Orbital tuning of a lower Cretaceous composite record (Maiolica Formation, central Italy)

Orbital tuning of a lower Cretaceous composite record (Maiolica

Formation, central Italy)

Mario Sprovieri,1 Rodolfo Coccioni,2 Fabrizio Lirer,1 Nicola Pelosi,1 and Francesca Lozar3

Received 20 September 2005; revised 1 June 2006; accepted 25 July 2006; published 7 December 2006.

[1] A high-resolution pelagic bulk carbonate stable isotope record from a central Tethyan lower Cretaceouscomposite section is presented. Three well-exposed sedimentary sequences (Chiaserna Monte Acuto, Bosso, andGorgo a Cerbara sections, central Italy) cropping out throughout the Maiolica Formation were correlated by adetailed magnetostratigraphy, lithostratigraphy, and calcareous plankton biostratigraphy in order to reconstruct acontinuous composite record from the middle Berriasian to the lower Aptian. The integrated stratigraphy of thethree sequences provided an accurate time framework for the new high-resolution C isotope curve which ispresented in this study. The composite d13C signal, recorded in the depth domain, was analyzed by combinedLomb-Scargle periodogram and weighted wavelet Z transform (WWZ) – weighted wavelet amplitudes (WWA)Foster wavelet spectral methodologies, both appropriate for unevenly sampled curves. These tools allowed us tounravel the main frequencies modulating the record and their hypothetical shift in depth, respectively. The long-term, �400,000 and �2,400,000 years, eccentricity cycles were consistently recorded throughout all thecomposite record. Once band-pass filtered in these two periodicity bands and compared to the lithologic patterncycles identified throughout the composite sequence, the d13C signal was used as a valuable proxy record for areliable construction of an orbital tuning of the early Cretaceous. An estimated age for all the differentstratigraphic events recognized throughout the composite record was reported. In particular, the reestimated agesof the paleomagnetic chrons, documented in the upper part of the record, show differences with those reportedby Gradstein et al. (2004) up to �2 m.y. The Valanginian carbon shift, present in the middle part of thecomposite sequence was estimated to be �2.3 m.y. long. The good agreement between the estimated age of thebase of this positive carbon isotope excursion (�136.34 m.y.) and the predominant volcanic phase associated tothe Parana-Etendeka large igneous province intrusion confirmed a possible cause-effect link between the twoevents.

Citation: Sprovieri, M., R. Coccioni, F. Lirer, N. Pelosi, and F. Lozar (2006), Orbital tuning of a lower Cretaceous composite record

(Maiolica Formation, central Italy), Paleoceanography, 21, PA4212, doi:10.1029/2005PA001224.

1. Introduction

[2] The evolution of paleoclimate and paleoceanographyduring the Early Cretaceous was described in several papers[e.g., Weissert, 1989; Lini et al., 1992; Lini, 1994; Weissertet al., 1998; Hennig et al., 1999; Grocke et al., 1999, 2003a,2003b; van de Schootbrugge et al., 2000; Wissler et al.,2001; Bersezio et al., 2002; Weissert and Erba, 2004; Erbaet al., 2004; Erba, 2004]. In particular, long-term carboncycle perturbations globally detected during the earlyCretaceous were analyzed using multiproxy stable isotope,faunal, floral and chemical tracers in order to better under-stand the physiological behaviour of the Earth system inresponse to external and/or internal forcing.[3] The lower Cretaceous pelagic deposits are preserved

with a remarkable continuity in the Umbria-Marche basin of

central Italy [Herbert and Fischer, 1986]. They displayrhythmic alternations of green/black marls and limestones,and/or cherty layers and limestone which are suitable forapplying cyclostratigraphic methodologies and reconstructaccurate orbital tunings of the studied records. In this paper,we present a high-resolution bulk carbonate oxygen andcarbon isotope record from three Umbria-Marche Creta-ceous (middle Berriasian– lower Aptian) sedimentarysequences. Moreover, the combination of the resultsachieved by cyclostratigraphic interpretation of the litho-logic record with the high-resolution analysis of the d13Csignal investigated with stationary (Lomb-Scargle periodo-gram) and nonstationary (WWZ-WWA Foster wavelet)power spectral methodologies, both appropriate for un-evenly sampled signals, provides a suitable strategy topropose a reliable orbital tuning for the lower Cretaceous.A reestimated age for the paleomagnetic chron boundariesand biostratigraphic events recognized throughout the recordis proposed and compared with literature data. Moreover,the age calculated for the base of the Valanginian and theduration estimated for the carbon isotope excursion providefurther evidence that volcanic events associated to the intru-sion of one of the world’s largest flood volcanic provinces

PALEOCEANOGRAPHY, VOL. 21, PA4212, doi:10.1029/2005PA001224, 2006ClickHere

for

FullArticle

1Istituto per l’Ambiente Marino Costiero, Consiglio Nazionale delleRicerche, Napoli, Italy.

2Istituto di Geologia e Centro di Geobiologia, Universita di Urbino,Urbino, Italy.

3Dipartimento di Scienze della Terra, Universita di Torino, Turin, Italy.

Copyright 2006 by the American Geophysical Union.0883-8305/06/2005PA001224$12.00

PA4212 1 of 19

(the Parana-Etendeka large igneous province, LIP) couldhave triggered the Valanginian C isotope excursion.

2. Analytical Methods

[4] A total of about 1000 bulk sample stable isotopeanalyses were carried out. They were measured by auto-mated continuous flow carbonate preparation GasBenchIIdevice [Spotl and Vennemann, 2003] and ThermoElectronDelta Plus XP mass spectrometer at the IAMC-CNR(Naples) isotope geochemistry laboratory. Acidification ofsamples was performed at 50�C. Each 6 samples, an internalstandard (Carrara Marble with d18O = �2.43 versus VPDBand d13C = 2.43 vs. VPDB) was run and, each 30 samples,the NBS19 international standard was measured. Standarddeviations of carbon and oxygen isotope measures wereestimated 0.1 and 0.08%, respectively, on the basis of�200 samples measured in triple.[5] The entire calibration is based directly on standard

materials that are part of each run (in our case the homoge-neous and certified Carrara Marble carbonate with isotopiccomposition determined by conventional offline dual-inlettechniques), rather than solving fractionation equations forthe acid-based reaction. In other words, there is no need toknow the stable isotopic composition of the CO2 referencegas a priori, nor the acid fractionation factor at the giventemperature of the reaction [Spotl and Vennemann, 2003].[6] All the isotope data are reported in per mil versus

VPDB.[7] Biostratigraphical analyses performed on the cal-

careous nannofossil assemblages were carried out on400 samples collected throughout all the Chiaserna MonteAcuto section. Smear slides were prepared from raw mate-rial, according to standard methods for highly induratelimestones [Monechi and Thierstein, 1985]; no concentra-tion or cleaning methods were applied, in order to retain theoriginal assemblage and mineralogical composition. Smearslides were analyzed with a polarizing light microscope at1250X magnification. From the Gorgo a Cerbara and Bossosections the available data reported by Channell et al.[1995] were used.[8] Biostratigraphical analyses on the calpionellid

assemblages were conducted on 400 samples collectedthroughout all the Chiaserna Monte Acuto section and onlyto recognize the Last Occurrence of the group, useful eventfor large-scale correlations. Analyses were performed onthin sections of limestone samples.

3. Geological and Stratigraphical Setting

[9] The Cretaceous pelagic sequence of the Umbria-Marche Basin was deposited in a complex basin and swelltopography along the continental margin of the Apulianblock, which moved with Africa relative to northernEurope. The basement of the Umbria-Marche Apenninesis continental and the Upper Jurassic through Paleocenepelagic succession overlies a subsiding Triassic to lowerJurassic carbonate platform. The Cretaceous pelagicsequence of the Umbria-Marche Basin is subdivided intofour formations as follows (from bottom to top): Maiolicap.p. (Tithonian–early Aptian), Marne a Fucoidi (early

Aptian– latest Albian), Scaglia Bianca (latest Albian–earliest Turonian), and Scaglia Rossa p.p. (earliest Turo-nian–early Lutetian) [Coccioni, 1996]. In particular, theMaiolica Formation which crops out over large areas of theSouthern Alps and central Italy, is characterized by thin-bedded white to gray pelagic limestones interbedded withblack shales in its upper part. In the uppermost Hauterivian,the Faraoni level is recognized at regional scale [Cecca etal., 1994a]. In the upper-lower Aptian, at the base of theMarne a Fucoidi Formation, the organic carbon-rich andcarbonate-free Selli level, 1 to 3 m thick, occurs which is thesedimentary expression of the oceanic anoxic event (OAE)1a [Coccioni et al., 1987, 1989; Arthur et al., 1990].

