Quantitative analysis of stylolite networks in different platform carbonate facies

Journal Pre-proof Quantitative analysis of stylolite networks in different platform carbonate facies Elliot Humphrey, Enrique Gomez-Rivas, Joyce Neilson, Juan Diego Martín-Martín, David Healy, Shuqing Yao, Paul D. Bons PII:

S0264-8172(19)30657-9

DOI:

https://doi.org/10.1016/j.marpetgeo.2019.104203

Reference:

JMPG 104203

To appear in:

Marine and Petroleum Geology

Received Date: 26 June 2019 Revised Date:

19 December 2019

Accepted Date: 20 December 2019

Please cite this article as: Humphrey, E., Gomez-Rivas, E., Neilson, J., Martín-Martín, J.D., Healy, D., Yao, S., Bons, P.D., Quantitative analysis of stylolite networks in different platform carbonate facies, Marine and Petroleum Geology (2020), doi: https://doi.org/10.1016/j.marpetgeo.2019.104203. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Ltd.

Elliot Humphrey: Conceptualisation, Methodology, Formal analysis, Resources, Data curation, Writing (Original Draft), Enrique Gomez-Rivas: Writing (Review & Editing), Supervision, Data curation, Joyce Neilson: Writing (Review & Editing), Supervision, Juan Diego Martín-Martín: Writing (Review & Editing), Supervision, David Healy: Writing (Review & Editing), Data curation, Shuqing Yao: Writing (Review & Editing), Resources, , Paul D. Bons: Writing (Review & Editing), Data curation.

1

Quantitative analysis of stylolite networks in different platform carbonate

2

facies

3 4

Elliot Humphrey1, Enrique Gomez-Rivas2,1, Joyce Neilson1, Juan Diego Martín-Martín2,

5

David Healy1, Shuqing Yao1 and Paul D. Bons3,4

6

(1) School of Geosciences, King’s College, University of Aberdeen, AB24 3UE Aberdeen, United Kingdom

7

(2) Departament de Mineralogia, Petrologia i Geologia Aplicada, Facultat de Ciències de la Terra,

8

Universidad de Barcelona (UB), Martí i Franquès s/n, 08028 Barcelona, Spain

9

(3) China University of Geosciences (Beijing), Xueyuan Road 29, Haidian district, 100083 Beijing, China

10

(4) Department of Geosciences, Tübingen University, Wilhelmstr. 56, 72074 Tübingen, Germany

11 12 13

Corresponding author: Humphrey, E.: Department of Geology and Petroleum Geology, School of Geosciences,

14

University of Aberdeen, Meston Building, King’s College, Aberdeen AB24 3UE, Scotland, United Kingdom; +

15

44 (0) 1224 273915; [email protected]

16 17

Keywords: Stylolite, Carbonate, Diagenesis, Lithofacies, Mechanical stratigraphy,

18

Maestrat Basin.

19 20

Abstract

21

Stylolites are rough surfaces that form by pressure solution, and present variable geometries

22

and spatial distributions. Despite being ubiquitous in carbonate rocks and potentially

23

influencing fluid flow, it is not yet clear how the type and distribution of stylolite networks

24

relate to lithofacies. This study investigates Lower Cretaceous platform carbonates in the

25

Benicàssim area (Maestrat Basin, Spain) to statistically characterize stylolite morphology and 1

26

stylolite network distributions in a selection of typical shallow-marine carbonate lithofacies,

27

from mudstones to grainstones. Bedding-parallel stylolite networks were sampled in the field

28

to quantify stylolite spacing, wavelength, amplitude, intersection morphology and

29

connectivity. Grain size, sorting and composition were found to be the key lithological

30

variables responsible for the development of rough anastomosing stylolite networks. Poorly-

31

connected stylolites with large vertical spacings were found to be dominant in grain-

32

supported lithofacies, where grains are fine and well sorted. Anastomosing stylolite networks

33

appear well developed in mud-supported lithofacies with poorly-sorted clasts that are both

34

heterogenous in size and composition. Mud-supported facies feature stylolites that are closely

35

spaced, have high amplitudes and intersection densities, and predominantly present suture

36

and sharp-peak type morphologies. Larger grains and poor sorting favour the formation of

37

stylolites with small vertical spacings, low wavelengths and high amplitudes. This statistical

38

analysis approach requires only limited information, such as that from drill core, and can be

39

used to characterise stylolite morphology and distributions in subsurface carbonate reservoirs.

40 41

1. Introduction

42 43

Carbonate reservoirs are notoriously complex to develop for hydrocarbon production due to

44

strong spatial and temporal variations in petrophysical properties which, when combined with

45

the presence of sub seismic-scale structures, inevitably impact reservoir quality (e.g., Agar

46

and Geiger, 2015). Sub-seismic-scale structures in carbonate rocks, such as fractures and

47

stylolites, can have profound impacts on the bulk rock permeability and fluid flow pathways

48

(Burgess and Peter, 1985; Guerriero et al., 2013; Larsen et al., 2010), requiring the use of

49

predictive guides to help understand their influence. Previous work on the characterisation of

50

sub seismic-scale structures in carbonate reservoirs that influence fluid flow has mostly 2

51

focused on fracture networks (e.g., Long and Witherspoon, 1985; Nelson, 2001; Di-Cuia et

52

al., 2005; Guerriero et al., 2013; Haines et al., 2016). Identifying the sedimentological and

53

diagenetic controls on mechanical stratigraphy (Laubach et al., 2009) enables predictions to

54

be made about fracture distribution based on lithofacies.

55 56

Stylolites are irregular dissolution surfaces with multiscale roughness created by

57

intergranular pressure solution (Koehn et al. 2007; Ebner et al. 2010). These dissolution

58

seams are primarily associated with strain localisation and form thin lateral drapes (Heap et

59

al. 2014), which commonly host relatively insoluble particles such as clay minerals, oxides,

60

organic matter, etc. (Nelson, 1981; Railsback, 1993; Ben-Itzhak et al., 2014). High solubility

61

and fast reaction kinetics make carbonate rocks the predominant lithology susceptible to

62

pressure-solution, although stylolites have been found to develop in sandstones (Baron and

63

Parnell, 2007; Nenna and Aydin, 2011), evaporites (Bäuerle et al., 2000) and other rock

64

types. Along with sub-seismic-scale fractures, understanding the controls on stylolite network

65

distribution is hampered by the high reactivity and complex mechanical behaviour of

66

carbonates (Fabricius, 2014). Stylolites can control petrophysical properties and fluid flow in

67

different ways (e.g., Paganoni et al., 2016; Martín-Martín et al., 2017; Heap et al., 2018;

68

Toussaint et al., 2018; Bruna et al., 2019). However, despite their abundance in carbonate

69

rocks, less attention has been paid to their study compared to fracture networks.

70 71

Stylolite morphologies and their orientation can be used to unravel the orientation and

72

magnitude of principal compressive stress in an area (Koehn et al., 2012). Bedding-parallel

73

stylolites form during vertical sediment compression due to compaction associated with

74

burial. The morphology of stylolites varies greatly according to their amplitude, wavelength,

75

spacing and connectivity, producing a variety of stylolite network geometries (Vandeginste

3

76

and John, 2013; Ben-Itzhak et al., 2014). Stylolite networks have been predominantly studied

77

in association with their influence on fluid flow, since stylolites have demonstrated a range of

78

behaviours when interacting with fluids. Some studies suggest that stylolites can act as

79

baffles to fluid flow (e.g., Burgess and Peter, 1985; Finkel and Wilkinson, 1990; Dawson,

80

1998; Alsharhan and Sadd, 2000; Gomez-Rivas et al., 2015; Martín-Martín et al., 2017)

81

based on field and petrographic observations. Alternatively, stylolites have been found to act

82

as conduits that improve permeability along their orientation as observed in some carbonate

83

reservoirs and outcrops (e.g., Carozzi and Bergen 1987; Bergen and Carozzi, 1990; Lind et

84

al., 1994; Harris, 2006; Chandra et al., 2014; Barnett et al., 2015; Paganoni et al., 2016;

85

Martín-Martín et al., 2017; Morad et al., 2018) in addition to laboratory permeability tests

86

(e.g., Heap et al., 2014, 2018; Rustichelli et al., 2015). The development of localised porosity

87

around stylolites has been associated with preferential dissolution along existing stylolite

88

planes (Bergen and Carozzi, 1990) and an increase in the average size of pore throats (Baud

89

et al., 2016).

90 91

The influence of fluids on the development and effect of stylolites is also uncertain.

92

Petroleum emplacement may inhibit stylolitisation (e.g., Neilson et al., 1998; Morad et al.,

93

2018). Laterally extensive stylolite networks, however, are present in hydrocarbon-bearing

94

limestones and can compartmentalise reservoirs causing variations in fluid flow (Hassan and

95

Wada, 1981; Ehrenberg et al., 2016). The scale of impact is also an important concept. Heap

96

et al. (2014) suggests that the discontinuous nature of stylolites only affects petrophysical

97

properties on a local scale. They (ibid.) suggest that stylolitic porosity creates elongated

98

‘finger-like’ pores aligned with stylolite teeth and larger pore throat radii along fluid flow

99

paths that are parallel to stylolites. Heap et al. (2018) also indicate that stylolites can cause

100

permeability anisotropy, where permeability is higher along the stylolite surface.

4

101 102

Whilst the role of stylolites on fluid flow at the local scale has been extensively studied and

103

remains a controversial topic, there is still uncertainty as to how sedimentary facies and

104

carbonate rock components influence stylolite morphology, type and growth. Along with this

105

there has been a growing interest for understanding the role of host rock heterogeneity on

106

stylolite morphology, location and extent (Toussaint et al., 2018; Morad et al., 2018). Whilst

107

previous authors identify key lithological controls on stylolitisation, Koehn et al. (2016)

108

highlight the importance of how these lithological controls interact with each other to impact

109

stylolite formation.

110 111

The complex nature of stylolites has led to an array of literature covering a range of topics

112

associated with stylolite formation, morphology, distribution and influences on fluid flow and

113

mechanical strength (e.g., Andrews and Railsback, 1997; Aharonov and Katsman, 2009;

114

Koehn et al., 2012; Baud et al., 2016; Koehn et al., 2016; Toussaint et al., 2018; Heap et al.,

115

2018). However, the diverse focus on different facets of stylolite behaviour has subsequently

116

led to the absence of a comprehensive established methodology for the collection of outcrop-

117

scale stylolite measurements.

118 119

This study aims to statistically characterise stylolite morphology and distribution in different

120

platform carbonates, based on outcrop-scale observations of the Aptian-Albian Benassal Fm

121

exposed in the Benicàssim half graben (Maestrat Basin, Spain) (Fig. 1). The specific

122

objectives are: i) to identify sedimentary stylolite populations and quantify their properties

123

(type, density, amplitude and connectivity) in seven typical Early Cretaceous shallow marine

124

lithofacies, including dolostones (Fig. 2), which crop out in the Orpesa Range; ii) to

125

statistically analyse the collected data sets to infer relationships between stylolite properties

5

126

and lithofacies characteristics; and iii) to discuss the most important lithological influences

127

on stylolite morphology and distribution. This well-studied area allows multiple stylolite

128

populations to be sampled in outcrops that have undergone very similar and well-known

129

burial histories, removing the influence of stress variations and instead allowing observations

130

to be directly related to lithofacies and diagenetic textures formed prior to chemical

131

compaction. The results of this study provide an estimation of pressure-solution signatures of

132

typical platform carbonate lithofacies, helping to predict the presence and properties of

133

stylolites when limited subsurface data are available.

134 135

2. Geological setting

136 137

2.1. Structure and stratigraphic architecture of the Benicàssim area

138 139

The Maestrat Basin is located in the east margin of the Iberian Chain, formed by the

140

Mesozoic inversion of the intraplate Iberian Rift System (Salas et al., 2001). Two main

141

cycles of syn- and post-rifting occurred between the Late Triassic and Late Cretaceous, of

142

which the second rift cycle was responsible for the breakup of the Iberian Basin and the

143

formation of the Maestrat basin (Salas and Casas, 1993; Salas et al., 2001; Nebot and

144

Guimerà, 2016). According to these authors, the Maestrat Basin was compartmentalised into

145

several sub-basins (Salas and Guimerà, 1996) and is bound to the north and south by listric

146

extensional faults (Nebot and Guimerà, 2016). Inversion associated with the Alpine Orogeny

147

in the Eocene to Early Miocene led to uplift of the Maestrat Basin (Liesa et al., 2006; Nebot

148

and Guimerà, 2016). Neogene extension opened the Valencia Trough on the present-day

149

eastern margin of the Maestrat Basin (Salas et al., 2001), which also partly reactivated

6

150

Cretaceous extensional features on the eastern part of the Iberian Chain (Guimerà et al.,

151

2004).

