· Web viewX-ray microtomography as a tool for investigating the petrological context of Precambrian cellular remains Keyron Hickman-Lewis1,2*, Russell J. Garwood3,4, Philip J. Withers5

X-ray microtomography as a tool for investigating the petrological context of

Precambrian cellular remains

Keyron Hickman-Lewis1,2*, Russell J. Garwood3,4, Philip J. Withers5 and David Wacey6,7

1: St Edmund Hall, Queens Lane, Oxford, OX1 4AR, United Kingdom 2: Department of Earth Sciences, The University of Oxford, South Parks Road, Oxford, OX1

3AN, United Kingdom.3: School of Earth, Atmospheric and Environmental Sciences, The University of Manchester,

Manchester, M13 9PL, United Kingdom.4: Department of Earth Sciences, The Natural History Museum, Cromwell Road, London

SW7 5BD, UK5: School of Materials, The University of Manchester, Manchester, M13 9PL, United

Kingdom.6: School of Earth Sciences, Life Sciences Building, University of Bristol, 24 Tyndall Avenue,

Bristol, BS8 1TH, United Kingdom.7: Centre for Microscopy Characterisation and Analysis, and Centre of Excellence for Core to Crust Fluid Systems, The University of Western Australia, 35 Stirling Highway, Perth, WA

6009, Australia.

*Corresponding author at: CNRS Centre de Biophysique Moléculaire, Rue Charles Sadron, 45071 Orléans, France

E-mail address: [email protected]

1

2

3

4

5

6789

101112131415161718192021222324

25

Abstract: A wide spectrum of tomographic techniques now exists for studying

palaeontological specimens, but the suitability of these methods for assessing Earth’s oldest

prokaryotic life has not been comprehensively investigated. Here we seek to evaluate the

ability of X-ray computed tomography – specifically X-ray microtomography (μCT) – to

reveal the morphology and petrological context of Precambrian microfossils, pseudofossils

and biosedimentary structures, all of which have significance to the origin and early evolution

of life of Earth. Material tested herein comes from the Pilbara Craton of Western Australia

(the 3.49 Ga Dresser Formation, 3.46 Ga Apex Chert and 3.43 Ga Strelley Pool Formation)

and the 1.88 Ga Gunflint Formation of Ontario, Canada. These units chart key developments

in palaeobiology: the oldest, contain profoundly controversial microfossil-like objects and

microbially induced sedimentary structures, whilst definitive prokaryotes are found in the

youngest. We demonstrate that the imaging of individual microfossils and pseudofossils

currently lies at the limits of lab-based μCT capabilities and requires beneficial taphonomy.

However, microtomography does provide a good overview of their petrological context at

flexible spatial scales, although the quality of data obtained from mesoscopic MISS and

stromatolites depends largely on their style of preservation.

Running title: X-ray CT of Precambrian fossils

Keywords: X-RAY COMPUTED TOMOGRAPHY–3D VISUALISATION–

STROMATOLITES–DRESSER FORMATION–APEX BASALT–STRELLEY POOL

FORMATION–GUNFLINT FORMATION–MICROFOSSILS–PSEUDOFOSSILS–

CHERT–PRECAMBRIAN

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

INTRODUCTION

Precambrian time (prior to ~ 0.54 Ga) comprises the majority of Earth history and

fundamental steps in the evolution and history of life. Thus, the Precambrian fossil record

provides key insights into important events, including the early evolution of the biosphere

and its impact on the planet, and into the patterns of evolution itself. Recent decades have

witnessed a developing recognition that exceptional preservation occurs in the Precambrian

fossil record (e.g. Westall et al., 2006; El Albani et al., 2010; Strother et al. 2011; Brasier et

al. 2015); the early cellular fossil record is one of complex taphonomy, but widespread three-

dimensional preservation. For example, this often results from the early encasement of

organisms in chert or phosphate, aided by a lack of bioturbation (a feature of sediments found

only since the Cambrian).

51

52

53

54

55

56

57

58

59

60

61

62

Recent decades have also seen the advent, development, and increasingly widespread

application of virtual palaeontology techniques, namely computer-aided three-dimensional

imaging and visualisation (Garwood et al. 2010). High-quality, three-dimensionally

preserved Precambrian fossils are potentially amenable to such approaches, which could

provide novel insights into both their morphology, and taxonomy. Whilst there are myriad

techniques for the three-dimensional reconstruction of fossils (see Sutton et al. 2014 for an

overview), a primary contributor to the popularity of virtual palaeontology has been

significant development in X-ray computed tomography (CT). Lab-based CT scanners

capable of micrometre to sub-micrometre spatial resolutions have been developed and

become increasingly widespread and easy to access. This high resolution form of CT — X-

ray microtomography or micro-CT (XMT/μCT) — has been applied to a wide range of

fossils, from large vertebrates, such as dinosaurs (Butler et al. 2008; Anné et al. 2015) and

marine reptiles (Neenan & Scheyera 2012), through plants (Spencer et al. 2012; Steart et al.

2015) and invertebrates (Garwood & Dunlop 2014; Streng et al. 2016), to microfossils

(Baumgartner-Mora et al. 2006). Concurrent developments in synchrotron tomography

(especially in ‘third-generation’ synchrotrons) have allowed specimens which are otherwise

difficult to study to be non-destructively imaged (e.g. Cunningham et al. 2015). This is

because the tunable, monochromatic beam, and associated optics, can allow better resolution,

and clearer differentiation of phases with similar attenuation contrast. Advances in digital

visualisation and increases in computer power have allowed tomographic data to be better

utilised. These factors combined have made X-ray CT a standard approach for the modern

palaeontologist in the analysis of three-dimensional fossils. Despite this, such techniques

have yet to be systematically applied to early cellular fossils. In part this is because μCT

scanning typically requires preparation of samples to ensure minimal host rock is outside the

field of view when scanning (Sutton et al. 2014); this is challenging for very small fossils.

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

Nevertheless, μCT could be used as a means of augmenting, rather than replacing, traditional

techniques for demonstrating the larger scale petrological context of these fossils. As such, it

could facilitate characterisation prior to destructive analytical methods such as serial

sectioning, or in instances where these are prohibited. Here we examine the use of μCT in

Precambrian palaeobiology, assessing its efficacy as a tool for the interrogative study of both

the petrological context of early cellular fossils, and potentially their morphology.

MATERIALS AND METHODS

Sample material

Sample material was chosen to represent the most common lithology in which the search for

early life on Earth has traditionally been focused: silicified sedimentary rocks. The samples

were chosen also to span a significant portion of Precambrian time, from some of the earliest

definitive sedimentary rocks on the planet (c. 3.5 Ga) containing controversial evidence for

life, through to 1.88 Ga sediments that contain conclusive evidence for a diverse level of

prokaryote cellular evolution. The materials from which our samples were chosen were

collected during a series of field campaigns to the Pilbara Craton of Western Australia and

Ontario, Canada between 1999 and 2011, coordinated through the Department of Earth

Sciences, Oxford University, The University of Western Australia and the Geological Survey

of Western Australia. These campaigns were extensive exercises in field mapping for these

areas, thus all samples are readily re-locatable by means of their co-ordinates, as determined

by Global Positioning System (G.P.S.) technology.

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

Samples were prepared in the Geological Facilities Laboratories, a Small Research Facility

(S.R.F.) within Oxford University Department of Earth Sciences and by one of us (DW) at

the University of Western Australia. Regions of interest within the samples were selected

following comprehensive hand sample and thin section observation. The Apex chert material

from Chinaman Creek, Pilbara, is stored as part of the Martin Brasier Collection in the

Oxford University Museum of Natural History (OUMNH) Earth Collections. Other material

is currently stored at the University of Western Australia.

3.49 Ga Dresser Formation, Pilbara Supergroup, Western Australia

This sample was collected from the Panorama greenstone belt in the North Pole Dome about

30 km west of Marble Bar in the Pilbara Craton of Western Australia. It is a portion of a

putative silicified stromatolite (sample DR1), similar to nodular or wavy-laminated structures

described by Walter et al. (1980) and Van Kranendonk (2006). The biogenicity of such

Dresser Formation stromatolites has been questioned, because distinguishing their

morphologies from soft sediment deformational features can be difficult (Lowe 1994).

3.46 Ga Apex Chert, Pilbara Supergroup, Western Australia

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

Four samples were chosen from this unit, all from the Chinaman Creek locality of Brasier et

al. (2005, 2011) in the Marble Bar greenstone belt, about 10 km west of Marble Bar. One

sample (CHIN3) comes from a hydrothermal black chert vein that contains filamentous

microstructures, once thought to be Earth’s oldest microfossils (Schopf 1993), but recently

reinterpreted as mineral artefacts caused by the hydrothermal alteration of flakes of mica

(Brasier et al. 2015; Wacey et al. 2015). The other samples (CCT5, CC43, CC164) come

from an overlying sedimentary stratiform bedded chert unit (Unit 4 of Brasier et al. 2011;

informally referred to as ’the Apex chert’) that contains a variety of carbonaceous

microstructures, some of which, laminated in morphology and intrinsic to the primary fabric

of the lithology, are potentially biogenic. They have been interpreted as probable microbially-

induced sedimentary structures (MISS; cf. Noffke 2009, 2010), formed in a shallow-water

marine environment (Hickman-Lewis et al. 2016). The selection of samples encompasses a

metalliferous, Fe-rich chert included to provide evidence of the ability of CT to delineate

sedimentary micro-facies (CCT5; see Brasier et al. 2011), a microgranular chert in which is

found a layer of primary laminated clasts (CC43; Hickman-Lewis 2015) and a microclastic

chert containing persistent, filament-like laminae which are intrinsic to the primary fabric of

the rock (CC164; Hickman-Lewis et al. 2016).

3.43 Ga Strelley Pool Formation, Pilbara Supergroup, Western Australia

This sample (SPV3) comes from the East Strelley greenstone belt of the Pilbara Craton about

80 km west of Marble Bar. It is a rock chip from the same hand sample that produced thin

sections containing tubular and spheroidal carbonaceous microfossils associated with pyrite

(Wacey et al. 2011b). The lithology is a silicified placer sandstone, rich in detrital heavy

minerals and rounded quartz grains. Microfossils typically occur in early silica cements in

between, and coating, the detrital grains (Wacey et al. 2011b, 2012).