4. Studied Sections

4.1. Chiaserna Monte Acuto Section

[10] The Chiaserna Monte Acuto section, about 240 mthick, is located along the road that from the village ofChiaserna climbs up to the top of the Monte Catria, windingalong the southern slopes of Monte Acuto (Figure 1). Thecomplete stratigraphic record of the section (Figure 2)extends from the middle Berriasian to the upper Hauteri-vian. Some faults and slumped intervals occur in the sectionthat, however, did not present any severe sampling prob-lems thanks to the careful lithostratigraphic control. Threethin black shale layers interbedded within limestone bedswere recognized at 137.70 m, 151.50 m, and 165.80 m,respectively.[11] Magnetostratigraphy and nannofossil biostratigra-

phy of the upper Valanginian–upper Hauterivian intervalwas reported by Channell et al. [1995]. Ammonite stra-tigraphy was provided by Cecca [1985, 1995] and Faraoniet al. [1997] following the ammonite zonal scheme ofHoedemaeker et al. [1993, 1995]. All the standard zonesand several biohorizons were recognized from the upper-most Berriasian to the lowermost Hauterivian.[12] Nannofossil assemblages are scarce to common and

characterized by poor to moderate preservation, with evi-dence of dissolution and/or recrystallization of delicatestructures of the central area of placoliths and nannoliths.Nannoconids and Conusphaera sp. are relatively moreabundant, together with Watznaueria sp., Diazomatolithuslehmanii, and Cyclagelosphaera margerelii. Despite theoverall moderate preservation of the sampled material,several bioevents were recognized and they allow for theidentification of the biostratigraphic framework, accordingto the schemes proposed by Bralower et al. [1989, 1995](Figure 2).[13] The following biozones can be identified (from

bottom to top).[14] 1. First is the Cretarhabdus angustiforatus zone

(NK2). The base of the zone was not recorded in the studiedmaterial, and the top is based on the first occurrence (FO) ofCalcicalathina oblongata.[15] 2. Second is the Ca. oblongata zone (NK3), based on

the FO of Ca. oblongata as lower boundary and the lastoccurrence (LO) of Tubodiscus verenae for the top. Thiszone was subdivided into two subzones, based on the LO ofRucinolithus wisei (top for the R. wisei subzone, NK3A)

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

2 of 19

PA4212

and the LO of T. verenae (top for the T. verenae subzone,NK3B).[16] 3. Third is the NC4 Cruciellipsis cuvillieri zone that

has been subdivided into two subzones, the NC4ACa. oblongata subzone, bounded by the FO of Lithraphi-dites bollii at the top, and the NC4B L. bollii subzone,bounded by the LO of Cruciellipsis cuvillieri at the top.[17] 4. Fourth is the NC5 Watznaueria oblonga zone of

which the two lowermost subzones are also recognized: theNC5A subzone (defined by the FO of Rucinolithus tere-brodentarius at the top) and the NC5B subzone above it.The upper boundary of the NC5B subzone, correspondingto the LO of L. bollii, was not recorded in this section.[18] In addition, the FO of T. verenae was recorded in

subzone NK3A (m 101.14 above the base). R. wisei isusually scattered in its lower range and the poor preserva-tion of the studied material does not allowed us a clear-cutidentification of its first occurrence. P. fenestrata was notfound in the studied material.[19] Calpionellid assemblages are characterized by poor

to moderate preservation. The last calpionellids were foundin the lowermost part of the lower Hauterivian Acanthodis-cus radiatus ammonite zone (Figure 2).

4.2. Bosso Section

[20] The sampled �50 m thick Bosso section is lateHauterivian–lower Barremian in age (Figure 2). It is located

along the road from Cagli to Pianello, km 9.900, about 4 kmwest of town of Cagli in a valley where the Bosso River cuts

through a northwest striking anticlinal fold (Figure 1).Detailed magnetostratigraphy and calcareous nannofossilbiostratigraphy was based on 1 sample/0.5 m by Channellet al. [1995]. Coccioni et al. [1998] correlated calcareousnannofossils, radiolarians, foraminifera, and magnetozones.The sedimentary record is characterized by high frequencyalternation of thin, well-bedded limestone and cherty layers.The Faraoni level occurs from 30.10 m to 30.36 m abovethe base of the section (Figure 2).

4.3. Gorgo a Cerbara Section

[21] The studied section is about 100 m thick and spansthe upper Hauterivian–lower Aptian interval (Figure 2).This section is located along the Candigliano River, 3 kmwest of the town of Piobbico. Magnetostratigraphy data ofLowrie and Alvarez [1984] and Channell et al. [2000] weretaken into account. All the ammonite standard zones ofHoedemaeker et al. [1993, 1995] were recognized by Ceccaand Pallini [1994] and Cecca et al. [1994a]. Cecca et al.[1994b] and Channell et al. [1995, 2000] provided correla-tion of polarity chrons to nannofossil and planktonic fora-miniferal events and ammonite zones.[22] The stratigraphic sampling resolution adopted by the

authors for magnetostratigraphic and biostratigraphic anal-yses was better than 1 sample/0.5 m. The sedimentaryrecord is characterized by high frequency alternation ofthin, well-bedded limestone and cherty layers. The Faraonilevel occurs from 22.00 m to 22.35 m above the base of thesection. About 60 m from the base upward black shale

Figure 1. Location map of the three studied sedimentary sequences. 1, Chiaserna Monte Acuto section;2, Gorgo a Cerbara section; 3, Bosso section.

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

3 of 19

PA4212

Figure

2

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

4 of 19

PA4212

layers start to occur. In the uppermost part of the section theorganic-rich sediments of the Selli level, 2 m thick, occur.

4.4. Composite Section

[23] The three sedimentary sequences were correlated bya detailed lithostratigraphy, magnetostratigraphy, and bio-stratigraphy (Figure 3). The magnetostratigraphic nomen-clature is that reported by Gradstein et al. [2004].[24] Top and base of CM5N were recognized in the Gorgo

a Cerbara (from 16.50 and 29.70 m from the base) andBosso (from 25.00 and 39.00 m) sections. Within the twosedimentary intervals a very similar lithologic organizationcan be observed suggesting a reliable correlation betweenthe two sections and a good preservation of the lithostrati-graphic records. Because of three slumped intervals recog-nized in the lowermost part of Gorgo a Cerbara, thecomposite sequence continues, back in time, from the baseof the CM5N magnetic chron, in the lower part of the Bossosection. There, top and base of the CM7R magnetic chronwere used to correlate this record with the sediments of theuppermost part of the Chiaserna Monte Acuto section. Onceagain, a very similar lithologic pattern, in the overlappingsedimentary interval, confirms the accurate correlation be-tween them. The recognition of the first occurrence ofR. terebrodentarius both in the Bosso and Chiaserna MonteAcuto sections, at the base of the CM7R paleomagneticchron, was used to further corroborate the correlationbetween the two records.[25] The obtained composite record has a thickness of

�330 m and it covers a continuous stratigraphic intervalfrom the middle–lower Berriasian to the lower part of theAptian. The isotope records were reported up to the base ofthe Selli level (recognized at 95 m from the base of theGorgo a Cerbara section) that is characterized by totalabsence of carbonate. Ammonite events reported byFaraoni et al. [1997], Cecca and Pallini [1994], and Ceccaet al. [1994b] for Chiaserna Monte Acuto and Gorgo aCerbara sections, respectively, were reported in Figure 2 asvaluable tool for large-scale correlations.[26] On the basis of the absolute ages reported by

Gradstein et al. [2004] for the paleomagnetic chron bound-aries recognized all along the middle-upper part of thecomposite section (Figure 2) we estimated an average sedi-mentation rate for that part of the record of �0.0144 m/kyr(Figure 4), assuming that no major sedimentary gaps exist inthe studied sections.[27] The average sampling rate adopted throughout the

composite record is about 1 sample/20–80 cm, although,due to occasional vegetation coverage of the sections and/orobjective difficulty of sediments collection, in some intervals

(such as an example between 11 and 16 m, 19 and 24 m, 73and 79 m at Chiaserna Monte Acuto and between 234 and240 m at Gorgo a Cerbara) the sampling distance increasesup to �4/5 m. This unevenly sampling rate reduces thepossibility to reliably detect short-term periodicities through-out the record and its careful check assisted us in the correctinterpretation of the results of spectral analysis.

5. Isotope Signals

5.1. Carbon and Oxygen Isotope Records

[28] Carbon isotope data of the composite section arereported in Figure 5. From the base of the compositesection up to the lower Valanginian (middle part of theBusnardoites campylotoxus ammonite zone and close to theLO of R. wisei) carbon isotopes oscillate around an averagevalue of �1.3%. The Valanginian-Hauterivian carbon iso-tope excursion reaching values of 3.1% is well preservedand corresponds to the well-known Valanginian carbonshift (the Weissert event of Erba et al. [2004]). In theHauterivian, the d13C values return to averages of �1.9%.Then a long and continuous trend to positive values is ob-served culminating in the Barremian with values of �2.5%.[29] The mid-Barremian event (MBE) reported by

Coccioni et al. [2003] in the uppermost part of the samesection precedes the large positive excursion recorded inthe lower Aptian [e.g., see Menegatti et al., 1998],corresponding to the Selli level deposition. The Selli levelat Gorgo a Cerbara is carbonate-free. Therefore no bulkcarbon isotope curve could be established through thisinterval.[30] The low-amplitude/long-term periodic modulations

of the d13C record, particularly evident during the middleBerriasian, the late Valanginian and the early Aptian (justbelow the Selli level), suggest that volcanism and tectonicsinfluenced the long-term carbon cycle.[31] In Figure 6, a comparison between our C isotope

composite record and the bulk rock composite d13C curvepresented by Weissert and Erba [2004] and based on acompilation of data from the Cismon section [Erba et al.,1999], Capriolo section [Channell et al., 1993] and Valledel Mis section [Weissert and Channell, 1989], is shown.The long-term trends and the two major excursion to d13Cheavy values in the Valanginian and Barremian show thesame structure and isotope values, thus confirming a reliablepreservation of the original oceanographic and environmen-tal signals in the proposed Umbria-Marche curve andexclude major sedimentary gaps in the composite section.However, this signal shows a more accurate and detailedregistration of higher frequency oscillations in the d13C

Figure 2. Lithostratigraphy, magnetostratigraphy, and biostratigraphy of the studied sections. Magnetostratigraphy andthe reported ages of the chron boundaries are according to Gradstein et al. [2004]. Magnetostratigraphy of Chiaserna MonteAcuto and Bosso sections are from Channell et al. [1995]. Magnetostratigraphy and biostratigraphy of the Gorgo a Cerbarasection are from Lowrie and Alvarez [1984] and Channell et al. [2000]. The following abbreviations are usedfor ammonites: S., Subthurmannia; Th., Thurmanniceras; B., Busnardoites; Sa., Saynoceras; N., Neocomites;A., Acanthodiscus; Ba., Balearites; P., Plesiospitidiscus; Ps., Pseudothurmannia; T., Taveraidiscus; Ss., Subsaynella;K., Kotetishvilia; Ho., Holcodiscus; An., Ancyloceras; H., Hemihoplites. The following abbreviations are used forcalcareous nannofossils: C., Cretarhabdus; Ca., Calcicalathina; Cr., Cruciellipsis; W., Watznaueria; Ch., Chiastozygus;R., Rucinolithus; T., Tubodiscus; N., Nannoconus; L., Lithraphidites; E., Ephrolithus; Rh., Rhagodiscus.

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

5 of 19

PA4212

record, mainly improving the reconstruction of the carbonisotope evolution during the upper Hauterivian and Barre-mian. A most accurate reconstruction of the Valanginiancarbon isotope excursion is achieved in terms of startingpoint and stratigraphic length.[32] In general, superimposed on the long-term trends,

higher-frequency oscillations (at different spatial scales) ofthe C isotope signal suggest a complex response of the carboncycle to different volcanotectonic and climate forcing.

[33] Oxygen isotope values (Figure 5) show long-termoscillations in a range of �1.0% around an average of��2.1% calculated throughout all the record. A longnegative trend from average d18O values of �1.3% tovalues of ��2.1% was observed from the base of theBerriasian to the upper Valanginian. This long trend ismodulated by shorter-term fluctuations in correspondenceof the top of the calcareous nannofossil C. angustiforatuszone, and the top of the Thurmanniceras otopeta ammonitezone and the base of the Thurmanniceras pertransiens

Figure 3. Correlation among the three studied sections (we reported only the coeval sedimentaryintervals) based on magnetostratigraphy, biostratigraphy, and cycle pattern organization. See text fordetails. Gray rectangles highlight the same lithologic pattern organization of coeval intervals recognizedin the three sedimentary sequences. Legend and abbreviations are as in Figure 2.

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

6 of 19

PA4212

ammonite zone shifting the d18O averages of �+0.7%.From middle Valanginian to the top of Hauterivian a trendto more positive oxygen isotope values is documented.Then, a rapid negative shift recorded at the base of theCM5N chron brings the d18O values to averages of �2.2%.This negative excursion is then followed by a long-termpositive trend (�0.7%) that ends with a final 1.2% negativeexcursion up to the top of the composite record. Super-imposed on this long-term record high-frequency d18Ooscillations, with a range of ±0.5%, were recognized.

5.2. Diagenetic Overprint

[34] The range of variability recorded in the measuredcarbon isotope curve corresponds to biogenic calcite pre-cipitated under open marine conditions during the earlyCretaceous [e.g., Weissert et al., 1985] whereas all d18Ovalues scattering between �3.5 and �0.5% appear depletedby �3% relative to diagenetically unaltered marine calcite[e.g., van de Schootbrugge et al., 2000, and referencetherein]. Our data suggest that the oxygen isotope compo-sition of the measured samples reflects elevated temperatureduring burial diagenesis and/or effects of meteoric diagen-esis, while the carbon isotopic composition seems lessaffected by these processes [e.g., Weissert, 1989]. Althoughabsence of covariance between d18O and d13C suggests alimited influence of secondary diagenesis on the isotoperecords, we preferred to restrict our discussion to the car-bon isotope signal that shows very similar absolute valuesand general trends previously recorded by coeval bulkd13C curves [e.g., Weissert, 1989; Lini et al., 1992; Lini,1994;Weissert et al., 1998;Grocke et al., 1999, 2003a, 2003b;Hennig et al., 1999; van de Schootbrugge et al., 2000;Wissler et al., 2001; Bersezio et al., 2002; Weissert and

Erba, 2004; Erba et al., 2004; Erba, 2004]. It firstly rulesout considerable meteoric or burial diagenesis affecting ourcarbon isotope signature. Obviously, postdepositional car-bonate precipitation resulting from the breakdown of or-ganic matter and consequently having very low carbonisotope values (��20%) may influence bulk carbon iso-tope composition. However, such carbonates are likelypresent in small amounts in the studied marine sediments,and in general, we suppose that their contribution to thed13C record can be ignored because of the overwhelmingdominance of pelagic carbonate. Finally, we recognize thatthe interpretation of the bulk carbonate isotopic analysescan be compromised by a combination of factors includingchanges in nannofossil species composition, changes in thesize distribution of planktonic foraminifera, and diagenesisand that isotope values may actually be monitoring varia-tions in diagenetic alteration within Milankovitch lithologiccycles, resulting from changes in carbonate abundance and/or composition. However, we do not believe that ourinterpretations are significantly compromised by these fac-tors and assume that the measured carbon isotope valuesreflect changes in the isotopic composition of the globaloceanic carbon reservoir. Finally, in order to reduce theeffect of differential diagenesis on the isotope record relatedto different lithology, samples were collected only from thecarbonate layers.

6. Power Spectral Analysis

[35] The strategy for application of spectral analysis to thed13C record followed two sequential steps. Firstly, thealgorithms of Lomb-Scargle [Lomb, 1976; Scargle, 1982]suitable for unevenly sampled records were applied to the

Figure 4. Age-depth profile for the upper part of the composite section on the basis of the age of thepaleomagnetic chron boundaries recognized throughout the composite record against composite depth.S.R. is estimated sedimentation rate based on a second-order best fit line. R2 is Parson’s correlationcoefficient. Magnetostratigraphy between 158 and 256.5 m of the composite section (corresponding tothe upper part of the Chiaserna Monte Acuto and Bosso sections) is from Channell et al. [1995].Magnetostratigraphy between 256.5 and 333 m of the composite section (corresponding to Gorgo aCerbara section) is from Lowrie and Alvarez [1984] and Channell et al. [2000].

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

7 of 19

PA4212

Figure 5. Bulk carbonate carbon and oxygen isotope data throughout the composite record. Thick graylines represent 15-point moving average calculated on the original isotope data. Legend and abbreviationsare as in Figure 2. MBE is mid-Barremian event [from Coccioni et al., 2003].

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

8 of 19

PA4212

Figure

6.

ComparisonbetweentheCisotopeUmbria-Marchecompositerecord

andthecomposited1

3CcurveofWeissert

andErba[2004].Thickgraybandsindicatecorrelationsbetweenpaleomagnetic

chronboundariesrecorded

inboth

the

compositesequences.SolidcirclesaredatafromtheCismonsection[Erbaetal.,1999].Open

circlesaredatafromCapriolo

section[Channellet

al.,1993].Crosses

aredatafrom

theValle

del

Missection[W

eissertandChannell,1989].

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

9 of 19

PA4212

composite signal. Power spectrum was estimated in thedepth domain and then interpreted, on the basis of theaverage sedimentation rate calculated for the middle-upperpart of the section, in terms of time periodicity. Then, afterdetection of the main periodicities modulating the signals, awavelet spectral analysis was performed on the d13C record,using the algorithm of Foster [1996]. It allowed us to verifythe variability of the major frequencies (due to potentialchanges in the sedimentation rate all along the threedifferent sections) throughout the composite sedimentaryrecord. The wavelet spectral analysis is the most recentsolution to overcome the shortcomings of the classicalFourier transform. In wavelet analysis the use of a fullyscalable modulated window function solves the signal-cutting problem. The window is shifted along the signaland for every position the spectrum is calculated. Then, thisprocess is repeated many times with a slightly shorter (orlonger) window for every new cycle. Results consist in anensemble of time-frequency representations of the signal, allwith different resolutions (multiresolution analysis). Tohandle irregularly sampled signals we needed an extensionof the classic wavelet formalism that was developed byFoster [1996]. The Foster’s algorithms define the wavelettransform as a suitable weighted projection onto three trialfunctions giving the weighted wavelet Z transform (WWZ)and the weighted wavelet amplitudes (WWA).[36] In Figure 7 the calculated Lomb-Scargle power

spectrum of the record evidences 7 main peaks of frequen-cies with an associated confidence interval higher than99.5%. Assuming an average sedimentation rate of�0.0144 m/kyr, calculated for the middle-upper part ofthe record, we estimated an average �400 kyr, 1200 kyr,1600 kyr, 2400 kyr, 2800 kyr, 3600 kyr and 9700 kyrlength for the detected cyclicities. The short-term compo-nent of eccentricity (corresponding to the periodicity of� 100 kyr) was not detected as statistically significant in

the d13C record, possibly due to the coarse sampling rateadopted in any segments of the composite record.[37] The Foster wavelet spectrum (Figure 8) shows that

the long-term (�400 kyr) eccentricity cycles enlarge from�0.0014 to �0.001 cycles/cm (corresponding to a wave-length variability between �6 and �10 m) from the base tothe top of the composite record. A similar frequency mod-ulation can be observed for the other two 1.2 and 2.4 m.y.periodicity bands shifting their frequencies from 0.0007 and0.0004 cycles/cm and 0.00025 and 0.0002 cycles/cmrespectively, from the base to the top of the compositerecord. Such a result assisted us for a proper selection ofthe central periodicity and respective bandwidths of differentband-pass filters to apply to the original d13C signal forextraction of selected periodicity bands from the isotoperecord.

7. Orbital Tuning of the Lower CretaceousComposite Record

[38] The superb lithologic expression of the three studiedsections, encouraged us to combine the results of the timeseries analysis applied to the d13C record with the classicapproach of lithologic cyclostratigraphy [e.g., Hilgen, 1991]in order to reconstruct a reliable orbital tuning of the earlyCretaceous.[39] The composite section is characterized by an evident

lithologic cyclicity with a periodic alternation of limestonebeds and cherty layers (in the Chiaserna Monte Acuto andBosso sections) and/or limestone beds and marls-blackshales (in the Gorgo a Cerbara section). The averagethickness of these basic cycles increases from �0.4 at thebase to �0.55 m in the upper part of the composite section.However, in the uppermost part of the sequence (upperBarremian), from about 300 m to the top of the compositesection, the average thickness of the basic cycles decreases

Figure 7. Lomb-Scargle power spectrum of the d13C signal. C.I. is confidence interval. Bandwidth is7.2 � 10�6. Highest peaks are labeled with average time periodicities estimated on the basis of thesedimentation rate calculated in the upper part of the composite record.

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

10 of 19

PA4212

again to about 0.3–0.35 m. The lithologic alternations aregenerally clustered in groups of 4/5 couplets in turn alter-nated to thick sedimentary intervals where chert and/orblack shale layers are rarer or totally absent (Figure 9).Two examples (one from the upper part of the ChiasernaMonte Acuto and one from the Bosso section) of lithologichierarchical organization of the sedimentary record arereported in Figure 10.[40] This hierarchical organization of lithologic cycles

was confirmed by results from Lomb-Scargle spectralanalysis performed on the digitalized sedimentary record(obtained by attributing two codes, 0 and 1 to the carbonateand cherty layers/black shales, respectively). The spectrum(Figure 11) shows several clear peridocities modulating thelithologic sequence, with frequency, related to precessionand eccentricity (cycles of 100, 400, and 2400 kyr) and1,200 kyr cycles, well represented. The chert/limestone andmarls-black shales/limestone couplets observed in theBosso and Gorgo a Cerbara sections were already inter-preted by Herbert [1992] to represent a sedimentary re-sponse to Milankovitch climate forcing. Particularly, resultsof spectral analysis based on the bedding thickness and twodifferent timescales [Kent and Gradstein, 1985; Harland etal., 1990; Herbert, 1992] indicated a duration of about18.5–23.5 kyr for the basic cycles (the precessional forcing)and of about 93–117 kyr for bundles of four/five couplets(the short-term eccentricity pattern). Moreover, followingthe work of Herbert [1992] and Fiet [1998] we assume thateach level of black shales and/or marls represents about one

half of the precessional cycle. The strong similarity ofthickness and lithologic pattern distribution between theblack shales-chert/limestone alternations recognized at theGorgo a Cerbara and Bosso sections and the cherty layers/limestone alternation at Chiaserna Monte Acuto allows thetime significance of the first cycle mode to be extrapolatedinto the older sedimentary composite succession.[41] Because of the lack of a reliable astronomical solu-

tion of the insolation curve for time intervals older than about45 Ma [Varadi et al., 2003; Laskar et al., 2004], we chose amultistep approach for our reconstruction. The carbon iso-tope record was filtered in the 0.001–0.0014 cycles/cmfrequency band (corresponding to the �400,000 yearslong-term eccentricity cycles). A direct comparison withthe lithologic sequence (Figure 9) shows that the lows in thefiltered signal generally correspond to sedimentary intervalswhere cherts and/or marls-black shales are rarer, whereashighs in the d13C filtered record correspond to sedimentaryintervals characterized by numerous and well clusteredlithologic couplets. A progressive numbering system, fromthe base of the section (number 1) up to the bottom of theCM0 paleomagnetic chron (number 46) was used to labelthe consecutive long-term eccentricity cycles. Cycles 1, 5,6, 9, 14, 15, 19, 27, 33, 34, 35, 41, 46, and 46 show anexcellent record of lithologic pattern organization such aspreviously described. Conversely, cycles 3, 8, 13, 17, 23,30, 40 characterize intervals where lithologic pairs are rareror totally absent.

Figure 8. Wavelet spectrum estimated for the carbon isotope record. On the right side the frequencyshift of the 0.4, 1.2, and 2.4 Ma periodicity bands throughout the composite record is evidenced. Contourline marks interval of > 99.5% confidence interval. Gray lines superimposed on the original carbonisotope curve represent a filtered record (cutoff frequency band of 0.0001 cycles/m).

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

11 of 19

PA4212

[42] This eccentricity response of the carbon isotoperecord does not necessarily imply an eccentricity-scaleEarth’s climate forcing, but rather suggests a robust preces-sional control of the record with a consequent significantamplitude modulation at the eccentricity periods of about100, 400, and 2,400 kyr [Herbert, 1996]. This eccentricity

modulation of the precessional cycles is testified by thepresence of the basic cycles more frequent and showingevident cluster pattern organization during maxima of the400 kyr d13C cycles.[43] The ratio of 13C to 12C in marine SCO2 varies as a

result of variations in the fraction of total carbon deposited

Figure 9. Orbital tuning of the composite record by comparison of lithologic pattern organization andband-pass filtered 0.4 (solid line) and 2.4 Ma (gray dotted line) d13C eccentricity cycles. Progressive 1–46numbering (for the 400 kyr cycles) and labeling A–Q (for the 2.4 Ma cycle) is discussed in detail in thetext. Legend and abbreviations are as in Figure 2. In the litho(1) column, all the lithological symbols arereported as in Figure 2. In the litho(2) column, only the cherty and/or black shale layers are reported toallow a more direct identification of the lithologic (short and long) eccentricity pattern cycles. Thebeginning and end of the Valanginian carbon shift, such as recognized in the carbon isotope curve anddefined by Erba et al. [2004], are reported along with the termination of the Valanginian/Hauteriviancarbon isotope perturbation (see text for definition).

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

12 of 19

PA4212

as carbonate or organic carbon and a net increase in organiccarbon deposition resulting in increasing d13C values forthe remaining SCO2 reflected in bulk carbonate isotopicmeasurements [Miller and Fairbanks, 1985; Vincent andBerger, 1985; Shackleton, 1987; Woodruff and Savin, 1991;Flower and Kennett, 1995; Zachos et al., 1997]. An in-creasing surface productivity and/or anoxic conditions and

consequent increasing organic carbon preservation in thesediments represent two possible sedimentary mechanismsto explain the positive d13C excursions during maxima ofeccentricity. The 100 and 400 kyr modulation of theprecession cycles in our composite record may have in-duced strongest 21 kyr insolation maxima with increasinglyhigh supply nutrients from continental runoff and/or expan-

Figure 10. Examples (from Chiaserna Monte Acuto and Gorgo a Cerbara sections) of lithologichierarchical organization of precession, short-term, and long-term eccentricity cycles. Magnetostrati-graphy of the Chiaserna Monte Acuto Bosso section is from Channell et al. [1995]. Magnetostratigraphyof the Gorgo a Cerbara section is from Lowrie and Alvarez [1984] and Channell et al. [2000].

Figure 11. Lomb-Scargle power spectrum of the lithologic record obtained by attributing two codes,0 and 1, to the carbonate and chert layers or black shales, respectively. C.I. is confidence interval.Bandwidth is 7.2 � 10�6. Horizontal axis is in logarithmic scale. Highest peaks are labeled with averagetime periodicities estimated on the basis of the sedimentation rate calculated in the upper part of thecomposite record.

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

13 of 19

PA4212

sion of more intense upwelling events. Enhanced produc-tivity during intervals of higher d13C is expressed in chertylayers deposition which testify with their high content ofradiolarians an evident increase in oceanic productivity[e.g., Leckie, 1989; Bralower et al., 1994; Erbacher et al.,1996].[44] The same phase relation between lithologic record

and d13C astronomically modulated eccentricity cycles wasassumed for the 2.4 Ma eccentricity cycles (extracted bythe original d13C record at the frequency of 0.0001 ±0.00005 cycles/m in Figure 7) that, compared with thelithologic pattern, allowed us to further constrain the orbitaltuning of the composite section. Once again, lows in the2.4 Ma filtered record correspond to long intervals (labeledin Figure 9 with A, C, E, G, I, M, O, and Q) characterizedby absence or rarer basic and/or 400 kyr lithologic cycles.Conversely, highs of the 2.4 Ma (labeled with letters B, D,F, H, L, N, and P) are generally characterized by moreregular alternation of lithologic pairs and well recognizablelithologic patterns attributable to 100 and 400 kyr cycles.[45] The last step in the orbital tuning was the identifica-

tion of the four clusters of 4/5 lithologic couplets withineach 400 kyr cycles, corresponding to the response of thesedimentary record to short-term eccentricity cycles. Unfor-tunately, the evenly sampling rate adopted for the compositerecord and the variable sedimentation rate did not allow usto reliably detect periodicity of 100 kyr in the d13Ccomposite signal thus limiting the possibility to constrainthe lithologic cyclostratigraphic interpretation with a 100 kyrfiltered isotope record. Nonetheless, where clearly evident,the 4/5 lithologic couplets, corresponding to 100 kyreccentricity cycles, were identified and reported inFigures 9 and 10.[46] The achieved orbital tuning of the studied record

provides a first tentative to astronomically calibrate theduration of the sedimentary cycles (Table 1) as well as ofthe magnetostratigraphic, biostratigraphic, and isotopeevents recorded throughout the composite section.

8. Orbital Calibration of Magnetostratigraphic,Biostratigraphic, and Isotope Events

[47] The orbital tuning resulted in an adjustment of theinitial age estimates of the polarity reversals. We used asstarting point an age of 125.0 m.y. [Gradstein et al., 2004]for the base of the youngest paleomagnetic chron (CM0)and consecutively added time calculated by multiplying thenumber of long-term eccentricity cycles singled out allalong the composite record by 400,000 years and thenumber of short-term eccentricity cycles (recognized onlyon the basis of lithological patterns) by 100,000 years.[48] In Figure 12 we reported the obtained curve of

sedimentation rate calculated plotting limits of the consec-utive 100,000 years cycles identified throughout the com-posite record. It shows a regular behaviour throughout therecord with an evident decrease in the sedimentation rateduring the upper part of the Barremian.[49] The estimated ages for the paleomagnetic chron

boundaries reported for the upper part of the compositerecord (Table 1) show differences ranging from about 0.2

and 2.0 m.y. (from the youngest to the oldest magneticchrons) when compared to the ages reported in the literaturefor the same events [e.g., Channell et al., 1995; Gradstein etal., 2004].[50] Assessment of the duration of the Stages documented

throughout the record and comparison with literature datacalls for a definition of base and top boundaries. Inparticular, three complete stages were recorded. From thebase of the composite section: Valanginian, Hauterivian,and Barremian. We used the definitions proposed byGradstein et al. [2004] to assess top and base of thesestages. Base of the Valanginian was defined by the firstoccurrence of Calpionellites darderi (base of Calpionellidzone E) [Bulot et al., 1996; Aguado et al., 2000; Gradsteinet al., 2004], followed by the lowest occurrence of ammo-nite Thurmanniceras pertransiens. Actually, taxonomic andcorrelation problem with ammonoid succession [e.g., Bulotet al., 1993] led Gradstein et al. [2004] to recommendplacing the base of the Valanginian Stage at the base ofCalpionellid zone E (approximately in the middle part ofmagnetic polarity chron CM14r), just below the FO of Ca.oblongata. Top of the Valanginian was defined by thelowest occurrence of A. radiatus [Gradstein et al., 2004].The duration of the Valanginian was here estimated to be�6.9 m.y. long (for a total of 69 short eccentricity cycles),�3 m.y. longer than that estimated by Gradstein et al.[2004] for the same stage (3.8 ± 1.0 m.y.).[51] The top of the Hauterivian was defined by the FO of

the Spitidiscus hugii ammonite zone [Gradstein et al., 2004]and within uppermost magnetic polarity zone CM4n (oruppermost part of chron CM5). The estimated duration ofthe stage was �3.5 m.y. (for a total of 35 short eccentricitycycles), about 3 m.y. shorter than that estimated byGradstein et al. [2004] who calculated a time span of 6.4± 1 m.y. for the same stage. The Barremian/Aptian bound-ary was suitably positioned at the base of the CM0 magneticchron according to Kent and Gradstein [1985] andGradstein et al. [1995, 2004]. Consequently, a duration of�4.4 m.y. for the Barremian was calculated more than0.6 m.y. shorter than the estimated 5.0 ± 0.5 m.y. durationreported by Gradstein et al. [2004]. A difference of - �0.1–0.3 m.y. exists with the cyclostratigraphic reconstructionreported by Fiet and Gorin [2000] for a different sedimen-tary sequence at Gorgo a Cerbara (sampled ‘‘in the calcar-eous promontory located on other side of the river’’, as citedby the authors). In particular, the most relevant lithologicdifference between the two logs can be observed in theupper part of the record (within the CM1n segment) where,slumping and/or faults could have repeated the correctsequence of sediments in the sequence of Fiet and Gorin[2000]. Comparison in the field between the two recon-structed sedimentary records with the sedimentary logpreviously reported by Cecca [1995] for the same section,confirmed the reliability of the lithologic reconstructionpresented in this paper. Finally, the application of the orbitaltuning to our section, where original paleomagnetic andbiostratigraphic data were collected [e.g., Cecca, 1995;Channell et al., 1995], ensures a faithful and coherentreconstruction of the sedimentary record. Comparison withcyclostratigraphic results obtained for the upper part of the

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

14 of 19

PA4212

Gorgo a Cerbara section by Herbert [1992] shows a generalgood agreement (differences in duration of the paleomag-netic chrons < 0.15 m.y.). The �0.5 m.y. difference esti-mated for the paleomagnetic chron CM1n could be due tothe lack of a detailed lithological log in the work by Herbert

[1992]. Moreover, slight differences in the results may bedue to the smoothing and interpolation procedures appliedto the of data in order to perform Fourier spectral analysis.Therefore we believe that the results here presented

Table 1. Synthesis of Major Stratigraphic Events Throughout the Composite Recorda

Datum Chron

Age, Ma

MetersCompositeSection

Duration of Magnetic Chrons, m.y.

Gradsteinet al.[2004]

ThisWork

Gradstein etal. [2004] -This Work

According toGradstein

et al. [2004]

Accordingto ThisWork

Accordingto Herbert[1992]

According toFiet and

Gorin [2000]

MagnetostratigraphyBase M0n 124.61 124.80 332.90 0.19 – – – –Base M0r 125.00 125.00 331.50 0.00 0.39 0.20 0.20 0.35 ± 0.04Base M1n 127.24 126.85 310.00 �0.39 2.24 1.85 2.45 2.28 ± 0.24Base M1r (or M1) 127.61 127.14 305.50 �0.47 0.37 0.29 0.45 0.43 ± 0.21Base M3n (or M2) 128.11 127.84 294.20 �0.27 0.50 0.70 0.70 0.70 ± 0.19Base M3r (or M3) 129.76 129.30 268.80 �0.46 1.65 1.46 partially

investigated1.75 ± 0.24

Base M5n (or M4) 130.80 129.92 255.70 �0.88 1.04 0.62Base M5r (or M5) 131.19 130.20 250.40 �0.99 0.39 0.28Base M6n 131.41 130.34 248.00 �1.07 0.22 0.14Base M6r 131.56 130.44 245.90 �1.12 0.15 0.10Base M7n 131.85 130.56 243.20 �1.29 0.29 0.12Base M7r 132.20 130.99 235.20 �1.21 0.35 0.43Base M8n 132.52 131.09 233.30 �1.43 0.32 0.10Base M8r 132.83 131.24 229.10 �1.59 0.31 0.15Base M9n 133.14 131.34 225.40 �1.80 0.31 0.10Base M9r 133.50 131.64 217.50 �1.86 0.36 0.30Base M10n 133.87 131.84 212.00 �2.03 0.37 0.20Base M10r 134.30 132.04 207.30 �2.26 0.43 0.20Base M10Nn.1n 134.62 132.29 200.80 �2.33 0.32 0.25Base M10Nn.1r 134.67 132.34 199.90 �2.33 0.05 0.05Base M10Nn.2n 134.98 132.64 194.70 �2.34 0.31 0.30Base M10Nn.2r 135.00 132.69 194.20 �2.31 0.02 0.05Base M10Nn.3n 135.28 132.99 188.00 �2.29 0.28 0.30Base M10Nr 135.69 133.44 182.50 �2.25 0.41 0.45Base M11n 136.44 134.14 168.50 �2.30 0.75 0.70Base M11r.1r 136.68 134.54 159.70 �2.14 0.24 0.40

Isotope StratigraphyStart Valanginin

C shift136.34 131.00

End ValanginianC shift

134.14 171.10

BiostratigraphyFO N. truitti 125.50 326.30FO R. irregularis 125.55 326.00LO Ca. oblongata 129.05 273.50Faraoni level 129.64 261.70FO R. terebrodentarius 131.09 233.30LO Cr. cuvillieri 131.24 229.10FO L. bollii 132.18 208.47LO Calpionellids 132.84 195.00FO N. bucheri 133.46 182.77LO T. verenae 134.14 171.10LO R. wisei 136.34 131.00FO T. verenae 137.94 101.30FO Ca. oblongata 139.94 60.00

aColumns are as follows: columns 1 and 2, magnetostratigraphic, biostratigraphic, and isotope events reported throughout the composite record; column3, magnetostratigraphic ages by Gradstein et al. [2004] for the magnetic chron boundaries recognized in the middle upper part of the composite record;column 4, orbitally calibrated age obtained in this work for the different magnetostratigraphic, biostratigraphic, and isotope events recognized all alongthe composite record; column 5, depth (m) of the stratigraphic events recognized through the composite section; column 6, difference (in m.y.) between theages reported by Gradstein et al. [2004] and those obtained by orbital calibration of the composite record for the magnetic boundaries recognized in themiddle upper part of the section; column 7, difference (in m.y.) between the consecutive magnetic boundaries recognized in the middle upper part ofthe section following Gradstein et al. [2004]; column 8, difference (in m.y.) between the consecutive magnetic boundaries recognized in the middle upperpart of the section based on the orbital calibration of the composite record (this work); column 9, difference (in m.y.) between the consecutive magneticboundaries recognized in the middle upper part of the section following Herbert [1992]; column 10, difference (in m.y.) between the consecutive magneticboundaries recognized in the middle upper part of the section following Fiet and Gorin [2000].

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

15 of 19

PA4212

generally confirm and at times refine those previouslyreported by Herbert [1992].[52] Finally, from the top of the Berriasian to the base of

the section a time length of 3.6 m.y. was estimated. Theduration of the Berriasian referred by Gradstein et al.[2004] is about 5.3 m.y. The base of the Cretaceous Periodcurrently lacks an accepted global boundary definition,despite over a dozen international conferences and workinggroups dedicated to the issue since 1974 [Zakharov et al.,1996]. However, traditional usage favors the lowest occur-rence of the ammonite Berriasella jacobi for defining thebase of the Cretaceous as the base of the Berriasian Stage[Gradstein et al., 2004]. This event was not recognized inthe lowermost part of the Chaserna Monte Acuto sectionexcluding the possibility to identify the base of theCretaceous.[53] Although the duration of the three investigated lower

Cretaceous stages (Valanginian, Hauterivian and Barremian)evidences major differences when our data are comparedwith those of Gradstein et al. [2004], the time span of thestratigraphic interval between the well constrained top of theBarremian and base of the Valanginian is absolutely com-parable (14.9 m.y. versus 15.0 m.y. of Gradstein et al.

[2004]). This result calls for a more appropriate approxi-mation of the intermediate boundaries of the top ofValanginian and Hauterivian, respectively, currently basedon ammonoid biostratigraphy.[54] The Valanginian carbon isotope shift (Figure 9), such

as defined by Erba et al. [2004] was estimated to be�2.3 m.y. long and starts from 136.34 m.y. The whole lateValanginian–early Hauterivian carbon isotope excursion(from the start of the Weissert event to the return of averagevalues of d13C � 1.9% typical of the following middleHauterivian–middle Barremian stratigraphic interval) wasestimated to last about 5.3 m.y. (Figure 9). Superimposed onthe carbon isotope excursion, we recognized long-termeccentricity cycles that do not show amplitude differenceswith the same 400 kyr oscillations recorded in the other partof the composite record. Thus climate forcing and relatedresponse of carbon cycle appear faithfully recorded alsoduring the carbon isotope Weissert event and all the lateValaniginian–middle Hauterivian carbon isotope excursion,although we exclude their primary control on this C cycleperturbation.[55] The obtained cyclostratigraphic results offer the

opportunity to test hypothesis that Parana volcanism

Figure 12. Age-depth profile of the composite section based on the orbital tuning of the record. Limitsof the consecutive 100,000 year cycles recognized throughout the composite record were reported. Thesedimentation rate estimated by a third-order polynomial fitting (R2 = 0.9996) curve (age (Ma) = 3 �10�8 � depth3 – 2 � 10�8 � depth2 + 0.0578 � depth + 0.0247) is, on average, 0.145 ± 0.008 m/kyr allalong the whole sedimentary record. Magnetostratigraphy between 158 and 256.5 m of the compositesection (corresponding to the upper part of the Chiaserna Monte Acuto and Bosso sections) is fromChannell et al. [1995]. Magnetostratigraphy between 256.5 and 333 m of the composite section(corresponding to the Gorgo a Cerbara section) is from Lowrie and Alvarez [1984] and Channell et al.[2000].

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

16 of 19

PA4212

triggered perturbation of C cycle [e.g., Channell et al.,1993; Erba et al., 2004; Weissert and Erba, 2004] and inparticular hypothesis of an important perturbation of the car-bon cycle in response to the intrusion of one of the world’slargest flood volcanic provinces (the Parana-Etendeka largeigneous province, LIP). This LIP intrusion is related inboth time and space to the opening of the South Atlantic,the disruption of western Gondwana, and more specificallythe extension across the Tristan da Cunha mantle plume[e.g., Hawkeswarth et al., 1992].[56] Although it is now generally accepted that Parana-

Etendeka volcanism occurred predominantly in the 135–130 Ma time span [Peate et al., 1992; Milner et al., 1995;Renne et al., 1996; Turner et al., 1996], recent 39Ar/40Ardating [Stewart et al., 1996] suggested increased eruptionrates between 138 and 131 m.y. from 0.031 km3/y toaverages (during the parossistic phase) of 0.21 km3/y witha peak at 132 m.y. for a total estimated duration of about10–12 m.y. for this volcanotectonic event.[57] The age of the base of the Valanginian carbon isotope

excursion was estimated in this work to be about 136.34 m.y.,which is in excellent agreement with the estimated age of theEtendeka-Parana starting event. This confirms that therecould have been a causal link between volcanism and Ccycle perturbation, which is documented in the Valanginian-Hauterivian C isotope excursions such as previously docu-mented by Channell et al. [1993] and Erba et al. [2004].

9. Conclusions

[58] The lower Berriasian–lower Aptian C and O isotopeUmbria-Marche composite record, about 19 m.y. long,

provided a nice opportunity to orbitally tune the record onthe basis of classic lithostratigraphic approach and applica-tion of different power spectral methodologies and filteringtechniques to the d13C signal. The cyclostratigraphic recon-struction of the composite section can be considered accu-rate to the scale of the short-term eccentricity cycles. Inparticular, orbital tuning of the record provided accuratetime calibration (1) for the paleomagnetic chron boundaries(CM0 to CM11 magnetic chrons) recognized throughout thesection that show a up to 2 m.y. difference (younger) onrespect to those reported by Gradstein et al. [2004], (2) forthe three complete Stages (Valanginian, Hauterivian andBarremian) recognized all along the composite recordand estimated to last �6.9, �3.5 and �4.4 m.y. respectively,and (3) for the starting point (136.34 m.y.) and duration(�2.3 m.y.) of the Valanginian carbon shift.[59] On the basis of this latter result, the estimated dating

of the start of the Valanginian carbon shift and the paross-istic phase of the Parana-Etendeka lavas eruption (between136 and 131 m.y. according to Stewart et al. [1996])suggests a more definitive and reliable causal link betweenthe two events.

[60] Acknowledgments. Many thanks are given to Michele Iavaronefor his heavy isotope acquisition job. We would like to thank A. Marini,M. Olivieri, and S. Galeotti for their help in the field. We are also grateful toE. Erba and H. Weissert for their helpful and meticulous reviews andcomments. This research was supported by the MIUR COFIN 2001 andMIUR ex 60% to R.C. Centro di Geobiologia publication 10.

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

17 of 19

PA4212

ReferencesAguado, R., M. Company, and J. M. Tavera(2000), The Berriasian/Valanginian boundaryin the Mediterranean region: New data fromthe Caravaca and Cehegin sections, SE Spain,Cretaceous Res., 21, 1–21.

Arthur, M. A., H. J. Brumsack, H. C. Jenkyns,and S. O. Schlanger (1990), Stratigraphy, geo-chemistry, and paleoceanography of organiccarbon-rich Cretaceous sequences, in Cretac-eous Resources, Events, and Rhythms, editedby R. N. Ginsburg, and B. Beaudoin, pp. 75–119, Springer, New York.

Bersezio, R., E. Erba, M. Gorza, and A. Riva(2002), Berriasian-Aptian black shales of theMaiolica formation (Lombardian Basin,southern Alps, northern Itay): Local to globalevents, Palaeogeogr. Palaeoclimatol. Palaeoe-col., 180, 253–270.

Bralower, T. J., S. Monechi, and H. R. Thierstein(1989), Calcareous nannofossil zonation of theJurassic-Cretaceous boundary interval and cor-relation with the geomagnetic polarity time-scale, Mar. Micropaleontol., 14, 153–235.

Bralower, T. J., M. A. Arthur, R. M. Leckie,W. V. Sliter, D. J. Allard, and O. Shlanger(1994), Timing and paleoceanography of ocea-nic disoxia/anoxia in the late Barremian toearly Aptian, Palaios, 9, 335–369.

Bralower, T. J., R. M. Leckie, W. V. Sliter, andH. R. Thierstein (1995), An integrated Cretac-eous microfossil biostratigraphy, in Geochro-

nology Time Scales and Global StratigraphicCorrelation, edited by W. A. Berggren, et al.,Spec. Publ. SEPM Soc. Sediment. Geol., 54,65–79.

Bulot, L. G., E. Blanc, J. P. Thieuloy, andJ. Remane (1993), La limite Berriasien-Valanginien dans le Sud-Est de la France:Donnees biostratigraphiques nouvelles, C.R.Acad. Sci. Paris, Ser. IIa, Sci. Terre Planetes,317, 387–394.

Bulot, L. G., et al. (1996), The ValanginianStage, Bull. Inst. R. Sci. Nat. Belg. Sci. Terre,suppl. 66, 11–18.

Cecca, F. (1985), Alcune Ammoniti provenientidalla ‘‘Maiolica’’ dell’Appennino Centrale(Umbria, Marche e Sabina), Boll. Serv. Geol.Ital., 103, 133–162.

Cecca, F. (1995), Late Valanginian ammonitesfrom Monte Catria (Umbria-Marche Apen-nines, Italy), Mem. Descr. Carta Geol. It., 51,47–57.

Cecca, F., and G. Pallini (1994), LatestHauterivian-Barremian ammonite biostratigra-phy in the Umbria-Marche Apennines (centralItaly), Geol. Alp., 20, 205–217.

Cecca, F., A. Marini, G. Pallini, F. Baudin, andV. Begouen (1994a), A guide-level of theuppermost Hauterivian (lower Cretaceous) inthe pelagic succession of Umbria-MarcheApennines (central Italy): The Faraoni level,Riv. Ital. Paleontol. Stratigr., 99, 551–568.

Cecca, F., G. Pallini, E. Erba, I. Premoli Silva,and R. Coccioni (1994b), Hauterivian-Barremian chronostratigraphy based on ammo-nites, nannofossils, planktonic foraminiferaand magnetic chrons from the Mediterraneandomain, Cretaceous Res., 15, 457–467.

Channell, J. E. T., F. E. Erba, and A. Lini (1993),Magnetostratigraphic calibration of the lateValanginian carbon isotope event in pelagiclimestones from northern Italy and Switzerland,Earth Planet. Sci. Lett., 85, 222–240.

Channell, J. E. T., F. Cecca, and E. Erba (1995),Correlations of Hauterivian and Barremian(early Cretaceous) stage boundaries to polaritychrons, Earth Planet. Sci. Lett., 134, 125–140.

Channell, J. E. T., E. Erba, G. Muttoni, andF. Tremolada (2000), Early Cretaceous mag-netic stratigraphy in the APTICORE drill coreand adjacent outcrop at Cismon (southernAlps, Italy), and correlation to the proposedBarremian-Aptian boundary stratotype, Geol.Soc. Am. Bull., 112, 1430–1443.

Coccioni, R. (1996), The Cretaceous of theUmbria-Marche Apennines (central Italy), inSymposium Cretaceous Stratigraphy, Paleo-biology and Paleobiogeography, edited byJ. Wiedmann, pp. 129 – 136, Tubingen,Germany.

Coccioni, R., O. Nesci, M. Tramontana, F. C.Wezel, and E. Moretti (1987), Descrizione diun livello guida ‘‘radiolaritico-bituminoso-

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

18 of 19

PA4212

ittiolitico’’ alla base delle Marne a Fucoidinell’Appennino umbro-marchigiano, Boll.Soc. Geol. It., 106, 183–192.

Coccioni, R., R. Franchi, O. Nesci, F. C. Wezel,F. Battistini, and P. Pallecchi (1989), Stratigra-phy and mineralogy of the Selli level (earlyAptian) at the base of the Marne a Fucoidi inthe Umbria-Marchean Apennines (Italy), inCretaceous of the Western Tethys, edited byJ. Wiedmann, pp. 563–584, Schweizebart,Stuttgart, Germany.

Coccioni, R., F. Baudin, F. Cecca, M. Chiari,S. Galeotti, S. Gardin, and G. Salvini (1998),Integrated stratigraphic, paleontological, andgeochemical analysis of the uppermost Hauter-ivian Faraoni level in the Fiume Bosso section,Umbria-marche Apennines, Italy, CretaceousRes., 19, 1–23.

Coccioni, R., S. Galeotti, and M. Sprovieri(2003), The Mid-Barremian event: The preludeto the OAE1a, Eos. Trans. AGU, 84(46), FallMeet. Suppl., Abstract F904.

Erba, E. (2004), Calcareous nannofossils andMesozoic oceanic anoxic events, Mar. Micro-paleontol., 52, 85–106.

Erba, E., J. E. T. Channell, and M. Claps (1999),Integrated stratigraphy of the Cismon APTI-CORE (southern Alps, Italy): A ‘‘reference’’section for the Barremian-Aptian interval atlow latitudes, J. Foraminiferal Res., 29,371–391.

Erba, E., A. Bartolini, and R. L. Larson (2004),Valanginian Weissert oceanic anoxic event,Geology, 32, 149–152.

Erbacher, J., J. Thurow, and R. Littke (1996),Evolution patterns of radiolarian and organicmatter variations: A new approach to identifysea-level changes in mid-Cretaceous pelagicenvironments, Geology, 24, 499–502.

Faraoni, P., D. Flore, A. Marini, G. Pallini, andN. Pezzoni (1997), Valanginian and early Hau-terivian ammonites successions in the Mt Ca-tria group (central Apennies) and in the LessiniMts (southern Alps), Italy, Palaeopelagos, 7,59–100.

Fiet, N. (1998), Black shales, a high-resolutionchronological tool: Case study of the Albianfrom the Umbria-Marche Basin (central Italy),Bull. Soc. Geol. Fr., 169, 221–231.

Fiet, N., and G. Gorin (2000), Lithological ex-pression ofMilankovitch cyclicity in carbonate-dominated pelagic Barremian deposits incentral Italy, Cretaceous Res., 21, 457–467.

Flower, B., and J. Kennett (1995), Middle Mio-cene deepwater paleoceanography in thesouthwest Pacific: Relations with East Antarc-tic Ice Sheet development, Paleoceanography,10, 1095–1112.

Foster, G. (1996), Wavelets for period analysis ofunevenly sampled time series, Astron. J., 112,1709–1729.

Gradstein, F. M., F. P. Agterberg, J. G. Ogg,J. Hardenbol, P. Van Veen, T. Thierry, andZ. Huang (1995), ATriassic, Jurassic and Cre-taceous time scale, in Geochronology, TimeScales and Global Stratigraphic Correlations:Unified Temporal Framework for a HistoricalGeology, edited by W. A. Berggren, et al.,Spec. Publ. SEPM Soc. Sediment. Geol., 54,95–128.

Gradstein, F. M., et al. (2004), A Geologic TimeScale 2004, 500 pp., Cambridge Univ. Press,New York.

Grocke, D. R., S. P. Hesselbo, and H. C. Jenkins(1999), Carbon-isotope composition of lowerCretaceous fossil wood: Ocean-atmosphere,chemistry and relation to sea-level change,Geology, 27, 155–158.

Grocke, D. R., G. D. Price, E. Baraboshkin,J. Mutterlose, and A. H. Ruffell (2003a), TheValanginian terrestrial carbon-isotope record,Geophys. Res. Abstr., 5, Abstract 13644.

Grocke, D. R., G. D. Price, A. H. Ruffell,J. Mutterlose, and E. Baraboshkin (2003b),Isotopic evidence for late Jurassic–early Cre-taceous climate change, Palaeogeogr. Palaeo-climatol. Palaeoecol., 202, 97–118.

Harland, W. B., R. L. Armstrong, A. V. Cox,L. E. Craig, A. G. Smith, and D. G. Smith(1990), A Geologic Time Scale 1989, 263pp., Cambridge Univ. Press, New York.

Hawkeswarth, C. J., K. Gallagher, S. Kelley,M. S.M.Mantovani, D.W. Peate, M. Regelous,and N. W. Rogers (1992), Parana magmatismand the opening of the South Atlantic, in Mag-matism and the Causes of Continental Break-up, edited by B. C. Storey, et al., Geol. Soc.Spec. Publ., 68, 2210–2240.

Hennig, S., H. Weissert, and L. Bulot (1999),C-isotope stratigraphy, a calibration tool be-tween ammonite and magnetostratigraphy:The Valanginian-Hauterivian transition, Geol.Carpathica, 50, 91–96.