152 153

The Late Aptian and Early Albian Benassal Fm was deposited during the second rift cycle of

154

the Maestrat Basin. Strong subsidence during rifting facilitated syn-rift deposition of Early

155

Cretaceous carbonates in excess of 2 km thick, resulting in the deposition of the Benassal Fm

156

with a present-day thickness of at least 1,600 m (Yao, 2019). Depositional facies are

157

characterised by abundant rudists, corals and orbitolinids, signifying an evolution from

158

basinal to inner-ramp settings on a shallow-marine carbonate ramp (Martín-Martín et al.,

159

2013; Yao, 2019). The Benicàssim half graben is defined by two Early Cretaceous large-scale

160

faults, inherited from the Variscan orogeny: the NNE-SSW striking Benicàssim Fault, and

161

the E-W striking Campello Fault (Martín-Martín et al., 2013; Martín-Martín et al., 2015;

162

Yao, 2019). These faults were also re-activated during the Alpine orogeny and the Neogene

163

extension periods (Gomez-Rivas et al., 2012). The Benassal Fm is partially dolomitised and

164

hosts Mississippi Valley-type (MVT) mineral deposits (Corbella et al., 2014; Gomez-Rivas et

165

al., 2014). Dolomitised geobodies range from massive patches next to fault zones to

166

connected stratabound geometries that extend for several kilometres away from them

167

(Martín-Martín et al., 2013; Corbella et al., 2014; Gomez-Rivas et al., 2014; Martín-Martín

168

et al., 2017; Yao, 2019). Dolomitisation occurred between 400-700 m burial depths and

169

predominantly replaced grain-supported lithofacies and beds located between strongly

170

stylolitised limestone facies (Martín-Martín et al., 2015). Chemical compaction (i.e.,

171

stylolitisation) predated dolomitisation, and was estimated to have taken place at relatively

172

shallow burial depths between 300-800m (Martín-Martín et al., 2017). Additional diagenetic

173

processes postdating both stylolitisation and dolomitisation included cementation by saddle

7

174

dolomite, and burial and meteoric calcite, as well as calcitisation of replacive and saddle

175

dolomites (for a complete review of the paragenesis see Martín-Martín et al., 2015).

176 177

2.2. Lithofacies

178 179

According to Martín-Martín et al. (2013) and Yao (2019), the Benassal Fm carbonates of the

180

Orpesa Range can be subdivided into 15 lithofacies and are stacked into three transgressive-

181

regressive sequences (I, II, III). Sequence I contains transgressive deposits of green basinal

182

marls containing brachiopods and echinoderms, interbedded between cross-bedded peloidal

183

and orbitolinid grainstones. These are overlain by regressive peloidal grainstones, orbitolinid

184

wackestones to rudstones and coralline limestones that are topped by a thick package of

185

rudist floatstones. Sequence II begins with transgressive tidally influenced grainstones,

186

overlain by sponge spicule wackestones. Regressive lithofacies initiate with coralline

187

limestones and orbitolinid wackestones that are capped by grainstones with hummocky-cross

188

stratification, bioclastic wackestones and packstones and peloidal grainstones. Sequence II is

189

also topped by a thick unit of rudist floatstones representing the most regressive deposits.

190

Sequence III has transgressive bioclastic wackestones and packstones, which are overlain by

191

a thick dolostone unit with internal limestone stringers. Above the dolostone are regressive

192

ooidal grainstones, bioclastic wackestones and packstones, peloidal grainstones, finally

193

topped by another dolostone unit.

194 195

3. Methods

196 197

A variety of stylolite measurements were performed in all lithofacies in order to characterise

198

stylolite network morphologies. Outcrop-scale measurements were collected from field

8

199

observations using the previously defined lithofacies by Martín-Martín et al. (2013) and Yao

200

(2019) in the Orpesa Range to statistically characterise stylolite populations. These

201

lithofacies include (in order of potential mechanical strength): spicule wackestone (A),

202

bioclastic wackestone/packstone (B), rudist floatstone (C), coralline limestone (D),

203

ooidal/peloidal grainstone (E), bioclastic grainstone (F) and dolostone (G). Lithofacies were

204

sampled using optical petrography to determine rock texture, facies and components.

205

Sampling windows with an area of 1 m2 were used to measure the density, type and size of

206

stylolites in different lithofacies, where windows were positioned adjacently along

207

stratigraphic beds to mitigate against vertical lithological changes. Windows were created at

208

the outcrop using markers which allowed them to maintain the same dimensions and scale

209

when being photographed, regardless of camera zoom and positioning. Stylolite spacing

210

measurements (Fig. 3) were collected in-situ whilst other parameters, such as stylolite

211

connection angles and intersection types, were collected from scaled window photographs.

212

Using vector graphics software, stylolites were traced in black and shown on a white

213

background for each window photograph, to reduce the impacts of vegetation/background

214

noise on stylolite data collection. As a result of photograph resolution limitations and

215

subsequent truncation bias (Zeeb et al., 2013), stylolite seam measurements typically

216

between 3 – 7 pixels were sampled. All spatial measurements (i.e., stylolite spacing,

217

amplitude and wavelength) were measured from photographs in pixels and subsequently

218

converted into centimetres using a conversion ratio defined by the sampling window.

219 220

To measure stylolite spacing, three 1 m-length vertical scanlines, with a horizontal offset of

221

50 cm, were created in each 1 m2 sampling window (Fig. 3). Scanlines were drawn

222

perpendicular to stylolites to avoid orientation bias. The vertical distances between stylolites

223

intersecting each scanline were recorded in the field using a tape measure, with vertical

9

224

spacing measurements from each scanline being totalled per window (Fig. 3). The detection

225

limit, i.e. the maximum vertical stylolite spacing measurement able to be recorded, is one

226

metre, as it is determined by the height of the field-sampling window. However, it is worth

227

noting that the spacing in all the facies analysed was always significantly lower than 1 m,

228

validating the size of the sampling window used here.

229 230

Stylolite wavelength and amplitude measurements were collected using a single vertical

231

scanline situated in the middle of the sampling window. All stylolites that intersected the

232

scanline were measured. Wavelength was recorded as the horizontal distance between the

233

nearest peak-peak or trough-trough that was situated on the scanline, whilst amplitude

234

measurements were represented by the vertical height of a stylolite tooth on or closest to the

235

scanline (Fig. 4a, b). Collecting the closest amplitude measurements to the scanline reduced

236

size bias, although censorship bias limited measurements greater than the 1 m-wide sampling

237

window from being recorded.

238 239

Distribution fitting was carried out on stylolite spacing, wavelength and amplitude

240

measurements from each lithofacies to assess which statistical distribution represented each

241

dataset. Stylolite data were fitted to three common statistical distributions – exponential,

242

power-law and log-normal. The quality of the fits was evaluated using a chi-squared test

243

alongside a Kolmogorov-Smirnoff (KS) normality test (Massey, 1951). Higher chi-squared

244

values indicate a higher goodness of fit between the fitted distributions and stylolite dataset.

245

For KS test results the null hypothesis is that stylolites do not fit the distribution being

246

assessed and therefore have a less than 0.05 p-value, whereas a p-value greater than 0.05

247

suggests that stylolite measurements are best represented by the assessed distribution and

248

therefore the null hypothesis is rejected. Histograms of stylolite of the 10-based logarithm of

10

249

spacing, wavelength and amplitude were constructed with the open-source software Past 3.11

250

(Hammer et al., 2001). Each dataset is divided in 20 equal-width bins, spanning the range of

251

the set. Different distributions are fitted to each dataset, including a normal distribution fitted

252

to each distribution of the log of the values. The goodness of fit is reported as the correlation

253

coefficient (r) of a straight line in the normal probability plot.

254 255

Stylolite populations were identified by Ben-Itzhak et al. (2014) as forming a range of

256

geometries based on the connectivity of stylolites, defined as isolated, long-parallel and

257

anastomosing. Measuring the angle of connectivity in stylolite populations has been

258

suggested as a method to understand whether anastomosing networks form through the

259

connection of isolated stylolites by cannibalisation, where individual stylolites roughen and

260

merge together (Ben-Itzhak et al., 2014). The dimensions of ‘islands’ or ‘lenses’, the length

261

and height of rock space situated between anastomosing stylolites, can allow estimating the

262

amount of dissolution required in order to generate anastomosing populations (Ben-Itzhak et

263

al., 2014) and measuring the intensity of anastomosis.

264 265

Connectivity measurements were recorded for anastomosing stylolite networks in only four

266

lithofacies in which these geometries were present (coralline limestone, bioclastic

267

wacke/packstone, rudist floatstone and spicule wackestone) using the method defined by

268

Ben-Itzhak et al. (2014). Measurements investigating connectivity focused on two aspects

269

based on the dimensions of the area between stylolites and the angle of stylolite junctions

270

(Fig. 4a, c). The dimensions of stylolite ‘islands’ (IL and IW for island length and width,

271

respectively) were acquired to provide a distribution of areal measurements and length to

272

width ratios that were examined per window and totalled for each lithofacies to provide a

273

distribution of ‘island’ dimensions. Stylolite islands were assumed to be rectangular when

11

274

measuring IL and IW (Fig. 4c). Connection angles (CA) were measured at stylolite junctions,

275

where the angle between two stylolites was recorded at different lengths (L), or distance,

276

measured from the stylolite intersection point. Angles were recorded at 1 cm intervals from

277

stylolite intersections to provide a distribution of connection angles per lithofacies. The

278

maximum distance which stylolite intersection angles were measured was defined as the

279

midpoint of a stylolite ‘island’ with anastomosing stylolites, or the ends of stylolites that were

280

laterally discontinuous.

281 282

The density of stylolite intersection types were measured as a function of stylolite length.

283

Intersections were classified, based on the scheme by Manzocchi (2002) and Sanderson et al.

284

(2015) used for fracture network connectivity, as either X or Y (Fig. 5). To calculate the

285

density of intersections per facies, the total number of intersections per window was divided

286

by the total horizontal stylolite length per window. This provided a ratio of stylolite

287

intersection density per metre (/m) for each window, which was then averaged to provide a

288

ratio for each lithofacies (Eqn. 1):

289 290

ID (/m) = (Iwn / (SLwn)

(Eqn. 1)

291 292

Where ID (/m) is the intersection density per metre, Iwn is the total number of intersections

293

per sampling window, and SLwn is the total stylolite length per window measured in pixels.

294

The total stylolite length per sampling window (SLwn) is affected by the size of the sampling

295

window and therefore is not representative of the actual horizontal length of stylolite

296

networks, but it is useful for normalising intersection density for each sampling window in

297

order to compare densities between lithofacies.

298

12

299

Stylolites in each photograph were classified based on morphology using definitions by

300

Koehn et al. (2016), where stylolites are classified as rectangular layer type, seismogram

301

pinning type, suture and sharp-peak type, and wave-like type (for an overview of stylolite

302

morphologies see Koehn et al., 2016 and Humphrey et al., 2019). Stylolite types were

303

counted per window and totalled to obtain the relative frequencies of stylolite types per

304

lithofacies.

305 306

Whilst Ben-Itzhak et al. (2014) devised techniques suited for evaluating stylolite

307

connectivity, our current sampling techniques are based on methods used for fracture network

308

characterisation and inadvertently contain one or more sampling biases which are orientation-

309

, censorship- (i.e., altering sampling area boundaries) or truncation-related (see discussion by

310

Zeeb et al., 2013 for fracture network analysis). Because of this, a combination of

311

measurement methods is required to effectively characterise stylolite networks whilst

312

mitigating against bias. Collecting measurements using vertical scanlines, orientation bias

313

may underestimate vertical stylolite spacings by failing to record oblique stylolites that did

314

not intersect the scanline. However, as the vast majority of stylolites in this study are

315

bedding-parallel, orientation bias is not significant. Censorship bias affects the coverage of

316

the sampling area as a result of vegetation or outcrop exposure, typically leading to the

317

overestimation of density measurements (at least in fracture networks) (Zeeb et al., 2013).

