Geophys. J. Int. (2009) 179, 125–143 doi: 10.1111/j.1365-246X.2009.04301.x

GJI

Geo

mag

netism

,ro

ckm

agne

tism

and

pala

eom

agne

tism

The Tortonian reference section at Monte dei Corvi (Italy): evidencefor early remanence acquisition in greigite-bearing sediments

S. K. Husing, M. J. Dekkers, C. Franke∗ and W. KrijgsmanPaleomagnetic Laboratory Fort Hoofddijk, Department of Earth Sciences, Utrecht University, Budapestlaan 17, 3584 CD Utrecht, The Netherlands.E-mail: huesing@geo.uu.nl

Accepted 2009 June 16. Received 2009 June 16; in original form 2008 November 10

S U M M A R YThe reliability of primary natural remanent magnetization (NRM) signals in greigite-bearingsediments has been frequently questioned. Here, we show that the stable NRM in the deepmarine Middle to Late Miocene sediments at Monte dei Corvi, northern Italy, is mainlycarried by greigite. Combined rock magnetic experiments and scanning electron microscopysuccessfully enabled discrimination between two greigite populations. One fine-grained andrelatively well-dispersed greigite population (grain size between 60 and 200 nm) is most likelyof magnetotactic origin. The second greigite population with larger grain sizes (typically700 nm to 1 μm) is most likely of authigenic (bacterially mediated) origin, and is related topost-depositional sulphidization processes. Greigite is the main magnetic remanence carrierin the older part of the section (12.8 to 8.7 Ma), whereas greigite and fine-grained (presumablymagnetotactic) magnetite are present in the younger part of the section (8.7 to 6.9 Ma).Similar remanent magnetization directions of the magnetite and greigite components, and thelikelihood of a magnetotactic origin, suggests that the NRM is of syn-depositional age. Wesuggest that moderate methane seepage from the underlying sediments may have occurredthat resulted in the preservation of pristine greigite. This corroborates the reliability of thepreviously established magnetostratigraphy at Monte dei Corvi.

Key words: Magnetostratigraphy; Rock and mineral magnetism; Europe.

1 I N T RO D U C T I O N

Over the last few decades, numerous stratigraphic studies have es-tablished high-resolution Neogene chronologies. These studies haveemployed biostratigraphy, radioisotopic dating, magnetostratigra-phy and astronomical tuning in different sedimentary environments.Global Stratotype Sections and Points (GSSPs) have been definedfor the Neogene to improve the standard geological timescale.

The Tortonian GSSP, i.e. the Serravallian–Tortonian boundary(Middle–Late Miocene boundary), has been defined at the Montedei Corvi section in northern Italy (Hilgen et al. 2003, 2005). Thissection has also been proposed as the Mediterranean reference sec-tion and unit stratotype for the Tortonian Stage (Husing et al. 2009).Its magnetostratigraphy is calibrated straightforwardly to the inter-val between 12.5 and 7.2 Ma of the geomagnetic polarity timescale(GPTS), with support from biostratigraphy, radioisotopic dating ofash layers and astronomical tuning of characteristic sedimentarycycle patterns to the precession index (∼21-kyr; Hilgen et al. 2003;Husing et al. 2007). Resulting astronomical ages for each rever-

∗Now at: Ecole des Mines de Paris, Centre des Geosciences, 35 rue Saint-Honore, 77305 Fontainebleau Cedex, France

sal boundary generally have small uncertainties (1 to 20 kyr). TheMonte dei Corvi section is the only marine Miocene section thathas such a high-resolution magnetostratigraphy and astronomicaldating, therefore its reversal ages have been included into the mostrecent Neogene geological timescale (Lourens et al. 2004) that pro-vides a basis for worldwide correlation and dating.

For any magnetostratigraphy to be reliable, the magnetic re-manence must be a primary palaeomagnetic signal that is stableover geological time. Detrital (fine-grained) magnetite is a reliablemagnetic remanence carrier because its magnetization is typicallylocked in shortly after deposition. Magnetite from bacterial magne-tosomes can also carry a reliable primary remanence. Early and latediagenetic processes can, however, alter the primary magnetic min-eral assemblage through dissolution and sulphidization processesthat lead to the formation of new magnetic minerals, particularlythe iron sulphides greigite and pyrrhotite (Berner & Raiswell 1983;Berner 1984; Roberts & Turner 1993; Wilkin et al. 1996; Kao et al.2004). Pyrrhotite is generally considered to form during late dia-genesis (e.g. Dinares-Turell & Dekkers 1999; Weaver et al. 2002;Horng & Roberts 2006), although it can occur as a detrital mag-netic phase (Horng & Roberts 2006). Authigenic pyrrhotite is con-sidered not suitable for palaeomagnetic studies where the syn- orearly depositional acquisition of the NRM is crucial. Greigite can

C© 2009 The Authors 125Journal compilation C© 2009 RAS

126 S. K. Husing et al.

form authigenically any time after deposition if the necessary re-actants are present (Roberts & Weaver 2005), which complicatesthe interpretation of sedimentary palaeomagnetic records. There-fore, greigite is also considered to be an unreliable magnetic carrierin many cases (Florindo & Sagnotti 1995; Thompson & Cameron1995; Horng et al. 1998; Richter et al. 1998; Jiang et al. 2001;Roberts et al. 2005; Roberts & Weaver 2005; Sagnotti et al. 2005;Rowan & Roberts 2006). In other cases, early greigite formationhas been demonstrated. For example greigite can form within astagnant water column (e.g. Cutter & Kluckhohn 1999) and sed-imentary greigite has been demonstrated to form within years ordecades (e.g. Pye 1981; Reynolds et al. 1999). Recently, greigitehas also been shown to carry a primary palaeomagnetic signal inMiocene-Pliocene brackish to fresh water environments (Vasilievet al. 2007, 2008).

In sediments of the Monte dei Corvi section, the behaviour ofthe NRM during thermal demagnetization and preliminary rockmagnetic experiments (acquisition of isothermal remanent mag-netization; IRM) suggested that beside magnetite the iron sulphidegreigite is most likely the main magnetic remanence carrier (Husinget al. 2007). Here, we investigate the nature and origin of this carrierby means of detailed rock magnetic experiments and electron mi-croscopic analysis. We discuss the NRM acquisition processes forgreigite. These NRM acquisition mechanisms have significant im-plications for the ages of reversal boundaries, which clearly dependon the type of magnetic remanence carriers present in a sediment.

2 B A C KG RO U N D

2.1 Lithology

The marine, pelagic to hemipelagic, Miocene sediments of theMonte dei Corvi section are exposed along the cliffs betweenAncona and Il Trave in the Conero Riviera, northern Italy(Montanari et al. 1997) (Fig. 1). The section encompasses the up-

Figure 1. Location of the Monte dei Corvi composite section in northernItaly. MdC: Monte dei Corvi Beach section. CB: Corvi Beach and extensionsection along the beach; Sar: La Sardella section along La Sardella track(Montanari et al. 1997; Hilgen et al. 2003; Husing et al. 2007, 2009).

per part of the Serravallian, the entire Tortonian and the lowerpart of the Messinian stage and is composed of cyclic alterna-tions of marls, marly limestones and sapropels (Fig. 2). Large-scalenon-repetitive stratigraphic changes allow discrimination betweena Lower, Brownish, Rossini and Euxinic Shale Interval (for moredetailed description of the lithology and sedimentary cycles, thereader is referred to Montanari et al. (1997), Hilgen et al. (2003)and Husing et al. (2007, 2009)).

Sapropel formation in the Mediterranean has been explained byincreased primary production, leading to a higher oxygen demand,and to development of more stagnant conditions resulting in reducedoxygen replenishment of bottom waters (Rossignol-Strick et al.1982; Calvert et al. 1992; Rohling 1994) related to climate-inducedincreases in freshwater input from the African and European conti-nents (Rossignol-Strick et al. 1982; Rossignol-Strick 1985; Hilgen1991; Rohling & Hilgen 1991; Rohling 1994). Distinct sapropelpatterns observed at Monte dei Corvi reflect astronomical cyclicitywith precession, obliquity and eccentricity all evident in the sedi-mentary succession (Hilgen et al. 2003; Husing et al. 2007). Thisclimatic link has led to correlation of individual ancient sapropelsto maxima in (65◦N) summer insolation and thus to minima in theprecession index using the same phase relations with the orbitalparameters as for younger Mediterranean sapropels (Hilgen 1991;Hilgen et al. 2003).

2.2 NRM demagnetization

Initial NRM intensities (at 20 ◦C) are generally low, ranging between0.001 and 1.804 mAm−1. Higher NRM intensities occur between78 and 95 m and between 112 and 128 m (Fig. 2). During stepwisethermal demagnetization, a viscous component was removed be-tween room temperature (20 ◦C) and 100 ◦C (Figs 2a–d), which isinterpreted to be a secondary post-tilt overprint (Hilgen et al. 2003;Husing et al. 2007). A second magnetic component was generallyremoved up to 260–280 ◦C (Fig. 2d), but where NRM intensitiesare higher, thermal demagnetization was useful up to ∼360 ◦C oreven up to 400 to 420 ◦C (Figs 2a and c). The high-temperaturecomponent can have either normal or reversed polarity and is in-terpreted to be the primary component (Husing et al. 2007, 2009).The decrease in remanence up to ∼360 ◦C (Fig. 2c) is similar to thatreported for greigite (Roberts 1995; Chang et al. 2008), with rema-nence decrease between 250 and 280 ◦C (Fig. 2d) being explainedby thermal alteration of greigite (Roberts 1995). To enable identi-fication of the primary magnetization before the onset of thermalbreakdown of greigite, selected samples were subjected to alter-nating field (AF) demagnetization after heating to 280 ◦C. In somesamples, the remanence diverged from a linear decay toward theorigin above 25 or 30 mT (Fig. 2e), which suggests that the greigitemay have acquired a gyroremanent magnetization (GRM) duringAF demagnetization (Snowball 1997a,b; Sagnotti & Winkler 1999;Stephenson & Snowball 2001). In other samples, the NRM decaywas linear toward the origin up to 100 mT (Fig. 2b), which indicatesthat greigite has apparently not acquired a GRM during AF demag-netization (see Rowan & Roberts 2006) and suggests the presenceof (fine-grained) magnetite.

Plotting the primary directions on an equal area projection re-veals two clusters that are located in the SE and NE quarters of thestereographic projection (Fig. 2f). The directional non-antipodalitycan be explained by overlapping blocking temperature spectra ofthe primary and secondary components, resulting in non-ideal

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 127

Figure 2. Lithologic and magnetic stratigraphy and correlation to the GPTS of the Monte dei Corvi section (Hilgen et al. 2003; Husing et al. 2007, 2009). The0 m level is based on the definition of Hilgen et al. (2003). Tortonian GSSP = Tortonian Global Stratotype Section and Point; located in the midpoint of thesapropel in basic cycle 76. T–M denotes the Tortonian–Messinian boundary. Black (white) circles denote reliable (unreliable) ChRM directions. In the polaritycolumn, black (white) zones indicate normal (reversed) polarity intervals. Thermal and AF demagnetization behaviour of selected samples are shown on theright, where temperatures used for thermal demagnetization are indicated in ◦C and AF demagnetization steps are indicated in mT. (a) COR493.A at 131.07 m;(b) COR667.A at 122.15 m; (c) MCOR1044.A at 81.87 m; (d) MCOR1009.B at 79 m; (e) MCOR1009.A at 79 m. Based on the trend of NRM intensities afterthermal treatment at 100 ◦C and the thermal and AF demagnetization behaviour of the samples (a to e), the section is subdivided into an ‘older interval’ and a‘younger interval’. A hiatus of ∼80 kyr at ∼133 m was confirmed by astronomical tuning (see Husing et al. 2009). Consequently, cycle 237 that contains thelast palaeomagnetic sample is astronomically dated at 6.919 Ma.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

128 S. K. Husing et al.

identification of the primary magnetization component (Hilgen et al.2003; Husing et al. 2007).

