Canadian Journal of Earth Sciences

IGCP591 special issue - The Homerian (late Wenlock, Silurian) carbon isotope excursion from Perunica: does dolomite control the magnitude of the carbon isotope excursion?

Journal: Manuscript ID Manuscript Type: Date Submitted by the Author:

Article 03-Jan-2016 Fryda, Jiri; Czech University of Life Sciences, Faculty of Environmental Sciences, Czech University of Life Sciences Prague, Kamýcká 129, Praha 6 – Suchdol, 165 21 Frydova , Barbora ; Czech University of Life Sciences Prague, 1Faculty of Environmental Sciences, Czech University of Life Sciences Prague, Kamýcká 129, Praha 6 – Suchdol, 165 21

af

Silurian, Homerian carbon isotope excursion, Perunica, dolomite, Chemostratigraphy

t

Keyword:

cjes-2015-0188.R1

Dr

Complete List of Authors:

Canadian Journal of Earth Sciences

https://mc06.manuscriptcentral.com/cjes-pubs

Page 1 of 18

Canadian Journal of Earth Sciences

1

The Homerian (late Wenlock, Silurian) carbon isotope excursion from Perunica: does

2

dolomite control the magnitude of the carbon isotope excursion?

3 4

Jiří Frýda1,2 and Barbora Frýdová1

5 6

1

7

129, Praha 6 – Suchdol, 165 21, Czech Republic, and 2Czech Geological Survey, Klárov

8

3/131, 118 21 Prague 1, Czech Republic

9

Corresponding author (e-mail: [email protected]).

Faculty of Environmental Sciences, Czech University of Life Sciences Prague, Kamýcká

10 11 12

Abstract: The δ13Ccarb records from two geographically close sections of the shallow-water

13

Kozel Limestone Member (late Wenlock Motol Formation; Perunica microplate) significantly

14

differ in the magnitude of the Homerian carbon isotope excursion as well as in their dolomite

15

content. The present paper tests a hypothesis as to whether a difference of about 2‰ in the

16

magnitude of the δ13Ccarb anomaly is caused by the different content of dolomite, which could

17

be enriched in both

18

experimental data. The new data obtained reveal that the δ18O composition of calcite and

19

dolomite was probably controlled by the pore fluid composition during limestone diagenesis

20

and that both carbonates seem to be close to equilibrium in oxygen isotope composition. On

21

the other hand, the δ13C values of dolomite are similar to those of calcite and thus the carbon

22

isotope composition of both carbonates was probably determined by the precursor carbonate

23

composition. Moreover, the values of δ13Cdolomite/ δ13Ccalcite ratios as well as their variability

24

suggest that both calcite and dolomite did not reach equilibrium in their carbon isotope

25

composition. Whole-rock, mineralogical and C and O isotope data clearly show that dolomite

26

is not the cause for the differences in magnitudes of the δ13C records observed between

27

dolomite-bearing and dolomite-lacking shallow-water limestone successions. The question

28

as to whether the observed differences in the δ13C records of the studied sections across the

29

Homerian carbon isotope excursion were controlled by the dependence of sea water

30

composition on water depth and/or proximity to shoreline, or if the δ13C values were later

31

affected by secondary processes during limestone diagenesis is still unsolved.

32

Résumé: xx

33

Key words: Silurian, Homerian, carbon isotope excursion, Perunica, dolomite

C and

18

O relative to coexisting calcite, as has been suggested by

t

af

Dr

13

34 35 36

Introduction

1 https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

Page 2 of 18

Our knowledge of the carbon isotope perturbations in Silurian marine ecosystems has

38

considerably increased during the last two decades (see summaries in Loydell 2007,

39

Munnecke et al. 2010, and Cramer et al. 2011). Evidence for their global distribution is now

40

indisputable as well as for their occurrence in different marine environments. The Silurian

41

carbon isotope anomalies have been documented in both inorganic and organic carbon

42

marine reservoirs. Nowadays all major Silurian carbon isotope excursions are recorded from

43

shallow-water platform environments to deeper pelagic depositional settings including

44

graptolite shales (see review in Cramer et al. 2011). Increasing data support earlier

45

observations (Kaljo et al. 1997, Noble et al. 2005, Melchin and Holmden 2006, and Loydell

46

2007) on the shoreward increase of the magnitude of carbon isotope anomalies. Numerous

47

conceptual models have been offered to explain major Paleozoic carbon isotope excursions

48

over the past two decades, stemming from detailed sedimentological investigations of the

49

potential role played by climate, sea level, and seawater circulation as factors influencing

50

stratigraphic variation in sedimentary δ13C values. Recently Kozłowski and Sobien (2012)

51

and Kozłowski (2015) published a new model which links the Ludfordian (late Silurian)

52

carbon isotope excursion, appearing to be the highest magnitude positive δ13Ccarb excursion

53

in the post-Cambrian Phanerozoic, with coeval regressions and development of evaporitic

54

areas in a dry climate, which would contribute elevated dolomite amounts to the basin.

55

Occurrence of dolomite is rather common in the Silurian limestone successions recording

56

carbon isotope excursions.

57

This is true also for the Homerian (late Wenlock, Silurian) carbon isotope positive anomaly

58

which is characterized by a double-peaked excursion and distinct faunal overturn (see

59

summaries in Loydell 1998, 2007, Cramer et al. 2011, 2012 and Jarochowska et al. 2015).

60

Two bioevents occurred just before or coincident with the onset of the carbon isotope

61

excursion: the first described affected graptolites and is known as the “Große Krise” or

62

lundgreni Event (Jaeger 1959, 1991, Koren’ 1991; Štorch 1995; Kaljo et al. 1996; Lenz &

63

Kozłowska-Dawidziuk, 2001, 2002; Porębska et al. 2004; Lenz et al. 2006; Noble et al. 2005,

64

2012) and the second to be described affected conodont faunas and has been named the

65

Mulde Event (Jeppsson et al. 1995, Calner & Jeppsson 2003, Jeppsson & Calner 2003,

66

Calner et al. 2006, but see also Jarochowska and Munnecke, 2015).