318

Whilst exposure in each sampling area did vary, sampling windows used for collecting

319

stylolite measurements were positioned adjacently and, where possible, at fixed intervals to

320

improve the coverage of sampling and mitigate against measurements being collected from

321

exclusively well-exposed areas with dense stylolite populations. The relationship between

322

stylolite morphology measurements and the sedimentological properties of sampled

323

lithofacies was assessed by using a correlation matrix. Categorical variables (i.e., degree of

13

324

sorting) were firstly converted into a numerical format and the Pearson correlation coefficient

325

was calculated for each pair of input parameters (both stylolite measurements and

326

sedimentological properties). Correlation coefficients ranged between -1 and +1 to indicate

327

negative and positive correlations respectively.

328 329

4. Results

330 331

4.1. Definition of lithofacies

332 333

This section describes the texture characteristics of the seven lithofacies to highlight potential

334

features that can potentially be correlated with stylolite properties (Fig. 2).

335 336

Spicule wackestone (Facies A) is composed of a micritic matrix that shows poor-moderate

337

sorting with less than 10% of grains larger than 2 mm, with minimal bioclasts (bivalve

338

fragments and foraminifera) (Fig. 6a, b). Calcitised spicules are evenly distributed throughout

339

the matrix and are randomly orientated (Fig. 6a, b). Spicule wackestone has metric-scale

340

bedding with minimal vertical fracturing.

341 342

Bioclastic wackestone/packstone (Facies B) has a poorly sorted micritic matrix with less than

343

10% of grains larger than 2 mm. Bioclastic wackestone and packstone are arranged in

344

alternating layers with abundant bioclasts. Primary grain components are orbitolinids,

345

Chondrodonta bivalve, miliolids and other foraminifera (Fig. 6c, d), whilst secondary

346

components include peloids, spicules and echinoderm fragments (Fig. 6d). Small fractures

347

appear partially cemented by calcite (Fig. 6c). Bioclastic wackestone/packstone has

348

decimetre-scale bedding.

14

349 350

Rudist floatstone (Facies C) is made of a micritic matrix with more than 10% of grains larger

351

than 2 mm. Primary bioclasts are rudists, which remain predominantly intact, alongside

352

poorly sorted orbitolinids, miliolids and bivalve fragments (Fig. 6e). Minor quantities of

353

echinoderm fragments are present and have been subject to micritisation (Fig. 6e), whilst

354

rudist fragments are calcitised (Fig. 6f). Rudist fossils remain intact with a diameter of 3-4

355

cm however some fossils can be >10 cm diameter. Rudist floatstone has metric-scale bedding

356

that is well exposed in outcrop.

357 358

Coralline limestone (Facies D) has a micritic texture with less than 10% of components

359

exceeding 2mm. Primary skeletal components consist of coral and bivalve fragments, both of

360

which have been replaced by calcite (Fig. 7a). The original structure of bivalves is preserved

361

(Fig. 7b). Minor quantities of peloids are distributed throughout the matrix and show no

362

evidence of preserved internal structures. The distribution of primary and secondary

363

components indicates a poor degree of sorting. Fine fractures cemented by calcite crosscut all

364

components (Fig. 7b). Coralline limestone has metric-scale bedding with a weathered nodular

365

appearance.

366 367

Ooidal/peloidal grainstone (Facies E) has a well sorted, grain-supported texture with less

368

than 10% of grains exceeding 2 mm in size. Grains are evenly distributed with no apparent

369

clustering or compaction of grains (Fig. 7c, d). Ooids and peloids are the primary lithofacies

370

components, with their relative abundance varying slightly across the sample to produce

371

marginally more ooid/peloid-rich areas. Ooids have poorly developed concentric cortices

372

(Fig. 7c, d), whilst peloids appear to have weakly defined internal structures. Bioclasts

373

present include miliolids and echinoderm fragments (Fig. 7d). There is a higher proportion of

15

374

calcite cement between bioclasts relative to the bioclastic grainstone lithofacies.

375

Ooidal/peloidal grainstone beds typically have centimetre-scale thicknesses.

376 377

Bioclastic grainstone (Facies F) has a well-sorted, grain-supported texture with more than

378

10% of components larger than 2 mm. There is a high density of peloids and bioclasts,

379

occasionally causing grainstones to be classified as peloidal rather than bioclastic, where

380

some areas feature grains that are compacted (Fig. 7e). Skeletal components are

381

predominantly Chondrodonta bivalves, which are vertically orientated in life position,

382

alongside rudists that have an approximate diameter of 10 cm. Orbitolinids, miliolids and

383

bioclastic fragments also occur in minor abundance. Intraclasts and abundant peloids are also

384

present. Fractures cemented by calcite crosscut the rock components (Fig. 7e).

385

Approximately 10-15% of micritised grains have yellow discolouration from iron staining.

386

Grainstones form decimetric- to metric-scale beds that are vertically fractured.

387 388

Dolostone (Facies G) comprises dolomite crystals with a planar-e to planar-s texture, with

389

some crystals being planar-anhedral (Fig. 7f). According to the detailed petrographic analysis

390

by Martín-Martín et al. (2015) and Martín-Martín et al. (2017), bedding-parallel stylolites

391

predate dolomitization in this area. However, stylolites were only occasionally preserved

392

(Fig. 7f) during dolomitisation and are therefore not widespread. Dolostone is well exposed

393

in outcrop and consists of metric-scale beds.

394 395

4.2. General characteristics of stylolite networks for each lithofacies

396 397

Stylolites in spicule wackestone (Facies A) (Fig. 8a) are well exposed with minor weathering

398

and are found to be evenly distributed throughout beds regardless of proximity to bedding

16

399

surfaces. Stylolite morphology is dominantly suture and sharp-peak type with occasional

400

wave-like type stylolites, which connect with each other both laterally and vertically to create

401

anastomosing networks. Stylolite intersections are typically Y-type and therefore stylolite

402

networks have a branching appearance.

403 404

Bioclastic wackestone/packstone (Facies B) (Fig. 8b) stylolites have a similar outcrop

405

appearance as spicule wackestone (Facies A), however the stylolites are less weathered and

406

harder to trace laterally. Stylolites are evenly distributed throughout beds and have a suture

407

and sharp-peak type morphology with minor occurrences of wavy- and rectangular layer type

408

morphologies. Stylolite networks appear isolated and laterally discontinuous in outcrop

409

however occasionally stylolite networks anastomose, producing Y- and X-type intersections.

410 411

Stylolites are well developed in rudist floatstone (Facies C) (Fig. 8c) samples and distributed

412

mostly along the grain-matrix contact, but also within the micritic matrix. Suture and sharp-

413

peak type morphology is dominant with minor occurrences of wave-like type stylolites.

414

Stylolites show evidence of anastomosis, with stylolite convergence typically along the grain-

415

matrix contact where rudists are present. Y-type intersections are common, with a minor

416

abundance of X-type intersections.

417 418

Stylolites in coralline limestone (Facies D) (Fig. 8d) have been weathered but can be

419

distinguished on the surface of outcrops. Stylolites are preserved and nucleate within the

420

matrix. However, stylolite interaction with grains/mineralogical contrasts is low as calcitised

421

grains are sparse (Fig. 6f). Stylolites are a dark yellow colour from iron staining. Stylolite

422

morphology is dominantly suture and sharp-peak type stylolites with occasional wave-like

17

423

type stylolites which can be observed as both anastomosing and forming long parallel

424

networks.

425 426

Stylolites in ooidal/peloidal grainstone (Facies E) (Fig. 8e) lithofacies are easy to distinguish

427

in outcrop due to extensive weathering and appear to develop along bedding surfaces, whilst

428

some beds display weak cross bedding. Similar to other lithofacies, suture and sharp-peak

429

type as well as wave-like type stylolites are the dominant stylolite morphology although

430

ooidal/peloidal grainstone contains a larger relative abundance of rectangular layer type

431

stylolites compared to all other studied lithofacies. Stylolites appear laterally discontinuous

432

and form isolated networks.

433 434

Bioclastic grainstone (Facies F) (Fig. 8f) have stylolites with similar outcrop exposure to

435

ooidal/peloidal grainstone (Facies E), where stylolites also develop proximal to bedding

436

surfaces. Stylolite morphologies are suture and sharp-peak type and wave-like type, which

437

appear to have similar relative abundances, alongside a minor abundance of rectangular layer

438

type stylolites. Stylolites terminate laterally and are isolated with minimal vertical

439

connectivity.

440 441

Dolostone (Facies G) (Fig. 8g, h) have stylolite populations that were difficult to observe due

442

to outcrop weathering associated with calcitisation and meteoric alteration of the rock.

443

Calcite and saddle dolomite precipitated in vugs which are often located parallel to stylolite

444

orientations. Stylolites have similar abundances and distributions to studied grainstone

445

lithofacies (Facies E and F), with suture and sharp-peak type and wave-like type

446

morphologies. Stylolites are isolated and have long-parallel geometries.

447

18

448

4.3. Stylolite statistical properties

449 450

Stylolite populations were initially characterised through outcrop-scale measurements to

451

define statistical properties of stylolite morphology within each studied lithofacies.

452

Morphological properties used to statistically characterise stylolites consisted of vertical

453

spacing, amplitude, wavelength, intersection type and intersection density (Table 1).

454 455

Bioclastic wackestone/packstone has the smallest vertical stylolite spacings, where 50% of

456

spacings are between 0-5 cm and 38% of spacings are between 6-12 cm when measured in 1

457

m2 sampling windows (Fig. 9). Comparatively, dolostones have the largest spacing, with 50%

458

of measurements progressively increasing from 17 to 61 cm. A small proportion of spacing

459

measurements (cumulative frequency of 15%) in ooidal/peloidal grainstone are of a similar

460

size from 31-58 cm, as in the dolostones. Most of the spacing sizes between stylolites in

461

ooidal/peloidal grainstones (cumulative frequency of 55%) range from 10 to 29 cm below

462

which the spacings converge with the majority of the remaining lithofacies. All other

463

lithofacies have similar spacing frequencies, with 50% of measurements approximately

464

between 0 and 12 cm. In these, the distribution of spacing measurements from 6 to 51 cm

465

varies. Bioclastic grainstone, spicule wackestone and coralline limestone show similar

466

spacing (cumulative frequency of approximately 50%) between 12 and 33 cm. Bioclastic

467

grainstone has a similar distribution of spacings to ooidal/peloidal grainstone, where 25% of

468

measurements are between 7 and 14 cm. Spicule wackestone has a higher frequency of

469

spacings where 64% of measurements are between 0 and 14 cm, creating a tighter frequency

470

distribution curve. Coralline limestone measurements are similar to bioclastic grainstone

471

where 50% of measurements are between 11 and 51 cm. Rudist floatstone spacings are

19

472

similar to spicule wackestone however have 17% more measurements between 0 and 9 cm,

473

and with 5% of measurements between 25 and 41 cm.

474 475

Bioclastic

wackestone/packstone

and

spicule

wackestone

have

similar

amplitude

476

measurements, where approximately 25% of measurements are between 0 and 1 cm and 65%

477

are between 1 and 2.4 cm. Rudist floatstone has 15% of measurements between 2.2 and 4.4

478

cm, similar to spicule wackestone, however has fewer measurements (40% compared to 50%

479

respectively) between 0 and 1.4 cm. Dolostone has a similar amplitude distribution to mud-

480

supported lithofacies, particularly bioclastic wackestone/packstone, where 50% of

481

measurements are between 1.2 and 2.6 cm. Grain-supported lithofacies, bioclastic grainstone

482

and ooidal/peloidal grainstone, have approximately 50% of amplitude measurements between

483

0 and 1.2 cm which is 12-30% higher than all other sampled lithofacies. Coralline limestone

484

has 50% of amplitude measurements between 0 and 1.8 cm which is the lowest of all sampled

485

lithofacies, however 10% of measurements are between 3.6 and 8.2 cm which is higher than

486

all sampled lithofacies.