The resulting characteristic polarity pattern (Fig. 2) was corre-lated straightforwardly to all polarity zones between normal polarityChron C5Ar.1n and Chron C3Br.1n, which spans the age intervalbetween 12.8 and 7.2 Ma. Even short polarity intervals of 30 to40 kyr duration were recorded (Husing et al. 2007).

Based on the different NRM characteristics during thermal andAF demagnetization, we refer to: (1) the older interval (26 to 112 m;12.8 to 8.7 Ma), which has generally lower NRM intensities and amarked decrease in remanence between 280 and 360 ◦C, and (2) theyounger interval (112 to 135 m; 8.7 to 6.9 Ma), which has generallyhigher NRM intensities, higher magnetic susceptibility values, andstable remanences above 360 ◦C (see Fig. 2).

3 M E T H O D S

Magnetic property analyses were carried out through IRM acquisi-tion, thermal demagnetization of a three-axis IRM (Lowrie 1990),high-temperature thermomagnetic runs and low-temperature mag-netic remanence measurements. These were complemented by scan-ning electron microscope (SEM) analyses.

IRM acquisition curves were obtained for 81 samples in sevenintervals (at ∼30.4, ∼54, ∼65.6, ∼79.0, 117, 122 and 125 m), eachcomprising two to four basic cycles. Most samples from the lowerfour depth intervals had 25 to 30 peak fields applied up to 2 T. Fewersteps were applied in the low-field range because of the inaccuracyof field settings with the PM4 pulse magnetizer. The IRM was sub-sequently measured using a horizontal 2G Enterprises DC SQUIDmagnetometer (noise level 3 × 10−12 Am2). All samples from theupper three intervals and eight additional samples from the lowerfour intervals were measured using a robotized SQUID magnetome-ter (noise level 1–2 × 10−12 Am2), which allows measurements of42 data points after IRM acquisition up to 700 mT. Data acquiredfrom the robotized system were corrected for the tray and sampleholder magnetization by (1) subtracting the magnetization of thetray, which was magnetized at 700 mT prior to measurements and(2) subtracting the magnetization of the sample holder includingthe residual remanence of the sample after AF demagnetization at300 mT. Each sample was AF demagnetized at 300 mT prior toIRM acquisition so that they were in the AF demagnetized startingstate (Heslop et al. 2004).

All data were subjected to component analysis of cumulativelog-Gaussian (CLG) curves (Kruiver et al. 2001), where every IRMcurve was decomposed into a number of CLG curves. These canbe individually characterized by their saturation IRM (SIRM), re-manent acquisition coercive force (B1/2), and dispersion parameter(DP). Results of the CLG analysis can be used to select suitable peakfields for three-axis thermal demagnetization of the IRM (Lowrie1990). Component fitting of IRM acquisition curves is best achievedusing three components: (1) between 0 and 35 mT, (2) between 35and 300 mT, and (3) between 300 mT and 2 T (Husing et al. 2007).Selected samples were subsequently imparted with an IRM using2 T, 300 mT and 35 mT fields along three perpendicular axes andwere stepwise thermally demagnetized from room temperature upto 600 ◦C.

Magnetic extracts were obtained from bulk sediment followingthe set up and separation procedure described by Dekkers (1988b).Heavy liquid extracts were obtained using the protocol of Frankeet al. (2007). High-temperature thermomagnetic runs were per-formed on bulk sediments, magnetic extracts and heavy liquid

extracts in air using a modified horizontal translation-type CurieBalance (Mullender et al. 1993). Samples were selected from eachlithology and were heated in air using an oscillating field between150 and 300 mT. The heating run started at room temperature. Heat-ing and cooling cycles were used in which the sample was heatedto successive maximum temperatures that were progressively in-creased by 50 to 100 ◦C. This was done to discriminate real magneticbehaviour from chemical alteration. Most samples were heated to720 ◦C—a few were heated to 640 ◦C only—with heating rates of5 ◦C per min and cooling rates of 10 ◦C per min.

Low-temperature magnetic remanence measurements were per-formed on a Quantum Design XL7 Magnetic Properties Measure-ment System (MPMS) with a noise level of ∼10−11Am2 at theUniversity of Bremen, Germany. Heavy liquid extracts of three rep-resentative samples were selected based on the following criteria: (1)behaviour during thermal demagnetization that suggests the pres-ence of greigite, (2) relatively high NRM intensities, (3) samplesfrom different lithologies, and (4) samples that straddle one basicprecession cycle. These selected samples are: MCOR1062 in lime-stone at 84.65 m, MCOR1064 in marl at 84.86 m and MCOR1066 ina sapropel at 85.02 m. Between 0.05 and 0.25 g of material was fixedin gelatin capsules (for sample preparation and instrument details,see Frederichs et al. 2003). Low-temperature remanence warmingcurves (10–300 K) were obtained at 1 K increments. For zero-fieldcooling (ZFC) remanence measurements, the sample was cooled inzero field to 10 K. A 2 T field was then applied and switched offbefore warming the sample back to room temperature (300 K). Forsubsequent field cooling (FC) remanence measurements, a 2 T fieldwas applied during initial cooling from room temperature to 10 K.The field was then switched off before warming back to 300 K. Thefirst derivatives of both ZFC and FC remanences versus temperaturewere calculated to visualize the changes in magnetic behaviour dur-ing sample warming. Finally, room temperature SIRM (RT-SIRM)cycles were performed by applying a 5 T field at room temperature(300 K) before the sample was cycled in zero field from 300 to10 K (cooling) and back to room temperature at 300 K (warming).Secondary electron (SE) and backscattered electron (BSE) imag-ing were carried out to visualize the magnetic mineral assemblageusing a FEI XL SFEG scanning electron microscope (SEM) at theCentre of Electron Microscopy Utrecht (EMU), Utrecht University,The Netherlands. Energy dispersive X-ray spectroscopy (EDS) wasused to determine the elemental composition of individual parti-cles in polished thin sections of bulk sediment and dispersed heavyliquid extracts. Particle sizes were estimated using the Scandiumimaging software. Atomic ratios of iron to sulphur were comparedto distinguish cubic pyrite (FeS2 = 0.5) and greigite (Fe3S4 = 0.75),monoclinic pyrrhotite (Fe7S8 = 0.875) and mackinawite (FeS = 1).

4 RO C K M A G N E T I C R E S U LT S

4.1 IRM acquisition component analysis

CLG analysis and interpretation of IRM acquisition curves aimsto introduce and fit the smallest number of components to avoidover-interpretation of data. Here, we use results of the thermal de-magnetization behaviour of NRM, which indicate that greigite ismost likely the main magnetic component. In the younger interval,(fine-grained) magnetite is present in addition to greigite. In allsamples, components are labelled the same way starting with thecomponent fitted to the low field range (Fig. 3, Table 1).

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 129

Fig

ure

3.R

epre

sent

ativ

eex

ampl

esof

CL

Gin

terp

reta

tion

ofIR

Mac

quis

itio

ncu

rves

.(a

)an

d(b

)ar

esa

mpl

esfr

omth

eol

der

inte

rval

wit

htw

ogr

eigi

teco

mpo

nent

s[C

ompo

nent

3=

(bio

geni

c)gr

eigi

tean

dC

ompo

nent

4=

auth

igen

ic,i

norg

anic

grei

gite

],(c

)an

d(d

)ar

esa

mpl

esfr

omth

eyo

unge

rin

terv

al,w

ith

the

addi

tion

alfi

ne-g

rain

ed(m

agne

tota

ctic

)m

agne

tite

Com

pone

nt2.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

130 S. K. Husing et al.T

able

1.R

esul

tsof

CL

Gan

alys

isof

IRM

acqu

isit

ion

curv

es.

Com

pone

nt1

Com

pone

nt2

Com

pone

nt3

Com

pone

nt4

Com

pone

nt5

The

rmal

activ

atio

n(M

agne

tota

ctic

)M

agne

tite

Bio

geni

c(m

agne

tota

ctic

)G

reig

ite

Aut

hige

nic

Gre

igit

eG

oeth

ite

Lev

elTo

talI

RM

B1/

2D

PC

ontr

ibut

ion

B1/

2D

PC

ontr

ibut

ion

B1/

2D

PC

ontr

ibut

ion

B1/

2D

PC

ontr

ibut

ion

B1/

2D

PC

ontr

ibut

ion

Sam

ple

code

(m)

(10−6

Am

2kg

−1)

(mT

)(l

og10

mT

)(p

erce

nt)

(mT

)(l

og10

mT

)(p

erce

nt)

(mT

)(l

og10

mT

)(p

erce

nt)

(mT

)(l

og10

mT

)(p

erce

nt)

(mT

)(l

og10

mT

)(p

erce

nt)

Cyc

le1

MC

OR

561.

B29

.89

9.1

10.0

0.33

1250

.10.

2041

50.1

0.40

4463

1.0

0.30

3M

CO

R56

2.B

30.0

18.

810

.00.

3013

55.0

0.19

3955

.00.

4044

398.

10.

305

MC

OR

563.

A30

.12

7.4

10.0

0.30

1151

.30.

2232

51.3

0.40

5479

4.3

0.30

3M

CO

R56

4.A

30.1

87.

810

.00.

3010

51.3

0.22

5051

.30.

4637

794.

30.

293

MC

OR

565.

A30

.410

.910

.00.

3012

50.1

0.21

3750

.10.

4048

631.

00.

304

MC

OR

566.

A30

.45

8.7

10.0

0.30

1549

.00.

2356

49.0

0.40

2339

8.1

0.30

6M

CO

R56

7.B

30.5

19.

012

.60.

3018

50.1

0.22

5650

.10.

4021

398.

10.

306

MC

OR

568.

B30

.56

7.9

12.6

0.30

1150

.10.

2239

49.0

0.42

4610

00.0

0.30

3M

CO

R56

9.A

30.5

98.

510

.00.

3012

51.3

0.23

4451

.30.

3840

794.

30.

305

MC

OR

570.

A30

.72

5.3

7.9

0.30

949

.00.

2246

49.0

0.40

4079

4.3

0.30

5M

CO

R57

1.A

30.8

49.

012

.60.

3017

51.3

0.22

4551

.30.

3533

446.

70.

304

MC

OR

577.

A31

.53

10.8

12.6

0.30

1550

.10.

2044

50.1

0.35

3750

1.2

0.30

4M

CO

R53

0∗

26.1

518

.515

.80.

255

50.1

0.21

5150

.10.

3634

218.

80.

309

MC

OR

533

∗26

.49.

315

.80.

246

51.3

0.21

5551

.30.

3532

281.

80.

306

Cyc

le2

MC

OR

751.

B53

.46

14.0

12.6

0.33

1852

.50.

2238

52.5

0.35

3750

1.2

0.35

6M

CO

R75

2.B

53.5

613

.810

.00.

3514

52.5

0.22

3852

.50.

3537

562.

30.

5011

MC

OR

753.

A53

.67

14.2

10.0

0.35

1351

.30.

2235

51.3

0.35

3910

00.0

0.60

14M

CO

R75

4.B

53.6

813

.212

.60.

3316

51.3

0.22

3051

.30.

3542

316.

20.

5512

MC

OR

755.

A53

.98

8.1

12.6

0.34

1652

.50.

2237

52.5

0.36

4279

4.3

0.30

5M

CO

R75

6.B

54.2

510

.512

.60.

3315

51.3

0.23

3951

.30.

3540

794.

30.

406

MC

OR

757.

B54

.32

8.7

12.6

0.33

1650

.10.

2243

50.1

0.35

3563

1.0

0.40

6M

CO

R75

9.A

54.4

25.

714

.10.

3321

51.3

0.22

4951

.30.

3521

354.

80.

479

MC

OR

760.

A54

.59

6.5

12.6

0.35

2152

.50.

2038

52.5

0.35

3131

6.2

0.55

10M

CO

R74

8∗

53.0

516

.919

9.5

0.30

752

.50.

2561

52.5

0.39

2116

.60.

2511

MC

OR

750

∗53

.32

16.9

199.

50.

309

52.5

0.25

6452

.50.

3920

16.6

0.25

7M

CO

R75

1.B

53.4

618

.715

.80.

257

53.7

0.24

4553

.70.

3637

199.

50.

3011

Cyc

le3

MC

OR

851.

A64

.43

7.0

14.1

0.32

1953

.70.

2037

53.7

0.35

3956

2.3

0.30

6M

CO

R85

4.B

64.7

313

.812

.60.

3515

50.1

0.21

4350

.10.

3837

631.

00.

304

MC

OR

855.

B65

.11

4.0

12.6

0.30

1750

.10.

2042

50.1

0.35

3531

6.2

0.30

5M

CO

R85

6.B

65.2

69.

110

.00.

3013

50.1

0.22

4950

.10.

3533

501.

20.

304

MC

OR

857.

B65

.35

15.9

12.6

0.26

2255

.00.

2047

55.0

0.35

2539

8.1

0.40

6M

CO

R85

8.B

65.4

611

.512

.60.

3017

51.3

0.22

4151

.30.

3537

398.

10.

305

MC

OR

859.

A65

.53

7.6

12.6

0.30

1751

.30.

2239

51.3

0.35

3950

1.2

0.33

5M

CO

R86

0.A

65.7

48.

810

.00.

3216

51.3

0.23

4751

.30.

3533

501.

20.

335

MC

OR

861.

A65

.85

8.8

12.6

0.34

2052

.50.

2050

52.5

0.35

2339

8.1

0.33

7M

CO

R86

2.A

65.9

114

.610

.00.

3312

50.1

0.22

3950

.10.

3542

501.

20.

306

MC

OR

863.

B65

.99

16.7

12.6

0.33

1652

.50.

2242

52.5

0.35

3544

6.7

0.30

7M

CO

R86

4.B

66.0

914

.612

.60.

3315

51.3

0.22

4451

.30.

3534

316.

20.

397

Cyc

le4

MC

OR

1003

.A78

.43

29.3

12.6

0.35

1763

.10.

2241

63.1

0.35

3863

1.0

0.33

4M

CO

R10

04.B

78.4

747

.012

.60.

3316

61.7

0.22

4361

.70.

3536

501.

20.

405

MC

OR

1005

.A78

.63

30.5

12.6

0.36

1358

.90.

2339

58.9

0.37

4463

1.0

0.30

3

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 131

Tab

le1.

(Con

tinu

ed.)

Com

pone

nt1

Com

pone

nt2

Com

pone

nt3

Com

pone

nt4

Com

pone

nt5

The

rmal

activ

atio

n(M

agne

tota

ctic

)M

agne

tite

Bio

geni

c(m

agne

tota

ctic

)G

reig

ite

Aut

hige

nic

Gre

igit

eG

oeth

ite

Lev

elTo

talI

RM

B1/

2D

PC

ontr

ibut

ion

B1/

2D

PC

ontr

ibut

ion

B1/

2D

PC

ontr

ibut

ion

B1/

2D

PC

ontr

ibut

ion

B1/

2D

PC

ontr

ibut

ion

Sam

ple

code

(m)

(10−6

Am

2kg

−1)

(mT

)(l

og10

mT

)(p

erce

nt)

(mT

)(l

og10

mT

)(p

erce

nt)

(mT

)(l

og10

mT

)(p

erce

nt)

(mT

)(l

og10

mT

)(p

erce

nt)

(mT

)(l

og10

mT

)(p

erce

nt)

MC

OR

1006

.A78

.77

42.3

12.6

0.32

1560

.30.

2237

60.3

0.38

4363

1.0

0.50

5M

CO

R10

07.B

78.7

943

.312

.60.

3313

58.9

0.24

4858

.90.

3832

398.

10.

406

MC

OR

1008

.B78

.81

39.9

12.6

0.33

1560

.30.

2342

60.7

0.38

3843

6.5

0.50

5M

CO

R10

09.A

7923

.212

.60.

3514

58.9

0.22

4258

.90.

3834

398.

10.

559

MC

OR

1010

.A79

.17

20.8

12.6

0.29

1458

.90.

2346

58.9

0.38

3566

0.7

0.30

5M

CO

R10

11.A

79.3

416

.812

.60.

3014

58.9

0.24

4258

.90.

3838

478.

60.

406

MC

OR

1012

.B79

.42

35.2

12.6

0.33

1760

.30.

2249

60.3

0.40

2850

1.2

0.50

5M

CO

R10

13.B

79.4

764

.112

.60.

3312

57.5

0.22

4257

.50.

3843

416.

90.

263

MC

OR

1014

.A79

.59

31.6

15.8

0.33

1756

.20.

2344

56.2

0.38

3236

3.1

0.47

7M

CO

R10

00C

∗78

.125

.815

.80.

257

69.2

0.22

4269

.20.

3024

257.

00.

227

MC

OR

1001

∗78

.23

30.9

15.8

0.20

457

.50.

2242

57.5

0.40

4116

6.0

0.40

8M

CO

R10

18∗

80.0

152

.515

.80.

255

55.0

0.24

4855

.00.

3427

199.

50.

4014

Cyc

le5

MC

OR

447

∗11

5.97

81.5

12.6

0.20

530

.20.

205

56.2

0.23

7457

.50.

367

316.

20.

3010

MC

OR

449

∗11

6.1

35.1

15.8

0.26

932

.40.

1916

64.6

0.20

5364

.60.

327

213.

80.

3216

MC

OR

451

∗11

6.4

102.

512

.60.

254

33.9

0.22

1055

.00.

2257

55.0

0.37

2031

6.2

0.37

10M

CO

R45

4∗

116.

7912

1.0

12.6

0.20

352

.50.

2046

52.5

0.36

4112

58.9

0.42

10M

CO

R45

7∗

117.

2111

6.7

15.8

0.34

830

.20.

207

56.2

0.22

4856

.20.

3030

398.

10.

307

MC

OR

460

∗11

7.48

451.

815

.80.

345

31.6

0.20

663

.10.

2153

63.1

0.34

3150

1.2

0.34

5M

CO

R46

3∗

117.

8136

1.0

15.8

0.28

831

.60.

207

57.5

0.21

5557

.50.

3422

398.

10.

348

MC

OR

466

∗11

8.06

351.

015

.80.

345

31.6

0.20

1764

.60.

2147

64.6

0.30

1339

8.1

0.30

19M

CO

R46

7∗

118.

0917

9.5

15.8

0.27

531

.60.

2011

60.3

0.22

4260

.30.

3025

354.

80.

3017

MC

OR

471

∗11

8.33

52.5

15.8

0.28

731

.60.

2015

64.6

0.23

4264

.60.

3527

316.

20.

3010

MC

OR

474

∗11

8.55

103.

015

.80.

285

31.6

0.20

1660

.30.

2243

60.3

0.35

2934

6.7

0.15

8M

CO

R47

5∗

118.

6867

.715

.80.

257

31.6

0.20

1358

.90.

2247

58.9

0.34

2131

6.2

0.34

12M

CO

R47

8∗

118.

8571

.515

.80.

256

31.6

0.20

2060

.30.

2145

60.3

0.35

1763

1.0

0.45

13

Cyc

le6

PC

OR

655

∗12

2.84

134.

015

.80.

207

31.6

0.20

2560

.30.

2238

60.3

0.30

1928

1.8

0.30

10P

CO

R66

1∗

122.

5817

22.0

15.8

0.20

231

.60.

2011

70.8

0.19

6770

.80.

3517

794.

30.

252

PC

OR

667

∗12

1.15

2120

.017

.80.

203

30.2

0.20

1374

.10.

2079

74.1

0.40

379

4.3

0.40

3M

CO

R51

8∗

122.

1112

63.0

15.1

0.20

236

.30.

2028

77.6

0.17

6177

.60.

352

251.

20.

306

MC

OR

519

∗12

2.2

938.

015

.10.

204

37.2

0.19

3279

.40.

1653

79.4

0.38

322

3.9

0.40

8M

CO

R52

1∗

122.

3911

52.0

15.8

0.26

335

.50.

1930

77.6

0.16

5677

.60.

353

177.

80.

328

MC

OR

522

∗12

2.45

240.

015

.80.

236

34.7

0.16

2070

.80.

1522

70.8

0.30

1310

00.0

0.75

40

Cyc

le7

PC

OR

599

∗12

4.37

203.

015

.80.

252

31.6

0.20

970

.80.

2142

70.8

0.39

1463

1.0

0.55

32P

CO

R60

2∗

124.

5211

40.0

15.8

0.20

330

.90.

2011

74.1

0.22

7474

.10.

363

794.

30.

3010

PC

OR

605

∗12

4.64

2365

.015

.80.

202

31.6

0.18

1272

.40.

1743

72.4

0.33

3312

58.9

0.40

10P

CO

R60

8∗

121.

7830

2.0

15.8

0.30

231

.60.

2013

83.2

0.20

6683

.20.

3414

631.

00.

305

PC

OR

611

∗12

4.93

868.

015

.80.

201

31.6

0.20

1272

.40.

2058

72.4

0.40

2672

4.4

0.20

3P

CO

R62

3∗

125.

4812

86.0

15.8

0.28

230

.20.

1816

72.4

0.17

4672

.40.

3428

891.

30.

367

PC

OR

626

∗12

5.69

1480

.015

.80.

203

31.6

0.20

1974

.10.

2067

74.1

0.34

363

1.0

0.30

7P

CO

R43

3∗

127.

6763

01.0

15.8

0.20

232

.40.

2017

75.9

0.21

7457

5.4

0.33

7

The

tota

lIR

Mis

calc

ulat

edby

addi

ngth

eS

IRM

ofin

divi

dual

com

pone

nts

ofre

spec

tive

sam

ples

.Ast

eris

k(∗

)de

note

sth

atsa

mpl

esw

ere

mea

sure

dby

the

robo

tize

dpr

oced

ure.

B1/

2is

the

rem

anen

tacq

uisi

tion

coer

cive

forc

ean

dD

Pis

the

dist

ribu

tion

para

met

er[s

mal

l(la

rge)

valu

esin

dica

tena

rrow

(wid

e)di

stri

buti

ons

ofgr

ains

inth

ere

pres

entiv

eco

mpo

nent

].C

ompo

nent

1(t

herm

alac

tivat

ion)

isph

ysic

ally

mea

ning

less

.C

ompo

nent

5th

atis

inte

rpre

ted

asgo

ethi

te,i

sm

ostl

ikel

ya

prod

ucto

fw

eath

erin

gpr

oces

ses.

See

text

for

deta

ils.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

132 S. K. Husing et al.

Component 1 was fit in the low field interval, with B1/2 =∼12 mT.The low B1/2 value and the relatively small contribution (about10 per cent of the total IRM) indicates that this component is re-lated to thermal activation resulting in a skewed distribution (Egli2004a,b; Heslop et al. 2004), and should therefore not be given anyphysical meaning. Nevertheless, this observation is consistent withthat of Rowan & Roberts (2006) who suggested that many greigitesamples contain substantial SP contributions.

Component 2 is present only in the younger interval (Fig. 3,Table 1). This component has a B1/2 value of ∼35 mT and a narrowgrain size distribution indicated by a small DP of ∼0.2 log10 mT.These values remain relatively constant throughout the youngerinterval of the section. The B1/2 values are typical of magnetite andthe narrow gain size distribution is indicative of a magnetotactic(Kruiver et al. 2001) rather than a detrital origin. This magnetotacticmagnetite contributes ∼15 per cent to the total IRM and is thereforeof minor importance.

Components 3 and 4 are both fitted with the same B1/2 values be-tween 49 and 83 mT, but the distribution parameters for component3 are smaller than those for component 4. The B1/2 values over-lap with remanent coercive force values reported for greigite andpyrrhotite (Dekkers 1988a; Rochette et al. 1990; Snowball 1991;Roberts 1995; Snowball 1997a,b; Horng et al. 1998; Sagnotti &Winkler 1999). We interpret this component as greigite rather thanpyrrhotite on the basis of the above described thermal behaviourof the NRM and thermomagnetic analysis (see later). Differenttypes of sedimentary greigite formation most likely result in differ-ent grain size distributions for different greigite populations (e.g.Vasiliev et al. 2007). We therefore chose to distinguish between twopopulations of greigite (this interpretation is supported by SEMobservations discussed in Section 4.4.). The first population (Com-ponent 3) is composed of fine-grained greigite with a narrow DP(∼0.2 log10 mT) and high B1/2 values. This indicates single do-main (SD) particle sizes (Roberts 1995; Snowball 1997a) that areexpected if this greigite component has a biogenic origin (Vasilievet al. 2008). The relatively constant and higher contribution ofComponent 3 to the total IRM (between 30 and 60 per cent)—independent of lithology—suggests that this greigite component isthe main magnetic carrier throughout the entire section.

The second population (Component 4) is characterized by a widergrain size distribution (DP = ∼0.36 log10 mT) that would be ex-pected for authigenic greigite resulting from bacterially mediatedformation. The contribution of Component 4 (<30 per cent) to thetotal IRM is smaller than that of greigite Component 3, particularlyin the younger interval. The survival of fine-grained (magnetotac-tic) magnetite Component 2 and lower concentrations of greigiteComponent 4 indicate that sulphidization processes were probablynot as severe in the younger sediments as in the older sediments.

Component 5 is fitted to the high-field interval and has B1/2

values between 300 mT and 1 T and a large DP (∼0.32 log10 mT).Its presence is also indicated by the high field ‘tail’ observed instandardized acquisition plots (SAP). This component is responsiblefor the non-saturation of IRM in all samples (e.g. Kruiver et al.2001) and could be carried either by goethite or haematite. Theweak NRM intensities and NRM decrease below 500 ◦C precludesthe presence of haematite. Goethite is likely present as a product ofweathering processes (Dunlop & Ozdemir 1997; France & Oldfield2000; Kruiver et al. 2001; Maher et al. 2004) and its contribution (2to 14 per cent) to the IRM is minor. The B1/2 values could indicate arelatively soft goethite component, but the applied maximum fieldsof only 700 mT and 2 T, respectively, are not sufficient for goethitesaturation.

4.2 3-axis thermal demagnetization of IRM

For all samples, the medium coercivity (35 and 300 mT) fraction isa constant contributor (Fig. 4). This coercivity range is indicative ofmagnetite, maghemite, titanomagnetite, greigite and/or pyrrhotite.Based on IRM acquisition curve analysis, thermomagnetic data andSEM observation (see following paragraphs), we can rule out thepresence of maghemite, titanomagnetite and pyrrhotite. The soft (0to 35 mT) and hard (300 mT to 2 T) fractions contribute only asmall percentage to the total remanence (Fig. 4).

For the older interval (Figs 4a and b) and some samples from theyounger interval (Fig. 4c), the medium and soft fractions have vir-tually the same thermal characteristics. The main inflection of thethermal demagnetization curves of the medium coercivity fractionat around 350 ◦C is indicative of the thermal alteration of greig-ite (Roberts 1995; Chang et al. 2008). The persistence of a re-manence to ∼500–550 ◦C indicates a contribution of fine-grained(titano)magnetite. Backward extrapolation of the thermal demagne-tization curves after the inflection point at ∼350 ◦C suggests that thecontribution of the greigite and magnetite component is 50 per centeach to the total IRM. This is in contrast with NRM intensities beingclose to the noise level of the magnetometer after thermal demag-netization above 360 ◦C. It indicates that after thermal alteration ofgreigite at about 360 ◦C, it appears that a magnetite contribution tothe NRM is minimal, for which we do not have a straightforwardexplanation. Note that the magnetite NRM is low with respect togreigite, but that the magnetite IRM is appreciable with respect togreigite. The sharp decrease in remanence of the hard coercivityfraction at ∼120 ◦C (Fig. 4b) indicates the presence of goethite.Not all samples contain a goethite component. We conclude thatthe main magnetic remanence in samples from the older intervalis therefore carried by greigite. This is consistent with the IRMcomponent analysis for greigite Components 3 and 4 (with a B1/2

between 50 and 80 mT) and NRM decrease between 280 and 360 ◦Cduring thermal demagnetizations.

For the younger interval (Figs 4c and d), inflection of the thermaldemagnetization curve for the intermediate and low coercivity frac-tion at around 350 ◦C indicates the presence of greigite (Figs 4a andd; relatively small decrease in samples MCOR525). The IRM ofthe intermediate coercivity fraction steadily decreases up to 600 ◦C,whereas the IRM of the low coercivity fraction decreases at ∼450 ◦C(Fig. 4d) or at 580 ◦C (Fig. 4c). The low coercivity fraction rangesfrom 0 to 35 mT and therefore includes contributions from mag-netite. The maximum unblocking temperature of both fractions justbelow 600 ◦C indicates the presence of relatively pure magnetite.The small increase at around 500 ◦C in the low coercivity fractionmay indicate pyrite oxidation (Passier et al. 2001). Magnetite andgreigite appear to be the main contributors to the remanence, andin samples like MCOR525 (Fig. 4d) the IRM contribution of mag-netite is higher than of greigite. This interpretation is supported bythe NRM decrease up to ∼400 ◦C during thermal demagnetizationand GRM acquisition up to 100 mT during AF demagnetization.

4.3 Thermomagnetic experiments

4.3.1 High-temperature Curie balance measurements

Heating curves for bulk sediment samples (Fig. 5a—older interval;5b—younger interval) are characterized by a hyperbolic decreaseup to 420 ◦C that indicates the dominance of paramagnetic mate-rial. Between 420 and 500 ◦C, the magnetization increases signif-icantly and is subsequently removed at 580 ◦C. The considerable

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 133

Figure 4. Representative examples of thermal demagnetization of composite three-axis IRM. (a) and (b) Samples from the older interval, (c) and (d) samplesfrom the younger interval. The significant NRM decrease at ∼360 ◦C indicates thermal alteration of greigite. The persistence of a remanence to 500–550 ◦Cindicates the presence of magnetite that is not observed for samples (a) and (b) during thermal demagnetization of the NRM (see Fig. 2 and text for discussion).The data are more noisy in the high temperature range.

magnetization increase between 420 and 500 ◦C is due to the for-mation of magnetic iron oxides and thermal breakdown of para-magnetic pyrite (Krs et al. 1992; Passier et al. 2001). The Curietemperature (Tc) of this newly formed iron oxide (Tc = 580 ◦C)is indicative of magnetite. Between 580 and 700 ◦C, the warmingand cooling curves do not undergo further changes and are re-versible. Thermal alteration to form magnetite has resulted in gen-erally higher magnetizations after heating compared to unheatedsamples.

Samples obtained from magnetic and heavy liquid extracts gen-erally have higher starting magnetizations than bulk sediments, buthave the same hyperbolic shape and magnetization increase above420 ◦C, which indicates that the contribution of paramagnetic min-erals in the extracts is still relatively high. The thermomagnetic dataare irreversible above 250 ◦C, as indicated by warming and cool-ing sub-cycles (see insets in Fig. 5). This irreversibility indicatesthe presence of greigite, and excludes the presence of pyrrhotite.The considerable magnetization increase between 420 and 500 ◦Cis less pronounced in the magnetic extract than in the heavy liquidextract. This is related to the extraction method, whereby pyrite isnot extracted in the magnetic procedure but it is present after heavyliquid separation.

4.3.2 Low-temperature measurements

Low-temperature magnetic experiments were performed to dis-tinguish between magnetite, pyrrhotite, and greigite. Magnetite

and pyrrhotite can be positively identified by the low-temperatureVerwey transition at 117 K for stoichiometric magnetite (Ozdemiret al. 1993; Kosterov 2001) and the 34 K transition for pyrrhotite(Dekkers et al. 1989; Rochette et al. 1990), whereas no low-temperature transition has been reported for greigite (e.g. Roberts1995; Chang et al. 2009). The heavy liquid extracts obtained fromthe sapropel, marl and limestone lithologies have similar ZFC andFC remanence curves (Fig. 6). This suggests that the magneticmineral assemblage is more or less the same for each lithology. Aplausible explanation for the rapid decrease in ZFC and FC rema-nence between 20 and 40 K (Fig. 6) is the presence of ultra-finesuperparamagnetic (SP) greigite particles (Roberts 1995; Changet al. 2009). Similar behaviour has been observed in oxidized syn-thetic magnetite with low unblocking temperatures (Ozdemir et al.1993) and natural magnetite (Smirnov & Tarduno 2000). Mag-netic interaction, surface effects in ultra-fine grains, or the presenceof surface layers around coarser magnetic particles may also ac-count for this rapid decrease in remanence (Pike et al. 2000; Passier& Dekkers 2002). Both ZFC and FC warming curves contain noevidence for any low-temperature transition, which indicates thatneither magnetite nor pyrrhotite is present (Dekkers et al. 1989;Rochette et al. 1990; Fig. 7). The difference between the ZFC andFC curves is largest at 10 K; it decreases upon warming and disap-pears completely at room temperature. This difference is largest inthe sapropel sample (Fig. 6a) and could be caused by the presenceof goethite (Smirnov & Tarduno 2000; Liu et al. 2006; Franke et al.2007).

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

134 S. K. Husing et al.

Figure 5. Selected examples of the high-temperature dependence of magnetization for (a) the older interval and (b) the younger interval, both for bulk sedimentsamples. The same experiment was carried out for (c) a magnetic extract from the older interval and (d) a heavy liquid extract from the older interval. Theirreversible behaviour upon heating to 360 ◦C indicates the presence of greigite (see insets).

The RT-SIRM curve for the sapropel sample (Fig. 7a) increaseslinearly upon cooling to 10 K, whereas the remanence of the marland limestone samples increases distinctively below 20 K (Figs 7band c). The increase in the low-temperature range (∼10 K) mightbe due to magnetic ordering of paramagnetic (clay-) minerals (Coey& Ghose-Subrata 1988). The considerably higher absolute SIRMvalues for the sapropel sample and the significant gain and lossof remanence during RT-SIRM cycling compared to the other twolithologies (marl and limestone) indicates that the sapropel samplecontains a higher concentration of magnetic minerals. This hasalso been observed in other deep-sea sapropels (e.g. Roberts et al.1999).

4.4 Scanning electron microscopy results

SEM observations coupled with EDS analysis can provide impor-tant information for identifying magnetic minerals. The magneticfraction usually represents less than 1 ‰ of the bulk sediment.Observed minerals are mostly clays and partially dissolved micro-fossils. Framboidal pyrite spherules (∼10 μm), which usually occurin the vicinity of fossil remnants (e.g. foraminifera), were observedin dispersed heavy liquid extracts (Fig. 8a) and in polished bulksediment sections (Fig. 9). Larger (between 1.5 and 2 μm) singlecrystals of euhedral pyrite were also observed (Figs 8a and d, Figs9a and b). Early pyrite formation is supported by the typical size ofthe framboids (<10 μm) (Figs 9a, b and c), which indicates syn-genetic formation of iron sulphides and rapid pyrite precipitationfrom solution (Wilkin & Barnes 1996). Euhedral pyrite also fills

space around framboids, which indicates a second phase of diage-netic pyrite formation (Suits & Wilkin 1998; Figs 9b and c). Thepresence of dispersed euhedral pyrite in the limestones and marlsindicates that the initial rate of pyritization was low and that thecrystals formed within the sediment where sulphide had more timeto react with the detrital minerals (Wilkin & Barnes 1996; Morse &Wang 1997).

The distinct monoclinic morphology of pyrrhotite and its typicaloccurrence as plates (Weaver et al. 2002; Larrasoana et al. 2007)was not observed in the studied sediments. In contrast, greigite is acubic mineral and occurs as octahedral crystals (Roberts & Weaver2005), which were clearly detected, particularly in the dispersedsamples (Figs 8b, c and d). EDS elemental analysis of small cu-bic minerals with grain sizes of ∼0.06–1 μm indicate Fe:S ratiosclose to the respective value for greigite (Fig. 8, Figs 9a and b).Discrepancies between measured and stoichiometric Fe:S ratios forgreigite (0.75) and pyrite (0.5) can be explained by uncertainties inthe position of the beam and the beam size exceeding the particlesize, which leads to spectral contributions from the surroundingminerals. For framboidal pyrite, clay mineral coatings can addition-ally account for Mg, Al, Si, K and Ca contaminations (Figs 8a).Irregular particle surfaces in the dispersed heavy liquid extractsalso lead to scattering of X-rays, whereas elemental analysis ofparticles in polished sections is more reliable. The representativeelemental spectra for greigite and pyrite from the polished sec-tion (Fig. 9) are therefore used as a reference for evaluating otheriron sulphide spectra, especially those in the dispersed heavy liquidextracts.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 135

Figure 6. Low-temperature magnetic remanence behaviour of heavy liquidextracts from the older, greigite-only-bearing sediments. No magnetic phasetransitions are evident, either for magnetite (∼120 K), or for pyrrhotite(∼34 K). ZFC and FC remanence curves were measured between 300 and 10K. For comparison, curves are normalized to the magnitude of the respectiveZFC value at 10 K for each sample.

The typical size of individual fine-grained greigite in the heavyliquid separates ranges between 0.06 and 0.1 μm (Fig. 8). Crystalsas small as 80 to 180 nm were also observed (Fig. 8b), which iswithin the expected SD size range (35 to 120 nm; Mann et al. 1990;

Figure 7. Room temperature saturation IRM cycling at low temperaturesfor three selected samples: (a) MCOR1066 from a sapropel, (b) MCOR1064from a marl and (c) MCOR1062 from a limestone. All three samples orig-inate from the same basic precession cycle at ∼85 m in the older, greigite-bearing sediments.

Snowball 1997b; Chang et al. 2008). Fe:S ratios of the dispersedultra-fine-grained particles are close to the composition of greigite(Figs 8b and 9). In polished thin sections, ultra-fine-grained greigiteparticles < 200 nm occur well-dispersed in the sediment (Fig. 9d).Larger crystals with sizes of ∼1 μm have Fe:S ratios typical of pyrite(Figs 8a, 9a, b and c). Massive accumulations of iron sulphides (e.g.Fig. 9c) are relatively uncommon in the studied samples. Greigiteparticles with sizes of ∼700 nm to 1 μm occur in close vicinity ofpyrite framboids (∼10 μm; Fig. 9c).

Two greigite populations can be distinguished. The first has ultra-fine-grained particles in the range of 60 to 200 nm, which arewidespread in all samples. The dispersed distribution of these grainssuggests that they formed early, either in the water column (whichwould have been anoxic during times of sapropel formation), at thesediment–water interface, or within relatively uncompacted shallow

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

136 S. K. Husing et al.

Figure 8. Scanning electron microscope (SEM) images derived from a heavy liquid extract of sample MCOR1062 (limestone) in the older interval. EDSelemental spectra for selected grains indicate compositions close to either pyrite (Py) or greigite (Gr). (a) Framboidal and euhedral pyrite and cubic pyrite withindividual grain size of ∼700 nm, (b) single greigite crystals with grain sizes between 1 μm and ∼60 nm, (c) close-up of greigite particles with grain sizes of∼600 and 700 nm and (d) ∼600 nm greigite particles next to a larger (∼1.4 μm) particle with uncertain iron sulphide (IS) composition; the irregular particlesurface leads to X-ray scattering and therefore uncertain mineralogical classification. Cross symbols indicate the spots where EDS spectra were taken.

sediment. These greigite particles could be biogenic or they couldhave resulted from early diagenetic sulphide precipitation. SEMobservations do not allow us to distinguish between these differentorigins. The second greigite population is characterized by parti-cles that have grown authigenically near pyrite framboids and inclose vicinity of fossil fragments and have observed particles sizesof typically 700 nm to 1 μm (Figs 9a, b and c). This populationis likely to have formed slightly later during post-depositional sul-phidization that resulted in inorganic precipitation of greigite andpyrite (Jiang et al. 2001; Roberts & Weaver 2005; Rowan & Roberts2006).

Particles (∼40 μm) with mixed composition containing iron ox-ide and iron sulphide minerals are also frequently present (Fig. 10).Detailed spot EDS analyses on the particle shown in Fig. 10(a)indicates an iron oxide (IO) composition for the lower area withdecreased electron backscatter, a transitional area with mixed com-position (IOIS) and iron sulphide (IS) composition in the brighterupper part of the particle. In contrast, the particle shown in Fig. 10(b)is entirely an iron sulphide. These particles might represent detri-

tus eroded from older material that was deposited in environmentswhere detrital iron oxides were sulphidized and replaced by ironsulphide minerals. Regardless, these detrital grains are so large thatthey are unlikely to be aligned by the geomagnetic field and aretherefore unlikely to be palaeomagnetically important.

5 D I S C U S S I O N

5.1 Testing for climatic control on magnetic mineralogy

The total IRM, B1/2 and their contributions to the total remanencewere plotted for seven intervals—each comprising three to six pre-cession cycles (Fig. 11)—to check for magnetic mineral variationswith lithology, and therefore with varying climatic conditions. AtMonte dei Corvi, the link between climatic changes and lithologyis given by the relation between sapropel formation and insolationmaxima (Fig. 11; Hilgen et al. 2003; Husing et al. 2007). Althoughthe varying climatic conditions presumably had profound effects on

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 137

Figure 9. SEM micrographs of polished sections from: (a) sample MCOR1065 (limestone) with greigite particles of ∼700 nm, Pyrite of ∼1 μm, (b) sampleMCOR1065 (limestone) with greigite (Gr∗) particle sizes of ∼700 nm and Gr of ∼900 nm, (c) sample MCOR1063 (marl) with fine greigite particles (∼700 nmto 1 μm) authigenically grown near pyrite framboids of ∼10 μm and (d) sample MCOR1063 (marl) showing ultra-fine-grained iron sulphides, likely greigiteparticles <200 nm that are well-dispersed; examples are indicated by white arrows (length ∼2 μm). All samples originate from the older greigite-bearingsediments. Py denotes pyrite and Gr denotes greigite. EDS spectra are shown for representive analyses of pyrite (left) and greigite (right), respectively.

the lithology at Monte dei Corvi, the rock magnetic parameters arerelatively constant throughout the entire section. The higher IRMand NRM intensities in the younger sediments can be explained bythe additional presence of fine-grained (magnetotactic) magnetitealong with greigite. These observations indicate that the magnetic

mineralogy is essentially independent of climate forcing and palaeo-ceanographic changes, and that it is not controlled by detrital input.This supports an authigenic origin of the greigite and magnetitecomponents, and confirms the importance of early diagenetic pro-cesses within the studied sediments.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

138 S. K. Husing et al.

Figure 10. SEM micrographs derived from a polished section of sample MCOR1063 in the older interval. (a) Large magnetite grain (∼40 μm) with evidencefor partial dissolution and replacement by iron sulphide in the lower part of the image, darker electron backscatter represent the iron oxide with iron sulphidereplacement (brighter backscatter) in the upper part, and mixed iron sulphide–iron oxide composition in the middle. Note the rounded grain edges, whichindicate a detrital origin for the particle. (b) A transformed detrital iron oxide particle that has resulted in preservation of the original grain with substantialreplacement of iron sulphide. IO denotes an iron oxide composition, IOIS denotes an intermediate composition between iron oxide and iron sulphide, ISdenotes an iron sulphide composition.

5.2 Greigite formation

Sapropel formation at Monte dei Corvi indicates that anoxic (sul-phidic) bottom water conditions existed (e.g. Passier et al. 1999).Anoxic bottom water has been linked to stratification of the watercolumn as a result of increased fresh waster discharge from riversduring insolation maxima (Rossignol-Strick et al. 1982; Rossignol-Strick 1985; Rohling & Hilgen 1991; Hilgen 1991; Rossignol-Strick1993; Rohling 1994) that increased the preservation of organic mat-ter. It is well known that diagenetic processes in organic-rich sedi-ments may produce new magnetic iron sulphide minerals, such asgreigite (Hilton & Lishman 1985; Snowball & Thompson 1988; Huet al. 1998; Roberts et al. 1999; Fu et al. 2008). Marls and limestoneswere generally deposited in more oxygenated bottom waters. How-ever, poor fossil preservation in all lithologies (marls, limestonesand sapropels) at Monte dei Corvi suggests that suboxic conditionsin the sediment resulted in at least partial dissolution of calcare-ous fossils that will also have caused iron oxide dissolution andiron sulphide formation. Reductive dissolution explains the reducedamount of iron oxides in the older sediments at Monte dei Corvi. Inthe younger interval, sapropels are less frequently developed, NRMintensities are often higher (Fig. 2) and AF and thermal demagneti-zation diagrams indicate the co-occurrence of fine-grained greigiteand magnetite, although anoxic conditions should have caused thedissolution of fine-grained magnetite. This suggests that anoxicconditions were less severely developed or that non-steady state di-agenesis occurred (cf. Passier et al. 2001; Larrasoana et al. 2003),which might have played a significant role in the preservation ofgreigite and fine-grained magnetite components.

5.2.1 Possible biogenic origin of greigite

Two greigite populations are identified in the Monte dei Corvi sedi-ments. Greigite component 3 (with a small DP and high B1/2 values)can be of biogenic origin. Biogenic greigite mineralization occurseither in the water column or near the sediment–water interfaceand would thus produce a syn-sedimentary recording of the ge-

omagnetic field (Kopp & Kirschvink 2008). Fossil magnetotacticgreigite has not been reported in ancient sediments until the recentdemonstration by Vasiliev et al. (2008). These authors also demon-strated that early diagenetic (inorganic) greigite records a stablesyn-depositional palaeomagnetic signal in Miocene–Pliocene sedi-ments of the Paratethys domain (Vasiliev et al. 2007, 2008). Greig-ite formation in the Monte dei Corvi section might be analogousto modern processes in the Black Sea (e.g. Reitner et al. 2005).We suggest that the smallest greigite particles (60 to 200 nm) ob-served during SEM analysis might represent the inferred biogenic(magnetotactic) greigite component (sizes between ∼30 and 80 nm;e.g. Bazylinksi et al. 1995; Posfai et al. 1998; Reitner et al. 2005;Vasiliev et al. 2008), although we cannot unambiguously confirm abiogenic origin for greigite Component 3.

5.2.2 Post-depositional magnetization

Inorganic early diagenetic and much later formation of greigite hasbeen frequently postulated in various marine sediments (e.g. Hornget al. 1998; Roberts et al. 1999; Roberts et al. 2005; Roberts &Weaver 2005). The second greigite population identified by SEManalysis and IRM acquisition experiments (with larger DP thangreigite Component 3) can be associated with authigenic greigitegrowth. It seems that later diagenetic sulphidization processes werelimited, as indicated by the lack of massive euhedral overgrowthof iron sulphides and of sulphides between cleavages of detritaliron-bearing phyllosilicates. Lack of evidence for these processessuggest that later diagenetic sulphidization events (e.g. Canfieldet al. 1992) and remagnetization events (e.g. Roberts & Weaver2005; Rowan & Roberts 2006) did not occur at Monte dei Corvi.

5.2.3 Methane seepage from underlying sediments

The traditional scenario of pyritization arrest (e.g. Berner 1970;Berner 1984; Wilkin & Barnes 1996; Kao et al. 2004) can accountfor inorganic greigite formation and its preservation in all threestudied lithologies at Monte dei Corvi. After greigite formation,

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 139

Figure 11. Rock magnetic parameters derived from IRM acquisition curve analysis of seven intervals spanning several basic cycles: Cycle 1 at ∼30.4 m,cycle 2 at ∼54 m, cycle 3 at ∼65.5 m, cycle 4 at ∼79 m, cycle 5 at ∼117 m, cycle 6 at ∼122 m and cycle 7 at ∼125 m of the stratigraphic section. TheB1/2 values and contributions of greigite Component 3, greigite Component 4 and magnetite Component 2 are plotted in stratigraphic order, which reveals nosignificant changes in values throughout the entire section or between the various lithologies. The insolation curve of La2004(1,1) (Laskar et al. 2004) is plottednext to lithology to indicate the relationship between climatic and lithological changes. Sapropel formation occurred during insolation maxima.

reducing conditions were maintained in the sediments that aidedits preservation. Methane seepage from underlying sediments isinferred to be an important factor that controls the anaerobic oxi-dation of methane (AOM) in the sediment and water column as isthe case in the Black Sea (e.g. Reitner et al. 2005; Kruger et al.2008). A source of significant additional organic matter to formwidespread and longstanding reducing conditions in the Monte deiCorvi sediments may be related to moderate and/or episodic up-ward migration of methane-rich interstitial fluids, or to diffusionof methane from the underlying sediment, particularly organic-richsapropels, into the bottom water (Michaelis et al. 2002; Kruger et al.2008). This flux might have been moderate or periodic, enough toestablish only slightly anoxic to suboxic conditions in the sedimentcolumn. We speculate that this process preserved the original iron

sulphides in the Monte dei Corvi sediments and that it accounts forpost-depositional dissolution of calcareous microfossils in marlsand limestones, although these sediments were deposited in oxy-genated bottom waters. However, further geochemical analyses arerequired to test this hypothesis.

5.3 Timing of remanence acquisition

The younger intervals of the Monte dei Corvi section can be consid-ered to carry a reliable primary NRM signal because fine-grainedmagnetite is present along with fine-grained greigite. Both com-ponents carry the same palaeomagnetic directions. It is thereforereasonable to assume that NRM acquisition for both minerals was

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

140 S. K. Husing et al.

roughly coeval. Although magnetite Component 2 seems to beabsent in the older interval, a fine-grained magnetite componentis indicated by other rock magnetic experiments. The greigite Com-ponent 3 is the dominant carrier of the palaeomagnetic signal forthese sediments.

The following observations and interpretations indicate syn-sedimentary or early diagenetic formation of pyrite and greigitein the sulphide-bearing sediments of the older interval at Monte deiCorvi.

(1) The occurrence of pyrite framboids indicates that pyriteformed rapidly during or shortly after deposition within the topfew centimetres of the sediment or in the water column (Sweeney &Kaplan 1973; Berner 1984; Canfield & Berner 1987; Wilkin et al.1996; Wilkin & Barnes 1997).

(2) Magnetite iron oxides, which react rapidly with dissolvedsulphide (the half-life of magnetite in anoxic marine sedimentsranges from 50 to 1000 yr; Canfield & Berner 1987) are not evidentlypresent, which suggests that they were rapidly dissolved to form ironsulphides during early diagenesis.

(3) Iron sulphides have not been observed between cleavagesof detrital sheet silicates (cf. Jiang et al. 2001; Roberts & Weaver2005). Phyllosilicates react much more slowly than iron oxides (thehalf-life of sheet silicates when in contact with 1 mM concentrationof pore water sulphide is ∼120 kyr; Canfield et al. 1992). Overall,slow diagenetic reactions that produce iron sulphides are not ob-served (Canfield et al. 1992; Raiswell & Canfield 1996; Passier &de Lange 1998), which provides indirect evidence against compli-cations from later magnetizations.

Based on the assumption that slightly anoxic or suboxic condi-tions prevailed after deposition in the Monte dei Corvi sediments, itis reasonable to expect preservation of an early diagenetic NRM car-ried by greigite. Similar scenarios have been suggested for youngersediments (e.g. Pye 1981; Tric et al. 1991; Roberts & Turner 1993;Reynolds et al. 1999; Vasiliev et al. 2007, 2008).

Reversal ages can be compared to those from other marine andcontinental sections spanning the same time interval as the Montedei Corvi section. Suitable sections include the marine Metochiasection on Gavdos (Hilgen et al. 1995) and the continental Orerasection in Spain (Abdul Aziz et al. 2003). Differences of reversalages derived from the Monte dei Corvi, Metochia and Orera sec-tions are only minor (between 3 and 41 kyr for the younger Montedei Corvi interval). These small discrepancies can be explained byuncertainties in the position of the reversal boundaries. These un-certainties are related to the large spacing between age calibrationpoints and to the precision with which reversal boundaries can beidentified, and to the use of different astronomical target curves forthe tuning (for a more detailed discussion see Husing et al. 2007,2009). If remanence acquisition at Monte dei Corvi had been de-layed for tens to hundreds of kyr, the recorded polarity pattern wouldhave been distorted (e.g. Horng et al. 1998) and would not be aseasily recognized as it is. Therefore, it seems much more likely thatremanence acquisition at Monte dei Corvi took place shortly afterdeposition, with a maximum delay of only a few thousand years(<10 kyr), particularly for the older greigite-bearing sediments (cf.Vasiliev et al. 2008).

6 C O N C LU S I O N S

The Monte dei Corvi section is one amongst few marine sectionsso far identified where greigite appears to be responsible for a syn-

sedimentary to early diagenetic NRM acquisition. Rock magneticanalysis and SEM observations indicate that two distinct popula-tions of greigite are present throughout the section. The greigitecomponent with a narrow (SD) grain size distribution is interpretedto be of biogenic (magnetotactic) origin, whereas the greigite com-ponent with a wider size distribution is attributed to larger authigenic(bacterially mediated) greigite crystals. Increased NRM and totalIRM intensities in the younger sediments are related to the elevatedamount of fine-grained (magnetotactic) magnetite. The consistentmagnetic mineralogy throughout precession-induced lithologicalvariations indicates no significant link between mineralogy and cli-matic changes on the precessional scale. This supports an authigenicorigin for the greigite and magnetite components and emphasizesthe importance of early diagenetic processes within the sediment.In addition to the traditional scenario for iron sulphide formationin marine sediments (sulphidization under reducing conditions),we suggest that moderate or occasional methane seepage from theunderlying sediment might have helped to maintain (slightly) reduc-ing conditions, which can account for greigite preservation. This issupported by the occurrence of greigite and pyrite, and the poorpreservation of calcareous microfossils throughout the sedimen-tary succession. Any delayed remanence acquisition resulting fromthe authigenic greigite component must have a maximum delay of<10 kyr because there is no evidence of distortion of the expectedpalaeomagnetic reversal pattern in the Monte dei Corvi sediments.The primary origin of the palaeomagnetic signal at Monte dei Corviattests to the reliability of the high-resolution magnetostratigraphyand supports the use of the determined astronomical ages of reversalboundaries (Husing et al. 2007, 2009).

A C K N OW L E D G M E N T S

We acknowledge financial support by the Netherlands ResearchCentre for Integrated Solid Earth Sciences (ISES), the NetherlandsGeosciences Foundation (ALW), and the Netherlands Organizationfor Scientific Research (NWO). This work was carried out in theframework of the Vening Meinesz Research School of Geodynamics(VMSG). We thank Andrew Roberts and Leonardo Sagnotti for theirconstructive reviews, which significantly improved the manuscript.

R E F E R E N C E S

Abdul Aziz, H., Hilgen, F.J., Krijgsman, W. & Calvo, J.P., 2003.An astronomical polarity time scale for the middle Miocene basedon cyclic continental sequences, J. geophys. Res., 108, 2159,doi:10.1029/2002JB001818.

Bazylinksi, D.A., Frankel, R.B., Heywood, B.R., Mann, S., King, J.W.,Donaghay, P.L. & Hanson, A.K., 1995. Controlled biomineralization ofmagnetite (Fe3O4) and greigite (Fe3S4) in a magnetotactic bacterium,Appl. Environ. Microbiol., 61, 3232–3239.

Berner, R.A., 1970. Sedimentary pyrite formation, Am. J. Sci., 268, 1–23.Berner, R.A., 1984. Sedimentary pyrite formation: An update, Geochim.

cosmochim. Acta, 48, 605–615.Berner, R.A. & Raiswell, R., 1983. Burial of organic carbon and pyrite sulfur

in sediments over Phanerozoic time: a new theory, Geochim. cosmochim.Acta, 47, 855–862.

Calvert, S.E., Nielson, B. & Fontugne, M.R., 1992. Evidence from nitro-gen isotope ratios for enhanced productivity during formation of easternMediterranean sapropels, Nature, 359, 223–225.

Canfield, D.E. & Berner, R.A., 1987. Dissolution and pyritization of mag-netite in anoxic marine sediments, Geochim. cosmochim. Acta, 51, 645–659.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 141

Canfield, D.E., Raiswell, R. & Bottrell, S., 1992. The reactivity of sedimen-tary iron minerals towards sulfide, Am. J. Sci., 292, 659–683.

Chang, L., Roberts, A.P., Tang, Y., Rainford, B.D., Muxworthy, A.R. & Chen,Q., 2008. Fundamental magnetic parameters from pure synthetic greigite(Fe3S4), J. geophys. Res., 113, B06104, doi:10.1029/2007JB005502.

Chang, L., Roberts, A.P., Rowan, C.J., Tang, Y., Pruner, P., Chen, Q. & Horng,C.-S., 2009. Low-temperature magnetic properties of greigite (Fe3S4),Geochem. Geophys. Geosys., 10, Q01Y04, doi:10.1029/2008GC002276.

Coey, J.M.D. & Ghose, S., 1988. Magnetic phase transitions in silicate min-erals, in Structural and magnetic phase transitions in minerals (Advancesin Physical Geochemistry), Vol. 7, pp. 162–184, eds Ghose, S., Coey,J.M.D. & Salje, E., Springer, New York.

Cutter, G.A. & Kluckhohn, R.S., 1999. The cycling of particulate carbon,nitrogen, sulfur, and sulfur species (iron monosulfide, greigite, pyrite, andorganic sulfur) in the water columns of Framvaren Fjord and the BlackSea, Mar. Chem., 67, 149–160.

Dekkers, M.J., 1988a. Magnetic properties of natural pyrrhotite Part I: be-haviour of initial susceptibility and saturation-magnetization-related rockmagnetic parameters in a grain-size dependent framework, Phys. Earthplanet. Inter., 52, 376–393.

Dekkers, M.J., 1988b. Some rockmagnetic parameters for natural goethite,pyrrhotite and fine-grained haematite, PhD thesis. Utrecht University,231 pp.

Dekkers, M.J., Mattei, J.-L., Fillion, G. & Rochette, P., 1989. Grain-sizedependence of the magnetic behaviour of pyrrhotite during its low-temperature transition at 34 K, Geophys. Res. Lett., 16, 855–858.

Dinares-Turell, J. & Dekkers, M.J., 1999. Diagenesis and remanence acqui-sition in the Lower Pliocene Trubi marls at Punta di Maiata (southernSicily): paleomagnetic and rock magnetic observations, in Paleomag-netism and Diagenesis in Sediments, Vol. 151, pp. 53–69, eds Tarling,D.H. & Turner, P., Geological Society, Special Publications, London.

Dunlop, D.J. & Ozdemir, O., 1997. Rock Magnetism. Fundamentals andFrontiers, Cambridge University Press, Cambridge, UK, 573 pp.

Egli, R., 2004a. Characterisation of individual rock magnetic componentsby analysis of remanence curves, 1. Unmixing natural sediments, Stud.Geophys. Geodaet., 48, 391–446.

Egli, R., 2004b. Characterisation of individual rock magnetic componentsby analysis of remanence curves, 2. Fundamental properties of coercivitydistribution, Phys. Chem. Earth, 29, 851–867.

Florindo, F. & Sagnotti, L., 1995. Palaeomagnetism and rock magnetism inthe upper Pliocene Valle Ricca (Rome, Italy) section, Geophys. J. Int.,123, 340–354.

France, D.E. & Oldfield, F., 2000. Identifying goethite and haematite fromrock magnetic measurements of soils and sediments, J. geophys. Res.,105, 2781–2795.

Franke, C., Frederichs, T. & Dekkers, M.J., 2007. Efficiency of heavy liquidseparation to concentrate magnetic particles, Geophys. J. Int., 170, 1053–1066.

Frederichs, T., von Dobeneck, T., Bleil, U. & Dekkers, M.J., 2003. Towardsthe identification of siderite, rhodochrosite, and vivianite in sedimentsby their low-temperature magnetic properties, Phys. Chem. Earth, 28,669–679.

Fu, Y., von Dobeneck, T., Franke, C., Heslop, D. & Kasten, S., 2008. Rockmagnetic identification and geochemical process models of greigite for-mation in Quaternary marine sediments from the Gulf of Mexico (IODPHole U1319A), Earth planet. Sci. Lett., 275, 233–245.

Heslop, D., McIntosh, G. & Dekkers, M.J., 2004. Using time andtemperature-dependent Preisach models to investigate the limitations ofmodelling isothermal remanent magnetization acquisition curves withcumulative log Gaussian functions, Geophys. J. Int., 156, 1–9.

Hilgen, F.J., 1991. Astronomical calibration of Gauss to Matuyama sapropelsin the Mediterranean and implication for the geomagnetic polarity timescale, Earth planet. Sci. Lett., 104, 226–244.

Hilgen, F.J., Krijgsman, W., Langereis, C.G., Lourens, L.J., Santarelli, A. &Zachariasse, W.J., 1995. Extending the astronomical (polarity) time scaleinto the Miocene, Earth planet. Sci. Lett., 136, 495–510.

Hilgen, F.J., Abdul Aziz, H., Krijgsman, W., Raffi, I. & Turco, E., 2003.Integrated stratigraphy and astronomical tuning of the Serravallian and

lower Tortonian at Monte dei Corvi (Middle-Upper Miocene, northernItaly): Palaeogeogr., Palaeoclimatol., Palaeoecol., 199, 229–264.

Hilgen, F.J. et al., 2005. The Global Boundary Stratotype Section and Point(GSSP) of the Tortonian Stage (Upper Miocene) at Monte dei Corvi,Episodes, 28, 6–17.

Hilton, J. & Lishman, J.P., 1985. The effect of redox change on the mag-netic susceptibilty of sediments from a seasonally anoxic lake, Limnol.Oceanogr., 30, 907–909.

Horng, C.-S. & Roberts, A.P., 2006. Authigenic or detrital origin ofpyrrhotite in sediments?: Resolving a paleomagnetic conundrum, Earthplanet. Sci. Lett., 241, 750–762.

Horng, C.-S., Torii, M., Shea, K.-S. & Kao, S.-J., 1998. Inconsistent mag-netic polarities between greigite- and pyrrhotite/magnetite-bearing ma-rine sediments from the Tsailiao-chi section, southwestern Taiwan, Earthplanet. Sci. Lett., 164, 467–482.

Hu, S.Y., Appel, E., Hoffmann, V., Schmahl, W.W. & Wang, S., 1998. Gy-romagnetic remanence acquired by greigite during static three-axis AFdemagnetization, Geophys. J. Int., 134, 831–842.

Husing, S.K., Hilgen, F.J., Abdul Aziz, H. & Krijgsman, W., 2007. Complet-ing the Neogene geological time scale between 8.5 and 12.5 Ma, Earthplanet. Sci. Lett., 253, 340–358.

Husing, S.K., Kuiper, K.F., Link, W., Hilgen, F.J. & Krijgsman, W., 2009. Theupper-Tortonian-lower Messinian at Monte dei Corvi (Northern Apen-nines, Italy): completing a Mediterranean reference section for the Torto-nian Stage, Earth planet. Sci. Lett., 282, 140–157.

Jiang, W.-T., Horng, C.-S., Roberts, A.P. & Peacor, D.R., 2001. Contradictorymagnetic polarities in sediments and variable timing of neoformation ofauthigenic greigite, Earth planet. Sci. Lett., 193, 1–12.

Kao, S.-J., Horng, C.-S., Roberts, A.P. & Liu, K.-K., 2004. Carbon–sulfur–iron relationships in sedimentary rocks from southwestern Taiwan: influ-ence of geochemical environment on greigite and pyrrhotite formation,Chem. Geol., 203, 153–168.

Kopp, R.E. & Kirschvink, J.L., 2008. The identification and biogeochem-ical interpretation of fossil magnetotactic bacteria, Earth-Sci. Rev., 86,42–61.

Kosterov, A., 2001. Magnetic properties of subaerial basalts at low temper-atures, Earth Planets Space, 53, 883–892.

Krs, M., Novak, F., Krsova, M., Pruner, P., Kouklikova, L. & Jansa, J., 1992.Magnetic properties and metastability of greigite–smythite mineralizationin brown-coal basins of the Krusne hory Piedmont, Bohemia, Phys. Earthplanet. Inter., 70, 273–287.

Kruger, M., Blumenberg, M., Kasten, S., Wieland, A., Kanel, L., Klock,J.-H., Michaelis, W. & Seifert, R., 2008. A novel, multi-layered methan-otrophic microbial mat system growing on the sediment of the Black Sea,Environ. Microbiol., 10, 1934–1947.

Kruiver, P.P., Dekkers, M.J. & Heslop, D., 2001. Quantification of magneticcoercivity components by the analysis of acquisition curves of isothermalremanent magnetisation, Earth planet. Sci. Lett., 189, 269–276.

Larrasoana, J.C., Roberts, A.P., Musgrave, R.J., Gracia, E., Pinero, E., Vega,M. & Martınez-Ruiz, F., 2007. Diagenetic formation of greigite andpyrrhotite in gas hydrate marine sedimentary systems, Earth planet. Sci.Lett., 261, 350–366.

Larrasoana, J.C., Roberts, A.P., Stoner, J.S., Richter, C. & Wehausen,R., 2003. A new proxy for bottom-water ventilation in the easternMediterranean based on diagenetically controlled magnetic properties ofsapropel-bearing sediments, Palaeogeogr., Palaeoclimatol., Palaeoecol.,190, 221–242.

Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A.C.M. & Levrard,B., 2004. A long-term numerical solution for the insolation quantities ofthe Earth, Astron. Astrophys., 428, 261–285.

Liu, Q.S., Yu, Y.J., Torrent, J., Roberts, A.P., Pan, Y.X. & Zhu, R.X., 2006.Characteristic low-temperature magnetic properties of aluminous goethite[alpha-(Fe,Al)OOH] explained, J. geophys. Res., 111(B12), B12S34,doi:10.1029/2006JB004560.

Lourens, L.J., Hilgen, F.J., Shackleton, N.J., Laskar, J. & Wilson, D., 2004.The Neogene Period, in A Geological Time Scale 2004, pp. 409–440, edsGradstein, F.M., Ogg, J.G. & Smith, A.G., Cambridge University Press,Cambridge, UK.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

142 S. K. Husing et al.

Lowrie, W., 1990. Identification of ferromagnetic minerals in a rock bycoercivity and unblocking temperature properties, Geophys. Res. Lett.,17, 159–162.