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

1.88 Ga Gunflint Formation, Animikie Group, Ontario, Canada

Three samples were chosen from the Gunflint Formation. Two come from the Schreiber

Channel locality made famous by the description of the first evidence of Precambrian

microfossils by Tyler & Barghoorn (1954) and Barghoorn & Tyler (1965). Of these two

samples, one is a black and white chert that is faintly laminated in hand specimen (SC10), the

other is a massive black chert (SC11). The third sample from the Gunflint Formation (G15)

comes from the Mink Mountain locality approximately 200 km west of the Schreiber

Channel locality. This sample is laminated in both hand specimen and thin section. Previous

examination of similar material from this location by Schelble et al. (2004) shows that it

comprises micro-quartz, iron oxide and minor amounts of iron silicate.

Methods

All scanning was conducted at the Henry Moseley X-ray Imaging Facility, School of

Materials, University of Manchester. Cone-beam μCT systems — as opposed to parallel

beam systems more commonly employed with synchrotron X-rays — employ geometric

magnification to provide smaller voxel sizes: the closer a sample is to the source, the greater

the magnification (Sutton et al. 2014). Furthermore, the filtered back projection algorithms

typically used for reconstructing volumes in a lab-based setting can deal poorly with large

amounts of material outside the field of view during a scan. Thus resolution is linked to

sample size, and field of view to detector panel size (and, in some scanners, optics employed

for any given scan). For this reason, we used two instruments for the current study – a Nikon

system capable of scanning larger samples, and a Zeiss scanner better suited to smaller

samples, but capable of achieving higher resolutions.

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

Nikon CT scanning

The largest samples were scanned using a Nikon 225/320 kV CT system fitted within a

customised bay with a 2K x 2K Perkin Elmer 1621-16-bit amorphous silicon flat-panel

detector and tungsten reflection target. This has a powerful enough source to penetrate large

samples, albeit providing lower resolution than the other scanner used for this study. The

largest specimen for which we present results was 8x3x.1.5 cm (Fig. 1), whilst the largest

specimen we scanned (which showed no features of interest) was 7.5x3.5x6.5 cm. The

specimens outlined in Table 1 were all scanned with an exposure of 708ms, for each of the

3142 projections, using the parameters shown in Table 1.

Sample Acceleration voltage (kv)

Current (µA)

Filter Voxel size (µm)

Dresser strom. DR1 180 100 1mm, Cu

42.3

Apex CC43_01 75 180 - 16.4Apex CC164C2 95 80 - 19.6Apex CC164C2_a 75 180 - 14.2Apex CCT5 70 175 - 8.2Strelley SPV3 70 175 - 4.16Gunflint Mink 70 175 - 7.7Gunflint SC10 50 130 - 5.8Gunflint SC11_5F 150 50 - 7.4

177

178

179

180

181

182

183

184

185

Table 1. Scanning parameters for samples scanned with the Nikon CT system.

The choice of scanning parameters reflects variations in the size and X-ray density of the

samples. Scanning was achieved by placing the sample within the field of view, and moving

the manipulator to ensure the longest axis remained in the field of view throughout a full

rotation. We then modified the voltage and current of the X-ray source, trying to keep the

former as low as possible to achieve maximal composition-based contrast, whilst allowing

enough penetration. For the largest sample, where the detector panel was saturating without

the sample in the field of view (precluding flat field corrections), we added a filter. For a full

overview of the considerations inherent in selecting scanning parameters, we refer the reader

to Sutton et al. (2014).

Zeiss Versa CT scanning

Smaller samples, typically rock chips or small cores, were scanned using a Zeiss Xradia

Versa 520. These employed the standard transmission target and the standard in-built, Zeiss

filters. All used 4x optical magnification to improve resolution and binning of 2 in XY and in

Z of the 2000 x 2000 detector to reduce noise. We collected 1001 projections, with other

adjustable scan parameters shown in Table 2.

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

Sample Acceleration voltage (kv)

Current (µA)

Voxel size (µm)

Exposure time (s)

Filter

Apex Chin03 80 88 1.12 10 LE3

Strelley SPV3

80 88 1.12 8 LE1

Gunflint SC11

80 88 1.12 8 LE1

Gunflint Mink

80 88 1.12 8 LE1

Gunflint SC10

80 87 1.12 8 LE1

Table 2. Scanning parameters for samples scanned with the Zeiss Versa system.

The samples scanned on this instrument were of a similar size, due to field of view

limitations. Hence all scans employed very similar source current and voltages, selected to

allow ~30% transmission through the densest region of the specimen. Choice of filter was

based on protocols provided by the manufacturer based upon transmission, and exposure time

was selected to minimise noise, based on visual inspections of the projections.

Drishti

All sample datasets were initially interrogated using volume rendering: this approach loads

the volume into RAM and then allows the user to dictate how voxels are displayed (Sutton et

al. 2014). By so doing, provided data is relatively clean and segmentation is not required, this

method can provide a useful visualisation of the major features visible in a sample in an

efficient manner. Volume renders for this work were created using the open source software

Drishti (sf.anu.edu.au/Vizlab/drishti; Limaye 2012) following the methods of Streng et al.

(2016). This was conducted on a PC with 256 Gb of RAM, a 64 core Intel I5 processor, and

an NVIDIA Quadro graphics card.

SPIERS

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

For samples which required segmentation — i.e. the isolation of individual structures within

a sample for separate rendering — we used a surfacing approach. This, whilst time

consuming, allows the user precise control over which features are rendered. This was

achieved using the SPIERS software suite (spiers-software.org; Sutton et al. 2012), following

the methods of Brasier et al. (2015). SPIERS processing does not require particularly

powerful hardware and was conducted on a standard MacBook Air computer with a 1.7 GHz

Intel Core i7 processor, 8 Gb of memory, and an Intel HD 5000 Graphics card.

RESULTS

Dresser Formation (sample DR1)

225

226

227

228

229

230

231

232

233

234

Sample DR1 comprises numerous distinct phases, visible in hand specimen (Fig. 1a-d),

including white, yellow-orange, red-brown and black phases, plus void space. Based on thin

section observation, the white phases are interpreted as silica and barite, yellow-orange and

red-brown are a mélange of various iron oxides and hydroxides, and the black phase is a

sulphide. At least three of these phases, plus void space can be distinguished in the CT slices

(e.g. Fig 1e), allowing these data to be reconstructed and visualised in order to show the

spatial distribution of several phases in 3D (Fig. 1f-i). The visualisations demonstrate the

overall morphology of, and some of the phase distributions within, the putative stromatolite.

They could therefore be used as a mechanism for digitally archiving such stromatolite

specimens. However, not all of the phases seen in the hand specimen and thin sections are

clear in the CT visualisations. Phases of similar density (iron oxides, hydroxides and

sulphide) are difficult to distinguish from one another in both the CT data and visualisations

(Fig. 1f-i). Hence only void space (black), silica (white, white arrows in Fig. 1), barite (grey,

grey arrows in Fig. 1), and Fe-rich phases (gold, gold arrows in Fig. 1) can be differentiated.

Nevertheless, CT data have the pivotal advantage of being easily manipulated to visualise the

stromatolite interior which is hidden from view in the hand specimen (Fig. 1g-h;

Supplementary movie), which confirms that there is three-dimensional complexity to the

laminae.

Apex Basalt, Chinaman Creek (sample CHIN3)

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

CT slices of CHIN3, acquired on the Versa instrument with 1.12μm voxels, possess a

uniform mid-grey background. Numerous wavering linear features cross the field of view,

and isolated denser grains of c. 10-100 μm diameter are also present. Furthermore, several

filament-like objects comprising material of lower density than quartz are of note. These are

often vermiform, curved or s-shaped (Fig. 2b, yellow arrow), are < 10 μm in diameter, reach

several tens of μm in length, and each persists through 4-6 successive slices.

255

256

257

258

259

260

261

Through comparisons with thin sections of identical material (Fig. 2a, c; Brasier et al. 2015;

Wacey et al. 2015) we interpret the uniform mid-grey background as microcrystalline quartz

matrix, and the wavering linear features as fractures. These fractures are frequently partially

carbon-filled (Fig. 2b, green arrow); the CT data is consistent with this observation showing

greater attenuation contrast between the fractures and the matrix than would be expected with

quartz-filled fractures. Unfilled fractures have not been seen in thin sections, thus they are an

unlikely explanation for the observed density contrast. Denser grains are likely to be either

pyrite or an oxidised product thereof. Pyrite grains are occasionally observed in thin sections,

as is a rusty red tinge indicative of iron oxide − a typical alteration style observed in hand

samples. The small filamentous objects visible in CT slices show strong morphological

resemblance to carbon-rich filaments observed in thin sections (Fig. 2c, yellow arrow; see

also Wacey et al. 2015). These were once thought to be microfossils of filamentous bacteria

and hence some of the oldest evidence for life on Earth (Schopf 1993). They have since been

reinterpreted as pseudofossils based on high-resolution electron microscopy data, formed

during the hydrothermal alteration of flakes of mica (Brasier et al. 2015; Wacey et al. 2015).

CT data suggest that there are at least 30 such objects in a c. 1mm3 volume of the CHIN 3

sample (Fig. 2e-f, Supplementary movie). A greater number are observed in thin sections

(Wacey et al. 2015), a discrepancy likely resulting from a combination of factors including:

(i) filament size will often be below the resolution of the scan (in this case c. 3μm); (ii) only

well preserved and particularly carbon-rich filaments are likely to have sufficient density

contrast with the siliceous matrix to be picked out by CT. Nevertheless, under favourable

conditions, CT can provide a way to rapidly examine rock chips for evidence of filamentous

objects as small as c.5 μm in diameter, and could thus be used as a preliminary non-

destructive tool in order to locate volumes of rock suitable for more detailed investigation.

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

Apex Basalt, Chinaman Creek (sample CCT5)

CCT5 is an iron-rich metalliferous chert from the stratiform Apex chert of the South Block at

Chinaman Creek; it is characterised by abiogenic linear fabrics such as stylolites and

precipitation fronts (Brasier et al. 2011). Figure 3 shows a comparison between a thin section

and μCT-based visualisation of the same rock chip. Lateral displacement between the

scanned area and the thin section accounts for the vertical displacement between equivalent

features in each. Petrographic observation (Fig. 3a) allows the thin-section to be divided into

four micro-facies: (1) an alternating dull brown sequence of metal oxides and sulphides

creating sub-millimetre horizontal laminations, interbedded with clast- and clot-rich

volcanogenic layers; (2) a stylolite-bounded layer with poorly-defined laminations; (3) a

metalliferous, Fe-stained laminated layer; and (4) an ash-dominated volcaniclastic layer.

287

288

289

290

291

292

293

294

295

296

297

298

Several of these features are apparent in the accompanying rendering of CT data; layer (3) is

clearly represented (outlined), as is a major vein through layers (1) and (2) which, in both thin

section and CT scan, is somewhat asymptotic to the bottom of the image (green arrow in Fig.

3b). The volcaniclastic grains in layer (4) are defined less clearly. They are less continuous in

the CT visualisation than when observed in the thin section. However, the thin section

demonstrates that an irregular region within the uppermost layer has been stained darker in

colour (indicated by a yellow arrow in Fig. 3a), which we here attribute to the flow of

metalliferous hydrothermal fluids. It is possible these originated in the central, laminated

layer (3), the fracture visible in hand specimen providing a possible conduit (though this does

not explain why other areas adjacent to the conduit would not be similarly stained). This area

in layer (4) is most apparent in the CT scan; a density consistent with layer (3) suggests a

genetic linkage for their attenuation contrast with the surrounding matrix. This is compatible

with density enhancement through the leaching of heavier elements from layer (2), which are

deposited in the above layer (3). The two vertical linear features in the CT scan do not

correspond to any object seen in either the thin section, or through macroscopic observation;

rather, they are artefacts produced during the scanning process (red arrows in Fig. 3b), and

picking out the corners of this rectangularly-sawed specimen.

Apex Basalt, Chinaman Creek (sample CC43)

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

This microgranular stratiform chert sample contains a dark layer of sub-rounded, coarse sand-

sized clasts (Fig 4a), approximately 8mm thick. Within only this layer, thin sections show

sub-millimetre-scale banded grains (Fig 4b–c) previously described as possible fragments

eroded from a microbial mat (Hickman-Lewis 2015). Volcanic ash clasts are otherwise

dominant. The aforementioned clast-rich layer is clear in CT data as a concentration of sub-

millimetre, sub-rounded grains, but the internal microstructure of these grains is not depicted

(Fig. 4d). This highlights the ability of μCT to demonstrate distinct sedimentological regions

within larger volumes of rock, thus allowing better focussed regions of study. The

laminations within the grains are likely not detected due to a lack of attenuation contrast

between weakly carbonaceous and silicified layers in this pervasively silicified sample. We

note that individual carbonaceous layers can be tens to hundreds of microns in thickness, and

thus their absence in scans is unlikely to be a resolution-based constraint.

Apex Basalt, Chinaman Creek (sample CC164)

Hickman-Lewis et al. (2016) reported a suite of potential microbially induced sedimentary

structures in the stratiform beds of the ‘Apex chert’, at Chinaman Creek. Amongst these, the

most biologically promising were mesoscopic, filamentous, repetitive laminations with

undulating, wrinkled topography containing detrital sediment grains. These were identified in

samples from two localities (CC164 and CCT23), and are morphologically comparable to a

MISS sub-type described by Noffke (2009, 2010) from ancient and modern siliciclastic

environments. The microgranular chert in which these laminations were identified was

atypical, containing a primary fabric of matrix-supported, coarse, volcaniclastic clasts and

composites of these clasts.

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

Tomography data presented herein (Fig. 5) shows three-dimensional features which can be

interpreted with reference to thin section observation. Sub-spherical and irregular objects

occur throughout the volume of the sample, and may represent: volcanic ash clasts;

composites of such clasts; and fragmented ash clasts. This tripartite delineation explains their

varying morphology, but consistent spatial associations and density. There is, however, a

region with a high concentration of irregular objects (arrowed in Fig. 5d) that do not bear

close resemblance to any features of the corresponding area in the thin section (compare Figs.

5a and 5d). Discontinuous packets of laminations correspond to the undulating, filament-like

layers of CC164, which occur horizontally and in persistent packages up to several hundred

µm thick (Figs. 5d and 6). String-like aggregates of putative filaments are obviously

identifiable in the μCT data (Fig. 6, arrowed). Previous petrographic study suggests that these

form multiple film-like laminae within each sub-millimetre-scale package (Hickman-Lewis et

al. 2016; and cf. Noffke, 2009, 2010; Noffke et al. 2003, 2013). This combination of

morphological features led to their interpretation as a sub-type of microbially induced

sedimentary structure (MISS) (Hickman-Lewis et al. 2016). The data presented show that

μCT scanning is capable of detecting these diffuse structures in three dimensions, but not for

their full horizontal extent as expected from a priori petrographic observation. This limitation

is likely to be because MISS fabrics are only delineated in regions where denser, detrital

sedimentary particles have been entrained. Entrainment of grains is a process permitted by

the presence of glutinous extracellular polymeric substance (EPS) in the original microbial

sediment (cf. Handley & Campbell, 2008), supporting a biological explanation for these

structures’ origin. Carbon is of lower atomic number than silicon, hence the signal we detect

(denser materials in gold colouration in Figs. 5–6) cannot be attributed solely to a

carbonaceous composition. We suggest that detrital grain entrainment must explain their

density. Through pervasive silicification of the stratiform chert (Hickman-Lewis et al., 2016)

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

original MISS fabrics have been partly replaced, obliterated or obfuscated elsewhere during

penecontemporaneous hydrothermal activity (cf. Oberger et al. 2006; Hofmann et al., 2014).

This removes attenuation contrast which might have otherwise allowed the more effective

imaging of multiple layers.

At higher resolution, a particular region of lamination stands out (a close-up of some layers in

this region is provided in Fig. 6). This package of laminae may stand out in comparison to at

least fifteen others known to be present from both petrography (Fig. 5a-c) and macroscopic

observation due to slightly denser grains entrained within these packets. The entrainment of

grains by filamentous laminae is a common biogenicity indicator for MISS (Gerdes et al.

2000; Noffke 2010; Hickman-Lewis et al. 2016). Spatially restricted visualisation of laminae

could also potentially result from the development of nanocrystals or nanoglobules of

biologically generated minerals, mediated by microbes (e.g., Sánchez-Román et al., 2014)

which may have colonised this volcaniclastic sediment.

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

Petrographic study of thin sections (Fig 5a) suggests that the horizontal filamentous structures

we resolve are unlikely to represent microcrystalline quartz veins; though these frequently

cross-cut the stratiform fabrics (see figures in Brasier et al. 2011 and Hickman-Lewis et al.

2016), they do so with orientations typically at a high angle to the filamentous fabric (Fig. 5a-

c). The enhanced density of some of these horizontal filamentous structures when compared

with the chert matrix (quartz), suggests that cross-cutting veins might have acted as conduits

for metalliferous fluids emanating from hydrothermal vents whose metal-rich composition

leaches into, perhaps replacively, filaments, and dictates a higher X-ray attenuation (as

previously hypothesised in sample CCT5). This is observed in further stratiform Apex chert

such as the proximal hydrothermal chert vein N4 of Brasier et al. (2011). The a priori

knowledge of the structures (filamentous MISS-like layers and cross-cutting veins) to be

expected in our samples meant that their identification did not pose a challenge. We note that

in three-dimensional reconnaissance studies without such knowledge, primary linear

microstructures may, at times, be challenging to differentiate from later stage veins. This is

particularly true when — as seen this sample — structures may be incompletely resolved.

Strelley Pool Formation (sample SPV3)

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

CT slices through SPV3, obtained using both the Nikon and the Versa instruments show at

least four different phases. A uniform mid-grey matrix is dominant (Fig. 7a), and a less

dense, dark-grey phase that frequently outlines sub-spherical grains of the mid-grey phase is

also present (Fig. 7a, green arrow). Two denser phases are also detected, one forming large

(up to about 200 μm) angular- to sub-spherical grains (Fig. 7a, yellow arrow), and the other

forming much more variable, and generally smaller grains. The latter sometimes outline

larger mid-grey grains (Fig. 7a, white arrow). These phases can be interpreted through

comparison with thin sections (Fig. 7b-c) and from previous studies (Wacey et al. 2010,

2011a,b). The dominant, uniform mid-grey phase is thus quartz, which occurs as both large

sub-rounded detrital grains, and as pore filling cement in thin sections (Fig. 7b-c). These

occurrences cannot be differentiated here using CT. The less dense phase that often appears

to coat quartz (and other denser grains) is here interpreted as organic material. In thin

sections, this commonly occurs as coatings to detrital grains (Fig. 7c) and clumps towards the

centre of pore-spaces. Organic material also occurs as spheroidal and tubular microfossils

(Wacey et al. 2011a) but our CT data do not have sufficient spatial resolution to differentiate

microfossil morphologies from simple mineral coatings in this sample. The denser phase of

large angular to sub-rounded grains is likely to represent detrital heavy mineral grains such as

chromite, rutile, and zircon observed in SPV3 thin sections (Fig. 7b). We suggest that the

more angular morphologies observed in the CT data correspond to zircon, whilst the more

rounded morphologies are more likely to represent chromite and rutile (cf. Wacey et al.

2011b). Finally, the moderately dense phase occurring as various-sized grains, including

small crystals coating other grains, is interpreted as pyrite. Pyrite has been observed in all

SPV3 thin sections and previous studies have identified both rounded detrital pyrite grains

(Wacey et al. 2011b) and syngenetic crystals formed by bacterial sulphate reduction (Wacey

et al. 2010) in the SPV3 rocks.

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

Renderings of the CT data show the three-dimensional distribution of these four interpreted

phases (organic material in green, pyrite in purple, chromite/rutile/zircon in gold, and quartz

transparent black; Fig. 7d-e, Supplementary movie). These data indicate that such CT scans

would be of utility for future preliminary evaluation of rock samples where one hypothesises

that potential microfossils or other organic material occur in association with heavy mineral

grains, as both phases can easily be resolved in such CT scans. This would allow subsequent

thin sectioning or more destructive analysis to be focussed on regions most likely to yield

microfossils. Furthermore, it would allow targeted cores to be created to study areas high in

organics using CT at higher resolution.

Gunflint Formation, Schreiber Channel (samples SC10 and SC11)

Much microfossil preservation in the Gunflint Formation is through pyritisation, and we note

this in both SC10 and SC11. CT slices through SC10, obtained using the Nikon instrument

(voxel size of 5.8 μm) show a rather uniform mid-grey background, with numerous light grey

grains of rhombic morphology, plus indistinct patches of bright white material (Fig. 8a). Data

acquired at greater spatial resolution using the Versa instrument (voxel size of 1.12 μm)

provide additional details of the internal structure of the rhombic grains and allow the

morphology of some of the bright white material to be resolved (Fig. 9a). Comparison of

figures 8 and 9 demonstrates the additional detail that high resolution scans using the Versa

can provide (see Supplementary movie).

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

As with other samples, objects seen in CT data can be interpreted with reference to features

observed in petrographic thin sections of the same material (Fig. 9b). The uniform mid-grey

background is again probably the microcrystalline quartz matrix, and the rhombic grains are

likely calcium carbonate phases (either calcite or a closely related mineralogy such dolomite).

Frequently, these rhombic grains are imperfectly formed, and appear to be partially replaced

by micro-quartz around their margins (Fig. 9a, green arrow). The patches of bright white

material, and densest phase present, are interpreted as pyrite. These occur in places as a semi-

continuous linear feature (Fig. 8a-b), where they likely represent a partially mineralised vein.

However, much of the pyrite occurs as isolated agglomerations, comprised of superficially

filamentous and spheroidal objects up to c. 20 μm in diameter (Fig. 9a, pink arrow). These

show close morphological similarity to patches of pyritised microfossils previously observed

in thin sections (Fig. 9b) and reported by Wacey et al. (2013).

The lower spatial resolution Nikon CT scanner dataset shows the 3D distribution of carbonate

rhombs in a c. 5mm3 volume of SC 10 (Fig. 8b, blue), plus the patches of pyrite (Fig. 8b,

gold) many in a semi-continuous vein-like structure. Visualisation of a representative part of

the higher spatial resolution Versa CT dataset shows the 3D distribution of numerous patches

of pyritised microfossils (Fig. 9c-d, pink), and several closely associated carbonate rhombs

(Fig. 9c, green). The CT data lack the spatial resolution to clearly resolve the morphology of

individual pyritised microfossils, but the overall morphology of the patches of pyrite closely

resemble clumps of microfossils seen in thin sections (compare Figs. 9b and 9d).

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

CT scanning of sample SC11 revealed similar information. Here, both the Nikon and Versa

data showed clear carbonate rhombs, standing out as lighter grey objects from the uniform

mid-grey micro-quartz background (Fig. 10a, c). The Versa scans also show that most

rhombs have small patches of less dense material within them (Fig. 10a). This could simply

be micro-quartz, but comparison with thin sections of the same material (Fig. 10b) suggest

that it is organic material. Pyritised material is also seen in SC11. This takes the form of sub-

spherical and partially concentrically laminated objects up to c. 600 μm in diameter. With the

aid of thin sections these are interpreted as partially pyritised ooids. Visualisation of the

Nikon dataset demonstrates the 3D distribution of carbonate rhombs (and ooids) in a 14.8

mm3 volume (Fig. 10d), showing they make up a significant proportion of this rock chip.

Since carbonate rhomb growth is known to be detrimental to microfossil preservation this

portion of the rock would be poorly suited as a target for microfossil studies.

468

469

470

471

472

473

474

475

476

477

478

479

480

Aside from fossils preserved in pyrite, there are also instances of carbonaceous preservation

in this material. An additional portion of sample SC10 was scanned using only the Versa

instrument in an attempt detect individual carbonaceous microfossils. CT slices through this

material show a uniform mid-grey background, interpreted to be micro-quartz, plus a small

number of less dense objects (Fig. 11). These objects have a maximum diameter of 10-15

pixels (i.e. 11.2 – 16.8 μm) and persist for between 10-15 successive CT slices. Visualisation

of these shows that they are spheroidal in morphology (Fig. 11c-g), while their ring-like

nature in CT slices (Fig. 11a) shows that they are hollow (quartz filled, based on thin section

observations). These show striking similarity to the microfossil Huroniospora (Fig. 11b) first

described by Barghoorn & Tyler (1965), and thought to be a rather thick-walled spore or

encysted cell (Strother & Tobin 1987). No other potential microfossils are observed in CT

scans of Schreiber Channel material. This is in contrast to thin sections of SC10 material,

where Huroniospora co-occurs with other microfossils, notably the filamentous form

Gunflintia, and both taxa are rather abundant (Fig. 11b; see also Wacey et al. 2012, 2013).

This is attributable to the combination of two factors: (i) limited spatial resolution of the CT

instrumentation, for example, Gunflintia filaments are generally <5 μm in diameter and are

unlikely to be detected in this experiment; (ii) some specimens of Huroniospora have been

noted to have thick dark walls (Strother & Tobin 1987), thicker than other Gunflint

microfossils. As such, their walls occupy more voxels and provide enhanced density contrast

compared to the quartz matrix, and hence present the greatest chance of detection by CT. It is

notable that relatively few Huroniospora are seen in the 1.1 mm3 volume sampled, so it is

likely, even within this one taxon, that CT can only detect the best preserved examples with

the thickest carbonaceous walls.

Gunflint Formation, Mink Mountain (sample G15)

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

Three distinct phases are present in CT scans of sample G15 (Fig. 12). The first is the dark

grey, micro-quartz matrix (least dense) as seen in the other Gunflint samples. The second is a

mid-grey phase (denser) occurring as more diffuse laminations, whilst the third a mid-light

grey phase (densest) occurring as isolated clasts. Based upon the pervasive red colouration of

the hand sample and thin sections (Fig 12a), plus previous observations (Schelble et al. 2004;

Shapiro & Konhauser 2015), the mid-grey denser phase can be interpreted as iron oxide,

possibly with minor iron silicate. Our visualisation of the CT data reveals the 3D distribution

of the iron-rich laminations, which form a lobate dome-like structure, ~8 mm across, with a

depression immediately adjacent (Figs. 12b–d, Supplementary movie). CT also demonstrates

that the largest dense spherical and sub-rounded iron oxide objects are associated with the

non-stromatolite layers (i.e. within micro-quartz inter-layers). consistent with petrographic

observations by Schelble et al. (2004). It is apparent in this putative stromatolite that the

thicknesses of laminae increase slightly toward this crest, a characteristic which has been

proposed as a criterion supportive of biogenicity (Pope & Grotzinger 2000). This morphology

could also suggest that the stromatolite grew in a relatively unstressed environment i.e., one

in which current events and strengths did not impose morphological constraints. Stronger

flow and scour in the traction zones between columns or domes would cause micro-digitate

morphologies to predominate (Bosak et al. 2013). Rather, here we observe an irregular

symmetry within the stromatolite, with low synoptic relief indicative of not having been

driven by random environmental forcing (Kardar et al. 1985; Batchelor et al. 2004, 2005).

This differs from the observations of Shapiro & Konhauser (2015), where digitate

morphologies are described and imaged in contemporaneous Gunflint stromatolites at Mink

Mountain; clearly, stromatolites are very sensitive to changes in environmental conditions,

and morphologies can change over even small distances. An abiotic origin remains a

possibility, since this simple form of stromatolite falls below the degree of complexity

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

generated abiotically in experiments by McLoughlin et al. (2008), and mathematically by

Grotzinger & Knoll (1999). The non-laminated dense objects present solely in the upper

portion of Figs. 12c–d may be Fe-mineralised ooids (cf. Schelble et al. 2004; Shapiro &

Konhauser 2015), or potentially microfossils (cf. Schelble et al. 2004). The spatial resolution

of this dataset is insufficient for a confident interpretation, but the size of these objects

favours the former interpretation. Conducting higher resolution imaging (e.g. through

preparation of very small samples for CT, synchrotron CT, or through FIB-SEM), could shed

further light on the biogenicity of this stromatolite if it were to reveal microfossils, and would

add weight to the promising SEM imagery of putative microbial colonies described by

Schelble et al. (2004).

DISCUSSION

In this study, we present extensive testing of the capabilities of μCT for appraising the

context and nature of Precambrian cellular remains. This discussion elucidates the strengths,

limitations and potential of the technique, highlighting potential applications of μCT for

future research into Precambrian fossiliferous material, the study of which necessitates a

multi-technique approach.

Effects of silicification

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

Silicified sedimentary rocks (cherts) are amongst our most promising windows into the

development of the biosphere throughout the Precambrian. They are almost ubiquitous

throughout the greenstone belts and terranes which preserve the least metamorphosed, and

thus least altered, rocks of the time. Cherts record a large number of significant evolutionary

transitions, potentially from soon after the origins of life itself to the origin of animals:

indeed, they provide geological witnesses for the least well-understood, but pivotal,

geobiological revolutions in Earth history. Silicification of these rocks is extensive, with

many Archaean cherts having SiO2 content exceeding 96-99% (Westall et al. 2015). Two

major Si sources were present in the Precambrian oceans; the first was seawater: Si

concentration was so elevated that cycling of seawater through sediment pore spaces would

naturally encourage rapid and early silicification (van den Boorn et al. 2007). The second was

hydrothermal activity: the Precambrian is generally accepted as having active and widespread

hydrothermal systems. Indeed, detailed fieldwork on many marine Precambrian units has

demonstrated associations with significant hydrothermal action (Paris et al. 1985; Eriksson et

al. 1994; Nijman et al. 1999; Brasier et al. 2002, 2005, 2011; Van Kranendonk et al. 2007,

Westall et al., 2006, 2015). Silicon isotope studies allow the relative contributions of these

siliceous inputs (silicate precursor or chemical origin) to be assessed, and their contributions

seem to vary between formations (André et al. 2006; van den Boorn et al. 2007).

Silicification — regardless of source — makes μCT analysis of such rocks challenging.

Chertification is a double-edged sword: it can markedly enhance the degree of preservation of

sedimentary and biological structures within the rock, but also contributes to their obliteration

or masking (Orberger et al. 2006), an effect that has been termed silica swamping (Lowe &

Knauth 1977). In terms of μCT, pervasive silicification can homogenise the mineralogy, and

thereby minimise X-ray attenuation variation, within rocks, since this variation is dictated by

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

differences in density and chemical composition. Together with limits of spatial resolution,

silicification accounts for the mixed success reported in our experiments.

Insights into X-ray CT as a tool for studying the Precambrian fossil record

We tested the potential of μCT on two putative stromatolites of different ages and

taphonomies: the 3.49 Ga Dresser Formation stromatolite (Fig. 1) is preserved as a complex

mixture of barite, Fe-oxides, hydroxides and sulphides, and silica, whilst the 1.88 Ga Mink

Mountain stromatolite (Fig. 12) is dominantly as silica and iron oxide with ferric iron likely

sourced from oxidised groundwaters (Shapiro & Konhauser 2015). We also tested one

example of a mesoscopic microbially induced sedimentary structure (MISS) sourced from the

stratiform member of the Apex chert at Chinaman Creek, and comprising filamentous

laminations (described by Hickman-Lewis et al. 2016; Figs. 5–6). μCT yielded considerably

more accurate renderings for the three-dimensional morphologies of the stromatolites than of

the MISS. Visualisations demonstrated that the stromatolites are conical-domal forms, the

biogenicity of which cannot be assured without growth models (e.g. Bosak et al. 2013) or

associations of biologically significant elements within promising biogenic morphologies (as

seen in the 3.43 Ga Strelley Pool stromatolites; Lepot et al. 2008; Wacey et al., 2010). This

study is of utility, however: the recognition of multiple discrete stromatolite morphologies in

a single formation bolsters biogenic arguments, particularly when variation can be plausibly

linked to environmental stresses in growth models (Bosak et al. 2013). Since Mink Mountain

stromatolites occur as both domed/conical forms (this study) and digitate forms (Shapiro &

Konhauser 2015), an environmentally-forced biogenic morphology is favourable.

Furthermore, the presence of probable microfossil colonies bound by EPS was noted by

Schelble et al. (2004). There are hence strong arguments for the biogenicity of this

stromatolite.

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

In contrast to the information-rich stromatolite datasets, the Apex chert MISS lacks detail.

The morphology of this filament-like MISS-type structure is poorly represented in

visualisations, which we interpret to result from the ubiquitous, probably hydrothermal,

silicification of the host lithology. This has masked their unique composition and, coupled

with their diffuse nature, hampers their detection in μCT. Furthermore, the coupling between

filament size and resolution makes their scale problematic. Scans of hand specimens lack the

required resolution, as these structures are between 5 µm and 20 µm in diameter (packets

having a thickness of 200-500 µm). Sample preparation for higher resolution scans of

individual filaments is increasingly destructive; to resolve 5µm layers would require a

minimum of two voxels per lamination to avoid issues with noise. Assuming a 2000 x 2000

detector and small enough source spot size, to achieve 2.5 µm voxels the hand specimen

should be 5mm, and thus is likely to require physical reduction in specimen size through

sawing.

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

While the three-dimensional renderings for the stratiform Apex MISS are imperfect, it is

encouraging that even geological materials with relatively minor compositional variation can

provide informative, if limited, results. Namely, these positive results are a petrological

context for the MISS, a spatial concord between structures seen in thin section or hand

specimen and in refined µCT data, and the ability to recognise the small-scale filamentous

nature of the MISS when preservation is advantageous. When tested on other MISS in

Precambrian rocks, the differences between features of interest (MISS) and the surrounding

matrix may be higher, for example: the varied examples of siliciclastic microbial structures

in Noffke et al. (2003, 2013); millimetre-scale roll-ups in Tice & Lowe (2004); and

phototrophic or chemotrophic laminations and clots in Westall et al. (2011, 2015). For these,

µCT may be a considerably more effective three-dimensional imaging technique capable of

resolving features in rock samples at the micron-, millimetre- and centimetre-scales. Where

successful µCT provides significantly more information than permitted by traditional

petrographic microscope observations. Nevertheless, combining the use of CT with

petrographic observations is essential for accurate interpretation of the former, such are the

complexities inherent in designating these biosignatures, especially when occurring in the

absence of their bacterial architects. When uniting petrographic and tomographic information,

geologists should consider:

(i) the two-dimensional fabrics of MISS i.e., whether they are found in packets of structures,

built of plausibly EPS-templated objects (Handley & Campbell 2008);

(ii) whether they pass the three-dimensional characteristics required for biogenicity, i.e. they

are constructed of multiple non-isopachous laminae (Noffke 2009), thicken toward the crests

of individual tufts (Pope & Grotzinger 2000), exhibit demonstrable initial plasticity (e.g. Tice

& Lowe 2004; Hickman-Lewis et al. 2016) and show the entrainment of detrital grains

(Noffke 2010; Noffke et al. 2013, Hickman-Lewis et al., 2016);

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630

631

632

633

634

635

636

637

638

(iii) whether the structures are likely to have enough compositional/density variation to be

mapped using X-ray attenuation, in a lab-based setting, or whether phase-contrast approaches

in a synchrotron are required;

(iv) the scale of the features of interest relative to the size of the hand specimen, and whether

potentially destructive sample preparation will be required to provide the resolution required.

639

640

641

642

643

644

We also tested the utility of µCT on samples of Gunflint and Apex Basalt formations known

to contain, respectively, microfossils and pseudofossils. Using scanners suited to sizes

common for both hand specimens (< 25 cm) and rock chips (0.25–2 cm), we were able to

identify a small proportion of microfossils present in the Gunflint chert, notably those that

had been pyritised or had particularly thick carbonaceous walls. Similarly, a proportion of the

pseudofossils present within sample CHIN03 of the Apex chert could be identified. We note

that further intensive, destructive preparation of known fossils may allow the taxonomy of

microfossils studied to be comprehensively assessed. In doing so, we would forsake the

advantage of X-ray CT being a non-destructive technique. We note, however, that for some

samples, scanning regions-of-interest within large samples at higher resolutions could

overcome these limitations. This relies on having a source strong enough to penetrate the

sample, and selecting scanning geometries where the sample does not collide with the source.

Such a setup can introduce significant artefacts in the form of a bright halo around the edge

of the reconstructed area. However, for samples with minimal material outside the field of

view, or where advanced (i.e. iterative) reconstruction techniques are available, this may hold

promise. The link of sample-size to resolution is also more relaxed within a synchrotron

setting, and in some cases could overcome some of the resolution (and indeed composition)

issues we report herein. Failing this, we advocate analysis of the morphology and chemistry

of microfossils down to the sub-micron-scale, using a combination of focused ion beam

milling and electron microscopy (cf. Wacey et al. 2012): a high resolution form of physical

optical tomography. For microfossils which have restrictions in place preventing destructive

imaging techniques, µCT provides an additional possible stratagem for their appraisal,

enhancing the currently available suite of non-destructive techniques (e.g. laser Raman and

confocal laser scanning microscopy) available to the Precambrian palaeobiologist. One

potential application could be for future analyses of the Apex chert holotype microfossil-like

645

646

647

648

649

650

651

652

653

654

655

656

657

658

659

660

661

662

663

664

665

666

667

668

669

objects described by Schopf (1993), though the dimensions of the thin sections in which they

occur could make CT scanning challenging.

From a broader geological standpoint, we believe that these results clearly demonstrate the

potential of µCT in illuminating of three-dimensional sedimentology. Indeed, geologists have

long been aware that the approach provides an excellent preliminary reconnaissance tool for

the investigation of the interior characteristics of hand samples prior to the preparation of

petrographic thin sections (e.g. Renter 1989). It overcomes the upper spatial range limits of

thin section observation, and facilitates hundreds to thousands of ‘virtual thin sections’

through centimetres of rock, if required. Attenuation contrast allowing, this permits the

spatial relationships of a plenitude of microfabrics to be mapped, up to hand sample scales.

For example, in the Pilbara and Barberton greenstone belts, there are multiple

sedimentological micro-facies reported, including at the hand sample scale (Lowe & Knauth

1977; Orberger et al. 2006; Westall et al. 2011). Whether the density contrasts between

microfabrics in this range of such silicified rocks are sufficient to permit their distinction,

however, remains to be tested. Hence we conclude that provided differences in X-ray

attenuation are sufficient between phases, µCT allows researchers to seek the locations of

structures of interest within rock samples before committing to destructive sampling, which is

particularly valuable where the sample may be irreplaceable.

X-ray CT as a complementary tool for palaeobiological study

670

671

672

673

674

675

676

677

678

679

680

681

682

683

684

685

686

687

688

689

690

There is a rich history in palaeontology of visualising three-dimensional fossils using other

techniques. These are often destructive, for example: serial grinding or sawing, for large

fossils (e.g. Sollas 1903, Briggs et al. 2012); through FIB-based serial sectioning and

SEM/TEM imaging studies for microfossils (e.g. Wacey et al. 2012, Brasier et al. 2015); and

even microtomy in some situations (e.g. Poplin & De Ricqles 1970). We further note that

high-resolution forms of serial sectioning are used in materials science (Uchic 2011), which

have the potential to investigate microfossils, but have not yet been used for this purpose. In

some samples which are not conducive to investigation via lab-based tomography, both

macro- (e.g. Steart et al. 2015) and micro- (e.g. Donoghue et al. 2006) fossils can be scanned

in a synchrotron setting. Here the monochromatic, tunable beam and high flux can allow

techniques such as phase-contrast tomography, which can overcome limited attenuation

contrast in samples, and the use of optics to study small specimens at very high resolutions

(voxels tens of nanometres in size are theoretically possible). A complete overview of the use

and capabilities of these techniques within palaeontology are provided by Sutton et al.

(2014).

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

We believe that the present study demonstrates that lab-based CT is highly complementary to

these other approaches. Serial sectioning can provide greater resolution, and important

information regarding the chemistry and thus taphonomy of fossils, but is destructive.

Synchrotron X-ray tomography is non-destructive, and can improve resolution and contrast –

but can be significantly more difficult to access than lab-scanners. Each technique has

advantages and disadvantages, and information from μCT can be used to plan these

approaches more strategically. For example, were it necessary to acquire focused ion beam

milled samples for very high spatial resolution SEM or TEM analysis, one could utilise X-ray

CT to scan volumes of approximately tens to hundreds of cubic microns in samples, which, in

subsequent renderings, can pinpoint the location of the microfossils (or other objects of

interest) within the sample’s “co-ordinate system”. FIB milling could proceed to intersect and

then mechanically section objects of interest at the correct depth within the thin section or

rock chip; this form of correlative tomography (Burnett et al. 2014) holds promise for

investigation of early cellular microfossils.

CONCLUSIONS

707

708

709

710

711

712

713

714

715

716

717

718

719

720

721

722

723

Here we present a preliminary survey to test the ability of X-ray microtomography (μCT) as

a tool for palaeobiologists to image and identify microfossils, microfossil-like objects,

stromatolites and other microbially induced sedimentary structures (MISS), in their

petrolological context. The success or failure of the CT scans primarily relied on attenuation

contrast within the sample. When this results from biological activity, i.e. it is caused by

either primary organic material, or the precipitation or alteration of (bio)mineral

(nonsedimentary) phases , μCT can delineate structures of Precambrian biological

significance. This is only true of some of the samples studied herein: this preliminary suite of

samples reflects a wide range in taxonomic division, morphology and preservational mode.

The samples also span over 1.5 Ga of Precambrian time, from units that contain controversial

evidence of early life (3.49 Ga Dresser Formation) to those where the presence of life is

certain but the types of life present remains debated (1.88 Ga Gunflint Formation). These

have demonstrated that whilst μCT has limitations for studying samples with Precambrian

biological significance, for some it can serve as a valuable initial means of investigation.

Fundamentally to palaeobiologists, the major advantage of μCT over currently used and

widespread techniques for imaging biological microstructures (e.g. secondary electron

microscopy, transmission electron microscopy, confocal laser scanning microscopy) is its

scope for giving context to the objects identified. It can do this in three dimensions, covering

a greater volume than traditional optical petrography, and provide a far higher resolution than

could ever be attained by hand sample analysis alone. It bridges a key area in the spectrum of

imaging techniques in palaeobiology, and is advantageously coupled with a non-destructive

nature.

724

725

726

727

728

729

730

731

732

733

734

735

736

737

738

739

740

741

742

743

744

745

746

747

Thus, despite the shortcomings identified, we can recommend μCT as a beneficial tool to the

palaeobiologist, particularly for obtaining petrological context from millimetre- to micron-

scale for potentially biological material from the Precambrian. Identifying and characterising

individual microfossils is at the very limits of the technique, but our study encourages its use

in identifying, characterising and digitally cataloguing biologically associated

(biosedimentary) structures, for instance stromatolites, MISS and biominerals. Future

developments in the technique may open up further opportunities for the use of μCT to

analyse these samples including the fossils therein: for example, the development of phase-

contrast-based techniques can improve contrast where attenuation alone remains ineffective

(Sutton et al. 2014). Whilst the technology currently exists to scan at higher resolution than

presented herein, this comes with increasingly challenging sample preparation protocols, the

application of which are increasingly destructive as spatial resolution is improved.

Synchrotron scans can ameliorate some of these issues, where access to the infrastructure is

possible. On the basis of the current study, μCT is currently strongest as a tool for providing

petrological context for microfossils, rather than details of microfossils themselves.

SUPPLEMENTARY INFORMATION

Supplementary information for this paper is available through the Zenodo data archive (DOI:

XXXXXXXXXXX). The datasets archived comprise:

i) A zipped Drishti volume for all CT scans reported in the paper, which can be used for both

3D visualisations and to inspect the underlying data.

ii) A .7z split zip file of one Drishti volume greater than 2GB in size.

iii) A HDMI supplementary movie showing digital visualisations for all of the scans reported

in the paper. 

748

749

750

751

752

753

754

755

756

757

758

759

760

761

762

763

764

765

766

767

768

769

770

771

ACKNOWLEDGMENTS

We acknowledge the facilities, scientific and technical assistance of the Henry Moseley X-

Ray Imaging Facility at the University of Manchester, which was funded in part by the

EPSRC (grants EP/F007906/1, EP/F001452/1 and EP/I02249X/1. DW was funded by the

European Commission and the Australian Research Council (FT140100321). Access to

samples from the Martin Brasier Collection at OUMNH was kindly organised by Eliza

Howlett and Monica Price. We thank Jon Wells at Oxford University Department of Earth

Sciences for preparing the Apex chert samples. Finally, Imran Rahman and an anonymous

reviewer are thanked for their insightful and constructive comments, which significantly

improved the manuscript.

REFERENCES

André, L., Cardinal, D., Alleman, L.Y., & Moorbath, S., 2006. Silicon isotopes in ~3.8 Ga

West Greenland rocks as clues to the Eoarchaean supracrustal Si cycle. Earth and Planetary

Science Letters, 245, 162–173. doi: 10.1016/ j.epsl.2006.02.046.

Anné, J., Garwood, R.J., Lowe, T., Withers, P.J. & Manning, P.L. 2015. Interpreting

pathologies in extant and extinct archosaurs using micro-CT. PeerJ, 3, e1130, doi:

10.7717/peerj.1130.

Barghoorn, E.S. & Tyler, S.A. 1965. Microorganisms from the Gunflint Chert.

Science, 147, 563-577.

772

773

774

775

776

777

778

779

780

781

782

783

784

785

786

787

788

789

790

791

792

793

794

795

Batchelor, M., Burne, R., Henry, B. & Jackson, M., 2004. A case for biotic morphogenesis of

coniform stromatolites. Journal of Physics A, 337, 319–26.

Batchelor, M., Burne, R., Henry, B. & Slatyer, T., 2005. Statistical physics and stromatolite

growth: new perspectives on an ancient dilemma. Journal of Physics A, 350, 6–11.

Baumgartner-Mora, C., Baumgartner, P.O. & Baumgartner, L. 2006. 3D-imaging of

foraminifera by X-ray microtomography. Anuário do Instituto de Geociências, 29, 323–324.

Bosak, T., Knoll, A.H., Petroff, A.P., 2013. The meaning of stromatolites. Annual Review of

Earth and Planetary Sciences, 41, 21–44.

Brasier, M.D., Green, O.R., Jephcoat, A.P., Kleppe, A.K., Van Kranendonk, M.J., Lindsay,

J.F., Steele, A., Grassineau, N.V., 2002. Questioning the evidence for Earth’s oldest fossils.

Nature, 416, 78–81.

Brasier, M.D., Green, O.R., Lindsay, J.F., McLoughlin, N., Steele, A. & Stoakes, C. 2005.

Critical testing of Earth’s oldest putative fossil assemblage from the 3.5Ga Apex chert,

Chinaman Creek, Western Australia. Precambrian Research, 140, 55-102.

Brasier, M.D., Green, O.R., Lindsay, J.F., McLoughlin, N., Stoakes, C., Brasier, A.T. &

Wacey, D. 2011. Geology and putative microfossil assemblage of the c. 3460 Ma ‘Apex

chert’, Chinaman Creek, Western Australia – a field and petrographic guide. Geological

Survey of Western Australia, Perth.

796

797

798

799

800

801

802

803

804

805

806

807

808

809

810

811

812

813

814

815

816

817

818

819

820

Brasier, M., Antcliffe, J., Saunders, M., & Wacey, D. 2015. Changing the picture of Earth’s

earliest fossils (3.5-1.9 Ga) with new approaches and new discoveries. Proceedings of the

National Academy of Sciences USA, 112, 4859-4864.

Briggs, D.E.G., Siveter, D.J., Siveter, D.J., Sutton, M.D., Garwood, R.J., Legg, D.A., 2012.

Silurian horseshoe crab illuminates the evolution of arthropod limbs. Proceedings of the

National Academy of Sciences of the United States of America, 109, 15702–15705, doi:

10.1073/pnas.1205875109.

Burnett, T.L., McDonald, S.A., Gholinia, A., Geurts, R., Janus, M., Slater, T., Haigh, S.J.,

Ornek, C., Almuaili, F., Engelberg, D.L., Thompson, G.E., Withers, P.J., 2014. Correlative

Tomography. Scientific reports 4, 4711. doi:10.1038/srep04711

Butler, R.J., Porro, L.B. & Norman, D.B. 2008. A juvenile skull of the primitive ornithischian

dinosaur Heterodontosaurus tucki from the ‘Stormberg’ of southern Africa. Journal of

Vertebrate Paleontology, 28, 702–711, doi: 10.1671/0272-

4634(2008)28[702:AJSOTP]2.0.CO;2.

Cunningham, J.A., Vargas, K., et al. 2015. Critical appraisal of tubular putative eumetazoans

from the Ediacaran Weng’an Doushantuo biota. Proceedings of the Royal Society of London.

Series B: Biological Sciences, 282, 20151169, doi: 10.1098/rspb.2015.1169.

821

822

823

824

825

826

827

828

829

830

831

832

833

834

835

836

837

838

839

840

841

842

El Albani, A., Bengtson, S., Canfield, D.E., Bekker, A., Macchiarelli, R.,Mazurier, A.,

Hammarlund, E.U., Boulvais, P., Dupuy, J.-J., Fontaine, C., Fürsich, F.T., Gauthier-Lafaye,

F., Janvier, P., Javaux, E., Ossa Ossa, F., Pierson-Wickmann, A.-C., Riboulleau, A., Sardini,

P., Vachard, D., Whitehouse, M., Meunier, A., 2010. Large colonial organisms with

coordinated growth in oxygenated environments 2.1 Gyr ago. Nature, 466, 100-104.

Eriksson, K.A., Krapez, B. & Fralick, P.W., 1994. Sedimentology of Archean greenstone

belts: Signatures of tectonic evolution. Earth-Science Reviews, 37, 1-88.

Garwood, R.J. & Dunlop, J. 2014. Three-dimensional reconstruction and the phylogeny of

extinct chelicerate orders. PeerJ, 2, e641, doi: 10.7717/peerj.641.

Garwood, R.J., Rahman, I.A. & Sutton, M.D. 2010. From clergymen to computers—the

advent of virtual palaeontology. Geology Today, 26, 96–100.

Gerdes, G., Klenke, T. & Noffke, N., 2000. Microbial signatures in peritidal siliciclastic

sediments: a catalogue. Sedimentology, 47, 279¬¬–308.

Grotzinger, J.P., Knoll, A.H., 1999. Stromatolites in Precambrian carbonates; evolutionary

mileposts or environmental dipsticks? Annual Review of Earth and Planetary Science Letters,

27, 313– 358.

843

844

845

846

847

848

849

850

851

852

853

854

855

856

857

858

859

860

861

862

863

864

Handley, K.M., & Campbell, K.A., 2008. Character, Analysis and Preservation of

Biogenicity in Terrestrial Siliceous Stromatolites from Geothermal Settings. In V.C. Tewari

and J. Seckbach (eds.), Stromatolites: Interaction of Microbes with Sediments, Cellular

Origin, Life in Extreme Habitats and Astrobiology 18, 359–381 DOI 10.1007/978-94-007-

0397-1_16. Springer.

Hickman-Lewis, K., 2015. A re-assessment of the nature of the sedimentary environment of

the ‘Apex chert’ and a critical analysis of structures of putative biogenic origin within.

Master of Earth Sciences (M.ESci) thesis, Oxford University.

Hickman-Lewis, K., Garwood, R.J., Brasier, M.D., Goral, T., Jiang, H., McLoughlin, N. &

Wacey, D. 2016. Carbonaceous microstructures from sedimentary laminated chert within the

3.46 Ga Apex Basalt, Chinaman Creek locality, Pilbara, Western Australia. Precambrian

Research, 278, 161-178.

Hofmann, A., Bolhar, R., Orberger, B., Foucher, F., 2014. Cherts of the Barberton

Greenstone Belt, South Africa: Petrology and trace element geochemistry of 3.5 to 3.3 Ga old

silicified volcaniclastic sediments. South African Journal of Geology, 116, 297–322.

Kardar, M., Parisi, G. & Zhang, Y., 1986. Dynamic scaling of growing interfaces. Physical

Review Letters, 56, 889–92.

Lepot, K., Benzerara, K., Brown, G.E., Philippot, P., 2008. Microbially influenced formation

of 2,724-million-year-old stromatolites. Nature Geoscience, 1, 118–121.

865

866

867

868

869

870

871

872

873

874

875

876

877

878

879

880

881

882

883

884

885

886

887

888

889

Limaye, A. 2012. Drishti: a volume exploration and presentation tool. Proc. SPIE 8506,

Developments in X-Ray Tomography VIII, 85060X. doi: 10.1117/12.935640.

Lowe, D.R., 1994. Abiological origin of described stromatolites older than 3.2 Ga. Geology,

22, 387–90.

Lowe, D.R. & Knauth, L.P., 1977. Sedimentology of the Onverwacht Group (3.4 billion

years), Transvaal, South Africa, and its bearing on the characteristics and evolution of the

early Earth. Journal of Geology, 85, 699–723.

McLoughlin, N., Wilson, L.A., Brasier, M.D., 2008. Growth of synthetic stromatolites and

wrinkle structures in the absence of microbes – implications for the early fossil record.

Geobiology, 6, 95–105.

Neenana, J.M. & Scheyer, T.M. 2012. The braincase and inner ear of Placodus gigas

(Sauropterygia, Placodontia)—a new reconstruction based on micro-computed tomographic

data. Journal of Vertebrate Paleontology, 32, 1350–1357, doi:

10.1080/02724634.2012.695241.

Nijman, W., de Bruijne, K.C.H. &Valkering, M.E., 1999. Growth fault control of Early

Archaean cherts, barite mounds and chert-barite veins, North Pole Dome, Eastern Pilbara,

Western Australia. Precambrian Research, 88, 25-52.

890

891

892

893

894

895

896

897

898

899

900

901

902

903

904

905

906

907

908

909

910

911

912

Noffke, N., 2009. The criteria for the biogeneicity of microbially induced sedimentary

structures (MISS) in Archean and younger, sandy deposits. Earth-Science Reviews, 96, 173–

180

Noffke, N., 2010. Microbial Mats in Sandy Deposits from the Archean Era to Today.

Heidelberg: Springer Verlag.

Noffke, N., Hazen, R.M. & Nhleko, N., 2003. Earth’s earliest microbial mats in a siliciclastic

marine environment (2.9 Ga Mozaan group, South Africa). Geology, 31, 673–676.

Noffke, N., Christian, D., Wacey, D. & Hazen, R.M. 2013. Microbially induced sedimentary

structures recording an ancient ecosystem in the ca. 3.48 billion-year-old Dresser Formation,

Pilbara, Western Australia. Astrobiology, 13, 1103-1124.

Orberger, B., Rouchon, V., Westall, F., de Vries., S.T., Wagner, C., Pinti, D.L., 2006.

Protoliths and micro-environments of some Archean Cherts (Pilbara, Australia). Geological

Society of America Special Paper 405, 133–152.

Paris, I., Stanistreet, I.G., Hughes, M.J., 1985. Cherts of the barberton greenstone belt

interpreted as products of submarine exhalative activity. Journal of Geology, 93, 111–129.

Pope, M.C., Grotzinger, J.P., 2000. Controls on fabric development and morphology

of tufa and stromatolites, uppermost Pethei Group (1.8 Ga), Great Slave Lake, northwest

Canada. In: Grotzinger, J., James, N. (Eds.), Carbonate Sedimentation and Diagenesis in the

Evolving Precambrian World. SEPM Special Publication, pp. 103–122.

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929

930

931

932

933

934

935

936

937

Renter, J.A., 1989. Applications of computerized tomography in sedimentology. Marine

Georesources & Geotechnology, 8(3), pp.201-211.

Sánchez-Román, M., Fernández-Remolar, D., Amils, R., Sánchez-Navas, A., Schmid, T.,

Martin-Uriz, P.S., Rodríeguez, N., McKenzie, J.A., Vasconcelos, C., 2014. Microbial

mediated formation of Fe-carbonate minerals under extreme acidic conditions. Scientific

Reports, 4, Article number: 4767. doi:10.1038/srep04767, 33, 1268-1273.

Schelble, R.T., Westall, F. & Allen, C.C. 2004. ~1.8 Ga iron-mineralized microbiota from the

Gunflint Iron Formation, Ontario, Canada: implications for Mars. Advances in Space

Research.

Schopf, J.W. 1993. Microfossils of the Early Archean Apex Chert: New Evidence of the

Antiquity of Life. Science, 260, 640-646.

Shapiro, R.S., Konhauser, K.O., Hematite-coated microfossils: primary ecological fingerprint

or taphonomic oddity of the Paleoproterozoic? Geobiology, 13, 209–224.

Sollas, W.J., 1903. A method for the investigation of fossils by serial sections. Philosophical

Transactions of the Royal Society of London. Series B, Biological Sciences, 196, 259–265.

938

939

940

941

942

943

944

945

946

947

948

949

950

951

952

953

954

955

956

957

958

959

Spencer, A.R.T., Hilton, J. & Sutton, M.D. 2012. Combined methodologies for three-

dimensional reconstruction of fossil plants preserved in siderite nodules: Stephanospermum

braidwoodensis nov. sp. (Medullosales) from the Mazon Creek lagerstätte. Review of

Palaeobotany and Palynology, 188, 1–17, doi: 10.1016/j.revpalbo.2012.09.001.

Steart, D.C., Spencer, A.R.T., Garwood, R.J., Hilton, J., Munt, M.C., Needham, J. & Kenrick,

P. 2014. X-ray Synchrotron Microtomography of a silicified Jurassic Cheirolepidiaceae

(Conifer) cone: histology and morphology of Pararaucaria collinsonae sp. nov. PeerJ, 2, e624,

doi: 10.7717/peerj.624.

Streng, M., Butler, A.D., Peel, J.S., Garwood, R.J. & Caron, J.-B. 2016. A new family of

Cambrian rhynchonelliformean brachiopods (Order Naukatida) with an aberrant coral-like

morphology. Palaeontology, 59, 269–293, doi: 10.1111/pala.12226.

Strother, P.K. & Tobin, K., 1987. Observations on the genus Huroniospora Barghoorn:

Implications for paleoecology of the Gunflint Chert. Precambrian Research, 36, 323-333.

Strother, P.K., Battison, L., Brasier, M.D. & Wellman, C.H., Earth’s earliest non-marine

eukaryotes. Nature, 473, 505–509.

Sutton, M.D., Garwood, R.J., Siveter, D.J. & Siveter, D.J. 2012. Spiers and VAXML: A

software toolkit for tomographic visualisation, and a format for virtual specimen interchange.

Palaeontologia Electronica, 15, 14p.

960

961

962

963

964

965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

Sutton, M.D., Rahman, I.A. & Garwood, R.J. 2014. Techniques for Virtual Palaeontology.

Wiley.

Tice, M.M., Lowe, D.R., 2004. Photosynthetic microbial mats in the 3416-Myr-old ocean.

Nature, 431, 549–552.

Tyler, S.A., Barghoorn, E.S., 1954. Occurrence of structurally preserved plants in

Precambrian rocks of the Canadian Shield. Science, 119, 606–608.

Uchic, M.D., 2011. Serial Sectioning Methods for Generating 3D Characterization Data of

Grain- and Precipitate-Scale Microstructures. In: Ghosh, S. & Dimiduk, D. (eds)

Computational Methods for Microstructure-Property Relationships. Springer US, 31–52.

van den Boorn, S.H.J.M., van Bergen, M.J., Nijman, W. & Vroon, P.Z., 2007. Dual role of

seawater and hydrothermal fluids in Early Archean chert formation: Evidence from silicon

isotopes. Geology, 35, 939-942.

Van Kranendonk, M.J. 2006. Volcanic degassing, hydrothermal circulation and the

flourishing of early life on Earth: A review of the evidence from c. 3490-3240 Ma rocks of

the Pilbara Supergroup, Pilbara Craton, Western Australia. Earth-Science Reviews, 74, 197-

240.

984

985

986

987

988

989

990

991

992

993

994

995

996

997

998

999

1000

1001

1002

1003

1004

1005

Van Kranendonk, M.J., Smithies, R.H., Hickman, A.H. & Champion, D.C., 2007.

Paleoarchean development of a continental nucleus: the East Pilbara Terrane of the Pilbara

Craton. In: Van Kranendonk, M.J., Smithies, R.H., Bennett, V.C. (Eds.), Earth’s Oldest

Rocks. Developments in Precambrian Geology, vol. 15. Elsevier, Amsterdam, The

Netherlands, pp. 307–337.

Wacey, D. 2011. Early Life on Earth: A Practical Guide. Springer.

Wacey, D., McLoughlin, N., Whitehouse, M.J. & Kilburn, M.R. (2010) Two co-existing

sulfur metabolisms in a ca. 3,400 Ma sandstone. Geology, 38, 1115-1118.

Wacey, D., Kilburn, M.R., Saunders, M., Cliff, J., & Brasier, M.D. 2011a. Microfossils of

sulfur metabolizing cells in ~3.4 billion year old rocks of Western Australia. Nature

Geoscience, 4, 698-702.

Wacey, D., Saunders, M., Brasier, M.D. & Kilburn, M.R. 2011b. Earliest microbially

mediated pyrite oxidation in ~3.4 billion-year-old sediments. Earth and Planetary Science

Letters, 301, 393-402.

Wacey, D., Menon, S., Green, L., Gerstmann, D., Kong, C., McLoughlin, N., Saunders, M. &

Brasier, M.D. 2012. Taphonomy of very ancient microfossils from the ~3400 Ma Strelley

Pool Formation and ~1900 Ma Gunflint Formation: new insights using focused ion beam.

Precambrian Research, 220-221, 234-250.

1006

1007

1008

1009

1010

1011

1012

1013

1014

1015

1016

1017

1018

1019

1020

1021

1022

1023

1024

1025

1026

1027

1028

1029

Wacey, D., Mcloughlin, N., Kilburn, M.R., Saunders, M., Cliff, J., Kong, C., Barley, M.E. &

Brasier, M.D. 2013. Nano-scale analysis reveals differential heterotrophic consumption in the

~1.9 Ga Gunflint Chert. Proceedings of the National Academy of Sciences USA, 110, 8020-

8024.

Wacey, D., Saunders, M., Kong, C., Brasier, A.T. & Brasier, M.D. 2015. 3.46 Ga Apex chert

‘microfossils’ reinterpreted as mineral artefacts produced during phyllosilicate exfoliation.

Gondwana Research, doi:10.1016/j.gr.2015.07.010.

Walter, M.R., Buick, R. & Dunlop, J.S.R. 1980. Stromatolites 3,400-3,500 Myr old from the

North Pole area, Western Australia. Nature, 284, 443-445.

Westall, F., de Ronde, C.E.J., Southam, G., Grassineau, N., Colas, M., Cockell, C., Lammer,

H., 2006. Implications of a 3.472–3.333 Gyr-old subaerial microbial mat from the Barberton

greenstone belt, South Africa for the UV environmental conditions on the early Earth.

Philosophical Transactions of the Royal Society B, 361, 1857-1875.

Westall, F., Foucher, F., Cavalazzi, B., de Vries, S., Nijman, W., Pearson, V., Watson, J.,

Verchovsky, A., Wright, I., Rouzaud, J.-N., Marchesini, D. & Anne, S., 2011. Volcaniclastic

habitats for early life on Earth and Mars: a case study from ~3.5 Ga-old rocks from the

Pilbara, Australia. Planetary and Space Science, 59, 1093–1106.

Westall, F., Campbell, K.A., Bréhéret, J.G., Foucher, F., Gautret, P., Hubert, A., Sorieul, S.,

Grassineau, N. & Guido, D., 2015. Archean (3.33 Ga) microbe-sediment systems were

diverse and flourished in a hydrothermal context. Geology, doi:10.1130/G36646.1

1030

1031

1032

1033

1034

1035

1036

1037

1038

1039

1040

1041

1042

1043

1044

1045

1046

1047

1048

1049

1050

1051

1052

1053

1054

FIGURE CAPTIONS

Fig 1. Imagery of the morphology of the Dresser Formation stromatolite DR1. (a–d)

Photographs depicting the sample DR1 from four orientations. (e) A single CT tomogram for

the slice indicated by the red line in (d); white colouration (gold arrow) represents higher-

density material, lower density material is in dark grey (grey arrow) and light grey (white

arrow). (f) Three-dimensional rendering of sample DR1, in which low-density phases are

identified in grey and white. (g) Composite rendering of the DR1 stromatolite, portraying the

spatial relationship between higher- and lower-density phases; higher-density phases seem to

preferentially occur at the tops of low-density layers. Arrows correspond to those in (e). (h)

Three-dimensional rendering of the higher-density phases, indicated by gold. (i) Three-

dimensional rendering of the CT data viewed from a different orientation, emphasising the

three dimensional nature of the domed layering.

1055

1056

1057

1058

1059

1060

1061

1062

1063

1064

1065

1066

1067

1068

1069

1070

1071

Fig 2. Lithological fabrics and microstructures of a hydrothermal chert vein (N1 of Brasier et

al., 2011) through the Apex Basalt at Chinaman Creek. (a) Thin section scan of CHIN3

giving context for the microstructures investigated herein. (b) Rendering of CT data, using

Drishti, of the three-dimensional texture of this lithology. Key features highlighted include a

denser, sub-rounded grain phase (gold) and less dense green phases, including elongate

objects (indicated by arrows in (a) and (b)), which may represent vein-like features created by

the infiltration of hydrothermal fluids. (c) Single CT tomogram, in which are indicated a

microfossil-like object (yellow arrow) and a persistent vein (green arrow), both discernible

through their density contrast with the microcrystalline matrix. (d) A similar field of view to

(c) in the corresponding thin section, again with microfossil-like object and vein arrowed. (e)

Three-dimensional rendering, using SPIERS, of the spatial distribution of microfossil-like

objects, now interpreted as altered phyllosilicates (cf. Wacey et al. 2015). (f) Close-up view

of six such microfossil-like objects, which depict a vermiform morphology. CT does not have

the spatial resolution to decode the fine details of these objects.

Fig 3. Sedimentological characteristics of stratiform Apex chert sample CCT5, an iron-rich

banded chert. (a) Thin section scan of CCT5 showing microfacies as follows: (1) dull brown,

structureless material, with poor delineation of laminations by small opaque grains,

interbedded with volcanogenic sediments (layers of larger ash clasts); (2) a stylolite-bound

poorly laminated microfacies; (3) a dark, iron-rich, well-laminated layer; and (4) a

volcaniclastic fall deposit layer. The Fe-stained region, into which a vein intrudes (see text

for description), is indicated by a yellow arrow. (b) CT rendering of the adjacent rock chip,

depicting most clearly the laminar layer (iii) (outlined) and the major asymptotic vein (green

arrow). Red arrows indicate a scanning artefact.

1072

1073

1074

1075

1076

1077

1078

1079

1080

1081

1082

1083

1084

1085

1086

1087

1088

1089

1090

1091

1092

1093

1094

1095

1096

Fig 4. Petrography and CT scanning of microclastic chert sample CC43. (a) Thin section scan

of CC43, in which the throughgoing layer characterised by pale, coarse sand-sized grains is

central. (b–c) Photomicrographs of these large grains demonstrate that they are often (~40%

of total grains) irregularly laminated. These laminations are non-isopachous and traverse only

the extent of the grain in which they are found. (d) CT scan for the rock chip from which the

thin section was cut. Note that although the layer of coarser grains is detected throughout the

sample, the internal lamination of grains is not seen in any instance.

Fig 5. Laminated textures within a microclastic chert from the North Block of the Apex chert

at Chinaman Creek (sample CC164). (a) Thin section scan (30 µm-thick section) showing

numerous feint filamentous laminations horizontally traversing the strata; these often occur in

small packages (e.g. boxed area). (b-c) Photomicrographs demonstrating the morphology of

these persistent laminae: they are diffuse, undulate and frequently entrain relict sediment

grains. In (c), laminations are highlighted and arrows indicate the entrained grains. (d)

Rendering of CT data depicting an area interpreted to be similar to that shown in (b). A cross-

cutting microquartz vein is clearly detected, while the two lamination packets are shown as

somewhat laterally extensive, though not to the spatial persistence observed in the thin

section. The rock chip scanned is the offcut from the thin section shown in (a) so one might

expect a higher abundance of packets of laminations to have been detected in the volume

rendered.

1097

1098

1099

1100

1101

1102

1103

1104

1105

1106

1107

1108

1109

1110

1111

1112

1113

1114

1115

1116

1117

Fig 6. Laminations from sample CC164, as also imaged in Fig. 5. (a-b) magnified views of

the packages of laminations rendered in Fig. 5d. The string-like, filamentous nature of

individual laminations is quite clear. We suggest that carbonaceous filaments stand out from

the matrix due to the entrainment of grains composed of heavier elements by EPS in the

original interpreted microbial mat environment. In both images, arrows indicate individual

filaments.

Fig 7. Petrological context of organic material and potential microfossils from the 3.43 Ga

Strelley Pool Formation (sample SPV3). (a) Single CT tomogram demonstrating the different

phases present, including quartz (dominant mid-grey matrix), darker grey organic material

(e.g. green arrow), bright white heavy minerals such as zircon (yellow arrow) and light grey

possible pyrite (white arrow). (b) Thin section photomicrograph of sample SPV3 showing

detrital heavy mineral grains. (c) Thin section photomicrograph of sample SPV3 showing

organic material coating detrital quartz grains and in pore spaces. (d) Rendering of the CT

data showing the interpreted distribution of heavy detrital minerals (gold), pyrite (purple) and

organic material (green) within a dominantly quartz matrix. (e) Higher resolution rendering

of CT data showing, in particular, the way organic material frequently coats rounded detrital

quartz and heavy mineral grains.

Fig 8. Petrological context of sample SC10 from the Schreiber locality of the 1.88 Ga

Gunflint chert. (a) Single CT tomogram showing carbonate rhombs (blue arrow) and pyrite

(yellow arrow) picked out as denser phases than the mid-grey microquartz matrix. (b)

Rendering of the CT dataset using Drishti showing the three-dimensional distribution of

carbonate rhombs (blue) and pyrite (gold).

1118

1119

1120

1121

1122

1123

1124

1125

1126

1127

1128

1129

1130

1131

1132

1133

1134

1135

1136

1137

1138

1139

1140

1141

1142

Fig 9. Higher spatial resolution analysis of sample SC10. (a) Single CT tomogram showing a

carbonate rhomb (green arrow) and clusters of pyritised objects (pink arrow). (b) Light

microscopy image of an area interpreted to be similar to that shown in the CT tomogram,

comprising part of a carbonate rhomb (bottom left) and a cluster of pyritised Gunflintia and

Huroniospora microfossils. (c) Rendering of the CT data using SPIERS showing three

dimensional organisation of carbonate rhombs (green) and pyritised objects interpreted as

microfossils. (d) Rendering of a single patch of pyritised objects. These show close similarity

to patches of pyritsied microfossils observed in thin sections, but this CT system does not

have the spatial resolution required to distinguish individual filaments and coccoids.

Fig 10. Analysis of sample SC11 from the Schreiber locality of the 1.88 Ga Gunflint

Formation. (a) Part of a single CT tomogram showing higher density carbonate rhombs light

grey) in a uniform microcrystalline quartz matrix (mid-grey). The speckled nature of the

interior of the carbonate rhombs is clearly visible. (b) Light microscopy image of a thin

section from the same sample, showing carbonate rhombs with inclusions of organic material,

interpreted to be comparable to those imaged in part (a). (c) Single CT tomogram showing

the carbonate rhombs (blue arrows) and rare pyritised objects interpreted as ooids (yellow

arrow). (d) Rendering of the CT dataset using Drishti showing abundant carbonate rhombs

(purple) in the studied volume of an otherwise quartz-rich rock. A denser phase, probably

pyrite, is represented in grey-gold.

1143

1144

1145

1146

1147

1148

1149

1150

1151

1152

1153

1154

1155

1156

1157

1158

1159

1160

1161

1162

1163

Fig 11. Analysis of an area of sample SC10 (Gunflint Fm.) containing carbonaceous

microfossils. (a) Single CT tomogram showing a low density ring shaped object (yellow

arrow), interpreted as a hollow carbonaceous microfossil (cf. Huroniospora) within a rather

uniform microquartz matrix. (b) Light microscopy image of a thin section from the same

sample showing three examples of spheroidal Huroniospora microfossils. (c) Rendering of

the CT data using SPIERS showing approximately 20 spheroidal objects interpreted to be

carbonaceous microfossils (cf. Huroniospora). (d-g) Four views of a single rendered

microfossil with parts (e) and (f) sectioned to show their hollow nature.

Fig 12. Imagery of the morphology of a stromatolite from the Mink Mountain locality of the

Gunflint Chert (sample G15). (a) Representative thin section of the stromatolite. (b) Single

CT tomogram depicting high- and low-density linear features which correspond to the

stromatolitic layers and inter-layers of the sample. In the lower portion of the image, several

high-density features are apparent, which we attribute to the constituent metalliferous grains

(haematite and pyrite, cf. Shaprio & Konhauser 2015). (c) Three-dimensional rendering of

the sample showing and overview of the low- and high-density lamination phases. (d) Three-

dimensional rendering of the sample showing the highest-density phases present (gold),

which are concentrated to the uppermost region of the sample. (e) Whole-sample rendering

depicting all features which were detected by CT: the outer surface of the sample is rendered

in grey, high-density laminations in purple, and the highest-density metalliferous grains in

gold. Note the association of metalliferous grains with the inter-layers of the stromatolite.

1164

1165

1166

1167

1168

1169

1170

1171

1172

1173

1174

1175

1176

1177

1178

1179

1180

1181

1182

1183

1184

1185