487 488

Stylolite populations across all lithofacies share similarities in the distribution of wavelength

489

measurements, where approximately 50% of all wavelength measurements across all

490

lithofacies range between 1.3 and 10 cm (Fig. 9). Rudist floatstone shows the largest

491

variation in wavelength compared to other lithofacies, with 50% of measurements varying

492

between 1.5 and 6 cm. Dolostone, ooidal/peloidal grainstone and bioclastic grainstone share

493

similar wavelength distributions compared to other lithofacies.

494 495

Based on the classification scheme by Koehn et al. (2016) stylolites predominantly have a

496

suture and sharp-peak type morphology throughout all sampled lithofacies (Fig. 10).

20

497

Stylolites with a suture and sharp-peak type morphology account for over 80% of sampled

498

stylolites per lithofacies, apart from bioclastic grainstone, whilst wave-like type stylolites

499

account for 2 – 48 % per lithofacies. Wave-like type stylolites are most abundant in spicule

500

wackestone and ooidal/peloidal grainstones compared to other lithofacies (15 % and 48 %

501

respectively). Rectangular layer type stylolites are the least abundant type of stylolite

502

morphology with relative abundances of 3 % - 16 %, with the highest recorded abundance in

503

ooidal/peloidal grainstone.

504 505

Stylolite distribution fitting shows that spacing, amplitude and wavelength measurements are

506

all best represented by a log-normal distribution (Figs. 9, 11), rejecting the null hypothesis

507

with p-values greater than 0.05. For spacing measurements (Figs. 9,11a), a log-normal

508

distribution produced chi-squared values between 3.60 and 7.56, whereas power-law and

509

exponential distributions have higher values of >220, signifying a reduced goodness of fit.

510

KS-test results with a p-value greater than 0.05 indicates a higher goodness of fit between the

511

fitted distribution and stylolite dataset. Log-normal distribution p-values are all greater than

512

0.05, whilst power-law and exponential are 0, indicating that spacing measurements are

513

significantly different compared to these distributions. Rudist floatstone and ooidal/peloidal

514

grainstones have the lowest p-values (0.36 and 0.26) for a log-normal distribution, whilst all

515

other lithofacies have values of >0.75, indicating a stronger fit. Stylolite amplitude

516

measurements have a similar range of chi-squared values (Fig. 9, 11b), from 3.47 to 11.85,

517

with grain-supported lithofacies sharing similar values between 5.2 and 6.2. Chi-squared

518

values from exponential and power-law distributions are >492, indicating a poor fit. Most

519

lithofacies have log-normal distributions with high p-values of >0.66. However, rudist

520

floatstone has a value of 0.36. No clear trend is evident between log-normal p-values and the

521

rock composition. Stylolite wavelength measurements have chi-squared values that are

21

522

typically low (Fig. 11c), between 2.13 and 6.83, but bioclastic grainstone has a relatively

523

large value of 21.50. Chi-squared values are all lower in mud-supported lithofacies in

524

comparison to grain-supported lithofacies. Power-law and exponential distributions have chi-

525

squared values of >98, indicating a poor fit. KS-test results have a distribution of p-values

526

between 0.23 and 0.97, where mud-supported lithofacies have higher values relative to grain-

527

supported lithofacies. Power-law and exponential distribution p-values are all zero, but values

528

for an exponential distribution in rudist floatstone are 0.052, indicating that rudist floatstone

529

wavelength data are not significantly different from those of an exponential distribution. The

530

p-value of a log-normal distribution for rudist floatstone is still significantly higher (0.97).

531

However, an exponential distribution can also be weakly representative.

532 533

For lithofacies intersections (Fig. 5, 12a) spicule wackestone features the highest proportion

534

of X-type intersections (16%) with other lithofacies having between 5 and 12% X-type

535

intersections. X-type intersection abundances tend to not fluctuate. Coralline limestone is the

536

anomaly in this trend. Lithofacies have an intersection density per metre of 0.7 to 5.2.

537

Intersection density is higher in mud-supported facies (Fig. 12b). Variations in density occur

538

between grain- and mud-supported lithofacies. Mud-supported facies have an intersection

539

density of 2.6 to 5.2 per metre whereas grain-supported facies have 0.7 to 1.2 per metre.

540 541

4.4. Stylolite connectivity

542 543

Stylolite island dimensions in the studied lithofacies range in length and width between 1-71

544

cm

545

wackestone/packstone have the smallest range of length and width measurements (50 cm and

546

14 cm) (Fig. 13a, b) and therefore the smallest stylolite islands. Comparatively, rudist

and

1-17

cm,

respectively

(Fig.

13).

22

Coralline

limestone

and

bioclastic

547

floatstone and spicule wackestone measurements are larger and with an increased spread of

548

measurements when stylolite lengths are above 20 cm (Fig. 13c, d). When comparing the

549

length to width ratio of stylolite islands it is apparent that coralline limestone has the largest

550

range of 14.4 whilst rudist floatstone has the lowest at 7.

551 552

Spicule wackestone has the largest island areas with a mean average of 50 cm2 (Fig. 13d),

553

compared to other lithofacies in which this value ranges between 8.1 cm2 and 13.9 cm2.

554

Coralline limestone and bioclastic wackestone/packstone and have similar mean average

555

areas (9.2 cm2 and 8.1 cm2), indicating minimal variation in island area (Fig. 13a, b). Rudist

556

floatstone and coralline limestone have similar ranges and standard deviations (Table 2),

557

whereas spicule wackestone has both the largest range and standard deviation indicating a

558

wide distribution of island area measurements.

559 560

Coralline limestone intersection angles show the strongest decrease with distance, with an

561

average angle of 46° (Fig. 14a). Bioclastic wackestone/packstone shares this trend by

562

showing a decrease in intersection angle with distance. However, intersection angles do not

563

begin as high as in coralline limestone resulting in an intersection angle of 42° (Fig. 14b).

564

Alternatively, both rudist floatstone and spicule wackestone stylolite networks have much

565

steeper intersection angles which weakly decrease with distance from stylolite intersection

566

(Fig. 14 c, d).

567 568

5. Discussion

569 570

5.1. Correlation Analysis

571 23

572

In order to characterise stylolite populations as a function of lithofacies through statistical

573

correlation analysis, it is important to firstly characterise the lithofacies parameters that

574

enable comparison between specific lithological and morphological features. Observations

575

from petrographic analyses and previous work by Martín-Martín et al. (2013) assist in both

576

the identification of different lithological parameters and their values, allowing each

577

lithofacies to be petrographically characterised. These parameters can be combined with

578

stylolite morphology parameters to create a correlation matrix (Fig. 15), identifying potential

579

relationships and influences of lithology on stylolite network geometries. The rock’s degree

580

of heterogeneity, in terms of component size, composition, sorting and bedding thickness,

581

show distinct correlations with key stylolite morphological parameters. Poorly sorted, mud-

582

supported lithofacies with metre-scale bedding thicknesses and heterogenous grain sizes

583

produces high amplitude stylolite networks with low vertical spacing and wavelengths,

584

whereas well-sorted grain-supported lithofacies with centimetre- to decimetre-scale bedding

585

thicknesses produce stylolites with low amplitudes and larger vertical spacings.

586 587

5.2. Statistical Analysis of Stylolite Networks

588 589

Stylolite spacing, amplitude and wavelength measurements are best represented by a log-

590

normal distribution (Figs. 9, 11). The data are far away from a power-law distribution and the

591

frequencies do not show power-law sections. Vandeginste and John (2013) use an

592

exponential distribution to represent stylolite spacing measurements, suggesting that

593

stylolites randomly nucleate in the host rock. Despite this, our data supports work by Merino

594

(1992) that stylolites are self-organising and influenced by spatial variations in porosity and

595

stress. These variations primarily relate to the overall heterogeneity in the host rock, in

596

particular the sorting of heterogenous grain compositions, which control the location of

24

597

pressure solution and subsequent stylolite distribution. All sampled lithofacies are best

598

represented by the same log-normal spacing distribution, suggesting that each lithofacies

599

contained sufficient spatial heterogeneity to influence the sites of stylolite nucleation.

600

Ooidal/peloidal grainstone (Facies E) has the weakest chi-squared and K-S test results for

601

spacing (Fig. 11a), which potentially correlates with the high level of sorting in the rock

602

resulting in less heterogeneity. However, it is worth noting that log-normal distributions of

603

stylolite spacing could also be related to truncation bias, a type of sampling bias determined

604

by the resolution of the sampling method (Zeeb et al., 2013). If a very thin stylolite is not

605

identified due to outcrop conditions or because it is below the resolution of photographs then

606

two small spacing values are removed and one larger spacing is added. This will lead to an

607

underrepresentation of small spacings while large spacings will be overrepresented,

608

potentially changing the spacing distribution from exponential to log-normal.

609 610

Koehn et al. (2007) and Ebner et al. (2009) associate log-normal amplitude measurements

611

with a non-linear growth facilitated by solubility variations and subsequent grain ‘pinning’

612

(i.e., where a stylolite plane is fixed to a grain or fossil). Bioclastic wackestone/packstone

613

(Facies B) has the strongest chi-squared and K-S test results for amplitude (Fig. 9, 11b),

614

associated with a mud-supported rock and a diverse range of grains with different

615

compositions. When comparing distribution-fitting results with key lithological components

616

of each lithofacies, it is apparent that there is only a weak correlation between grain size

617

heterogeneity and the degree of sorting on stylolite amplitude. Grain-supported lithofacies

618

generally show a reduced fit compared to mud-supported lithofacies when examining both

619

chi-squared and KS results for log-normal stylolite distributions. These grain-supported

620

lithofacies, including ooidal/peloidal grainstone (Facies E), bioclastic grainstone (Facies F)

621

and dolostone (Facies G) (which is interpreted to have had a grain-supported precursor

25

622

texture before dolomitisation), are all well sorted with lower compositional heterogeneity.

623

The lack of heterogeneity may therefore re-emphasise the importance of solubility variations

624

on stylolite amplitude as previously suggested by Ebner et al. (2009), where the absence of

625

sufficient grain ‘pinning’ reduces non-linear amplitude growth. Stylolite wavelength

626

measurements have stronger chi-squared and K-S test results in lithofacies with a mud-

627

supported rock (Fig. 9, 11c) which could be related to clay content, providing enhanced

628

dissolution as proposed by Aharonov and Katsman (2009), and therefore spatial solubility

629

variations which also influence amplitude growth.

630 631

5.3. Lithological Influences on Stylolite Morphology

632 633

Key relationships influencing stylolite morphology can be derived from the correlation

634

matrix of host rock lithological parameters (Fig. 15). Where previous work has focused on

635

identifying the influences of different types of rock heterogeneity on the degree of stylolite

636

suturing, this study can effectively evaluate specific heterogeneities and their influences on a

637

variety of stylolite morphological parameters.

638 639

5.3.1. Mud- versus grain-dominated facies

640 641

The presence of micrite and clay minerals has often been identified as a major cause of

642

heterogeneity and subsequent stylolite roughening (Wanless, 1979; Koehn et al., 2012;

643

Paganoni et al., 2016; Morad et al., 2018). The high surface area of micrite enhances the

644

rock’s solubility. This study shows weak correlations where stylolite amplitudes (i.e.,

645

roughening) vary based depending on the presence and volume of micrite and produce higher

646

amplitudes in mud-supported facies, as recently reported by Morad et al. (2018). Higher

26

647

amplitude stylolites facilitate more anastomosis shown by positive correlations with X- and

648

Y-type intersections, along with smaller vertical spacing and wavelengths. Interestingly, the

649

correlation matrix shows the opposite morphological trends when stylolites develop in grain-

650

supported facies (Fig. 15). Clay minerals promote heterogeneity and an increase in stylolite

651

amplitude, which is likely to cause stylolites to nucleate and intersect. This is further

652

facilitated by the feedback loop of clay minerals and pressure-solution, as proposed by

653

Aharonov and Katsman (2009), where the accumulation of clay minerals enhances

654

dissolution to form stylolites which subsequently accumulate more clays.