67

The Homerian carbon isotope excursion has hitherto been recorded from Baltica (Samtleben

68

et al. 2000; Calner & Jeppsson 2003; Calner et al. 2006; Kaljo et al. 2007; Calner et al. 2012;

69

Jarochowska et al. 2015), Avalonia (Corfield et al. 1992; Marshall et al. 2012), Laurentia

70

(Cramer et al. 2006) and peri-Gondwana (Frýda and Frýdová 2014). The first peak of the

71

double-peaked excursion typically reaches a δ13Ccarb value of about +3.5‰ as summarized

72

by Cramer et al. (2011) and Saltzman and Thomas (2012), and the second peak is always

73

slightly lower. Nevertheless there is increasing evidence for a much higher magnitude first

t

af

Dr

37

2 https://mc06.manuscriptcentral.com/cjes-pubs

Page 3 of 18

Canadian Journal of Earth Sciences

74

peak from limestone successions deposited in shallow-water marine environments often

75

containing dolomite (e.g., Kaljo et al. 2007, Marshall et al. 2012).

76

Here we report a new δ13Ccarb record from shallow-water limestones of the Prague Basin

77

across the first peak of the Homerian carbon isotope excursion reaching up to +4.5‰. The

78

new isotope data come from the Kozel Syncline area which is located only about 1.5 km from

79

the previously described Kozel section (no. 760) recording a much lower δ13Ccarb anomaly

80

(Frýda and Frýdová 2014). Unpublished sedimentological data suggest that the limestones

81

with higher δ13Ccarb values (Kozel Syncline section no. 244JF) were deposited in slightly

82

shallower depth than those having lower δ13Ccarb values (Kozel section no. 760). Both

83

sections record the same carbon isotope anomaly but its magnitude differs by about 2‰ δ13C

84

(Fig. 1). The only distinct difference between the two sections is the high dolomite content in

85

the section with the higher δ13C values. The effect of dolomite on carbon isotope anomalies

86

in marine ecosystems is not yet well evaluated (see discussion in Kozłowski 2015). In

87

addition, numerous theoretical and experimental studies have revealed that dolomite in

88

equilibrium with calcite is enriched in both

89

summary in Horita 2014). The main aim of the present paper is to test a hypothesis: whether

90

the distinct difference in the magnitude of the first peak of the Homerian carbon isotope

91

excursion observed in these two shallow-water limestone successions from the Prague Basin

92

is caused by a difference in dolomite content.

13

C and

18

O relative to coexisting calcite (see

af

Dr

93 Geological settings

95

The Homerian (late Wenlock) strata of the Barrandian form the upper part of the Motol

96

Formation [for definition and stratigraphical range of the Motol Formation see Kříž (1975)].

97

According to its present definition, the Motol Formation (thickness 80 m to 300 m) comprises

98

the uppermost Telychian (Llandovery), the entirety of the Wenlock, and possibly the

99

lowermost Gorstian (Kříž 1992, 1998, Frýda and Frýdová 2014). The lower part of the Motol

100

Formation is sedimentologically uniform and is represented mostly by calcareous clayey

101

shales. Facies distribution in the upper part of the Motol Formation was strongly influenced

102

by synsedimentary tectonic movements and volcanic activity (Bouček 1934, 1953; Horný

103

1955a,b, 1960; Kříž 1991).

104

During the Homerian the high activity of several volcanic centres (i.e., Svatý Jan Volcanic

105

Centre, Řeporyje Volcanic Centre and Nová Ves Volcanic Centre) created shallow areas

106

generating a complex facies pattern from very shallow intertidal limestones to deeper water

107

shales deposited in an anoxic environment (see for details Bouček 1953; Horný 1955a,b,c,

108

1960, 1962; Fiala 1970; Kříž 1991, 1992, 1998; Štorch 1998). The central area of the Svatý

109

Jan Volcanic Centre was emergent during the latest Wenlock and Ludlow forming a volcanic

110

island. Shallow-water environments surrounding the island gave rise to mostly crinoidal–

t

94

3 https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

Page 4 of 18

111

coral–brachiopod–bryozoan grainstones, rudstones and packstones containing a variable

112

amount of tuffitic and/or dolomitic components. The latter shallow-water limestone unit was

113

formally established as the Kozel Limestone Member of the Motol Formation (Frýda and

114

Frýdová 2014).

115

The present study focuses on two sections within the Kozel Limestone Member which are

116

about 1.3 km apart. Both sections are situated on the left (north) bank of the Berounka River.

117

The first section (49.9561064N, 14.0977517E) known as “Kozel” was named by Kříž et al.

118

(1993) as the Kozel section (no. 760) and it has been intensively studied since the beginning

119

of the last century (see Kříž 1991, 1992, 1998; Kříž et al. 1993 for references). The most

120

detailed description of the section was given by Kříž (1992), Kříž et al. (1993) and Dufka

121

(1995), who discussed its lithology, faunal communities and biostratigraphy. The second

122

section (49.9536697N, 14.1149714E) named here as section no. 244JF occurs in the so-

123

called "Kozel Syncline" situated about 1.3 km E of section no. 760, and 1 km W of the

124

estuary of the Kačák brook (Fig. 1). The Kozel Syncline area was mentioned in the past as

125

an important fossil locality but has never been studied in detail.

126 Methods

128

Limestone samples (about 5 kg each) from the Kozel Syncline section (no. 244JF) were

129

sampled in order to investigate the chemostratigraphical record across the first peak of the

130

Homerian (late Wenlock, Silurian) carbon isotope excursion. This new sampling campaign

131

included primarly 56 limestone samples for carbon and oxygen isotope analyses, 36 of which

132

were selected for thin-section study and whole-rock analyses. Grainstones, rudstones, and

133

packstones represent the dominant limestone lithologies of the lower part of the Kozel

134

Limestone Member in the Kozel Syncline section.

135

A few milligrams of rock powder from each sample were recovered with a dental drill from cut

136

and polished slabs. The mineralogical composition of the rock powder was controlled by X-

137

ray powder diffraction analysis using a Philips X´Pert diffractometer. Only calcite samples

138

having no dolomite or traces of dolomite were used for carbon and oxygen isotope analyses.

139

In three samples calcite and dolomite grains were mechanically separated under the

140

microscope. About 1 kg of each of the 36 whole-rock samples was, after cleaning in the

141

ultrasonic bath, dissolved in cold and very weak HCl (1:20) to separate fine grains of

142

dolomite from the limestone. The purity of the dolomite was tested by X-ray powder

143

diffraction analysis and only calcite-free dolomite samples were used later for carbon and

144

oxygen isotope analyses.

145

The carbonate powder (calcite or dolomite) was reacted with 103% phosphoric acid at 25°C

146

for about 24 hours (calcite powders) or 48 hours (dolomite powders). The carbon and oxygen

147

isotopic composition of the released CO2 was meassured with a Thermo Delta 5 mass

t

af

Dr

127

4 https://mc06.manuscriptcentral.com/cjes-pubs

Page 5 of 18

Canadian Journal of Earth Sciences

spectrometer in dual inlet configuration. The meassured δ18O value of dolomite was

149

corrected according to the equations published by Becker and Clayton (1976) and

150

Rosenbaum and Sheppard (1986). All values are reported in ‰ relative to the V-PDB by

151

assigning a δ13C value of +1.95 ‰ and a δ18O value of +2.20 ‰ to NBS 19. Accuracy and

152

precision were controlled by replicate measurements of laboratory standards and were better

153

than ±0.1‰ for both carbon and oxygen isotopes.

154

Whole-rock samples (about 2 kg) were, after cleaning in the ultrasonic bath, powdered using

155

an agate mill and then homogenized. Two small samples of about 5g from each powdered

156

whole-rock sample were selected for further chemical analyses. The first sample was

157

leached by weak HCl (1:1) at room temperature to dissolve preferably carbonates, and the

158

second sample was dissolved using a mixture of H2SO4, HNO3 and HF at a temperature of

159

200°C in a closed Teflon box. Both solutions were later analysed for their Ca, Mg, Fe, and

160

Mn contents by the atom absorption spectrometry method using a Perkin-Elmer AAnalyst

161

100 spectrometer. Analyses of the chemical composition of calcite and dolomite were

162

performed by CAMECA SX100 microprobe under 15 kV and 4 nA using spectrometers with

163

LIF, TAP and PET crystals and certified carbonate standards for Ca, Mg, Fe, Sr and Mn.

164

The non-parametric rank-order Spearman´s as well as Kendall´s tests (including calculation

165

of the Spearman´s and Kendall´s correlation coefficients and t-test of their significance) were

166

used for testing the correlation of meassured geochemical data. Because of the latter data

167

are not normally distributed, the non-parametric Kruskal-Wallis test was used for testing that

168

samples were taken from population with the same mean.

t

af

Dr

148

169 170 171

Results

172

δ13C chemostratigraphy

173

The δ13Ccarb data from limestone samples from the Kozel section (no. 760) were reported by

174

Frýda and Frýdová (2014). The new δ13Ccarb data from the Kozel Syncline section (no.

175

244JF) are summarized in Figures 1 and 2. The first three samples from loose limestone

176

blocks enclosed by tuffitic beds below the base of the Kozel Limestone Member range from

177

1.3‰ to 1.6‰. By contrast all δ13Ccarb values from the overlying limestone beds belonging to

178

the lower part of the Kozel Limestone Member are much higher. The δ13Ccarb values rise

179

rapidly from +2.7‰ in the first bed of the Kozel Limestone Member to +4.5‰ about 5 m

180

above the base of the Kozel Limestone Member. The δ13Ccarb values of fourteen samples

181

from the about 2.4 m thick interval just above the highest δ13C value of the the Kozel

182

Syncline section (no. 244JF) have a decreasing trend ranging from +4.5‰ to +3.3‰ (Fig. 1).

183

In the subsequent interval, about 0.3 m thick, the δ13C values again increase to +4.1‰. The

5 https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

Page 6 of 18

184

δ13Ccarb values in the overlying limestone beds with a total thickness of about 6 m fall to a

185

value of +2.5‰ (Figs 1 and 2).

186 187

Whole rock chemistry

188

Whole rock analyses of the limestone samples from the Kozel Syncline section (no. 244JF)

189

revealed a rather high variability in Mg content. Increased Mg content was recorded from the

190

interval about 5.5 to 8 m from the base of the Kozel Limestone Member and in the two

191

stratigraphically highest samples (Fig. 2A) in which the Mg content reached values of about 8

192

and 5.5 wt.% respectively. Chemical analyses of solutions formed after dissolving of whole

193

rock powder in 1:1 HCl gave slightly lower Mg contents (Fig. 2A). X-ray analyses of residua

194

after dissolution revealed that dolomite was also dissolved. The Mg/Ca molar ratio of most of

195

whole rock samples varies from about 0.01 to 0.1 (Fig. 2B). Only one Mg/Ca ratio value is

196

below this range (0.005) and it belongs to one of limestone clasts enclosed by the tuffites

197

below the base of the Kozel Limestone Member (Fig. 1). The highest Mg/Ca ratio values,

198

0.333 and 0.237 respectively, were found in the two stratigraphically highest samples.

199 Calcite and dolomite

201

The new carbon and oxygen isotope data from calcite and dolomite from the Kozel Syncline

202

(section no. 244JF) are summarized in Figure 3. There is a statistically significant difference

203

between the δ13C values of the calcite samples (Kruskal-Wallis test; p0.05) (Fig.

207

3A). On the other hand, all three δ18O datasets [i.e., δ18Ocalcite values from both sections (no.

208

760 and no. 244JF) as well as δ18Odolomite values from section no. 244JF] are significantly

209

different (Fig. 3B). Detailed comparison of carbon isotope fractionation between coexisting

210

calcite and dolomite grains from three whole-rock samples (including three replicates), which

211

were mechanically separated under the microscope, revealed that the δ13C values of

212

dolomite may be slightly lighter or heavier than the δ13C values of coexisting calcite (Fig. 3C).

213

In addition, available data suggest that the carbon isotope fractionation between calcite and

214

dolomite depends on the Mg content of the whole rock and thus on the content of dolomite

215

(Fig. 3D). However, because of the low number of studied samples (n=9) the latter

216

relationship is not quite robust.

217

Microprobe analyses showed that calcite contains on average about 1 molar percent of

218

MgCO3, thus having Mg/Ca ratio of about 0.01. Analyses also showed that dolomite is Ca

219

enriched and thus far from ideal stoichiometry CaMg(CO3)2. Average dolomite composition

t

af

Dr

200

6 https://mc06.manuscriptcentral.com/cjes-pubs

Page 7 of 18

Canadian Journal of Earth Sciences

220

may be expressed by the formula Ca1.10Mg0.82Fe0.08(CO3)2, which corresponds to a Mg/Ca

221

molar ratio of about 0.75 (Fig. 2B).

222 Discussion

224

The main aim of the present paper is to test the hypothesis that a high dolomite content may

225

considerably increase the δ13C values of limestone samples coming from the dolomite-

226

bearing Kozel Sycline section (no. 244JF). X-ray diffraction data, whole-rock analyses (Fig.

227

2A) and thin section studies revealed that dolomite is the main Mg-bearing phase and

228

therefore the Mg content of whole-rock samples may be used for an estimation of dolomite

229

content. If average compositions of calcite and dolomite are used (Fig. 2B), then the dolomite

230

content in each whole-rock sample may be calculated. The latter estimation revealed that the

231

dolomite content in the majority of studied whole-rock samples is less than 15 molar % with

232

average and median values of 5% and 4% respectively. The highest dolomite content was

233

found in the two stratigraphically highest samples which contain about 44% and 30%

234

dolomite respectively. The relationships of the δ13C values and the Mg content in the whole-

235

rock samples (Fig. 2D) as well as of the δ13C values and Mg/Ca molar ratios (i.e., estimated

236

dolomite content) (Fig. 2C) were analysed to test whether dolomite is responsible for the high

237

δ13C values. The analyses showed that there are no statistically significant correlations

238

amongst the above-mentioned variables (for both non-parametric rank-order Spearman´s as

239

well as Kendall´s tests is p>>0.05; Figs 2B and 2D). In addition, the highest δ13C values

240

come from samples with very low Mg and thus dolomite contents (Fig. 2). These results

241

suggest that the high dolomite content in samples from the Kozel Syncline section (no.

242

244JF) does not increase the δ13C record across the first peak of the Homerian carbon

243

isotope excursion (Fig. 1).

244

Study of δ13Ccalcite and δ13Cdolomite compositions resulted in a similar conclusion. There is no

245

statistically significant difference between the δ13Ccalcite and δ13Cdolomite values of samples from

246

the Kozel Syncline section (no. 244JF) (Fig. 3A). As shown in Figure 3C the dolomite has

247

similar δ13C values to coexisting calcite but its δ13C values may be slightly lighter or heavier

248

than the δ13C values of coexisting calcite. The first peak of the double-peaked Homerian

249

excursion reaches a δ13Ccarb value of about +3.3‰ in the Kozel section (no. 760), similar to

250

the values noted by Cramer et al. (2011). The magnitude of the first peak of the Homerian

251

carbon isotope excursion at the Kozel Syncline section (no. 244JF) is, however, about 2‰

252

higher than typically recorded values. The small difference in the δ13C values of calcite and

253

dolomite from the Kozel Sycline section cannot be responsible for this increase of about 2‰

254

found in these dolomite-bearing limestones. Determined dolomite/calcite carbon isotope

255

fractionation values (Fig. 3D) suggest that even if any whole-rock sample was formed of pure

256

dolomite, then the maximum increase of its δ13C value should be less than 0.7‰ relative to

t

af

Dr

223

7 https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

Page 8 of 18

dolomite-free limestone, and thus not the observed about 2‰ difference (see Fig. 1).

258

Moreover, the samples with the highest δ13C values have a very low dolomite content

259

(compare Figs 2A and C). Taken together, the whole-rock chemistry and carbon isotope data

260

clearly reject the hypothesis that a high dolomite content is responsible for the 2‰ difference

261

in the magnitude of the first peak of Homerian carbon isotope excursion found in the two

262

studied sections.

263

The δ13C and δ18O composition of calcite and dolomite may also help to understand the

264

origin of dolomite, but one has to keep in mind that both calcite and dolomite isotope

265

signatures could be influenced by diagenetic processes during limestone burial. Diagenetic

266

carbonates formed during burial tend to show decreasing δ18O values with burial as a result

267

of many different factors, including increased temperature and evolution of the δ18O values of

268

the pore fluids. In contrast, the δ13C values of the diagenetic carbonate should not be

269

generally significantly different from the δ13C values of the original marine carbonate

270

sediment if its organic matter content is low (see summary in Swart 2015). However, both the

271

δ13C and δ18O values of carbonates are controlled by several factors. The δ13C and δ18O

272

values of marine carbonate minerals depend upon the temperature of their formation, the

273

δ13C and δ18O values of the sea water, the mineralogy of precipitating carbonate, the pH of

274

the solution, as well as on kinetic effects (e.g., Urey, 1947, Epstein et al., 1953, Emrich et al.,

275

1970 and Deines et al., 1974, Zeebe and Wolf-Gladrow, 2001, and McConnaughey, 2003,).

276

In this context, the δ18O values of calcite from the two studied sections (no. 780 and no.

277

244JF) fall within the range for Palaeozoic limestones having a well-preserved primary δ13C

278

signature. For example, the δ18O values of about 40 limestone samples from the flat-topped

279

peak of the mid-Ludfordian carbon isotope excursion (i.e., from the δ13C chemostratographic

280

S-zone) having constantly rather high δ13C values of about +8‰ (see Frýda and Manda

281

2013), range from -6.5 to -4‰. The latter δ18O values are thus rather variable probably

282

because the limestone was affected locally by a different intensity of reaction with diagenetic

283

fluids. On the other hand, their constant δ13C values suggest that the carbon isotope

284

composition was not considerably affected by diagenesis. The δ18O values of limestone

285

samples coming from the Kozel section (no. 760) and the the Kozel Syncline section (no.

286

244JF) are significantly different (Fig. 3B). Samples from the Kozel (section no. 670) have

287

lighter δ18O values than those from the Kozel Sycline (section no. 244JF) which may suggest

288

a more intensive influence of secondary (diagenetic) processes on the buried marine

289

carbonates. The latter processes could theoretically also lower their δ13C values (e.g., by

290

reaction with pore fluids contaning light carbon isotope values from oxidation of organic

291

matter).

292

The δ13C and δ18O values of calcite and dolomite samples from the the Kozel Syncline

293

section (no. 244JF) revealed that there is no statistically significant difference between the

t

af

Dr

257

8 https://mc06.manuscriptcentral.com/cjes-pubs

Page 9 of 18

Canadian Journal of Earth Sciences

294

δ13Ccalcite and δ13Cdolomite values (Fig. 3A), but, their δ18O values are significantly different (Fig.

295

3B). Published experimental as well as natural data showed that dolomite should be enriched

296

in both

297

of the studied dolomite and calcite (Fig. 3B) suggest that both carbonates were modified

298

during diagenetic processes (e.g., Hudson, 1977). The dolomite from the Kozel Syncline

299

section is enriched in

300

agreement with experimental and natural data (see Horita 2014 for references). The δ18O

301

composition of calcite and dolomite was probably controlled by the pore fluid composition

302

and both carbonates seem to be close to equilibrium in δ18O. On the other hand, the δ13C

303

values of dolomite are similar to those of calcite and thus the carbon isotope composition of

304

both carbonates was probably determined by the precursor carbonate composition, because

305

pore diagenetic fluids generally have low carbon content. Therefore the δ13C values of

306

limestones could be considered as a good proxy for the isotopic compositions of carbonate

307

precipitated from marine water (Fig. 3). Nevertheless, the observed dependence of carbon

308

isotope fractionation between calcite and dolomite on the Mg whole-rock content and thus on

309

the content of dolomite (Fig. 3D) may suggest a weak influence of the dolomitization process

310

on the δ13C values of limestones.

13

C and 18O in respect to coexisting calcite (Horita 2014). The negative values of δ18O

O relative to calcite by about 3 to 4‰ (Fig. 3), which is in good

Dr

311

18

Conclusions

313

The main aim of the present paper was to test the hypothesis whether the distinct difference

314

in magnitude of the first peak of the Homerian carbon isotope excursion observed in two

315

shallow-water successions of the Kozel Limestone Member was caused by differences in

316

dolomite content. Whole-rock elemental composition and isotope data clearly rejected this

317

hypothesis. On the other hand, the question as to whether the observed differences in the

318

δ13C records of the two studied sections across the Homerian carbon isotope excursion were

319

controlled only by dependence of sea water composition on water depth and/or proximity to

320

land, or whether the δ13C values were later partly overprinted by some secondary process

321

during limestone diagenesis is still unsolved.

t

af

312

322 323

Acknowledgements

324

This paper is a contribution to IGCP 591 and IGCP 596. The research was supported by a

325

grant from the Grant Agency of the Czech Republic (GAČR 15-133105), and a grant from the

326

Czech Geological Survey (338800). We are very grateful to D. K. Loydell who improved the

327

English and provided valuable comments on the manuscript. This paper also benefited from

328

the constructive reviews of two anonymous referees.

329 330

References

9 https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

Page 10 of 18

331

Becker, R.H. and Clayton, R.N. 1976. Oxygen isotope study of a Precambrian Banded Iron-

332

Formation, Hamersley Range, Western-Australia. Geochimica et Cosmochimica Acta

333

40(10): 1153-1165.

334

Bouček, B. 1934. Bemerkungen zur Stratigraphie des böhmischen Gotlandien und seinen

335

Faziesverhältnissen. Centralblatt für Mineralogie, Geologie und Palaeontologie, Abt. B,

336

11: 477–494.

337

Bouček, B. 1953. Biostratigraphy, development and correlation of the Želkovice and Motol

338

Beds of the Silurian of Bohemia. Sborník Ústředního Ústavu geologického, Oddíl

339

paleontologický 20: 421–484.

340

Calner, M. and Jeppsson, L. 2003. Carbonate platform evolution and conodont stratigraphy

341

during the middle Silurian Mulde Event, Gotland, Sweden. Geological Magazine 140, (2):

342

173–203.

343

Calner, M., Kozłowska, A., Masiak, M. and Schmitz, B. 2006. A shoreline to deep basin

344

correlation chart for the middle Silurian coupled extinction–stable isotopic event. GFF 128:

345

79–84.

346

Mulde Event interval (Silurian, Wenlock), Gotland, Sweden. GFF 134: 65–67.

Dr

347

Calner, M., Lehnert, O. and Jeppsson, L. 2012. New chemostratigraphic data through the Corfield, R.M., Siveter, D.J., Cartlidge, J.E. and McKerrow, W.S. 1992. Carbon isotope

349

excursion near the Wenlock–Ludlow, (Silurian) boundary in the Anglo–Welsh area.

350

Geology 20: 371–374.

af

348

Cramer, B.D., Brett, C.E., Melchin, M.J., Männik, P., Kleffner, M.A., McLaughlin, PI., Loydell,

352

D.K., Munnecke, A., Jeppsson, L., Corradini, C., Brunton, F.R., and Saltzman, M.R. 2011.

353

Revised correlation of Silurian provincial series of North America with global and regional

354

chronostratigraphic and δ13Ccarb chemostratigraphy. Lethaia 44: 185–202.

t

351

355

Cramer, B.D., Condon, D.J., Söderlund, U., Marshall, C., Worton, G.J., Thomas, A.T., Calner,

356

M., Ray, D.C., Perrier, V., Boomer, I., Patchett, P.J. and Jeppsson, L. 2012. U-Pb

357

(zircon) age constraints on the timing and duration of Wenlock

358

paleocommunity collapse and recovery during the “Big Crisis”. Geological Society of

359

America Bulletin 124: 11-12, 1841-1857.

360 361

(Silurian)

Cramer, B.D., Kleffner, M.A. and Saltzman, M.R. 2006. The late Wenlock Mulde positive carbon isotope excursion in North America. GFF 128, 2: 85–90.

362

Dufka, P. 1995. Upper Wenlock miospores and cryptospores derived from a Silurian volcanic

363

island in the Prague Basin (Barrandian area, Bohemia). Journal of Micropalaeontology 14:

364

67–79.

10 https://mc06.manuscriptcentral.com/cjes-pubs

Page 11 of 18

Canadian Journal of Earth Sciences

365

Deines, P., Langmuir, D. and Harmon, R.S. 1974. Stable carbon isotope ratios and

366

exisitence of a gas-phase in evolution of carbonate ground waters. Geochim. Cosmochim.

367

Acta, 38: 1147–1164.

368 369 370 371 372 373

Emrich, K., Ehhalt, D.H. and Vogel, J.C. 1970. Carbon isotope fractionation during the precipitation of calcium carbonate. Earth Planet. Sci. Lett., 8: 363–371. Epstein, S., Buchsbaum, R., Lowenstam, H.A. and Urey, H.C. 1953. Revised carbonatewater isotopic temperature scale. Geol. Soc. Am. Bull., 64: 1315–1326. Fiala, F. 1970. Silurian and Devonian diabases of the Barrandian Basin. Journal of Geological Sciences, Geology 17: 7–71.

374

Frýda, J. and Frýdová, B. 2014. First evidence for the Homerian (late Wenlock, Silurian)

375

positive carbon isotope excursion from peri-Gondwana: new data from the Barrandian

376

(Perunica). Bulletin of Geosciences 89: 617–634.

377

Frýda, J. and Manda, Š. 2013. A long–lasting steady period of isotopically heavy carbon in

378

the late Silurian ocean: evolution of the δ13C record and its significance for an integrated

379

δ13C, graptolite and conodont stratigraphy. Bulletin of Geosciences 88: 463–482.

380

382 383

(Kosov u Berouna). Věstník Ústředního Ústavu Geologického 30: 81–86. Horný, R. 1955b. Studie o vrstvách budňanských v západní části Barrandienu. Sborník Ústředního Ústavu geologického, Oddíl geologický 21: 315–447.

t

385

Horný, R. 1955a. Base vrstev kopaninských eß 1 na jihozápadním okraji vulkanické facie

af

384

to elevated temperatures. Geochimica et Cosmochimica Acta, 129: 111–124.

Dr

381

Horita, J. 2014. Oxygen and carbon isotope fractionation in the system dolomite–water–CO2

386

Horný, R. 1960. Stratigraphy and tectonics of the western closures of the Silurian–Devonian

387

synclinorium in the Barrandian area. Sborník Ústředního Ústavu Geologického, Oddělení

388

geologické 1: 495–524.

389

Horný, R. 1962. Das mittel böhmische Silur. Geologie 11 (8): 873–916.

390

Hudson, J.D. 1977. Stable isotopes and limestone lithification. Journal of the Geological

391

Society of London 133: 637–660.

392

Jaeger, H., 1959. Graptolithen und Stratigraphie des jüngsten Thüringer Silurs.

393

Abhandlungen der Deutschen Akademie der Wissenschaften zu Berlin, Klasse für

394

Chemie, Geologie und Biologie 2: 1–197.

395

Jaeger, H., 1991. New standard graptolite zonal sequence after the ‘Big Crisis’ at the

396

Wenlockian/Ludlowian

boundary

(Silurian).

397

Paläontologie, Abhandlungen 182: 303–354.

Neues

Jahrbuch

für

Geologie

und

398

Jarochowska, E. and Munnecke, A. 2015. Late Wenlock carbon isotope excursions and

399

associated conodont fauna in the Podlasie Depression, eastern Poland: a not-so-big

400

crisis? Geological Journal, published online, DOI: 10.1002/gj.2674

11 https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

Page 12 of 18

401

Jarochowska, E., Munnecke, A., Frisch, K., Ray, D. C. and Castagner, A. 2015. Faunal and

402

facies changes through the mid Homerian (late Wenlock, Silurian) positive carbon isotope

403

excursion in Podolia, western Ukraine. Lethaia, published online, DOI: 10.1111/let.12137

404 405

Jeppsson, L., Aldridge, R.J., & Dorning, K.J. 1995. Wenlock (Silurian) oceanic episodes and events. Journal of the Geological Society of London 152: 487–498.

406

Jeppsson, L. & Calner, M. 2003. The Silurian Mulde Event and a scenario for secundo–

407

secundo events. Transactions of the Royal Society of Edinburgh: Earth Sciences 93: 135–

408

154.

409

Kaljo, D., Boucot, A.J., Corfield, R.M., Koren’, T.N., Kříž, J., Le Hérrisé, A., Männik, P.,

410

Märss, T., Nestor, V., Shaver, R.H., Siveter, D.J. and Viira, V. 1996. Silurian bio–events.

411

173–224. In Walliser, O.H. (ed.) Global Events and Events Stratigraphy in the

412

Phanerozoic. Springer–Verlag, Berlin.

413

Kaljo, D., Grytsenko, V., Martma, T. and Mõtus, M.–A. 2007. Three global carbon isotope

414

shifts in the Silurian of Podolia (Ukraine): Stratigraphical implications. Estonian Journal of

415

Earth Sciences 56: 205–220. Kaljo, D., Kiipli, T. and Martma, T. 1997. Carbon isotope event markers through the Wenlock-

417

Pridoli sequence at Ohesaare (Estonia) and Priekule (Latvia). Palaeogeography,

418

Palaeoclimatology, Palaeoecology 132: 211–223.

Dr

416

Koren’, T. N. 1991. The lundgreni extinction event in central Asia and its bearing on graptolite

420

biochronology within the Homerian. Proceedings of the Estonian Academy of Sciences,

421

Geology 40: 74–78.

t

af

419

422

Kozłowski, W. 2015. Eolian dust influx and massive whitings during the kozlowski/Lau Event:

423

carbonate hypersaturation as a possible driver of the mid-Ludfordian Carbon Isotope

424

Excursion. Bulletin of Geosciences 90(4): 807–840.

425

Kozłowski, W., and Sobien, K. 2012, Mid-Ludfordian coeval carbon isotope, natural gamma

426

ray and magnetic susceptibility excursions in the Mielnik IG-1 borehole. Palaeogeography,

427

Palaeoclimatology, Palaeoecology 339: 74-97.

428 429

Kříž, J. 1975. Revision of the Lower Silurian stratigraphy in Central Bohemia. Věstník Ústředního Ústavu Geologického 50: 275–282.

430

Kříž, J. 1991. The Silurian of the Prague Basin (Bohemia) – tectonic, eustatic and volcanic

431

controls on facies and faunal development. Special Papers in Palaeontology 44: 179–203.

432

Kříž, J. 1992. Silurian field excursions: Prague Basin (Barrandian), Bohemia. National

433 434 435

Museum Wales, Geological Series 13: 1–111. Kříž, J., 1998. Silurian, 79–101. In Chlupáč, I., Havlíček, V., Kříž, J., Kukal, Z. and Štorch, P., Paleozoic of the Barrandian (Cambrian to Devonian). Český geologický ústav, Praha.

12 https://mc06.manuscriptcentral.com/cjes-pubs

Page 13 of 18

Canadian Journal of Earth Sciences

436

Kříž, J., Dufka, P., Jaeger, H., and Schönlaub, H.P. 1993. The Wenlock/Ludlow Boundary in

437

the Prague Basin (Bohemia). Jahrbuch der Geologischen Bundesanstalt, A 136 (4): 809–

438

839.

439

Lenz, A.C. and Kozłowska–Dawidziuk, A. 2001. Upper Wenlock (Silurian) graptolites from

440

Arctic Canada: Pre–extinction, lundgreni biozone fauna. Palaeontographica Canadiana

441

20: 1–61.

442

Lenz, A.C. and Kozłowska–Dawidziuk, A. 2002. Late Wenlock and early Ludlow graptolite

443

extinction, evolution and diversification: A reassessment. Special Papers in Palaeontology

444

67: 171–183.

445

Lenz, A.C., Noble, P.J., Masiak, M., Poulson, S.R. and Kozłowska–Dawidziuk, A. 2006. The

446

lundgreni Extinction Event: integration of paleontological and geochemical data from

447

Arctic Canada. GFF 128 (2): 153–158.

448

Loydell, D.K. 1998. Early Silurian sea–level changes. Geological Magazine 135: 447–471.

449

Loydell, D.K. 2007. Early Silurian positive δ13C excursions and their relationship to

450

glaciations, sea–level changes and extinction events. Geological Journal 42: 5, 531–546. Marshall, C., Thomas, A.T., Boomer, I. and Ray, D.C. 2012. High resolution δ13C stratigraphy

452

of the Homerian (Wenlock) of the English Midlands and Wenlock Edge. Bulletin of

453

Geosciences 87(4): 669–679.

454

McConnaughey, T.A. 2003. Sub-equilibrium

18

O and

13

af

455

Dr

451

C levels in biological carbonates:

carbonate and kinetic models. Coral Reefs, 22: 316–327. Melchin, M.J., and Holmden, C. 2006. Carbon isotope chemostratigraphy of the Llandovery

457

in Arctic Canada: implications for global correlationand sea-level change. GFF 128: 173–

458

180.

t

456

459

Munnecke, A., Calner, M., Harper, D.A.T. and Servais, T. 2010. Ordovician and Silurian sea–

460

water chemistry, sea level, and climate: a synopsis. Palaeogeography, Palaeoclimatology,

461

Palaeoecology 296: 389–413.

462

Noble, P.J., Lenz, A.C., Holmden, C., Masiak, M., Zimmerman, M.K., Poulson, S.R. and

463

Kozłowska, A. 2012. Isotope geochemistry and plankton response to the Ireviken (Earliest

464

Wenlock) and Cyrtograptus lundgreni Extinction Events, Cape Phillips Formation, Arctic

465

Canada. 631–652. In Talent, J.A. (ed.), Earth and Life – Global Biodiversity, Extinction

466

Intervals and Biogeographic Perturbations Through Time. Series: International Year of

467

Planet Earth, Springer, 1100 pp.

468

Noble, P.J., Zimmerman, M.K., Holmden, C. and Lenz, A.C. 2005. Early Silurian

469

(Wenlockian) δ13C profiles from the Cape Phillips Formation, Arctic Canada and their

470

relation to biotic events. Canadian Journal of Earth Sciences 42: 1419–1430.

13 https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

Page 14 of 18

471

Porębska, E., Kozłowska–Dawidziuk, A. and Masiak, M. 2004. The lundgreni event in the

472

Silurian of the East European platform, Poland. Palaeogeography, Palaeoclimatology,

473

Palaeoecology 213: 271–294.

474

Rosenbaum, J. and Sheppard, S.M.F. 1986. An isotopic study of siderites, dolomites and

475

ankerites at high-temperatures. Geochimica et Cosmochimica Acta, 50 (6): 1147-1150.

476

Saltzman, M. R. and Thomas, E. 2012. Carbon isotope stratigraphy. In The Geologic Time

477

Scale (Gradstein, F., Ogg, J., Schmitz, M. D. & Ogg, G., eds), pp. 207-232. Elsevier,

478

Amsterdam.

479

Samtleben, C., Munnecke, A. and Bickert, T. 2000. Development of facies and C/O-isotopes

480

in transects through the Ludlow of Gotland: evidence for global and local influences on a

481

shallowmarine environment. Facies 43: 1–38.

482 483 484 485

Swart, P. K. 2015. The geochemistry of carbonate diagenesis: The past, present and future. Sedimentology, 62(5): 1233–1304. Štorch, P. 1995. Biotic crises and post–crisis recoveries recorded by graptolite faunas of the Barrandian area, Czech Republic. Geolines 3: 59–70. Štorch, P. 1998. Volcanism, 149–164. In Chlupáč, I., Havlíček, V., Kříž, J., Kukal, Z. and

487

Štorch, P., Palaeozoic of the Barrandian (Cambrian to Devonian). Český geologický

488

ústav, Praha.

489

492

562–581.

Zeebe, R.E. and Wolf-Gladrow, D. 2001. CO2 in seawater: equilibrium, kinetics, isotopes. Elsevier, Amsterdam, 346 pp.

t

491

Urey, H. 1947. The thermodynamic properties of isotopic substances. J. Chem. Soc., 99:

af

490

Dr

486

493 494 495 496

14 https://mc06.manuscriptcentral.com/cjes-pubs

Page 15 of 18

Canadian Journal of Earth Sciences

497

Explanation of figures:

498

Figure 1. The δ13C records across the Homerian (late Wenlock, Silurian) carbon isotope

499

positive anomaly from the Barrandian area (Perunica). A - Distribution of Silurian rocks in the

500

Barrandian area, including the locations of both studied sections, the Kozel section (no. 760)

501

and the Kozel Syncline section (no. 244JF). B - The δ13C record from the Kozel section

502

(modified from Frýda and Frýdová, 2014). C - The δ13C record from the Kozel Syncline

503

section.

504 505

Figure 2. Whole-rock elemental composition and isotope data from the Kozel Syncline

506

section. A - Stratigraphical distribution of Mg whole-rock content. B - Relationship of the δ13C

507

values and Mg/Ca molar ratios. Grey bands represent composition of calcite and dolomite. C

508

- Stratigraphical distribution of δ13C values. Grey curve represents locfit regression. D -

509

Relationship of the δ13C values and the Mg content in the whole-rock samples. See text for

510

discussion.

Dr

511

Figure 3. A, B - The δ13C and δ18O composition of calcite and dolomite from the two studied

513

sections, the Kozel section (no. 760) and the Kozel Syncline section (no. 244JF). C - The

514

δ13C composition of coexisting calcite and dolomite from three whole-rock samples of the

515

Kozel Syncline section. Grains separated mechanically are connected by a line. D -

516

Relationship of the carbon isotope fractionation between coexisting calcite and dolomite and

517

the Mg content in the whole-rock samples. See text for discussion.

t

af

512

518

15 https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

t

af

Dr https://mc06.manuscriptcentral.com/cjes-pubs

Page 16 of 18

Page 17 of 18

Canadian Journal of Earth Sciences

t

af

Dr https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

5

Kozel Syncline

Kozel

A

Kozel

-1

4

Page 18 of 18

B

Kozel Syncline

-2

Ocarb

-4

18

13

C carb

-3 3

-5

2

-6 -7

1

calcite

calcite

-8

dolomite

calcite

calcite

dolomite

-9

0 5

1.3

C

D C calcite

4

dolomite

3

13

2

C dolomite /

13

13

C carb

calcite

1.2

1

isotopically heavier dolomite

1.1

1

isotopically lighter dolomite

0.9

Kozel Syncline

Kozel Syncline

0

0.8

-6

-5

-4

-3 18

-2

-1

0

0

1

2

3

4

5

6

7

8

t

af

Dr

Mg (wt. %) d Ocarb

https://mc06.manuscriptcentral.com/cjes-pubs