Herbert, T. D. (1992), Paleomagnetic calibrationof Milankovitch cyclicity in lower Cretaceoussediments, Earth Planet. Sci. Lett., 112, 15–28.

Herbert, T. D. (1996), A long marine history ofcarbon cycle modulation by orbital-climaticchanges, Proc. Natl. Acad. Sci. U.S.A., 94,8362–8369.

Herbert, T. D., and A. G. Fischer (1986), Milan-kovitch climatic origin of mid-Cretaceousblack shale rhythms in central Italy, Nature,321, 739–743.

Hilgen, F. J. (1991), Extension of the astronomi-cally calibrated (polarity) time scale to theMiocene-Pliocene boundary, Earth Planet.Sci. Lett., 107, 349–368.

Hoedemaeker, P. J., et al. (1993), Ammoniteszonation for the lower Cretaceous of the Med-iterranean region: Basis for the stratigraphiccorrelations within IGCP-Project 262, Rev.Esp. Paleontol., 8, 117–120.

Hoedemaeker, P. J., et al. (1995), Report on the3rd International Workshop on the StandardLower Cretaceous Ammonite Zonation of theMediterranean Region, Mem. Descr. CartaGeol. Ital., 51, 213–215.

Kent, D. V., and F. M. Gradstein (1985), A Cre-taceous and Jurassic geochronology, Geol.Soc. Am. Bull., 96, 1419–1427.

Laskar, J., F. Joutel, and F. Boudin (2004), Along term numerical solution for the insolationquantities of Earth, Astron. Astrophys., 428,261–285.

Leckie, R. M. (1989), A paleoceanographic mod-el for the early evolution history of planktonicforaminifera, Palaeogeogr. Palaeoclimatol.Palaeoecol., 73, 107–138.

Lini, A. (1994), Early Cretaceous carbon isotopestratigraphy of the Maiolica Formation, south-ern Alps (northern Italy and southern Switzer-land): Stratigraphic and paleoenvironmentalsignificance, Ph.D. thesis, 259 pp., Eidg. Tech.Hocsch. Zurich, Zurich, Switzerland.

Lini, A., H. Weissert, and E. Erba (1992), TheValanginian carbon isotope event: A firstepisode of greenhouse climate conditionsduring the Cretaceous!, Terra Nova, 4,374–384.

Lomb, N. R. (1976), Least-square frequency ana-lysis of unequally spaced data, Astrophys.Space Sci., 29, 447–462.

Lowrie, W., and W. Alvarez (1984), Lower Cre-taceous magnetostratigraphy in Umbrian pela-

gic limestone sections, Earth Planet. Sci. Lett.,71, 315–328.

Menegatti, A. P., H. Weissert, R. S. Brown,R. V. Tyson, P. Farrimond, A. Strasser, andM. Carnon (1998), High-resolution d13C strati-graphy though the early Aptian ‘‘Livello Selli’’of the Alpine Tethys, Paleoceanography, 13,530–545.

Miller, K., and R. Fairbanks (1985), Oligocene toMiocene carbon isotope cycles and abyssal cir-culation changes, in The Carbon Cycle andAtmospheric CO2: Natural Variations, Archeanto Present, Geophys. Monogr. Ser., vol. 32,edited by E. Sundquist, and W. Broecker,pp. 469–486, AGU, Washington, D. C.

Milner, S. C., A. R. Duncan, A. M. Whittingham,and A. Ewart (1995), Trans-Atlantic correla-tion of eruptive sequences and individual si-licic volcanic units in the Parana-Etendekaigneous province, J. Volcanol. Geotherm.Res., 69, 137–157.

Monechi, S., and H. R. Thierstein (1985), LateCretaceous-Eocene nannofossil and magnetos-tratigraphic correlations near Gubbio, Italy,Mar. Micropaleontol., 9, 419–440.

Peate, D.W., C. J. Hawkeswarth, andM.Mantovani(1992), Chemical stratigraphy of the Paranalavas, South America: Classification of themagma types and their spatial distribution,Bull. Volcanol., 55, 119–139.

Renne, P. R., J. M. Glen, S. C. Milner, and A. R.Duncan (1996), Age of Etendeka flood volcan-ism and associated intrusions in southwesternAfrica, Geology, 24, 659–662.

Scargle, J. D. (1982), Studies in astronomicaltime series analysis, II Statistical aspects ofspectral analysis of unevenly spaced data,Astrophys. J., 263, 835–853.

Shackleton, N. (1987), The carbon isotope recordof the Cenozoic: History of organic carbonburial and of oxygen in the ocean and atmo-sphere, in Marine Petroleum Source Rocks,edited by J. Brooks and A. Fleet, Geol. Soc.Spec. Publ., 26, 423–434.

Spotl, C., and T. W. Vennemann (2003), Contin-uous-flow isotope ratio mass spectrometricanalysis of carbonate minerals, Rapid Com-mun. Mass Spectrom., 17, 1004–1006.

Stewart, K., S. Turner, S. Kelley, C. Hawkeswart,L. Kirstein, and M. Mantovani (1996), 3-D40Ar–39Ar geochronology in the Parana con-tinental flood basalt province, Earth Planet.Sci. Lett., 143, 95–109.

Turner, S., C. Hawkesworth, K. Gallagher,K. Stewart, D. Peate, and M. Mantovani(1996), Mantle plumes, flood basalts, and ther-mal models for melt generation beneath con-tinents: Assessment of a conductive heatingmodel, J. Geophys. Res., 101, 11,503–11,518.

van de Schootbrugge, B., K. B. Follmi, L. G.Bulot, and S. J. Burns (2000), Paleoceano-graphic changes during the early Cretaceous(Valanginian-Hauterivian): Evidence fromoxygen and carbon stable isotopes, Earth Pla-net. Sci. Lett., 181, 15–31.

Varadi, F., B. Bunnegar, and M. Ghil (2003),Successive refinements in long-term integra-tions of planetary orbits, Astrophys. J., 592,620–630.

Vincent, E., and W. Berger (1985), Carbon diox-ide and polar cooling in the Miocene: TheMonterey hypothesis, in The Carbon Cycleand Atmospheric CO2: Natural VariationsArchean to Present, Geophys. Monogr. Ser.,vol. 32, edited byE. Sundquist, andW.Broecker,pp. 455–468, AGU, Washington, D. C.

Weissert, H. (1989), C-isotope stratigraphy asmonitor of paleoenvironmental changes: A

PA4212 SPROVIERI ET AL.: LOWER CRETACEOUS ORBITAL TUNING

19 of 19

PA4212

case study from the early Cretaceous, Surv.Geophys., 10, 1–16.

Weissert, H., and J. E. T. Channell (1989), Teth-yan carbonate carbon isotope stratigraphyacross the Jurassic-Cretaceous boundary: Anindicator of decelerated carbon cycling, Paleo-ceanography, 4, 483–494.

Weissert, H., and E. Erba (2004), Volcanism,CO2 and paleoclimate: A late Jurassic–earlyCretaceous carbon and oxygen isotope record,J. Geol. Soc. London, 161, 695–702.

Weissert, H., J. A.McKenzie, and J. E. T. Channell(1985), Natural variations in the carbon cycleduring the early Cretaceous, in The Carbon Cy-cle and Atmospheric CO2: Natural VariationsArchean to Present, Geophys. Monogr. Ser.,vol. 32, edited by E. Sundquist, andW. Broecker,pp. 455–468, AGU, Washington, D. C.

Weissert, H., K. B. Lini, K. B. Follmi, andO. Kuhn (1998), Correlation of early Cretac-

eous carbon isotope stratigraphy and platformdrowning events: A possible link?, Palaeo-geogr. Palaeoclimatol. Palaeoecol., 137,189–203.

Wissler, L., H. Funk, and H. Weissert (2001),Response of early Cretaceous carbonate plat-forms to changes in atmospheric carbon diox-ide levels, Palaeogeogr. Palaeoclimatol.Palaeoecol., 200, 187–205.

Woodruff, F., and S. M. Savin (1991), Mid-Miocene isotope stratigraphy in the deep-sea:High resolution correlations, paleoclimatic cy-cles, and sediment preservation, Paleoceanogra-phy, 6, 755–806.

Zachos, J., B. Flower, and H. Paul (1997),Orbitally paced climate oscillations across theOligocene/Miocene boundary, Nature, 388,567–570.

Zakharov, V. A., P. Bown, and P. F. Rawson(1996), The Berriasian stage and the Jurassic-

Cretaceous boundary, Bull. Inst. R. Sci. Nat.Belg. Sci. Terre, suppl. 66, 7–10.

�������������������������R. Coccioni, Istituto di Geologia e Centro di

Geobiologia, Universita di Urbino, CampusScientifico, Localita Crocicchia, I-61029 Urbino,Italy.F. Lirer, N. Pelosi, and M. Sprovieri, IAMC,

CNR, Calata Porta di Massa, Interno Porto diNapoli, I-80100 Napoli, Italy. ([email protected])F. Lozar, Dipartimento di Scienze della Terra,

Universita di Torino, Via Valperga Caluso 37,I-10123 Torino, Italy.