655 656

5.3.2. Grain size, type, bedding and sorting

657 658

Other lithological parameters, aside from the presence of micrite and clay minerals, have

659

been found to promote multifaceted heterogeneity in carbonate host rocks to subsequently

660

impact stylolite morphology. Compositionally, bimodal grain sizes lead to increased suturing

661

due to variations in solubility and susceptibility to pressure-solution (Railsback, 1993;

662

Andrews and Railsback, 1997; Koehn et al., 2012). Whilst stylolites are prone to develop in

663

fine-grained carbonates (Rustichelli et al., 2012), this study supports pre-existing work where

664

lithologies with 10% of grains larger than 2 mm correlate positively with higher amplitudes

665

(Figure 15). Additionally, there is an inverse relationship when grains are smaller than 2 mm

666

and consequently more uniform in grain size.

667 668

Grain size variation within the stylolitised rock must also be related to the degree of sorting,

669

where a combination of grain size variation and poor sorting will enhance the rock’s

670

heterogeneity and consequent roughness (Koehn et al., 2012). Poorly-sorted lithofacies are

671

positively correlated with higher amplitude stylolites, which are more inclined to have

27

672

smaller vertical spacings, abundant intersections and a higher degree of anastomosis (Fig.

673

15). These conditions are facilitated by variations in solubility created by grain size and

674

distribution, causing grain pinning and differences in dissolution rates, which establish sites

675

for stylolites to develop and roughen (Aharonov and Katsman, 2009; Koehn et al., 2016).

676

Increased roughening and dissolution caused by poor sorting can also facilitate anastomosis

677

by stylolite ‘cannibalism’, as suggested by Ben-Itzhak et al. (2014). The third factor

678

influencing stylolite morphology within the rock is the type of grains present, as the

679

composition of grains leads to solubility variations and the promotion of the stylolite

680

development and suturing. Wanless (1979) identifies Mg as an impurity that causes solubility

681

variations in the rock, where the degree of heterogeneity in grain composition can lead to

682

either a uniform or varied distribution of solubility. Measurements of Mg content in fossils by

683

Tucker and Wright (2009) can be used to establish which grains and lithofacies are likely to

684

feature variations in Mg and therefore solubility. For example, foraminifera typically have an

685

Mg content up to 25% whilst other fossils such as bivalves and sponges have limited ranges

686

of 0-5% and 10-15% respectively. High or low Mg content in calcite can also facilitate

687

differences in solubility and pressure solution, alongside additional factors influencing

688

solubility such as calcite/aragonite content in bioclasts, the presence of non-soluble grains

689

and silica, and the solubility of mud (Dewers and Ortleva, 1994).

690 691

The relative scale of bedding thickness was also considered for each lithofacies to identify

692

potential influences on stylolite morphology. Stylolites are found to have smaller amplitudes

693

and larger wavelengths as bedding changes from the metre to the centimetre scale (Fig. 9,

694

15), and no apparent correlation exists between vertical spacing or stylolite intersection

695

type/density. Smaller stylolite amplitude measurements from lithofacies with centimetre-

696

scale bedding differ from results by Sheppard (2002) and Koehn et al. (2012), who suggested

28

697

that a smaller bedding scale would promote lithological contrasts to promote stylolite

698

nucleation at bedding planes. Whilst this study shows that higher amplitude stylolites are

699

found in lithofacies with metre-scale bedding, results may differ from the work by Sheppard

700

(2002), since a lithofacies with centimetre-scale bedding may not have sufficient lithological

701

contrasts which are necessary to promote stylolite nucleation at bedding planes. The

702

depositional texture of studied lithofacies was generalised across sampling windows and

703

supported by observations from Martín-Martín et al. (2013) and Yao (2019). Therefore, the

704

influence of textural variations between beds on stylolite amplitude remains uncertain.

705 706

Rectangular layer type stylolites are abundant in grain-supported lithofacies with centimetre-

707

scale bedding, such as ooidal/peloidal grainstones (Facies E). This supports work by Koehn

708

et al. (2016), suggesting that rectangular layer type stylolites propagate and pin at layer

709

interfaces. Wanless (1979) identifies grain size as the main factor that determines whether

710

stylolites have a jagged or undulose waveform (similar to the suture and sharp-peak and

711

wave-like types of the stylolite classification scheme), arguing that undulose waveforms tend

712

to develop in coarse-grained lithologies. Our study observes an inverse correlation between

713

the abundance of suture and sharp-peak and wave-like type stylolites in coarser facies (i.e.,

714

10% > 2 mm), where wave-like type stylolites develop in finer facies. A predominantly

715

inverse correlation is also seen when evaluating the abundance of stylolite classes in relation

716

to different allochems, where suture and sharp-peak type stylolites are more inclined to

717

develop in facies with larger fossil grains, which again is different from observations by

718

Koehn et al. (2016).

719 720

5.3.3. Stylolite Connectivity and Nucleation

721

29

722

The stylolites island aspect ratios quantified here show similar results to Ben-Itzhak et al.

723

(2014), where islands are predominantly lenticular in shape and influenced by laterally

724

decreasing connection angles at stylolite junctions. Spicule wackestone (Facies A) and rudist

725

floatstone (Facies C) have higher connection angles at stylolite junctions, which Ben-Itzhak

726

et al. (2014) suggest is caused by cannibalised bedding-parallel stylolite networks.

727

Cannibalisation would be promoted by stylolite amplitude growth, in order to facilitate

728

stylolite connectivity, and as a consequence would be expected to be more prominent in

729

lithofacies with a large grain size heterogeneity and poor sorting. Bioclastic

730

wackestone/packstone (Facies B) and coral limestone (Facies D) have lower connection

731

angles, with angles more indicative of anastomosis by linking isolated discontinuous

732

stylolites during increased dissolution (Ben-Itzhak et al., 2014). A combination of clay

733

content and grain size heterogeneity in these lithofacies would facilitate dissolution, as

734

proposed by Aharonov and Katsman (2009) and Koehn et al. (2012). Bioclastic

735

wackestone/packstone (Facies B) would possess both of these qualities whereas coral

736

limestone (Facies D) would be primarily attributed to coral fragments facilitating grain size

737

heterogeneity.

738 739

Stylolite cannibalisation requires the connection of long-parallel stylolites, which can be

740

achieved as stylolites progressively increase in amplitude while growing. However, field

741

observations have shown that stylolite geometry can be altered by the presence of large

742

fossils (Fig. 16). The contact between fossils and the matrix creates grain composition

743

heterogeneity resulting in the formation of nucleation surfaces, in a similar way to stylolites

744

found along bedding surfaces, to cause localised variations in stylolite geometry. A higher

745

proportion of large fossils may facilitate the propagation of long-parallel stylolites in a

30

746

perpendicular orientation, and subsequently providing a mechanism of stylolite linkage aside

747

from amplitude growth.

748 749

Higher angle connectivity in stylolite networks results in the development of more X and Y-

750

type intersections. Results from the correlation matrix (Fig. 15) indicates X and Y-type

751

intersections are promoted by mud-supported facies with 10% grains larger than 2 mm and

752

poor-to-moderate sorting. The role of larger fossils in altering stylolite geometries, as

753

previously discussed, could additionally promote the formation of more X and Y-type

754

intersections. However, spicule wackestone (Facies A) features similar intersection density

755

measurements to rudist floatstone (Facies C), despite having no evident large fossils. This

756

suggests that the presence of larger fossils cannot be primarily responsible for influencing

757

intersection type/density and instead requires the interplay between multiple drivers of

758

heterogeneity across multiple scales, as proposed by Koehn et al. (2016).

759 760

5.3.4. Influence of diagenesis

761 762

Aside from Mg content variations in fossil grains, the influence of diagenetic processes may

763

alter grain composition to subsequently modify heterogeneities initially created by

764

depositional processes. Micritisation, two early calcite cementation phases, dissolution and

765

early stages of fracturing occurred prior to stylolitisation in the Benassal Fm (Martín-Martín

766

et al., 2015, 2017). According to these authors, stylolitisation occurred at relatively shallow

767

burial depths and mostly predated dolomitisation of the Benassal Fm in the study area.

768

Therefore, stylolite morphology cannot be attributed to dolostone lithofacies and is

769

predominantly a result of the depositional texture before chemical compaction that Martín-

770

Martín et al. (2017) interprets as being mostly grainy with minimal calcite cementation.

31

771

Statistical trends from this study support this interpretation, as stylolites sampled from

772

dolostone lithofacies share similar statistical properties with grain-supported lithofacies.

773

Fracturing and calcite cementation were observed from petrographic analysis, although their

774

interaction with stylolite formation could not be fully characterised as 1) the absence of clear

775

cross-cutting relationships between stylolites and fractures makes it difficult to ascertain

776

relative timing and therefore influences on stylolite morphology, and 2) calcite cementation

777

in this study was not characterised using the paragenetic sequence nomenclature of Martín-

778

Martín et al. (2017) which again makes it difficult to estimate relative timings and influences

779

on stylolites. Fracturing initiated during the onset of Early Cretaceous rifting (Gomez-Rivas

780

et al., 2012; Martín-Martín et al., 2015) and therefore did not coincide with stylolitisation

781

that Martín-Martín et al., (2015) estimate started in the Late Cretaceous.

782 783

6. Conclusions

784 785

Stylolite networks were statistically characterised according to morphological variations in

786

various typical shallow-marine carbonate platform lithofacies. The spacing, wavelength and

787

amplitude of stylolites follow a log-normal distribution for all the studied lithofacies.

788

Multiple sources of lithological heterogeneity were identified and related to stylolite

789

morphology. Specifically, the interplay between grain sizes above 2 mm, heterogenous grain

790

compositions, poor sorting and metre-scale bed thicknesses resulted in the formation of high-

791

amplitude stylolites which are closely spaced and feature a high level of anastomosis (Fig.

792

17a). These stylolite networks are typically found in bioclastic wackestone/packstone, rudist

793

floatstone and spicule wackestone, whereas isolated stylolites with larger spacings and low

794

amplitudes are found in lithofacies with limited heterogeneity, such as bioclastic grainstone,

795

ooidal/peloidal grainstone and dolostone (Fig. 17b). Stylolite networks produced in

32

796

lithofacies that contain large grains (e.g., coralline limestone) have moderate amplitudes and

797

wavelength with wavelengths, and have closer spacings than other grain-dominated

798

lithofacies (Fig. 17c). The extent of stylolite network connectivity, and their subsequent

799

potential control on fluid flow, can therefore be primarily determined by the characteristics of

800

the host lithofacies.

801 802

Acknowledgements

803

This research was funded by the Natural Environment Research Council (NERC) Centre for

804

Doctoral Training (CDT) in Oil & Gas, through a PhD grant to EH. Additional funding was

805

provided by the Grup Consolidat de Recerca “Geologia Sedimentària” (2017SGR-824) and

806

the DGICYT Spanish Projects CGL2015-66335-C2-1-R, CGL2015-69805-P and PGC2018-

807

093903-B-C22. EGR acknowledges the support of the Beatriu de Pinós programme of the

808

Government of Catalonia's Secretariat for Universities and Research of the Department of

809

Economy and Knowledge (2016 BP 00208). We are grateful to anonymous reviewers, whose

810

constructive comments have improved the article, together with the editorial guidance of

811

Miroslaw Slowakiewicz.

812 813

References

814

Aharonov, E. and Katsman, R., 2009. Interaction between pressure solution and clays in

815

stylolite development: Insights from modeling. American Journal of Science, 309(7), 607-

816

632.

817

Agar, S.M. and Geiger, S., 2015. Fundamental controls on fluid flow in carbonates: current

818

workflows

to

emerging

819

Publications, 406(1), 1-59.

technologies. Geological

33

Society,

London,

Special

820 821

Alsharhan, A.S., Sadd, J., 2000. Stylolites in lower Cretaceous carbonate reservoirs, U.A.E.. Soc. Econ. Paleontol. Mineral. Special Publication 69, 185–207.

822

Andrews, L.M. and Railsback, L.B., 1997. Controls on stylolite development: morphologic,

823

lithologic, and temporal evidence from bedding-parallel and transverse stylolites from the

824

US Appalachians. The Journal of Geology, 105(1), 59-73.