Maher, B.A., Karloukovski, V.V. & Mutch, T.J., 2004. High-field rema-nence properties of synthetic and natural submicrometer haematites andgoethite: significance for environmental context, Earth planet. Sci. Lett.,226, 491–505.

Mann, S., Sparks, N.H.S., Frankel, R.B., Bazylinksi, D.A. & Jannasch, H.W.,1990. Biomineralization of ferrimagnetic greigite (Fe3S4) and iron pyrite(FeS2) in a magnetotactic bacterium, Nature, 343, 258–261.

Michaelis, W. et al., 2002. Microbial reefs in the Black Sea fueled byanaerobic oxidation of methane, Science, 297, 1013–1015.

Montanari, A. et al., 1997. Integrated stratigraphy of the Middle and Up-per Miocene pelagic sequence of the Conero Riviera (Marche region,Italy), in Miocene Stratigraphy: An Integrated Approach (Dev. Palaeon-tol. Stratigr.), Vol. 15, pp. 409–450, eds Montanari, A., Odin, G.S. &Coccioni, R., Elsevier, Amsterdam.

Morse, J.W. & Wang, Q., 1997. Pyrite formation under conditions approxi-mating those in anoxic sediments, Mar. Chem., 57, 187–193.

Mullender, T.A.T., van Velzen, A.J. & Dekkers, M.J., 1993. Continuousdrift correction and separate identification of ferromagnetic and param-agnetic contribution in thermomagnetic runs, Geophys. J. Int., 114, 663–672.

Ozdemir, O., Dunlop, D.J. & Moskowitz, B.M., 1993. The effect of oxidationon the Verwey transition in magnetite, Geophys. Res. Lett., 20, 1671–1674.

Passier, H.F. & de Lange, G.J., 1998. Sedimentary sulfur and iron chemistryin relation to the formation of Eastern Mediterranean sapropels, Proc.ODP, Sci. Res., 160, 249–259.

Passier, H.F. & Dekkers, M.J., 2002. Iron oxide formation in the activeoxidation front above sapropel S1 in the eastern Mediterranean Sea asderived from low-temperature magnetism, Geophys. J. Int., 150, 230–240.

Passier, H.F. et al., 1999. Sulphidic Mediterranean surface waters duringPliocene sapropel formation, Nature, 397, 146–149.

Passier, H.F., de Lange, G.J. & Dekkers, M.J., 2001. Magnetic propertiesand geochemistry of the active oxidation front and the youngest sapropelin the eastern Mediterranean Sea, Geophys. J. Int., 145, 604–614.

Pike, C.R., Roberts, A.P. & Verosub, K.L., 2000. The effect of magneticinteractions on low temperature saturation remanence in fine magneticparticle systems, J. Appl. Phys., 88, 967–974.

Posfai, M., Buseck, P.R., Bazylinksi, D.A. & Frankel, R.B., 1998. Reactionsequence of iron sulfide minerals in bacteria and their use as biomarkers,Science, 280, 880–883.

Pye, K., 1981. Marshrock formed by iron sulphide and siderite cementationin saltmarsh sediments, Nature, 294, 650–652.

Raiswell, R. & Canfield, D.E., 1996. Rates of reaction between silicate miner-als and dissolved sulfide in Peru Margin sediments, Geochim. cosmochim.Acta, 60, 2777–2787.

Reitner, J., Peckmann, J., Blumenberg, M., Michaelis, W., Reimer, A. &Thiel, V., 2005. Concretionary methane-seep carbonates and associatedmicrobial communities in Black Sea sediments, Palaeogeogr., Palaeocli-matol., Palaeoecol., 227, 18–30.

Reynolds, R.L., Rosenbaum, J.G., van Metre, P., Tuttle, M., Callender, E.& Goldin, A., 1999. Greigite (Fe3S4) as an indicator for drought—the1912–1994 sediment magnetic record from White Rock Lake, Dallas,Texas, USA, J. Paleolim., 21, 193–206.

Richter, C., Roberts, A.P., Stoner, J.S., Benning, L.G. & Chi, C.T., 1998.Magnetostratigraphy of Pliocene–Pleistocene sediments from the easternMediterranean Sea, Proc. ODP, Sci. Res., 160, 61–74.

Roberts, A.P., 1995. Magnetic properties of sedimentary greigite (Fe3S4),Earth planet. Sci. Lett., 134, 227–236.

Roberts, A.P. & Turner, G.M., 1993. Diagenetic formation of ferrimagneticiron sulphide minerals in rapidly deposited marine sediments, South Is-land, New Zealand, Earth planet. Sci. Lett., 115, 257–273.

Roberts, A.P. & Weaver, R., 2005. Multiple mechanisms of remagnetizationinvolving sedimentary greigite (Fe3S4), Earth planet. Sci. Lett., 231, 263–277.

Roberts, A.P., Stoner, J.S. & Richter, C., 1999. Diagenetic magnetic en-hancement of sapropels from the eastern Mediterranean Sea, Mar. Geol.,153, 103–116.

Roberts, A.P., Jiang, W.T., Florindo, F., Horng, C.-S. & Laj, C., 2005. As-sessing the timing of greigite formation and the reliability of the UpperOlduvai polarity transition record from the Crostolo River, Italy, Geophys.Res. Lett., 32, L05307, doi:10.1029/2004GL022137.

Rochette, P., Fillion, G., Mattei, J.-L. & Dekkers, M.J., 1990. Magnetic tran-sition at 30–34 Kelvin in pyrrhotite: insight into a widespread occurrenceof this mineral in rocks, Earth planet. Sci. Lett., 98, 319–328.

Rohling, E.J., 1994. Review and new aspects concerning the formation ofeastern Mediterranean sapropels, Mar. Geol., 122, 1–28.

Rohling, E.J. & Hilgen, F.J., 1991. The eastern Mediterranean climate attimes of sapropel formation: a review, Geol. Mijnbouw, 70, 253–264.

Rossignol-Strick, M., 1985. Mediterranean Quaternary sapropels, an imme-diate response of the African monsoon to variation of insolation, Palaeo-geogr., Palaeoclimatol., Palaeoecol., 49, 237–263.

Rossignol-Strick, M., 1993. African monsoons, an immediate climate re-sponse to orbital forcing, Nature, 304, 46–49.

Rossignol-Strick, M., Nesteroff, W., Olive, P. & Vergnaud-Grazzini, C.,1982. After the deluge: Mediterranean stagnation and sapropel formation,Nature, 295, 105–110.

Rowan, C.J. & Roberts, A.P., 2006. Magnetite dissolution, diachronousgreigite formation, and secondary magnetizations from pyrite oxidation:unravelling complex magnetizations in Neogene marine sediments fromNew Zealand, Earth planet. Sci. Lett., 241, 119–137.

Sagnotti, L. & Winkler, A., 1999. Rock magnetism and palaeomagentismof greigite-bearing mudstones in the Italian Peninsula, Earth planet. Sci.Lett., 165, 67–80.

Sagnotti, L., Roberts, A.P., Weaver, R., Verosub, K.L., Florindo, F., Pike,C.R., Clayton, T. & Wilson, G.S., 2005. Apparent magnetic polarity re-versals due to remagnetization resulting from late diagenetic growth ofgreigite from siderite, Geophys. J. Int., 100, 89–100.

Smirnov, A.V. & Tarduno, J.A., 2000. Low-temperature magnetic propertiesof pelagic sediments (Ocean Drilling Program Site 805C): tracers ofmaghemitization and magnetic mineral reduction, J. geophys. Res., 105,16 457–16 471.

Snowball, I.F., 1991. Magnetic hysteresis properties of greigite (Fe3S4) anda new occurrence in Holocene sediments from Swedish Lappland, Phys.Earth planet. Inter., 68, 32–40.

Snowball, I.F., 1997a. The detection of single domain greigite (Fe3S4) usingrotational remanent magnetization (RRM) and the effective gyro field(Bg): mineral magnetic and palaeomagnetic applications, Geophys. J. Int.,130, 704–716.

Snowball, I.F., 1997b. Gyroremanent magnetization and the magnetic prop-erties of greigite-bearing clays in southern Sweden, Geophys. J. Int., 129,624–636.

Snowball, I.F. & Thompson, R., 1988. The occurrence of greigite in sedi-ments from Loch Lomond, J. Quat. Sci., 3, 121–125.

Stephenson, A. & Snowball, I.F., 2001. A large gyromagnetic effect ingreigite, Geophys. J. Int., 145, 570–575.

Suits, N.S. & Wilkin, R.T., 1998. Pyrite formation in the water column andsediment of a meromictic lake, Geology, 26, 1099–1102.

Sweeney, R.E. & Kaplan, I.R., 1973. Pyrite framboid formation: laboratorysynthesis and marine sediments, Econ. Geol., 5, 618–634.

Thompson, R. & Cameron, T.D.J., 1995. Palaeomagnetic study of Cenozoicsediments in North Sea boreholes: an example of a magnetostratigraphicconundrum in a hydrocarbon-producing area, in Palaeomagnetic Appli-cations in Hydrocarbon Exploration, Vol. 98, pp. 223–236, eds Turner, P.& Turner, A., Spec. Publ. Geol. Soc. London.

Tric, E., Laj, C., Jehanno, C., Valet, J.-P., Kissel, C., Mazaud, A. & Iaccarino,S., 1991. High resolution record of the upper Olduvai transition from PoValley (Italy) sediments: support for dipolar transition geometry?, Phys.Earth planet. Inter., 65, 319–336.

Vasiliev, I., Dekkers, M.J., Krijgsman, W., Franke, C., Langereis, C.G. &Mullender, T.A.T., 2007. Early diagenetic greigite as a recorder of thepalaeomagnetic signal in Miocene–Pliocene sedimentary rocks of theCarpathian foredeep (Romania), Geophys. J. Int., 17, 613–629.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS

Early remanence acquisition in Tortonian greigite-bearing sediments 143

Vasiliev, I., Franke, C., Meeldijk, J.D., Dekkers, M.J., Langereis, C.G. &Krijgsman, W., 2008. Putative greigite magnetofossils from the Plioceneepoch, Nature Geosci., 1, 782–786.

Wilkin, R.T. & Barnes, H.L., 1996. Pyrite formation by reactions ofiron monosulfides with dissolved inorganic and organic sulfur species,Geochim. cosmochim. Acta, 60, 4167–4179.

Weaver, R., Roberts, A.P. & Barker, A.J., 2002. A late diagenetic (syn-folding) magnetization carried by pyrrhotite: implications for paleo-

magnetic studies from magnetic iron sulphide-bearing sediments, Earthplanet. Sci. Lett., 200, 371–386.

Wilkin, R.T. & Barnes, H.L., 1997. Formation processes of framboidalpyrite, Geochim. cosmochim. Acta, 61, 323–339.

Wilkin, R.T., Barnes, H.L. & Brantley, S.L., 1996. The size dis-tribution of framboidal pyrite in modern sediments: an indica-tor of redox conditions, Geochim. cosmochim. Acta, 60, 3897–3912.

C© 2009 The Authors, GJI, 179, 125–143

Journal compilation C© 2009 RAS