic.201201029

Kellermeier,M.,Glaab,F.,Melero-García,E.,&García-Ruiz,J.M.(2013).Experimentaltechniquesforthegrowthandcharacterizationofsil-icabiomorphsandsilicagardens.Methods in Enzymology,532,225–256.https://doi.org/10.1016/B978-0-12-416617-2.00011-4

Kellermeier, M., Melero-García, E., Glaab, F., Eiblmeier, J., Kienle, L.,Rachel,R.,…García-Ruiz,J.M.(2012).Growthbehaviorandkineticsofself-assembledsilica-carbonatebiomorphs.Chemistry,18,2272–2282.https://doi.org/10.1002/chem.201102407

Knoll, A. H., & Golubic, S. (1979). Anatomy and taphonomy of aPrecambrian algal stromatolite. Precambrian Research,10,115–151.https://doi.org/10.1016/0301-9268(79)90022-6

Koch, A. L. (1966). Distribution of cell size in growing cultures ofbacteria and the applicability of the collins-richmond princi-ple. Journal of General Microbiology, 45, 409–417. https://doi.org/10.1099/00221287-45-3-409

Konhauser,K.O.,Jones,B.,Reysenbach,A.-L.,&Renaut,R.W.(2003).Hotspringsinters:KeystounderstandingEarth’searliestlifeforms.Canadian Journal of Earth Sciences, 40, 1713–1724. https://doi.org/10.1139/e03-059

Lang,S.Q.,Butterfield,D.A.,Schulte,M.,Kelley,D.S.,&Lilley,M.D.(2010). Elevated concentrations of formate, acetate and dissolvedorganic carbon found at the Lost City hydrothermal field. Geochimica et Cosmochimica Acta, 74, 941–952. https://doi.org/10.1016/j.gca.2009.10.045

Lide,D.(1947).CRC handbook of chemistry and physics.VilledeQuébec,QC,Canada:ULaval.

Ludwig, K. A., Kelley, D. S., Butterfield, D. A., Nelson, B. K., & Früh-Green,G.(2006).Formationandevolutionofcarbonatechimneysat

     |  295ROUILLARD et AL.

theLostCityHydrothermalField.Geochimica et Cosmochimica Acta,70,3625–3645.https://doi.org/10.1016/j.gca.2006.04.016

Macleod,G.,McKeown,C.,Hall,A.J.,&Russell,M.J.(1994).HydrothermalandoceanicpHconditionsofpossiblerelevancetotheoriginoflife.Origins of Life and Evolution of the Biosphere,24,19–41.https://doi.org/10.1007/BF01582037

Maliva,R.G.,Knoll,A.H.,&Simonson,B.M.(2005).SecularchangeinthePrecambriansilicacycle: Insightsfromchertpetrology.Bulletin of the Geological Society of America, 117, 835–845. https://doi.org/10.1130/B25555.1

Marshall,C.P.,Emry,J.R.,&Marshall,A.O.(2011).Haematitepseudo-microfossilspresent inthe3.5-billion-year-oldApexChert.Nature Geoscience,4,240–243.https://doi.org/10.1038/ngeo1084

Martin,W.,&Russell,M.J.(2007).Ontheoriginofbiochemistryatanalkaline hydrothermal vent. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 362, 1887–1926.https://doi.org/10.1098/rstb.2006.1881

McCollom, T. M., & Bach, W. (2009). Thermodynamic constraintson hydrogen generation during serpentinization of ultramaficrocks.Geochimica et Cosmochimica Acta, 73, 856–875. https://doi.org/10.1016/j.gca.2008.10.032

McCollom,T.M.,&Seewald, J.S. (2006).Carbon isotopecompositionoforganiccompoundsproducedbyabiotic synthesisunderhydro-thermal conditions. Earth and Planetary Science Letters,243,74–84.https://doi.org/10.1016/j.epsl.2006.01.027

McLoughlin,N.,Wilson,L.A.,&Brasier,M.D.(2008).Growthofsyntheticstromatolites andwrinkle structures in the absenceofmicrobes–Implicationsfortheearlyfossilrecord.Geobiology,6,95–105.https://doi.org/10.1111/j.1472-4669.2007.00141.x

Melero-García, E., Santisteban-Bailón, R., & García-Ruiz, J.M. (2009).Role of bulk pH during witherite biomorph growth in silica gels.Crystal Growth and Design,9, 4730–4734. https://doi.org/10.1021/cg9005967

Milesi,V.,Guyot,F.,Brunet,F.,Richard,L.,Recham,N.,Benedetti,M.,… Prinzhofer, A. (2015). Formation of CO2, H2 and condensedcarbon from siderite dissolution in the 200–300°C range and at50MPa.Geochimica et Cosmochimica Acta,154,201–211.https://doi.org/10.1016/j.gca.2015.01.015

Munoz-Saez, C., Saltiel, S.,Manga,M.,Nguyen, C., &Gonnermann,H. (2016). Physical and hydraulic properties of modern sin-ter deposits: El Tatio, Atacama. Journal of Volcanology and Geothermal Research, 325, 156–168. https://doi.org/10.1016/j.jvolgeores.2016.06.026

Nijman,W.,DeBruijne,K.C.,&Valkering,M. E. (1998).Growth faultcontrol of Early Archean cherts, barite mounds and chert-bariteveins, North Pole Dome, Eastern Pilbara, Western Australia.Precambrian Research, 88, 25–52. https://doi.org/10.1016/S0301-9268(97)00062-4

Nývlt, J.,Söhnel,N.,Matuchová,M.,&Broul,B. (1985).The kinetics of industrial crystallization.NewYork:Elsevier.

Opel, J., Wimmer, F. P., Kellermeier, M., & Cölfen, H. (2016).Functionalisationofsilica–carbonatebiomorphs.Nanoscale Horizons,1,144–149.https://doi.org/10.1039/C5NH00094G

Parman,S.W.,Shimizu,N.,Grove,T.L.,&Dann,J.C.(2003).Constraintson the pre-metamorphic trace element composition of Barbertonkomatiites from ion probe analyses of preserved clinopyroxene.Contributions to Mineralogy and Petrology,144,383–396.https://doi.org/10.1007/s00410-002-0406-1

Philippot,P.,VanZuilen,M.,Lepot,K.,Thomazo,C.,Farquhar,J.,&VanKranendonk,M.J.(2007).EarlyArchaeanmicroorganismspreferredelementalsulfur,notsulfate.Science (New York, N.Y.),317,1534–1537.https://doi.org/10.1126/science.1145861

Pinti,D. L.,Mineau,R.,&Clement,V. (2009).Hydrothermal alterationandmicrofossil artefactsof the3,465-million-year-oldApexchert.Nature Geoscience,2,640–643.https://doi.org/10.1038/ngeo601

Proskurowski,G.,Lilley,M.D.,Seewald,J.S.,Fruh-Green,G.L.,Olson,E. J., Lupton, J. E., … Kelley, D. S. (2008). Abiogenic hydrocarbonproduction at lost city hydrothermal field. Science, 319, 604–607.https://doi.org/10.1126/science.1151194

Ruff, S.W., & Farmer, J. D. (2016). Silica deposits onMars with fea-turesresemblinghotspringbiosignaturesatElTatioinChile.Nature Communications,7,13554.https://doi.org/10.1038/ncomms13554

Russell, M. J., Hall, A. J., & Martin, W. (2010). Serpentinization asa source of energy at the origin of life: Serpentinization andthe emergence of life. Geobiology, 8, 355–371. https://doi.org/10.1111/j.1472-4669.2010.00249.x

Schopf,J.W.,Kudryavtsev,A.B.,Agresti,D.G.,Wdowiak,T.J.,&Czaja,A.D.(2002).Laser-RamanimageryofEarth’searliestfossils.Nature,416,73–76.https://doi.org/10.1038/416073a

Schopf, J.W.,Kudryavtsev,A.B., Sugitani,K.,&Walter,M.R. (2010).Precambrian microbe-like pseudofossils : A promising solutionto the problem. Precambrian Research, 179, 191–205. https://doi.org/10.1016/j.precamres.2010.03.003

Schulte,M.,Blake,D.,Hoehler,T.,&McCollom,T.(2006).SerpentinizationanditsimplicationsforlifeontheearlyEarthandMars.Astrobiology,6,364–376.https://doi.org/10.1089/ast.2006.6.364

Shibuya,T.,Komiya,T.,Nakamura,K.,Takai,K.,&Maruyama,S.(2010).Highly alkaline, high-temperature hydrothermal fluids in the earlyArchean ocean. Precambrian Research, 182, 230–238. https://doi.org/10.1016/j.precamres.2010.08.011

Siever, R. (1992). The silica cycle in the Precambrian. Geochimica et Cosmochimica Acta, 56, 3265–3272. https://doi.org/10.1016/ 0016-7037(92)90303-Z

Singer,E.,Emerson,D.,Webb,E.,Barco,R.,Kuenen, J.,Nelson,W.,…Edwards,K. (2011).MariprofundusferrooxydansPV-1thefirstge-nomeofamarineFe(II)oxidizingZetaproteobacterium.PLoS ONE,6,e25386.https://doi.org/10.1371/journal.pone.0025386

Stefurak, E. J. T., Lowe, D. R., Zentner, D., & Fischer, W. W. (2014).Primarysilicagranules—AnewmodeofPaleoarcheansedimentation.Geology,42,283–286.https://doi.org/10.1130/G35187.1

Sugitani,K.,Grey,K.,Nagaoka,T.,Mimura,K.,&Walter,M.R. (2009).TaxonomyandbiogenicityofArchaean spheroidalmicrofossils (ca.3.0Ga)fromtheMountGoldsworthy-MountGrantareainthenorth-eastern Pilbara Craton, Western Australia. Precambrian Research,173,50–59.https://doi.org/10.1016/j.precamres.2009.02.004

Sugitani, K., Lepot, K., Nagaoka, T., Mimura, K., Kranendonk, M. V.,Oehler, D. Z., &Walter, M. R. (2010). Biogenicity of morphologi-cally diverse carbonaceousmicrostructures from the ca. 3400Mastrelley pool formation, in the Pilbara Craton, Western Australia.Astrobiology,10,899–920.https://doi.org/10.1089/ast.2010.0513

Sugitani,K.,Mimura,K.,Nagaoka,T.,Lepot,K.,&Takeuchi,M. (2013).Microfossil assemblage from the3400MaStrelley pool formationinthePilbaraCraton,WesternAustralia:Resultsformanewlocal-ity. Precambrian Research, 226, 59–74. https://doi.org/10.1016/j.precamres.2012.11.005

Sugitani,K.,Mimura,K.,Takeuchi,M.,Suzuki,K.,Senda,R.,Asahara,Y.,…VanKranendonk,M.J.(2015).APaleoarcheancoastalhydrother-mal field inhabitedbydiversemicrobial communities :TheStrelleyPool formation, Pilbara Craton,Western Australia.Geobiology,13,522–545.https://doi.org/10.1111/gbi.12150

Ueno, Y., Yamada, K., Yoshida,N.,Maruyama, S., & Isozaki, Y. (2006).Evidence from fluid inclusions for microbial methanogenesis in the early Archaean era. Nature,440,516–519.https://doi.org/10.1038/nature04584

Uysal,Z.(2001).ChroococcoidcyanobacteriaSynechococcusspp.intheBlackSea:Pigments,size,distribution,growthanddiurnalvariability.Journal of Plankton Research,23,175–189.https://doi.org/10.1093/plankt/23.2.175

VandenBoorn,S.H.J.M.,VanBergen,M.J.,Nijman,W.,&Vroon,P.Z. (2007). Dual role of seawater and hydrothermal fluids in Early

296  |     ROUILLARD et AL.

Archeanchertformation:Evidencefromsiliconisotopes.Geology,35,939–942.https://doi.org/10.1130/G24096A.1

Van Kranendonk, M. J. (2006). Volcanic degassing, hydrothermal cir-culation and the flourishing of early life on Earth: A review of the evidence from c. 3490–3240Ma rocks of thePilbara Supergroup,Pilbara Craton,Western Australia. Earth- Science Reviews,74, 197–240.https://doi.org/10.1016/j.earscirev.2005.09.005

Van Kranendonk, M. J. V., Hickman, A. H., Williams, I. R., & Nijman,W. (2001).Archaean geology of the east Pilbara Granite -greenstone Terrane Western Australia – A field guide.Perth:GeologicalSurveyofWesternAustralia.

VanKranendonk,M.J.,&Pirajno,F. (2004).Geochemistryofmetaba-saltsandhydrothermalalterationzonesassociatedwithc.3.45Gachertandbaritedeposits: Implicationsforthegeologicalsettingofthe Warrawoona Group, Pilbara Craton, Australia. Geochemistry: Exploration, Environment, Analysis,4,253–278.

Voinescu, A. E., Kellermeier, M., Carnerup, A. M., Larsson, A. K.,Touraud, D., Hyde, S. T., & Kunz, W. (2007). Co-precipitation ofsilica and alkaline-earth carbonates using TEOS as silica source.Journal of Crystal Growth,306, 152–158. https://doi.org/10.1016/j.jcrysgro.2007.03.060

VonDamm,K.(1995).Controlsonthechemistryandtemporalvariabilityofseafloorhydrothermalfluids.In:S.E.Humphris,R.A.Zierenberg,L.S.Mullineaux&R.E.Thomson, (Eds.),Seafloor hydrothermal sys-tems: Physical, chemical, biological, and geological interactions (pp.222–247).Oxford:Wiley.

Wacey,D.,Kilburn,M.R.,Saunders,M.,Cliff,J.,&Brasier,M.D.(2011).Microfossils of sulphur-metabolizing cells in 3.4-billion-year-oldrocksofWesternAustralia.Nature Geoscience,4,698–702.https://doi.org/10.1038/ngeo1238

Wacey, D., Saunders, M., Kong, C., Brasier, A., & Brasier, M. (2015).3.46GaApexchert“microfossils”reinterpretedasmineralartefacts

produced during phyllosilicate exfoliation.Gondwana Research,36,296–313.

Walker,J. (1985).CarbondioxideontheearlyEarth.Origins of Life and Evolution of the Biosphere, 16, 117–127. https://doi.org/10.1007/BF01809466

Wolf,E.T.,&Toon,O.B.(2014).ControlsontheArcheanclimatesysteminvestigated with a global climate model. Astrobiology,14,241–253.https://doi.org/10.1089/ast.2013.1112

Workman,R.K.,&Hart,S.R.(2005).Majorandtraceelementcomposi-tionofthedepletedMORBmantle(DMM).Earth and Planetary Science Letters,231,53–72.https://doi.org/10.1016/j.epsl.2004.12.005

Zhang,G.(2015).Morphogenesis of self-assembled crystalline materials of calcium carbonate and silica.

Zhang, G.,Morales, J., &García-Ruiz, J.M. (2017). Growth behaviourof silica/carbonate nanocrystalline composites of calcite and ara-gonite. Journal of materials Chemistry B,5, 1658–1663. https://doi.org/10.1039/C6TB02612E

SUPPORTING INFORMATION

Additional Supporting Information may be found online in thesupportinginformationtabforthisarticle.

How to cite this article:RouillardJ,García-RuizJ-M,GongJ,vanZuilenMA.Amorphogramforsilica-witheritebiomorphsanditsapplicationtomicrofossilidentificationintheearlyearthrockrecord.Geobiology. 2018;16:279–296. https://doi.org/10.1111/gbi.12278