825

Barnett, A.J., Wright, V.P., Chandra, V.S., Jain, V., 2015. Distinguishing between eogenetic,

826

unconformity-related and mesogenetic dissolution: a case study from the Panna and Mukta

827

fields, offshore Mumbai, India. Geological Society, London, Special Publications, 435(1):

828

67.

829

Baron, M. and Parnell, J., 2007. Relationships between stylolites and cementation in

830

sandstone reservoirs: Examples from the North Sea, UK and East Greenland. Sedimentary

831

Geology, 194(1), 17-35.

832

Baud, P., Rolland, A., Heap, M., Xu, T., Nicolé, M., Ferrand, T., Reuschle, T., Toussaint, R.

833

and

Conil,

N.,

2016.

Impact

834

limestone. Tectonophysics, 690, 4-20.

of stylolites

on

the mechanical

strength

of

835

Bäuerle, G., Bornemann, O., Mauthe, F. and Michalzik, D., 2000. Origin of stylolites in

836

Upper Permian Zechstein anhydrite (Gorleben salt dome, Germany). Journal of

837

Sedimentary Research, 70(3). 726-737.

838

Ben-Itzhak, L.L., Aharonov, E., Karcz, Z., Kaduri, M. and Toussaint, R., 2014. Sedimentary

839

stylolite networks and connectivity in limestone: Large-scale field observations and

840

implications for structure evolution. Journal of Structural Geology, 63, 106-123.

841 842

Bergen, D.V. and Carozzi, A.V., 1990. Experimentally simulated stylolitic porosity in carbonate rocks. Journal of Petroleum Geology, 13(2), 179-192.

34

843

Burgess, C.J. and Peter, C.K., 1985. Formation, distribution, and prediction of stylolites as

844

permeability barriers in the Thamama Group, Abu Dhabi. Middle East Oil Technical

845

Conference and Exhibition. Society of Petroleum Engineers (SPE-13698-MS).

846 847 848 849

Bruna, P.O., Lavenu, A.P., Matonti, C. and Bertotti, G., 2019. Are stylolites fluid-flow efficient features?. Journal of Structural Geology, 125, 270-277. Carozzi, A.V. and Bergen, D.V., 1987. Stylolitic porosity in carbonates: a critical factor for deep hydrocarbon production. Journal of Petroleum Geology, 10(3), 267-282.

850

Corbella, M., Gomez-Rivas, E., Martín-Martín, J.D., Stafford, S.L., Teixell, A., Griera, A.,

851

Travé, A., Cardellach, E. and Salas, R., 2014. Insights to controls on dolomitization by

852

means of reactive transport models applied to the Benicàssim case study (Maestrat Basin,

853

eastern Spain). Petroleum Geoscience, 20(1), 41-54.

854

Chandra, V., Wright, P., Barnett, A., Steele, R., Milroy, P., Corbett, P., Geiger, S. and

855

Mangione, A., 2015. Evaluating the impact of a late-burial corrosion model on reservoir

856

permeability and performance in a mature carbonate field using near-wellbore

857

upscaling. Geological Society, London, Special Publications, 406(1), 427-445.

858 859

Dawson, W.C., 1988. Stylolite porosity in carbonate reservoirs (No. CONF-880301-). American Association of Petroleum Geologists, Tulsa, OK.

860

Dewers, T., Ortoleva, P., 1994. Formation of stylolites, marl/limestone alternations, and

861

dissolution (clay) seams by unstable chemical compaction of argillaceous carbonates. In:

862

Wolf, K.H., Chilingarian, G.V. (Eds.), Diagenesis: IV. Developments in Sedimentology,

863

51, 155–216.

864

Di-Cuia, R., Gout, C., Balzagette, L., Masse, P. and Vieban, F., 2005. Reservoir Connectivity

865

in a Complex Fractured Reservoir: The Relationship Between Fracturing and Facies in the

866

Upper Cretaceous Apulian Platform of the Maiella Mountain, Southern Italy. In

35

867

International Petroleum Technology Conference. International Petroleum Technology

868

Conference.

869

Ebner, M., Koehn, D., Toussaint, R., Renard, F., 2009. The influence of rock heterogeneity

870

on the scaling properties of simulated and natural stylolites. Journal of Structural

871

Geology, 31(1), 72-82.

872

Ebner, M., Piazolo, S., Renard, F. and Koehn, D., 2010. Stylolite interfaces and surrounding

873

matrix material: Nature and role of heterogeneities in roughness and microstructural

874

development. Journal of Structural Geology, 32(8), 1070-1084.

875

Ehrenberg, S.N., Morad, S., Yaxin, L. and Chen, R., 2016. Stylolites and Porosity, In A

876

Lower Cretaceous Limestone Reservoir, Onshore Abu Dhabi, UAE. Journal of

877

Sedimentary Research, 86(10), 1228-1247.

878 879 880 881

Fabricius, I.L., 2014. Burial stress and elastic strain of carbonate rocks. Geophysical Prospecting, 62(6), 1327-1336. Finkel, E.A. and Wilkinson, B.H., 1990. Stylolitization as Source of Cement in Mississippian Salem Limestone, West-Central Indiana (1). AAPG Bulletin, 74(2), 174-186.

882

Gomez-Rivas, E., Corbella, M., Martín-Martín, J.D., Stafford, S.L., Teixell, A., Bons, P.D.,

883

Griera, A. and Cardellach, E., 2014. Reactivity of dolomitizing fluids and Mg source

884

evaluation of fault-controlled dolomitization at the Benicassim outcrop analogue (Maestrat

885

Basin, E Spain). Marine and Petroleum Geology, 55, 26-42.

886

Gomez-Rivas, E., Martín-Martín, J.D., Bons, P.D. and Koehn, D., 2015. Can stylolite

887

networks control the geometry of hydrothermal alterations? Geotectonic Res, 97, 34-36.

888

Gomez-Rivas, E., Warber, K., Kulzer, F., Bons, P.D., Koehn, D. and Martín, J.D.M., 2012.

889

Structural evolution of the Benicàssim area (Maestrat basin, NE Spain): insights from

890

fracture and vein analysis. Geogaceta, (51), 79-82.

36

891

Guerriero, V., Mazzoli, S., Iannace, A., Vitale, S., Carravetta, A. and Strauss, C., 2013. A

892

permeability model for naturally fractured carbonate reservoirs. Marine and Petroleum

893

Geology, 40, 115-134.

894

Guimerà, J., Mas, R. and Alonso, Á., 2004. Intraplate deformation in the NW Iberian Chain:

895

Mesozoic extension and Tertiary contractional inversion. Journal of the Geological

896

Society, 161(2), 291-303.

897 898 899 900

Haines, T.J., Michie, E.A., Neilson, J.E. and Healy, D., 2016. Permeability evolution across carbonate hosted normal fault zones. Marine and Petroleum Geology, 72, 62-82. Hammer, Ø., Harper, D.A.T., Ryan, P.D., 2001. PAST: Paleontological statistics software for education. folk.uio.no/ohammer/past.

901

Harris, N.B., 2006. Low-porosity haloes at stylolites in the feldspathic Upper Jurassic Ula

902

sandstone, Norwegian North Sea: an integrated petrographic and chemical mass-balance

903

approach. Journal of Sedimentary Research, 76(3), 444-459.

904 905 906 907 908 909

Hassan, T.H. and Wada, Y., 1981. Geology and development of Thamama zone 4, Zakum field. Journal of Petroleum Technology, 33(07), 1-327 Heap, M.J., Baud, P., Reuschlé, T. and Meredith, P.G., 2014. Stylolites in limestones: Barriers to fluid flow?. Geology, 42(1), 51-54. Heap, M., Reuschlé, T., Baud, P., Renard, F. and Iezzi, G., 2018. The permeability of stylolite-bearing limestone. Journal of Structural Geology, 118, 81-93.

910

Humphrey, E., Rivas, E.G., Koehn, D., Bons, P.D., Neilson, J., Martin-Martin, J. and

911

Schoenherr, J., 2019. Stylolite-controlled diagenesis of a mudstone carbonate reservoir:

912

A case study from the Zechstein_2_Carbonate (Central European Basin, NW

913

Germany). Marine and Petroleum Geology, 109, 88-107.

914

Koehn, D., Ebner, M., Renard, F., Toussaint, R. and Passchier, C.W., 2012. Modelling of

915

stylolite geometries and stress scaling. Earth and Planetary Science Letters, 341, 104-113.

37

916

Koehn, D., Renard, F., Toussaint, R. and Passchier, C.W., 2007. Growth of stylolite teeth

917

patterns depending on normal stress and finite compaction. Earth and Planetary Science

918

Letters, 257(3), 582-595.

919

Koehn, D., Rood, M.P., Beaudoin, N., Chung, P., Bons, P.D. and Gomez-Rivas, E., 2016. A

920

new stylolite classification scheme to estimate compaction and local permeability

921

variations. Sedimentary Geology, 346, 60-71.

922

Larsen, B., Grunnaleite, I. and Gudmundsson, A., 2010. How fracture systems affect

923

permeability development in shallow-water carbonate rocks: an example from the

924

Gargano Peninsula, Italy. Journal of Structural Geology, 32(9), 1212-1230.

925 926

Laubach,

S.E.,

Olson,

J.E.

and

Gross,

M.R.,

2009.

Mechanical

and

fracture

stratigraphy. AAPG bulletin, 93(11), 1413-1426.

927

Liesa, C.L., Soria, A.R., Meléndez, N. and Meléndez, A., 2006. Extensional fault control on

928

the sedimentation patterns in a continental rift basin: El Castellar Formation, Galve sub-

929

basin, Spain. Journal of the Geological Society, 163(3), 487-498.

930 931

Lind, I., Nykjaer, O., Priisholm, S. and Springer, N., 1994. Permeability of stylolite-bearing chalk. Journal of Petroleum Technology, 46(11), 986-993.

932

Long, J. and Witherspoon, P.A., 1985. The relationship of the degree of interconnection to

933

permeability in fracture networks. Journal of Geophysical Research: Solid Earth, 90(B4),

934

3087-3098.

935 936

Manzocchi, T., 2002. The connectivity of two dimensional networks of spatially correlated fractures. Water Resources Research, 38(9).

937

Martín-Martín, J.D., Gomez-Rivas, E., Bover-Arnal, T., Travé, A., Salas, R., Moreno-

938

Bedmar, J.A., Tomás, S., Corbella, M., Teixell, A., Vergés, J. and Stafford, S.L., 2013.

939

The Upper Aptian to Lower Albian syn-rift carbonate succession of the southern Maestrat

38

940

Basin (Spain): Facies architecture and fault-controlled stratabound dolostones. Cretaceous

941

Research, 41, 217-236.

942

Martín-Martín, J.D., Travé, A., Gomez-Rivas, E., Salas, R., Sizun, J.P., Vergés, J., Corbella,

943

M., Stafford, S.L. and Alfonso, P., 2015. Fault-controlled and stratabound dolostones in

944

the Late Aptian–earliest Albian Benassal Formation (Maestrat Basin, E Spain): petrology

945

and geochemistry constrains. Marine and Petroleum Geology, 65, 83-102.

946

Martín-Martín, J.D., Gomez-Rivas, E., Gómez-Gras, D., Travé, A., Ameneiro, R., Koehn, D.

947

and Bons, P.D., 2017. Activation of stylolites as conduits for overpressured fluid flow in

948

dolomitized platform carbonates. Geological Society, London, Special Publications,

949

459(1), 157-176.

950 951

Massey, F.J., 1951. The Kolmogorov-Smirnov test for goodness of fit. Journal of the American statistical Association, 46(253), 68-78.

952

Merino, E., 1992. Self-organization in stylolites. American Scientist, 80(5), 466-473.

953

Morad, D., Nader, F.H., Morad, S., Al Darmaki, F. and Hellevang, H., 2018. Impact of

954

Stylolitization On Fluid Flow and Diagenesis in Foreland Basins: Evidence from an Upper

955

Jurassic Carbonate Gas Reservoir, Abu Dhabi, United Arab Emirates. Journal of

956

Sedimentary Research, 88(12), 1345-1361.

957

Nebot, M. and Guimerà, J.J., 2016. Structure of an inverted basin from subsurface and field

958

data: the Late Jurassic-Early Cretaceous Maestrat Basin (Iberian Chain). Geologica Acta,

959

14(2), 155-177.

960 961

Nenna, F. and Aydin, A., 2011. The formation and growth of pressure solution seams in clastic rocks: A field and analytical study. Journal of Structural Geology, 33(4), 633-643.

962

Neilson, J.E., Oxtoby, N.H., Simmons, M.D., Simpson, I.R. and Fortunatova, N.K., 1998.

963

The relationship between petroleum emplacement and carbonate reservoir quality:

39

964

examples from Abu Dhabi and the Amu Darya Basin. Marine and Petroleum Geology,

965

15(1), 57-72.

966 967

Nelson, R.A., 1981. Significance of fracture sets associated with stylolite zones: geologic notes. AAPG Bulletin, 65(11), 2417-2425.

968

Nelson, R.A., 2001. Geologic analysis of naturally fractured reservoirs. Elsevier, 159.

969

Paganoni, M., Al Harethi, A., Morad, D., Morad, S., Ceriani, A., Mansurbeg, H., Al Suwaidi,

970

A., Al-Aasm, I.S., Ehrenberg, S.N. and Sirat, M., 2016. Impact of stylolitization on

971

diagenesis of a Lower Cretaceous reservoir from a giant oilfield, Abu Dhabi, United Arab

972

Emirates. Sedimentary Geology, 335, 70-92.

973

Railsback, L.B., 1993. Lithologic controls on morphology of pressure-dissolution surfaces

974

(stylolites and dissolution seams) in Paleozoic carbonate rocks from the mideastern United

975

States. Journal of Sedimentary Research, 63(3), 513-522.

976

Rustichelli, A., Tondi, E., Agosta, F., Cilona, A. and Giorgioni, M., 2012. Development and

977

distribution of bed-parallel compaction bands and pressure solution seams in carbonates

978

(Bolognano Formation, Majella Mountain, Italy). Journal of Structural Geology, 37, 181-

979

199.

980

Rustichelli, A., Tondi, E., Korneva, I., Baud, P., Vinciguerra, S., Agosta, F., Reuschlé, T. and

981

Janiseck, J.M., 2015. Bedding-parallel stylolites in shallow-water limestone successions of

982

the Apulian Carbonate Platform (central-southern Italy). Italian Journal of Geosciences,

983

134(3), 513-534.

984

Salas, R. and Casas, A., 1993. Mesozoic extensional tectonics, stratigraphy and crustal

985

evolution during the Alpine cycle of the eastern Iberian basin. Tectonophysics, 228(1-2),

986

33-55.

987 988

Salas, R. and Guimerà, J., 1996. Main structural features of the Lower Cretaceous Maestrat Basin (Eastern Iberian Range). Geogaceta, 20 (7), 1704-1706.

40

989

Salas, R., Guimerà, J., Mas, R., Martín-Closas, C., Meléndez, A. and Alonso, A., 2001.

990

Evolution of the Mesozoic central Iberian Rift System and its Cainozoic inversion (Iberian

991

chain). Peri-Tethys Memoir, 6, 145-185.

992 993

Sanderson, D.J. and Nixon, C.W., 2015. The use of topology in fracture network characterization. Journal of Structural Geology, 72, 55-66.

994

Sheppard, T.H., 2002. Stylolite development at sites of primary and diagenetic fabric contrast

995

within the Sutton Stone (Lower Lias), Ogmore-by-Sea, Glamorgan, UK. Proceedings of

996

the Geologists' Association, 113(2), 97-109.

997

Tomás, S., Löser, H. and Salas, R., 2008. Low-light and nutrient-rich coral assemblages in an

998

Upper Aptian carbonate platform of the southern Maestrat Basin (Iberian Chain, eastern

999

Spain). Cretaceous Research, 29(3), 509-534.

1000

Toussaint, R., Aharonov, E., Koehn, D., Gratier, J.P., Ebner, M., Baud, P., Rolland, A. and

1001

Renard, F., 2018. Stylolites: A review. Journal of Structural Geology, 114, 163-195.

1002

Tucker, M.E. and Wright, V.P., 2009. Carbonate sedimentology. John Wiley & Sons, 1-14.

1003

Vandeginste, V. and John, C.M., 2013. Diagenetic implications of stylolitization in pelagic

1004

carbonates, Canterbury Basin, Offshore New Zealand. Journal of Sedimentary Research,

1005

83(3), 226-240.

1006 1007

Wanless, H.R., 1979. Limestone response to stress: pressure solution and dolomitization. Journal of Sedimentary Research, 49(2).

1008

Yao, S. (2019). Static modelling of a shallow-marine carbonate reservoir analogue:

1009

prediction of facies distribution, sequence stratigraphy and fault-associated dolostone

1010

geometry.

1011

https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.774048

Ph.D.

thesis,

University

of

Aberdeen.

1012

Zeeb, C., Gomez-Rivas, E., Bons, P.D. and Blum, P., 2013. Evaluation of sampling methods

1013

for fracture network characterization using outcrops. AAPG bulletin, 97(9), 1545-1566.

41

1014 1015

Figures:

1016 1017

Fig. 1. (A) Simplified map of the Iberian Peninsula showing the location the Maestrat Basin

1018

in the Iberian Chain (square). (B) Paleogeographic map of the Maestrat Basin showing the

1019

thickness distribution of Late Jurassic-Early Cretaceous succession along the basin. Key

1020

sub-basins are labelled as Morella (Mo), Salzedella (So), Penyagolosa (Pe), Oliete (Ol)

1021

and Galve (Ga). The star represents the location of the main study area relative to

1022

Benicàssim and Orpesa. Key sampling location coordinates in this study are 40.110413°

1023

N, 0.123381° W and 40.111254° N, 0.125295° W. Modified after Salas et al. (2001) and

1024

Martín-Martín et al. (2017).

1025 1026

Fig. 2. (A) Panoramic photograph of the Benassal Fm. Photo plates of studied lithofacies in

1027

order of high to low energy; spicule wackestone (B), bioclastic wackestone/packstone (C),

1028

rudist floatstone (D), coralline limestone (E), ooidal/peloidal grainstone (F), bioclastic

1029

grainstone (G) and dolostone (H).

1030 1031

Fig. 3. (A) Example of sample window dimensions and locations of three scanlines used for

1032

sampling stylolite spacing measurements. (B) Annotated photograph categorising stylolite

1033

intersection type as either X-type or Y-type.

1034 1035

Fig. 4. (A) Annotated field image of a sampling window with scanline used for stylolite

1036

wavelength and amplitude measurements. (B) Example measurements for stylolite

1037

wavelength (W) and amplitude (A). (C) Example of collection stylolite island dimensions

1038

(IL & IW) and connection angle (CA) measured as a function of length (L) using

42

1039

methodology by Ben-Itzhak et al. (2014).

1040 1041

Fig. 5. Annotated field image showing stylolite intersection types. (A) X-type. (B) Y-type.

1042 1043

Fig. 6. Photomicrographs showing textural characteristics of lithofacies. (A) Spicule

1044

wackestone (plane polarised light; PPL) with sponge spicules (black arrow) distributed in

1045

a mass made of fine-grained skeletals and micrite alongside with minor foraminifera

1046

(white arrow). (B) Spicule wackestone (diffused light; DL) with a weakly defined stylolite

1047

(black arrows) that shows weak iron staining. (C) Bioclastic wackestone/packstone (DL)

1048

with

1049

wackestone/packstone (cross polarised light; XPL) with abundant bivalve fragments

1050

(black arrow) and a minor quantity of echinoderm fragments (white arrows). (E) Rudist

1051

floatstone with preserved growth laminations (black arrow) and an orbitolinid fragment

1052

(white arrow). (F) Rudist floatstone with open anastomosing stylolites (black arrows)

1053

propagating along grain-matrix contact.

orbitolinid

(white

arrow)

and

stylolite

(white

arrow).

(D)

Bioclastic

1054 1055

Fig. 7. Photomicrographs showing textural characteristics of lithofacies. (A) Coralline

1056

limestone (PPL) showing fragments of corals within the micritic matrix (black arrow). (B)

1057

Coralline limestone (XPL) with a bivalve fragment showing preserved growth laminations

1058

(black arrow). (C) Ooidal/peloidal grainstone (XPL) with stylolite showing minor

1059

anastomosis (black arrow). (D) Ooidal/peloids grainstone (PPL) with a minor quantity of

1060

dolomite rhombs (black arrow). (E) Peloidal grainstone (PPL) with abundant peloids

1061

(black arrow) and benthic foraminifera (white arrow). (F) Crystalline dolomite texture

1062

(DL) with remnant of a stylolite (black arrow).

1063

43

1064

Fig. 8. Photo plate of stylolite network morphologies in sampled lithofacies with annotated

1065

stylolites (white arrows). (A) Spicule wackestone. (B) Bioclastic wackestone/packstone.

1066

(C) Rudist floatstone. (D) Coralline limestone. (E) Ooidal/peloidal grainstone. (F)

1067

Bioclastic grainstone. (G) Dolostone. (H) Dolostone.

1068 1069 1070

Fig. 9. Histograms of log-normal stylolite spacing, amplitude and wavelength measurements per sampled lithofacies.

1071 1072 1073

Fig. 10. Percentage abundances of stylolite morphology using the classification scheme by Koehn et al. (2016).

1074 1075

Fig. 11. Log-normal distribution fitting results for stylolite measurements using chi-squared

1076

and Kolmogorov-Smirnoff (K-S) normality test. (A) Spacing measurements. (B)

1077

Amplitude measurements. (C) Wavelength measurements.

1078 1079 1080

Fig. 12. (A) Stacked chart of stylolite intersection abundance in studied lithofacies. (B) Stylolite intersection density plot for each lithofacies.

1081 1082 1083

Fig. 13. Stylolite island dimension plots per sampled lithofacies. (A) Coralline limestone. (B) Bioclastic wackestone/packstone. (C) Rudist floatstone. (D) Spicule wackestone.

1084 1085 1086

Fig. 14. Stylolite intersection angle plots per sampled lithofacies. (A) Coralline limestone. (B) Bioclastic wackestone/packstone. (C) Rudist floatstone. (D) Spicule wackestone.

1087 1088

Fig. 15. Correlation matrix evaluating measuring the Pearson correlation coefficient between 44

1089

studied stylolite morphology and geological variables. Variables are classified as stylolite

1090

morphology (A), matrix (B), allochems (C), grain sorting (D) and bedding thickness scale

1091

(E). Key correlations have been highlighted (red boxes).

1092 1093

Fig. 16. Annotated field images of stylolite interactions with allochems in studied lithofacies.

1094

(A) Stylolite anastomosis between cm-scale rudists (black arrows) in a rudist floatstone.

1095

(B) stylolite nucleating along contact between Chondrodonta bivalves (black arrows) and

1096

matrix in a bioclastic wackestone/packstone.

1097 1098

Fig. 17. Diagram of typical stylolite network morphologies observed in studied lithofacies.

1099

(A) Example of stylolite networks produced in mud-dominated facies (spicule

1100

wackestone, bioclastic wackestone, rudist floatstone) where stylolites are high amplitude,

1101

high wavelength, closely spaced and highly anastomosing with frequent X- and Y-type

1102

intersections. (B) Example of stylolite networks produced in grain-dominated facies

1103

(ooidal/peloidal grainstone, bioclastic grainstone and including dolostone) where stylolites

1104

are low amplitude and wavelength with large vertical spacings. Stylolites develop isolated

1105

long-parallel networks. (C) Example of stylolite networks produced in lithofacies with

1106

large grain components (coralline limestone) where stylolite networks have moderate

1107

amplitudes and wavelength with closer spacings than other grain-dominated lithofacies.

1108

Minor amounts of X- and Y-type intersections which creates long-parallel to slightly

1109

anastomosing networks with a branching appearance.

1110 1111

Table 1. Summary table of stylolite network morphology statistics in sampled lithofacies.

1112 1113

Table 2. Summary of stylolite island area statistics for four sampled lithofacies. 45

Mean (cm)

Range (cm)

Spicule wackestone

49.8

216.7

Standard Deviation (cm) 53.9

Bioclastic wackestone / packstone

9.2

38.9

9.7

Rudist floatstone

13.9

98.8

17.0

Coralline limestone

8.1

144.4

16.6

Lithofacies Type (LFT)

Characteristic Components

Sedimentary Features

Stylolite Spacing (cm)

Stylolite Amplitude (cm)

Stylolite Wavelength (cm)

Intersection Density (X& Y-type p/m)

Anastomosis Angle (°)

Dolostone

Miliolids

Idiotopic texture, metrescale bedding

18.6

1.6

10.7

5

n/a

Bioclastic Grainstone

Peloids, foraminifera miliolids, orbitolinids, bivalve fragments

Good sorting, decimetrescale bedding

13.9

1.2

11.1

7

n/a

Ooidal / Peloidal Grainstone

Ooids, peloids, miliolids, echinoderms

Good sorting, centimetrescale bedding

17.0

1.2

11.6

6

n/a

Coral Limestone

Corals, peloids, bivalves

Poor sorting, metre-scale bedding

132.0

2.0

12.0

11

46

Bioclastic Wacke / Packstone

foraminifera orbitolinids, miliolids, bivalves, peloids, echinoderms

Poor sorting, decimetrescale bedding

6.1

1.6

10.9

8

42

Rudist Floatstone

Rudists, other bivalves, echinoderms

Poor sorting, metre-scale bedding

10.9

1.6

8.5

6

80

Spicule Wackestone

Spicules, bivalves, foraminifera

Moderate sorting, metrescale bedding

12.6

1.5

11.5

4

92

Abundance (%)

100 90 80 70 60 50 40 30 20 10 0

Lithofacies

A

B Suture & sharp

C

D

E

Rectangular

F

G Wavy

1 0.8 13.3 16.1 7.6

5.3

6.6

Chi-Squared goodness of fit Chi-Squared goodness of fit

0.4 0.2

1

0 A

B

C D E F Lithofacies type

G

KS P-value

100

1 0.8

10

0.6

11.9 7.3

6.7 3.5

5.5 5.2 6.2

0.4 0.2

1

0 A

B

C D E F G Lithofacies type Chi

C

6.0 3.6

Chi

B

0.6

Kolmolgorov-Smirnoff P- value

10

Kolmolgorov-Smirnoff P- value

100

KS P-value

100

1 0.8

21.5 10 2.4

5.4

3.8

6.8

0.6 6.1

0.4 0.2

2.1

1

0 A

B

C D E F G Lithofacies type Chi

KS P-value

Kolmolgorov-Smirnoff P- value

Chi-Squared goodness of fit

A

A

X - type

Y - type

Intersection Abundance

100% 80% 60% 40% 20% 0%

A

B

C

D

E

F

G

E

F

G

Lithofacies type

Intersection Density (n / m)

B 12 10 8 6 4 2 0

A

B

C

D Lithofacies type

A

Coralline Limestone (Facies D)

B

Bioclastic wackestone/packstone (Facies B)

C

Rudist floatstone (Facies C)

D

Spicule wackestone (Facies A)

Coralline limestone (Facies D)

B B

Bioclastic wackestone/packstone (Facies B)

180

180

160

160

140

140

Connection angle (°)

Connection angle (°)

A A

120

100 80 60 40

80 60

40

0

0

0

Rudist floatstone (Facies C)

180

0

20 40 60 Lateral distance from stylolite junction (cm)

D D

160

160

140

140

120 100 80 60 40 20

20 40 60 Lateral distance from stylolite junction (cm) Spicule wackestone (Facies A)

180

Connection angle (°)

Connection angle (°)

100

20

20

C C

120

120 100 80

60 40 20

0

0 0

20 40 60 Lateral distance from stylolite junction (cm)

0

20 40 60 Lateral distance from stylolite junction (cm)

Pearson Correlation Coefficient

Echinoderms Echonoderms

A

B

C

Poor sorting Moderate sorting Good sorting

Echonoderms

Echinoderms

Poor sorting Moderate sorting Good sorting

D

E

A

B

15 mm

60 mm

A Lithofacies Spicule Wackestone (A) Bioclastic Wackestone / Packstone (B) Rudist Floastone (C) Coralline Limestone (D) Ooidal / Peloidal Grainstone (F) Bioclastic Grainstone (E) Dolostone (F)

B

Matrix type

Grain size

Grain Composition

Grain Sorting

Bedding Scale

Mud

10 % < 2 mm

High

Poor

Meter

Mud

10 % < 2 mm

High

Poor

Meter

Mud

10 % < 2 mm

Moderate

Poor

Meter

Grain

10 % > 2 mm

Moderate

Moderate

Meter

Grain

10 % > 2 mm

Moderate

Well

Decimeter

Grain

10 % > 2 mm

Moderate

Well

Decimeter

Crystalline

10 % < 2 mm

Low

Well

Centimeter

C

D

B 180

180

160

160

140

140

Connection angle (°)

Connection angle (°)

A

120

100 80 60 40

100 80 60

40 20

20

0

0

0

20 40 60 Lateral distance from stylolite junction (cm)

C

0

20 40 60 Lateral distance from stylolite junction (cm)

0

20 40 60 Lateral distance from stylolite junction (cm)

D 180

180

160

160

140

140

Connection angle (°)

Connection angle (°)

120

120 100 80 60 40 20

120 100 80

60 40 20

0

0 0

20 40 60 Lateral distance from stylolite junction (cm)

S-N

A

B

Lithofacies E Lithofacies D Lithofacies C Lithofacies G

Lithofacies B

15 cm

6m

C

E

D

H

G

F

15 cm

15 cm

15 cm

15 cm

15 cm

15 cm

1m B

Vertical Scanline

1m

A

Spacing

Y-type

X-type

B

A

Amplitude Wavelength B

C IW

C

5 cm

10 cm

10 cm

IL

L

CA

A. A

B. B

1 mm 1 mm

C. C

1 mm

D

1 mm 1 mm

E

1 mm

F. F

1 mm

1 mm

A. A

B

1 mm

1 mm

D. D

C

1 mm

0.75 1 mm mm

F. F

E. E

1 mm

11mm mm

A

B

5cm

5cm

frequency frequency frequency frequency

frequency frequency frequency frequency frequency

frequency frequency frequency frequency frequency

frequency frequency frequency frequency frequency

frequency frequency frequency frequency frequency

frequency frequency frequency

frequency frequency frequency

10 0 0 0 40 40 40

Spacing (mm) = 0.9857 0.9818 rr =

30 30 30 20 20 20 10 10 10 0 0 0 40 40 40 40

Spacing (mm) r= = 0.9947 0.9947 rr = 0.9880

30 30 30 30 20 20 20 20 10 10 10 10 0 0 0 0 40 40 40 40 40 40 40 30 30 30 30 30 30 30 20 20 20 20 20 20 20 10 10 10 10 10 10 10 0 0 0 0 0 0 0 40 40 40 40 40 40 40 30 30 30 30 30 30 30 20 20 20 20 20 20 20 10 10 10 10 10 10 10 0 0 0 0 0 0 0 40 40 40 40 40 30 30 30 30 30 20 20 20 20 20 10 10 10 10 10 0 0 0 0 0 40 40 40 40 40 30 30 30 30 30 20 20 20 20 20 10 10 10 10 10 0 0 0 0 0 40 40 40 40 30 30 30 30 20 20 20 20 10 10 10 10 0 0 0 0

Spacing (mm)

0.9950 (mm) = Spacing 0.9857 rr = = 0.9982 0.9818 rr = 0.9982

0.9950 r = 0.9857 0.9880 0.9950 = 0.9857 0.9947 rr =

10 10

rr = 0.9818 = 0.9950 0.9880 = rr = 0.9880

100 100

= 0.9818 0.9947 rr = r = 0.9818

10 = 0.9947 0.9950 rr =

100

10 10 0 0 50 0 50 50 r 40 r 40 40 30 30 30 20 20 20 10 10 10 0 0 50 0 50 50 50 r r 40 r 40 40 40 30 30 30 30 20 20 20 20 10 10 10 10 0 0 0 50 0 50 50 50 50 50 50 rrr 40 40 40 40 r 40 40 40 30 30 30 30 30 30 30 20 20 20 20 20 20 20 10 10 10 10 10 10 10 0 0 0 0 0 50 0 50 50 50 50 0 r 50 50 r 40 40 40 40 r 40 40 40 30 30 30 30 30 30 30 20 20 20 20 20 20 20 10 10 10 10 10 10 10 0 0 0 00 1000 50 0 0 1000 0 50 50 0 r 50 50 rr 40 40 r 40 40 40 30 30 30 30 30 20 20 20 20 20 10 10 10 10 10 0 0 0 0 50 50 50 0 50 50 rr 40 40 r 40 40 40 30 30 30 30 30 20 20 20 20 20 10 10 10 10 10 0 0 0 0 50 50 00 1000 50 50 rr 40 40 40 40 30 30 30 30 20 20 20 20 10 10 10 10 0 0 0 50 50 50 0

10 10 0 0 0 50 50 r 50 r 40 40 40 30 30 30 20 20 20 10 10 10 0 0 0 50 r 50 50 r 50 r 40 40 40 40 30 30 30 30 20 20 20 20 10 10 10 10 0 0 0 0 50 r 50 50 50 rr 50 50 40 r 40 40 40 40 40 30 30 30 30 30 30 20 20 20 20 20 20 10 10 10 10 10 10 0 0 0 0 0 0 r 50 50 50 50 r 50 50 40 r 40 40 40 40 40 30 30 30 30 30 30 20 20 20 20 20 20 10 10 10 10 10 10 0 0 0 00 1000 0 0 1000 0 r 50 50 rr 50 50 r 40 40 40 40 30 30 30 30 20 20 20 20 10 10 10 10 0 0 0 0 50 50 rr 50 50 r 40 40 40 40 30 30 30 30 20 20 20 20 10 10 10 10 0 0 0 00 1000 50 50 rr 50 50 40 40 40 40 30 30 30 30 20 20 20 20 10 10 10 10 0 0 0 0

Wavelength (mm) = 0.9985 0.9980 =

Wavelength (mm) = 0.9953 0.9953 = = 0.9957

Wavelength (mm) Wavelength (mm) 0.9955 = 0.9985 = 0.9980 0.9938 = = 0.9938

0.9909 = 0.9985 0.9957 0.9909 = 0.9985 0.9953 =

10 10

= 0.9980 = 0.9955 0.9957 = = 0.9957

100 100

= 0.9980 0.9953 = = 0.9980

10 = 0.9953 0.9955 =

100

Amplitude (mm) = 0.9968 0.9940 =

Spicule Spicule Coral wackestone wackestone limestone (Facies A)

Amplitude (mm) = 0.9970 0.9970 = = 0.9949

Dolostone Bioclastic Dolostone Ooidal/ wacke/ peloidal Packstone grainstone (Facies B)

Amplitude (mm)

(mm) 0.9964 =Amplitude 0.9968 = 0.9982 = 0.9940 = 0.9982

Rudist Coral Rudist Bioclastic Spicule Bioclastic floatstone limestone wacke/ floatstone wackestone wacke/ (Facies C) Packstone Packstone

0.9980 = 0.9968 0.9949 0.9980 = 0.9968 0.9970 =

= 0.9940 = 0.9964 0.9949 = = 0.9949

Bioclastic Ooidal/ Bioclastic Coral Dolostone Coral grainstone peloidal grainstone limestone limestone grainstone (Facies D)

10 10

100 100

Spicule Ooidal/ Rudist Ooidal/ wackestone peloidal floatstone peloidal grainstone grainstone (Facies E)

= 0.9940 0.9970 = = 0.9940

= 0.9970 0.9964 10 =

Dolostone Spicule Bioclastic Spicule wackestone grainstone wackestone (Facies F)

100

Rudist Dolostone Dolostone floatstone (Facies G)

A

B

25 cm

C

10 cm

D

10 cm

E

25 cm

F

5 cm

G

8 cm

H

5 cm

30 cm

•

Stylolite networks in carbonate lithofacies have significant morphological variations

•

Stylolites in grain-supported lithofacies have large vertical spacings with small amplitudes

•

Anastomosing stylolite networks appear well developed in mud-supported lithofacies

•

Statistical techniques can be used to characterise stylolite network morphologies

Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: