13. upper cretaceous cyclic sediments from hole 762c

von Rad, U., Haq, B. U., et al., 1992Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 122

13. UPPER CRETACEOUS CYCLIC SEDIMENTS FROM HOLE 762C, EXMOUTH PLATEAU,NORTHWEST AUSTRALIA1

Zehui Huang,2 Ron Boyd,2 and Suzanne O'Connell3

ABSTRACT

Well-developed Campanian to Maestrichtian pelagic cyclic sediments were recovered from Hole 762C on theExmouth Plateau, off northwest Australia, during Ocean Drilling Program Leg 122. The cycles consist of nannofossilchalk (light beds) and clayey nannofossil chalk (dark beds). Both light and dark beds are strongly to moderatelybioturbated, alternate on a decimeter scale, and exhibit gradual boundaries. Bioturbation introduces materials froma bed of one color into an underlying bed of another color, indicating that diagenesis is not responsible for thecyclicity. Differences in composition between the light and dark beds, revealed by calcium carbonate measurementand X-ray diffraction analysis, together with trace fossil evidence, indicate that the cycles in the sediments are adepositional feature. Diagenetic processes may have intensified the appearance of the cycles.

Spectral analysis was applied to the upper Campanian to lower Maestrichtian cyclic sediments to examine theregularity of the cycles. Power spectra were calculated from time series using Walsh spectral analysis. The mostpredominant wavelengths of the color cycles are 34-41 cm and 71-84 cm. With an average sedimentation rate of1.82 cm/k.y. in this interval, we found the time durations of the cycles to be around 41 k.y. and 21 k.y., respectively,comparable to the obliquity and precession periods of the Earth's rotation, which strongly suggests an orbital originfor the cycles.

On the basis of sedimentological evidence and plate tectonic reconstruction, we propose the followingmechanism for the formation of the cyclic sediments from Hole 762C. During the Late Cretaceous, when there wasno large-scale continental glaciation, the cyclic variations in insolation, in response to cyclic orbital changes,controlled the alternation of two prevailing climates in the area. During the wetter, equable, and warmer climaticphases under high insolation, more clay minerals and other terrestrial materials were produced on land and suppliedby higher runoff to a low bioproductivity ocean, and the dark clayey beds were deposited. During the drier andcolder climatic phases under low insolation, fewer clay minerals were produced and put into the ocean, wherebioproductivity was increased and the light beds were deposited.

INTRODUCTION

Since Hays et al. (1976) successfully related the climaticcyclicity in Quaternary sediments to the orbital cycles pro-posed by Milankovitch, many researchers have attempted torelate pre-Pleistocene cyclic sediments, both from land out-crop and from deep-sea drilling, to orbital cycles (e.g.,Schwarzacher and Fischer, 1982; Arthur et al., 1984; Cotillonand Rio, 1984; Barron et al., 1985; Weedon, 1986; Herbert andFischer, 1986; Schwarzacher, 1987; Hart, 1987; Bottjer et al.,1986; Research on Cretaceous Cycles (R.O.C.C.) Group,1986; Ogg et al., 1987; Jarrard and Arthur, 1989) with variousdata sets and approaches. However, controversy often fol-lowed. The question of diagenetic origin vs. climatic origin ofthe basal Jurassic cyclic sediments (Blue Lias) of south Britainis one example (Weedon, 1986; Hallam, 1986). This case is theresult of the relatively old geological age of the strata, insuf-ficient stratigraphic control, errors in the time scale, andevident diagenetic overprint. Another example is the origin ofshort-term sea-level fluctuations vs. origin of climatic fluctu-ations in the Upper Cretaceous marl-limestone sequences inAlabama (King, 1990; Bottjer et al., 1986; R.O.C.C. Group,1986). The origin of sea-level change here seems plausiblebecause the sequences are the product of shelf sedimentation.The other reason for questioning the conclusion of the

1 von Rad, U., Haq, B. U., et al., 1992. Proc. ODP, Sci. Results, 122:College Station, TX (Ocean Drilling Program).

2 Department of Geology, Dalhousie University, Halifax, Nova Scotia B3H3J5, Canada.

3 Department of Earth and Environmental Sciences, Wesleyan University,Middletown, CT 06457, U.S.A.

R.O.C.C. Group (1986) on the Alabama Upper Cretaceousmarl-limestone sequence is that the sequence contains too fewbeds for a valid statistical test. Pelagic cyclic sedimentsyounger than Early Cretaceous are better suited for thepurpose of testing the theory of orbital origin of sedimentcyclicity in pre-Pleistocene time in a quantitative fashion. Inthis paper we report our study on an Upper Cretaceous cyclicsequence from the Indian Ocean (the Exmouth Plateau). Incomparison to Lower Cretaceous cyclic sequences from theAtlantic-Tethyan region (Cotillon and Rio, 1984; Barron et al.,1985; Herbert and Fischer, 1986; Ogg et al., 1987; ten Kateand Sprenger, 1989) and Upper Cretaceous cyclic sequencesfrom the South Atlantic (Herbert and D'Hondt, 1990) andNorth America (R.O.C.C. Group, 1986), Upper Cretaceouscyclic sequences from the Indian Ocean are less reported.

The Upper Cretaceous cyclic sediments of this study wererecovered from Hole 762C, which is located on the westernpart of the central Exmouth Plateau (19°53.24'S, 112°15.24'E)in a water depth of 1360 m and about 260 km from shore (Fig.1). The Cretaceous succession from Hole 762C begins withblack to very dark gray siity claystone and clayey siltstonedeposited in a shelf-margin prodelta environment during Ber-riasian to Valanginian time. Uppermost Valanginian to Bar-remian is represented by a major hiatus (Haq, von Rad,O'Connell, et al., 1990). Lower Aptian sediments are black todark gray calcareous claystones deposited in a hemipelagicepicontinental shelf environment. The rest of the Cretaceoussediments are mostly pelagic nannofossil chalk and clayeynannofossil chalk with minor claystone. The cyclicity in theform of color band alternation mainly on a decimeter scalebegins in Upper Cretaceous sediments (younger than Cenom-anian), but is best developed in Cores 122-762C-47X through

259

Z. HUANG, R. BOYD, S. O'CONNELL

90 100

Figure 1. Location of Site 762.

110" 120

122-762C-56X, 592.5-687.5 meters below seafloor (mbsf).The sediments of this interval (nannofossil chalk) indicate apelagic environment. The burial history analysis by Gradsteinand von Rad (pers. comm., 1990) estimates that the paleowa-ter depth in Late Cretaceous time was at least 550 m. Thisprecludes the possibility that short-term, low-amplitude fluc-tuations in sea level caused the cyclicity. There is also enoughevidence against a diagenetic origin for this sequence. Withhigh core recovery (78% for Cores 122-762C-43X through-76X), this interval of cyclic sediments provides us with agood opportunity to test the orbital origin of sediment cyclic-ity in pre-Pleistocene time.

Milankovitch cycles are quasi-periodic variations in the Earth'sorbital parameters of precession, obliquity, and eccentricity. Obliq-uity is the tilt of the polar axis relative to the plane of the orbit,varying from 22° to 24.5° with a period 41 k.y. The precession is thechange in the seasonal position of the Earth on its way round theellipse, varying with periods of 19 k.y. and 23 k.y. Eccentricitydescribes the irregular variation in the ellipticity of the orbit rangingfrom 0.0005 to 0.0607, with periodicities of 95 k.y, 123 k.y., and 413k.y. (Berger, 1977). These parameters affect the seasonal andannual latitudinal distribution of solar insolation and long-termglobal and hemispheric climates (Imbrie and Imbrie, 1980). Obliq-uity increases its importance with higher latitudes (Berger andPestiaux, 1984). Precession affected the contrast between summerand winter seasons in each hemisphere. Eccentricity primarilymodulates the importance of precession. Because of the long-termchange in Earth-Moon distance during geological time, the periodsof these orbital elements should be shorter in the geological pastthan present. Recently Berger and Loutre (1989) calculated thelong-term variations of the main orbital parameters during the past500 m.y., and demonstrated that the periods of these parameterswere shorter in the geological past. According to their results,which may be inaccurate for various reasons, the differencein the periods between the present and the Cretaceous is notvery significant. For example, the present 41-k.y. obliquitycycle was about 39.5 k.y. in the Late Cretaceous (around 74Ma). The difference for obliquity periodicity between theCretaceous and the present is 2,000 yr at most. The differ-ence for precession periodicity is smaller. The precessionperiodicities were 22.5 k.y. and 18.6 k.y. in the LateCretaceous (around 74 Ma). These small differences will notpose a difficulty for testing the Milankovitch theory in UpperCretaceous strata.

In this study we first examined the features of the cyclicsediments. Sedimentological study indicates that the coloralternation is a depositional feature instead of a diageneticresult. Then time series were built on the basis of coreinspection. The time series were analyzed with the Walshspectral analysis method (Beauchamp, 1975) to reveal theperiodicity in the sediments, to provide insight into theprocess of cyclic sedimentation, and to search for the ultimateorigin of the cyclicity.

FEATURES OF CYCLIC SEDIMENTSCyclicity in the form of color variation is observed in the

Upper Cretaceous pelagic sediments of Cores 122-762C-43Xthrough -76X, 554.5-819.5 mbsf. The cyclicity is best devel-oped in Cores 122-762C-47X through -56X (592.5-687.5 mbsf,upper Campanian to lower Maestrichtian). Figure 2, a photo-graph of Core 122-762C-49X, is an example of upper Cam-panian to lower Maestrichtian cyclic sediments. The sedi-ments have been classified on the basis of smear slide exam-ination and calcium carbonate abundances using the OceanDrilling Program (ODP) sediment classification scheme (Haq,von Rad, O'Connell, et al., 1990).

The Upper Cretaceous age in this hole was determinedwith nannofossils, which are more abundant and better pre-served than other minor fossil groups. The nannofossil eventsobserved by Bralower and Siesser (this volume), together withlithological records, indicate a stratigraphically complete Up-per Cretaceous in Hole 762C. The bio stratigraphy of theinterval of interest is shown in Figure 3, following the work byBralower and Siesser (this volume). The Santonian-Cam-panian boundary is placed at the first occurrence of Broinso-nia parca (J14.14 mbsf). The lower Campanian-upper Cam-panian boundary is placed at the first occurrence of Quadrumgothicum (687.80 mbsf). The Campanian-Maestrichtianboundary is placed at the last occurrence of Eiffellithuseximius (611.50 mbsf). The lower Maestrichtian-upper Mae-strichtian boundary is placed at the first occurrence ofLithraphidites quadratus (578.30 mbsf). There is a hiatus atthe Cretaceous-Tertiary boundary. The top of Upper Creta-ceous strata is placed at the top of Core 122-762C-43X (554.80mbsf) where the last Cretaceous species occurs.

In Figure 3, the numeric ages of the stage boundary inmillions of years are from the time scale by Kent andGradstein (1985). For the interval of well-developed cyclicity,we used the lower Maestrichtian-upper Maestrichtian bound-ary (71 Ma) and the lower Campanian-upper Campanianboundary (77 Ma) as the age control to estimate the averagesedimentation rate. The sediment thickness between thesetwo boundaries is 109.50 m and the duration is 6 m.y. Theestimated average sedimentation rate for this interval is 1.82cm/k.y. This average sedimentation rate may contain someerrors from both the placement of the age boundaries and fromthe time scale and is not necessarily constant.

The completeness in both stratigraphy and recovery for theinterval of well-developed color alternations enables a de-tailed study of the nature of cyclic sediments. Because of highgas pressure in the drilling area, the cores were moderatelydisturbed and slightly expanded, resulting in recovery higherthan 100%, but this disturbance is not serious enough toprevent recognizing color variation couplets in most of thecores. Our statistical analysis on the cycles, however, isaffected by whole-round sampling for organic geochemistry(OG) analysis and interstitial water (IW) analysis. We areunable to tell if a color variation boundary occurs in awhole-round sample (25 cm long for an OG sample and 10 cmfor an IW sample) or, if so, where the boundary is.

260

UPPER CRETACEOUS CYCLIC SEDIMENTS

cc

Figure 2. Typical core of Upper Cretaceous pelagic sediments fromHole 762C displaying alternation of light nannofossil chalk and dark,clayey nannofossil chalk to nannofossil chalk. The boundaries be-tween the light beds and the dark beds are gradual. Core 122-762C-49X.

General Character of the Cycles

The 95-m-thick interval of well-developed cyclic sedimentsdisplays alternations of light-colored and dark-colored beds(Fig. 2). The color tends to become darker downhole in thisinterval. Above 603.5 mbsf the light beds are white (2.5Y 8/0)to very light greenish gray (10Y 8/1) nannofossil chalk whilethe dark beds are light greenish gray (5GY 7/1) clayey nanno-fossil chalk. Below 603.5 mbsf the light beds are white (10Y8/1) to light greenish gray (10Y 7/1) nannofossil chalk and thedark beds are light reddish brown (5Y 6/3) to reddish brown(5Y 5/3) clayey nannofossil chalk with some nannofossilchalk.

The thickness of color alternations ranges from 10 to 120cm but most of the alternations are on a decimeter scale.Boundaries between the light and dark beds are generallygradual with transitional zones varying from 2 to 12 cm. It isfeasible to judge the regularity of the cycles visually.

The thickness of the light and dark beds was measured inCores 122-762C-47X through -56X, but beds that are notcomplete (e.g., whole-round sample intervals) are not in-cluded. Light beds (n = 195) vary in thickness from 4.0 to 80cm with an average of 18 cm and a standard deviation of 14cm. Dark beds (n = 194) vary from 3.5 to 89 cm in thicknesswith an average of 22 cm and a standard deviation of 18 cm,indicating that the dark beds are thicker and more variablethan light beds. In addition, only dark beds show lamination,and where present it is very faint. The lamination is not causedby current activity as there are no ripple cross-bedding or cutand fill features.

BioturbationBoth the light and dark beds that constitute the cycles are

moderately to strongly bioturbated with many recognizabletrace fossils such as Planolites, Zoophycos, Chondrites, Te-ichichnus, and Heliminthoida. Teichichnus and Heliminthoidaare relatively rare and the latter is found only above 603.5mbsf. The burrows are predominantly horizontal. However,downward burrowing mixed lighter sediments locally into thedarker underlying beds. Section 122-762C-56X-2 contains atypical example (Fig. 4). These burrows indicate that the colorvariation is primarily a depositional product instead of adiagenetic outcome. Inoceramus fragments are more commonin light beds.

Calcium Carbonate and Total Organic Carbon ContentCalcium carbonate measurements (see Appendix for the

details of the method) were carried out on 186 samples fromCores 122-762C-47X through -56X (Table 1). The 92 calciumcarbonate measurements from light beds vary from 74.71% to93.72% with an average of 86.99% and a standard deviation of3.17%. The 94 measurements from dark beds vary from57.22% to 91.5% with an average of 74.46% and a standarddeviation of 7.03%. Therefore, on average, the calcium car-bonate content is about 13% lower and more variable in thedark beds than in the light beds. Detailed measurements inCores 122-762C-48X, -49X, -51X, and -52X show that cal-cium carbonate content of light beds is always higher than thatof the nearby dark beds (Fig. 5A), and the calcium carbonatecontent is more variable in the dark beds than in the light beds.

Total organic carbon (TOC) content is low in both light anddark beds. There is no significant difference in TOC betweendark and light beds. The TOC values vary from 0.0% to 0.05%in eight dark beds and from 0.0 to 0.02% in four light beds(Table 2).

Other Compositional DifferencesWhole-rock X-ray diffraction (XRD) analysis (refer to

Wilkens, this volume, for details of the method) was carriedout, using the same samples for which we have coulometri-cally determined calcium carbonate content, in Cores 122-762C-47X through -53X (fable 1). Only quartz and calciteconsistently show reliable peaks on the X-ray spectra. Peaksof clay minerals and other accessory minerals such as feld-spar, zeolite, and pyrite were identified when possible. Con-tent of total clay minerals was calculated, and with rareexception, indicated that more clay minerals were present indark beds than in the adjacent light beds (Fig. 5B). The sametendency is found in the variation of quartz content (Fig. 5C).We also observed that (1) more calcium carbonate occurs in

261


550—

600 —

c/)-O

E

Q.α>Q

650

700—

750 —

Nannofossil Datum

Last Cretaceousspecies

M.murus

L.quadratus

R.levis

B.parca & E.eximius

Q. trifidum

Q.gothicum

I Lgrillii

P.regularis & B. parca

Lithological Descreption

White to greenish gray nannofossil chalk.

Nannofossil chalk and clayey nannofossilchalk. Color variation is white to very lightgreenish gray with light greenish gray.

Nannofossil chalk and clayey nannofossilchalk with some claystone. Cyclic colorvariation is white to light greenish graywith reddish brown to light reddishbrown.

White to very light greenish gray nanno-fossil with greenish gray nannofossilclaystone.

Age

Tertiary66.4 Ma

.71 MaMaestrichtian

74.5 Ma

Campanian

.77 Ma

— 84 Ma -Santonian

Figure 3. Lithologic and stratigraphic summary of the Campanian to Maestrichtian cyclic sediments in Hole 762C, showing corenumber, core recovery (in black), nannofossil datum, lithology, and age (Haq, von Rad, O'Connell, et al., 1990). Shaded areaindicates the interval of well-developed cyclic color variations. The numeric ages at the stage boundaries are from the time scaleby Kent and Gradstein (1985).

light beds than in dark beds, as is supported by the coulomet-ric measurements, (2) more kaolinite and illite peaks thansmectite peaks appear in dark beds, but more smectite peaksthan kaolinite and illite peaks appear in light beds, and (3)more zeolite and feldspar peaks are detected in dark beds.

Compositional differences between the light and dark bedswere also detected in the shipboard smear slide examinations(Table 3) of Cores 122-762C-47X through -56X. The differ-ences presented by the smear slide record can be outlined asfollows: (1) quartz, clay, and opaques are more common indark beds than in light beds, (2) zeolites and dolomite are morecommon in dark beds than in light beds, (3) sponge spiculesand ostracodes are more abundant in dark beds than in lightbeds, (4) biogenic carbon grains and bioclasts are morecommon in dark beds than in light beds, and (5) nannofossils andforaminifers are more abundant in light beds than in dark beds.

Until now, the source of micropaleontological informationfor the Upper Cretaceous cyclic sequences from the ExmouthPlateau has been limited to smear slide examination. Morework is needed to have a systematic micropaleontologicalcomparison between the dark and light beds. The smear slideobservation that there are more nannofossils in the light bedsis in agreement with what has been observed in similar cyclicsequences (i.e., Atlantic Lower Cretaceous cyclic sequences,Ogg et al., 1987; and Lower Cretaceous cyclic sequences fromVocontian Basin, Darmedru et al., 1982). This observationcan be explained by higher bioproductivity in the surfacewater during the phase in which the light beds were deposited.However, the smear slide observation that there are moreforaminifers in the light beds is not in accordance with othersimilar cyclic sequences.

Diagenetic Overprint

The sediments recovered in the studied interval are relativelysoft. Diagenetic alteration appears as small, thin seams cutting tracefossils. These seams occur irregularly and more often in dark bedsthan in light beds. These seams cut trace fossils and result fromearly diagenetic compaction, local calcium carbonate dissolution,and reprecipitation during burial (Fig. 6). Scanning electron micro-scope (SEM) examination on seven samples from three and a halfcycles in Sections 122-762C-49X-2 through -49X-3 exposes etchednannofossils and some secondary deposited calcium carbonate insamples from dark beds and fewer dissolved nannofossils also withsome secondary deposited calcium carbonate in samples from lightbeds (Fig. 7). It is observed that the light beds are more porous andtherefore they cannot have a large volume of cement. The effects ofthe diagenetic process are not profound enough to alter the overallcyclic color alternation produced by depositional processes, asredeposition of calcium carbonate is not restricted to the light beds.Scattered pyrite crystals and a burrow filled by pyrite crystals arefound only above 603.5 mbsf.

Physical properties measurements (Table 4) show that lightbeds have higher water content, lower bulk density, slightlylower grain density, and higher porosity than dark beds. Thistable shows that for both the light and dark beds porositytends to decrease with increasing depth. As the cyclic se-quences in the Indian Ocean are not fully indurated, theinfluence from cementation is small and the dissimilarity inphysical properties (e.g., porosity) may be more directlyrelated to and reflect the differences in compaction betweenthe light and dark beds, caused by the original differences incomposition. It is still feasible, however, to quantify the

262


cm

50 1—

Figure 4. Typical bioturbation that introduces the light sediments fromthe upper light bed to the lower dark bed. The lower, sharply inclinedboundary represents a microfault plane. Sample 122-762C-56X-3,26-50 cm.

difference in compaction. If comparing the mean values ofwater content and porosity between the light and dark beds,the difference is small, both less than 4.5%. Supposing theoriginal porosity of the light and dark sediments is nearly thesame, according to a mean difference of 4.03% in presentporosity between these two sediment, the difference in com-paction should also be small. If the original porosity of thelight bed is higher than the dark beds, the actual difference incompaction may be smaller; otherwise there might be a greaterdifference in compaction. In any case, we expect that theinfluence of differential compaction, if there is any, on statisticalanalysis of Indian Ocean cyclic sediments would be muchsmaller than that on older and more indurated cyclic sequences.

ImplicationsA summary of the differences between the light and dark beds is

given in Table 5. The geological implications from sedimentologicalexamination of the cyclic interval and a comparison of the light anddark beds can be summarized as follows:

The cyclic color alternation in the sediments is a deposi-tional feature and not a diagenetic product. A primary originfor color cyclicity is supported by (1) burrows that introducethe light-colored sediment of an overlying bed into a dark-colored underlying bed and (2) consistently higher calciumcarbonate content in light beds.

Strong to moderate bioturbation throughout the cyclic intervaland the generally low TOC value in both light and dark bedsindicate that the bottom water had sufficient oxygen supply forbenthic organisms to survive. The occasionally faint lamination indark beds may imply bottom water with slightly less oxygen supplyand hence the reduction of benthic organisms. The changes inoxygen supply to the bottom water are subtle and are not crucial tothe formation of the cycles. Compositional differences between thelight beds and dark beds imply that the change lies mainly insediment supply. When the light beds were deposited, biogenicmaterial was the major component, with less terrestrial input.When the dark beds were deposited, terrestrial input was in-creased, with some decrease in biogenic supply. The smallerstandard deviation of bed thickness and less-variable calciumcarbonate content of light beds may suggest that the amount ofbiogenic supply is less variable than the amount of terrestrialsupply. More variability in terrestrial supply indicates that land-based changes may have played a more important role in cyclicsedimentation.

Diagenetic processes may have enhanced the color cyclic-ity, producing small seams that locally cut trace fossils andlocal calcium carbonate dissolution/reprecipitation.

A background change in the cyclic sequence is indicated bythe overall color change from a brownish-hued interval below603.5 mbsf to a greenish-hued interval with pyrite crystalsabove that depth. This background change is also reflected byexclusive occurrence of the trace fossil Heliminthoida abovethis depth. The background change is believed to have beencaused by a relative sea-level change (Haq, von Rad, O'Con-nell, et al., 1990) from highstand to lowstand. Whatever thecause, the background change does not seem to have affectedthe formation of the cycles and it is not a controlling factor.

QUANTITATIVE SEARCH FOR PERIODICITY OFCYCLIC SEDIMENTS

Method and Time Series GenerationAs it is difficult to estimate the regularity of the color cycles

visually, a statistical method should be utilized for this purpose.Spectral analysis is frequently employed in the studies of Qua-ternary cyclic sediments. However, the conventional Fourierspectral analysis method is best suitable for smooth time series

263


Table 1. Calcium carbonate, clay, and quartz contents in Cores122-762C-47X through 122-762C-56X.

Table 1 (continued).

Core, section,interval (cm)

122-762C-

47X-1, 9-1147X-1, 95-97

d47X-2, 81-8347X-2, 128-130

d47X-2, 140-14247X-3, 15-1747X-3, 40-4247X-4, 40-42

d47X-4, 71-73d47X-4, 106-10847X-5, 23-2547X-5, 59-6047X-6, 23-25

d47X-6, 66-6847X-6, 85-8648X-1, 4-548X-1,40-42

d48X-l, 90-9248X-1, 100-10148X-1, 138-13948X-2, 19-2048X-2, 35-3648X-2, 45-4748X-2, 61-6148X-2, 71-7248X-2, 80-8148X-2, 93-9448X-2, 105-10648X-2, 114-11548X-2, 132-13348X-3, 20-2148X-3, 57-5848X-3, 74-75

d48X-3, 94-%48X-3, 129-13048X-4, 10-1148X-4, 17-1848X-4, 30-3148X-4, 39-4048X-4, 54-5548X-4, 73-7548X-4, 85-8648X-4, 103-10548X-4, 120-12148X-4, 138-140

d48X-5, 55-5748X-5, 78-8048X-5, 111-11348X-5, 122-12448X-5, 145-146

d48X-6, 68-7048X-6, 87-8948X-6, 97-9948X-6, 104-10648X-6, 117-11848X-6, 147-14949X-1.7-9

d49X-l, 9-1149X-1, 23-2549X-1, 61-6249X-1, 95-9649X-1,116-11749X-2, 20-2149X-2, 82-8349X-2, 100-10149X-2, 112-11349X-3, 20-2149X-3, 38-3949X-3, 55-5649X-3, 65-67

d49X-3, 78-8049X-3, 89-9049X-3, 105-10649X-3, 120-12149X-3, 130-131

Depth(mbsf)

592.59593.35594.81595.28595.40595.65595.90597.40597.71598.06598.73599.09600.23600.66600.83602.04602.40602.90603.00603.38603.69603.85603.95604.11604.21604.30604.43604.55604.64604.82605.20605.57605.74605.94606.29606.60606.67606.80606.89607.04607.23607.35607.53607.70607.88608.55608.78609.11609.22609.45610.18610.37610.47610.54610.67610.97611.57611.59611.75612.11612.45612.66613.20613.82614.00614.12614.70614.88615.05615.15615.28615.39615.55615.70615.80

CaCO3

(%)a

82.8191.6982.7488.8271.9978.8086.9086.7984.3387.4173.4388.6566.3682.7488.9281.4885.7888.4180.2688.8385.0280.8580.8367.7785.5265.2684.1079.1185.7668.2885.7178.9288.4884.1687.1279.5085.0361.3683.8275.1185.5073.4784.0776.9186.6876.3391.3062.6185.6784.1789.4161.8879.8885.3289.1579.0574.7190.4163.6191.9285.8089.5675.1484.3070.1984.4073.9382.4069.9675.8869.8384.5476.4489.3877.15

CaCO3

(%)b

87.390.3

89.4

82.985.391.4

77.495.2

95.083.193.0

83.588.086.481.983.372.487.775.586.385.985.470.286.681.587.9

90.287.684.069.586.080.988.276.685.376.493.277.691.668.287.483.6

65.975.987.090.983.588.1

73.389.386.491.678.981.174.282.675.183.772.178.3

87.181.496.380.7

Clay(%)b

9.38.5

8.1

10.710.86.2

13.82.5

2.9114.2

11.59.79.9

11.511.919.08.6

14.78.66.9

10.916.79.8

12.49.4

6.77.29.6

16.89.3

11.67.9

15.710.714.31.4

15.06.7

17.28.5

12.0

25.012.38.25.48.58.7

13.99.28.65.6

11.512.212.310.414.010.616.413.6

7.410.10.0

10.2

Quartzpeaksc

3.31.2

2.5

6.53.92.4

8.82.3

2.15.92.8

5.02.43.76.64.88.63.79.85.17.23.7

13.13.66.12.7

3.15.26.4

13.84.77.43.97.74.09.35.47.01.7

14.64.14.5

9.011.84.83.68.03.2

12.81.55.02.89.56.7

13.57.0

10,95.7

11.58.1

5.58.53.79.1

Mineral0

K, I, SK, S, I

S

K, I, SKI

K, I, SK,S

S

sK, S

K, SKIKK, SK, SK, I, SK, S, FK, S, IK, SK, I, FK, I, SK,SKS

K, SK, SK, SK, I, S, FK . I . SK, I, S, FSK, IS, K, Z?K, I, SK, I, SK, S, Z, F

K, I

K

K, SS, 1K

K, 1

K, I, SS, K, Z?K

K, I, SK,IK, I, SSK, I, SIS, I, KK, S

K, I, SK, I, S

K, I, S

Color

dark (g)light (g)dark (g)light (g)dark (g)dark (g)light (g)light (g)dark (g)light (g)dark (g)light (g)dark (g)dark (g)light (g)dark (g)light (g)light (g)dark (g)light (g)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)light (b)light (b)dark (b)light (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)


d49X-4, 19-2049X-4, 93-9449X-4, 105-10649X-4, 117-11849X-4, 134-13549X-4, 20-2149X-5, 35-3649X-5, 50-5149X-5, 80-8149X-5, 95-9649X-5, 124-125

d49X-5, 132-13449X-5, 135-13650X-1, 80-81

d50X-2, 48-5050X-2, 95-9650X-2, 114-11550X-3, 45-4650X-3, 91-92

d50X-4, 94-%50X-4, 116-11750X-5, 20-2150X-5, 55-5650X-6, 20-21

d50X-6, 116-1185OX-7, 7-850X-7, 20-2151X-1, 38-4051X-1, 114-11651X-1, 140-14151X-2, 5-6

d51X-2, 66-6851X-2, 68-6951X-2, 81-8251X-2, 120-12151X-3, 50-5151X-3, 94-9551X-3, 109-11051X-3, 120-12151X-3, 139-14051X-4, 5-651X-4, 25-2651X-4, 45-4651X-4, 60-6151X-4, 78-8051X-4, 90-9151X-4, 107-10851X-4, 125-127

d51X-4, 142-14451X-4, 147-14851X-5, 7-851X-5, 17-1851X-5, 70-7251X-5, 120-12251X-5, 142-14452X-1,49-5152X-1, 66-6952X-1, 90-9152X-1, 100-10252X-1,115-11752X-1, 135-13752X-2, 49-50

d52X-2, 50-5252X-2, 77-7952X-2, 113-11552X-2, 121-12352X-2, 139-14152X-3, 29-3052X-3, 44-4652X-3, 77-7852X-3, 103-10552X-3, 124-12652X-3, 140-14152X-4, 14-1652X-4, 71-73

d52X-4, 76-7852X-4, 111-11352X-4, 139-140

Depth(mbsf)

616.19616.93617.05617.17617.34617.70617.85618.00618.30618.45618.74618.82618.85621.80622.98623.45623.64624.45624.91624.44626.66627.20627.55628.70629.66630.07630.20630.88631.64631.90632.05632.66632.68632.81633.20634.00634.44634.59634.70634.89635.05635.25635.45635.60635.78635.90636.07636.25636.42636.47636.57636.67637.20637.70637.92640.49640.66640.90641.00641.15641.35641.99642.00642.27642.63642.71642.89643.29643.44643.77644.03644.24644.40644.64645.21645.26645.61645.89

CaCO3

(%)a

92.2484.5290.2765.4189.8987.8588.9472.3986.1078.5087.6279.2481.8577.4481.3387.4275.1387.0568.6887.4177.8674.3290.6690.6385.1688.6073.4874.5489.6084.2388.2273.0872.8478.8075.%75.7486.2369.5890.8475.5393.7274.7287.3287.1587.4780.1886.3581.7085.4987.5787.9184.4687.1375.72

73.7484.8480.7188.1780.0693.6885.3285.6678.5685.8482.1887.2277.7086.7788.5984.0387.0768.9786.7681.4481.9188.4783.82

CaCO3

(%)b

%.682.791.976.591.487.295.776.085.081.388.7

74.879.9

89.382.888.573.4

80.479.695.989.2

91.179.879.789.985.688.1

74.581.580.978.583.773.990.276.689.377.395.879.995.981.286.586.083.393.286.487.194.671.7

74.892.480.886.884.487.782.7

82.486.183.786.477.083.685.083.387.270.987.980.3

87.585.7

Clay(%)b

1.213.65.5

11.96.08.50.0

14.511.07.46.9

18.29.6

6.37.27.3

13.5

9.711.11.37.6

4.711.111.06.68.07.9

16.210.29.8

11.710.714.57.0

12.17.8

12.60.0

15.60.0

11.07.77.4

11.511.09.48.20.0

18.1

13.60.09.86.46.65.79.8

8.38.07.58.9

13.110.49.79.07.4

13.45.9

10.5

7.88.8

Quartzpeaksc

2.23.72.5

11.62.74.34.39.44.0

11.24.4

7.010.5

4.49.94.2

13.1

9.99.32.83.2

4.29.19.33.56.43.9

9.38.39.39.95.6

11.62.8

11.32.8

10.24.24.74.17.85.86.65.25.84.24.75.4

10.1

11.77.69.46.89.06.67.5

9.45.98.84.79.96.05.47.65.4

15.76.19.1

6.45.5

Mineral0

ZI, SK, 1K, II, SK

K,I , FK, SK, IK,S

K, IK, I, S

K, I, SK, I, S

I, SS, K, I, F

K, I , SK, I

S, IS, K, Z?

I, SK, SK, I, SK, S, Z?IK, I, SSK, I, S, Z?s, z?K, S

K, I, SK,SKS, I, K, Z

K, I, S

K, I

K, IK,S

S, KK, SK, IK, SKK, I, SI, SK, I, Z?I, SK, 1SK, I, SS

sK, I

Color

light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)dark (b)dark (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark(b)dark (b)light (b)light (b)light (b)light (b)dark(b)dark (b)light (b)dark (b)light (b)dark (b)dark (b)dark (b)dark (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)dark(b)dark(b)light (b)light (b)dark (b)light (b)dark (b)dark (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)light (b)light (b)dark (b)light (b)dark (b)light (b)dark(b)light (b)light (b)dark(b)light (b)dark (b)light (b)dark (b)dark (b)light (b)dark (b)

264


Table 1 (continued).


52X-5, 70-7252X-5, 132-13452X-6, 14-1652X-6, 29-3052X-6, 48-5052X-6, 69-7152X-6, 104-10652X-7, 29-3053X-1, 109-110

d53X-2, 20-2253X-2, 90-9253X-2, 134-13553X-3, 10-1253X-3, 33-3553X-3, 99-10053X-4, 89-90

d53X-4, 94-9753X-4, 108-11053X-5, 110-111

d53X-6, 8-1053X-6, 10-1153X-6, 30-3253X-6, 74-7653X-6, 124-126

d54X-2, 31-33d54X-4, 49-52d54X-4, 86-88d54X-6, 104-106d55X-2, 67-69d55X-4, 7-9d56X-l, 63-65d56X-3, 90-92d56X-5, 56-58

Depth(mbsf)

646.70647.32647.64647.79647.98648.19648.54649.29650.59651.20651.90652.34652.60652.83653.49654.89654.94655.08656.60657.08657.10657.30657.74658.24660.81663.99664.36667.54670.67673.07678.63681.90684.56

CaCO3

(%)a

81.4386.9580.4888.3977.6488.5662.5679.7857.2284.4188.5268.3185.8764.7086.2678.1681.4991.4469.5388.6687.5778.%88.1579.3587.2480.8376.7484.6688.4985.3385.3384.4185.41

CaCO3

(%)b

86.087.482.184.376.284.368.479.364.5

91.571.690.467.584.977.5

86.273.9

86.479.986.679.5

Clay(%)b

5.38.19.3

10.812.99.6

13.013.015.7

2.214.31.9

19.17.1

10.9

7.49.6

8.510.48.7

12.0

Quartzpeaks0

8.74.58.64.9

10.86.2

18.67.7

19.7

6.314.17.7

13.38.1

11.5

6.416.6

5.29.74.78.6

Mineral0

K, IS, KK, IKK, I, SS, IK, I, SK, I, SK, I, S

KK, I, SKK, IK,I, SK

K, SK, I, S

K, SK, SK

Color

dark (b)light (b)dark (b)light (b)dark (b)light (b)dark(b)light (b)dark (b)dark (b)light (b)dark (b)light (b)dark (b)light (b)light (b)light (b)light (b)dark (b)light (b)light (b)dark (b)light (b)dark (b)light (b)dark (b)dark (b)light (b)light (b)dark (b)dark (b)dark (b)light (b)

Note: Color abbreviations are (g) = greenish hue with scattered pyrite crystalsand (b) = brownish hue.

a From coulometric method.b From XRD analysis.c Identified on XRD spectra. K = kaolinite, I = illite, S = smectite, Z = zeolite,

and f = feldspar.d Shipboard measurements (Haq, von Rad, O'Connell, et al., 1990).

such as geophysical logging data or curves built by a largenumber of closely spaced laboratory-derived data points, such asoxygen/carbon isotopes, calcium carbonate content, and othergeochemical measurements. The geophysical logging tool ofhigher vertical resolution used by ODP is the natural gamma-rayintensity log (0.3-m vertical resolution) (Schlumberger, 1987). Itwould be unable to resolve color alternations thinner than 30 cm.Although we take one sample from every bed for calciumcarbonate measurements in some of the cores, the number ofdata points is still too sparse to build a smooth time series. In thissituation it is better to build step-like (discrete) time series on thebasis of core inspection and measurement of bed thickness, andapply the Walsh spectral analysis method to the discrete timeseries (Schwarzacher, 1987). As with Fourier analysis, any timeseries can be resolved into Walsh functions with different se-quences, zero crossings of the time base, an equivalence offrequency in Fourier spectral analysis (Weedon, 1986). Walshfunctions are discontinuous waves that take only two values, +1or - 1 . Therefore, Walsh spectral analysis is well suited foranalyzing step-like time series (for more details on the method,refer to Beauchamp, 1975). Walsh spectral analysis was appliedto geoscience for studying the cyclicity of magnetic reversals(Negi and Tiwari, 1983). Later Weedon (1986) and Schwarzacher(1987) introduced this technique to study cyclic sediments inEurope. One advantage of this method is the speed in transformcomputation, as only addition and subtraction is required.

The application of spectral analysis to geological problemsinvolves four aspects: time series generation, transform oper-ation, spectrum generation, and interpretation. In our timeseries generation for Site 762 sediments, we code the lightbeds with +1 and the dark beds with - 1 . Figure 8 is anexample of the step-like time series we generated. Time seriesgeneration in this way is efficient. However, it is no more thanan approximation of the cycles, because in reality the bound-ary between light and dark beds is gradual with a transitionalzone varying from 2 to 12 cm, a graduation which may also bereflected in calcium carbonate content. For the purposes ofthis study we assume that a single bed was produced underrelatively constant depositional conditions. To generate thestep-like time series, we placed the color boundary in themiddle of the transition zone. Therefore, this approximationcan represent the cyclic changes in the sediments. It is thebest choice for an interval without a sufficient number oflaboratory measurements but with a good lithological record.An advantage of building time series in this way is that it willnot introduce a superimposed trend, such as a gradual de-crease or increase of a certain parameter (e.g., calciumcarbonate content) over the interval.

High recovery in all of the cores in the interval, exceptCore 122-762C-55X (5 m long), aids our time series genera-tion. Other factors, however, prevent time series analysis ofevery core. Microfault planes were observed in Sections122-762C-47X-2 and -56X-3 (Fig. 4). The microfault planes,which offset the cyclic succession, can distort the time seriesand this distortion is difficult to correct. Whole-round samplesare taken in Sections 122-762C-47X-5 (115-140 cm, 140-150cm) and 122-762C-53X-4 (115-140 cm, 140-150 cm). Core122-762C-54X is not only more expanded (>100% recovery),but also more disturbed than the other cores. For thesereasons, Sections 122-762C-47X-1 through -47X-5 and Cores122-762C-53X through -56X were rejected for time seriesgeneration. The three time series we have built in the cyclicinterval are:

1. Sections 122-762C-47X-6 through -49X-CC (600.0-621.0 mbsf);

2. Sections 122-762C-50X-1 through -50X-6 (621.0-629.74mbsf);

3. Sections 122-762C-51X-1 through -52X-CC (630.5-649.5mbsf).

In Section 122-762C-50X-2 there is a 10-cm-long IWwhole-round sample, but we interpolated the color of thewhole-round sample based on the surrounding color. Twocores were merged in time series 1 and 3 to increase the lengthof time base for analysis. An 18-m time series is more than 10times longer than the longest cycle length observed in thecores. Longer time series would improve the accuracy of theposition of frequency peaks. With longer time series it is alsopossible to examine the stability of the frequency peaks bydividing the time series into two subsections and applyingspectral analysis on each of them. However, the cost of usinglonger time series is the introduction of an additional compli-cation caused by unknown changes in sedimentation rate.

The Power SpectraIn spectral analysis computation, the step-like time series

is first sampled with a user-defined uniform interval. Thenumber of data points for a fast Walsh transform must beequal to a power of 2. In this study the sampling interval variesslightly with the length of every time series, but is less than 1cm in order to guarantee sampling the thinnest bed observedin the cores. The program subsequently performed the Walsh

265


A Calcium carbonate (%) B Clay minerals (%) C Quartz (%)

60 70 80 90 100 0 10 20 30 0 10 20

602 -i

604 -

606 -CO-Q

E

Q_0Q 608

610 -

612 J Key: • dark bed D light bed

Figure 5. Compositional difiFerences in a succession of cyclic color alternations, Core 122-762C-48X. A. Content ofcalcium carbonate (%) measured with coulometric method. The content of calcium carbonate in a light bed is alwayshigher than that of the adjacent dark beds. B. Content of clay minerals (%) from XRD analysis. There are more clayminerals in dark beds than in light beds, with rare exception. C. Content of quartz (%) from XRD analysis. Quartz ismore abundant in dark beds than in light beds, with rare exception.

transform and power value calculations. The power valueswere normalized for an application of Fisher's test for whitenoise (Nowroozi, 1967), and the Walsh spectrum wassmoothed using a 3-point Hanning window. The Walsh powerspectrum illustrates the relative importance of the frequencypeaks. Confidence levels (90%, 95%, 98%, and 99%) werecalculated. In this study the 90% confidence level was used toinspect the validity of the frequency peaks. A component ofparticular frequency, in terms of zero crossings per meter(cycles/m), with greater power value indicates a more impor-tant role if it is used to recreate the time series. The frequencycan also be converted to wavelength in length (cm), thereciprocal of the frequency, or wavelength in time (k.y.) withestimated average sedimentation rate.

Walsh spectral analysis was carried out on the three timeseries 1,2, and 3. The stability of the frequency peaks in aspectrum was tested when analyzing time series 1 and 3 bydividing a time series into two subsections and calculatingsubsection spectra from them.

The power spectra calculated from time series 1,2, and 3are shown in Figures 9, 10, and 11, respectively. On each

spectrum the lengths of the time series are indicated as timebase, and the bandwidth by a short horizontal bar. A narrowerbandwidth ensures higher resolution. The 90% confidencelevel is drawn as a horizontal line on both whole-section andsubsection spectra. To help interpretation, the ratios of thewavelength of the significant frequency peaks are also indi-cated. As the subsection is only one-half the length of thewhole section, the peaks in the very low-frequency region ofthe whole-section spectra are difficult to compare with thosein subsection spectra in Figures 9 and 11. Therefore, peaks inthe low-frequency region (less than 0.5 cycles/m) are not to beinterpreted and discussed in this study. We regard a frequencypeak in a whole-section spectrum to be stable if a strongcounterpart is found in one or both of the subsection spectra.The wavelengths of the stable peaks in the higher frequencyregion are indicated in the figures. In Figures 9 and 11, thelocation of frequency peaks in the spectra from the subsec-tions deviates slightly from their corresponding stable peaks.The subsection spectra of time series 3 display more variation.The variations may be attributed to change in sedimentationrate and influence from differential compaction. There is no

266


Table 2. Total organic carbon in the light anddark beds in Cores 122-762C-47X through 122-762C-56X.


Light beds122-762C-

47X-4, 106-10847X-5, 140-14253X-4, 94-9656X-5, 113-115Mean

Dark beds122-762C-

47X-2, 140-14250X-2, 48-5050X-2, 140-14251X-4, 142-14452X-4, 76-7853X-5, 0-255X-4,49-5156X-3, 90-92Mean

Depth(mbsf)

598.06599.90654.90685.10

595.40622.98623.90636.42645.26655.50663.99681.90

TOC(%)

0.02 (g)0.02 (g)0.00 (b)0.02 (b)0.015

0.03 (g)0.01 (b)0.02 (b)0.04 (b)0.01 (b)0.05 (b)0.02 (b)0.00 (b)0.025

Note: This table is compiled from shipboard measure-ments (Haq, von Rad, O'Connell, et al., 1990).Color abbreviations are (g) = greenish hue withscattered pyrite crystals and (b) = brownish hue.

subsection spectra calculated for time series 2 because of itsshorter length. The location of the frequency peaks in Figure10 is comparable to the location of those appearing in Figures9A and 11 A. Thus, the frequency peaks with wavelengths 84.0cm, 38.2 cm, and 35.0 cm are interpreted to be stable. Thesampling intervals used in analyzing these three time seriesare less than 1 cm, much smaller than the thickness of thethinnest bed. The Nyquist frequency is at about a 10-k.y.period, judging from the estimated average sedimentation rate(1.82 cm/k.y.). Therefore, the frequency peaks in the spectraare not subjected to an aliasing problem (see Pisias and Mix,1988). The structure of the spectra does not suggest there is aharmonic problem. We therefore conclude that the stablefrequency peaks reveal the predominant frequencies of thecolor alternations in the cyclic sediments. In Figure 9A, thepeak with a 177.8-cm wavelength does not have a counterpartin the spectrum from the upper half of time series 1 (Fig. 9B),but has a counterpart in that from the lower half of the timeseries wavelength (Fig. 9C), which is longer than 177.8 cm(196.6 cm) for the inaccuracy in the lower frequency portion.The peak in Figure 9A with a 40.8-cm wavelength has itscounterparts in both subsection spectra (Figs. 9B and 9C), andthe peak with a wavelength of 34.3 cm has a counterpart onlyin the lower subsection spectrum (Fig. 9C). In Figure 11 A, thepeaks with 80.1-cm and 70.9-cm wavelengths have only onestrong counterpart in the lower subsection spectrum (Fig.11C). This counterpart is strong and has a shoulder because of

Table 3. Smear slide examination on the composition of the light and dark beds in Cores 122-762C-48X through 122-762C-53X, inpercentages.

Core, sectioninterval (cm)

Light beds122-762C-

47X-1, 4548X-1, 9049X-1, 8452X-3, 7552X-6, 7553X-6, 6754X-1, 8654X-5, 8455X-2, 2256X-1, 13956X-3, 110

Dark beds122-762C-

47X-3, 3647X-3, 6647X-CC, 1748X-2, 4848X-4, 5548X-6, 3948X-CC, 3849X-3, 1949X-5, 9249X-CC, 5152X-3, 11052X-6, 10553X-1, 3553X-2, 5553X-3, 12254X-3, 6755X-1, 1955X-3, 8656X-5, 50

Quartz

5.0

2.02.02.0

tr1.0tr

3.03.02.02.03.04.02.04.0

5.09.05.02.0tr

8.0

trtr

Feldspar

1.0tr

2.0tr

20.0

Clay

5.02.05.05.03.05.0

10.05.07.0

15.010.0

5.07.03.02.02.02.01.0

10.010.010.07.0

13.011.05.08.0

20.027.024.0

Volcanicglass

1.0

2.01.02.0tr

1.0

Zeolite +dolomite

1.0

1.0

tr

5.03.02.05.0

1.02.0

1.01.0

Opaque +accessoryminerals

2.0

2.0

tr

2.04.0

tr

tr5.0

Foraminifers

15.07.0

80.013.016.010.07.0

10.08.07.0

13.0

15.015.05.03.03.06.0

10.08.0

10.05.07.05.07.0

12.013.08.0

5.071.0

Nannofossils

55.080.010.075.067.082.076.081.073.075.073.0

62.060.080.085.076.073.077.059.080.065.070.063.050.065.052.069.0

8.060.0

3.0

Carbonategrains8

25.010.0

2.02.01.07.03.0

10.03.0

15.025.0

3.012.07.05.0

15.0

15.03.03.0

28.012.018.03.0

52.010.0

Othersb

2.010.0

1.0

4.0

6.0

2.0

5.0tr2.01.0

12.0

Note: Compiled from shipboard observations (Haq, von Rad, 0'Connell, et al., 1990); tr = trace.aBiogenic carbonate grains + bioclast.bSponge spicules + fish remains + ostracodes.

267

Z. HUANG, R. BOYD, S. 0'CONNELL

cm70

75

80

•

B5L_

Figure 6. Diagenetic alteration of the sediments: small, thin seamscutting trace fossils (Sample 122-762C-56X-5, 70-85 cm).

the lower resolution of this spectrum. The peak in Figure 11Awith a 30.7-cm wavelength has its counterparts in both sub-section spectra. In Figure 11B, there are some weaker peaksaround the peak with a 30.7-cm wavelength.

In the examination and interpretation of the power spectra,we shall first use the wavelength ratios of the frequency peaksto test the hypothesis that the cyclic sediments resulted fromprocesses with Milankovitch frequencies. The wavelengthratios are examined and compared with the ratios of present-day periods of the orbital elements (Table 6). The reason forcomparison with the ratio of present-day periods of the orbitalelements is that the orbital element periods in the Cretaceouswere not much shorter than present-day periods, which aremore accurately calculated by Berger (1977). The use of ratiosfor interpretation is independent of absolute time and canavoid some errors brought in by sedimentation rate estima-tion. The ratio method was used by Molinie and Ogg (pers.comm., 1990) in analyzing Upper Jurassic and Lower Creta-

ceous sedimentary cycles of the equatorial Pacific wherestratigraphic constraints are uncertain. However, there arestill some interfering factors in and limitations to this method.First, some of the ratios of orbital periodicities are very near.For example, the O/Pl ratio (2.15) is near the El/O ratio(2.31), and therefore a definite interpretation sometimes can-not be derived solely in the light of the wavelength ratio.Second, there are errors in the wavelength ratios which mayresult from the lack of precision in the position of thelow-frequency peaks. Third, possible differential compactionbetween the more calcareous portion and the marly portioncan also introduce errors to the ratio of two compactedsedimentary cycles. Furthermore, when a spectrum yieldsonly one frequency peak, though a very rare situation, theratio method cannot be applied. As the ratio comparison canbe affected by the above factors, the time duration in thou-sands of years of frequency peaks converted from the esti-mated average sedimentation rate, which may also haveerrors, should also be examined. Mutual constraint of ratiocomparison and time duration examination is expected toachieve more certain insight into the periodicities of the cyclicsequence.

In Figure 9A, the wavelength ratio of peak a to b is 4.36, near theE1/P2 ratio (4.07, see Table 6), that of peak a to c is 5.18, near theEl/Pl and E2/P2 ratios (4.99 and 5.20, see Table 6), and that ofpeaks b to c is 1.19, near the P2/P1 ratio (1.25, see Table 6). Thetime durations of peaks a, b, and c, from the estimated averagesedimentation rate (1.82 cm/ky), are 97.7 k.y., 22.4 k.y., and 18.8k.y. These time durations are very near the modern periodicity ofone of the eccentricity cycles (94.9 k.y.) and precession cycles (23.7k.y. and 19.0 k.y.). Therefore, peak a in Figure 9A (time series 1) iscomparable to the short eccentricity cycle and peaks b and c toprecession cycles. In this case the wavelength ratios fit very wellwith the time duration from the estimated sedimentation rate ininterpreting the spectrum.

In Figure 10, the wavelength ratio of peak a to b is 2.20,near the O/Pl ratio (2.15, see Table 6), and that of peak a to cis 2.40, near the El/O ratio (2.31, see Table 6). The wave-length ratio of peak b to c is 1.09, near the P2/P1 ratio (1.25,see Table 6). The time duration from the estimated averagesedimentation rate for peaks a, b, and c are 46.2 k.y., 21.0k.y., and 19.2 k.y. The time duration for peak a (46.2 k.y.) isapparently much shorter that the periodicity of eccentricity.Combining both the wavelength ratio and the time durationestimated, peak a is correlative to the obliquity cycle andpeaks b and c to precession cycles.

In Figure 11 A, the ratio of peak a to c (2.61) is between theE2/O ratio (3.01, see Table 6) and El/O ratio (2.31, see Table6) and the ratio of peak b to c (2.31) is equal to the El/O ratio(2.31, see Table 6). If interpreted just from the ratios, peaks aand b in this figure may have been related to the eccentricitycycle and peak c to the obliquity cycle. However, the timedurations from the estimated sedimentation rate for peaks a,b, and c are 44.0 k.y., 38.9 k.y., and 16.9 k.y., respectively.The former two are much shorter than the periodicities ofeccentricity and the latter is much shorter than the periodicityof obliquity. Also the positions of peaks a and b in this figureis comparable to those of peak a in Figure 10, which is relatedto the obliquity cycle, and the position of peak c in this figureis comparable to that of peaks b and c in Figure 10, which arerelated to precession cycles. Mainly based on the estimatedtime durations, we tentatively compare peaks a and b to theobliquity cycle and peak c to the precession cycle. It may havebeen small fluctuations in sedimentation rate that producedtwo peaks out of one orbital element. This illustrates thatinterpretation of the power spectra of cyclic sediments just onthe basis of wavelength ratio can sometimes be misleading.

268


5µm

5µm

Figure 7. SEM photos of dark, clayey nannofossil chalk and light nannofossil chalk(×5000). A. In the dark beds, the nannofossils are etched, with some secondarycalcium carbonate deposition. Sample 122-762C-49X-3, 20-21 cm. B. In the lightbeds, the nannofossils are less dissolved, also with some secondary calcium carbon-ate deposition. The sediments are more porous. Sample 122-762C-49X-3, 65-67 cm.

We found that the wavelength ratios fit less well with the timedurations downhole (from Figs. 9 to 11). This may be attributed tothe increasing effect of differential compaction on the sedimentcolumn with depth, and also on the wavelength ratio method. Theexamination and interpretation of the power spectra with bothwavelength ratios and time duration converted from the estimated

average sedimentation rate demonstrate that the cyclic sequencefrom the Exmouth Plateau can be best related to the cyclicvariations in orbital elements, rather than other processes. Thecolor variation in the Upper Cretaceous cyclic sediments recoveredfrom Hole 762C was most likely caused by long-term orbitalvariations. This is another example from Cretaceous strata support-

269


Table 4. Physical properties of the light and dark beds in Cores122-762C-48X through 122-762C-53X.


Light beds

122-762C-

48X-1, 90-9248X-6, 68-7049X-1,9-1149X-4, 19-2150X-4, 94-9653X-6, 8-10Mean

Dark beds

122-762C-

48X-3, 94-9648X-5, 55-5749X-3, 78-8049X-5, 132-13450X-6, 116-11851X-2, 66-6851X-4, 142-14451X-5, 142-14452X-4, 76-7853X-2, 20-22Mean

Depth(mbsf)

602.90610.18611.59616.19626.44657.08

605.94608.55615.28618.82629.66632.66636.42637.92645.26651.20

Watercontent

{%)

16.4120.9718.9621.3917.4223.3119.69

16.9116.9712.5415.4119.9814.0218.3315.9518.0119.1516.73

Wet-bulkdensity(g/cm?)

2.142.032.092.042.232.022.09

2.132.162.292.232.192.312.212.262.192.122.21

Graindensity(g/cm*)

2.702.702.702.712.772.682.71

2.682.702.722.712.822.692.702.762.732.772.73

Porosity(%)

33.5941.6238.6642.5438.0145.9040.05

35.0935.9928.0433.6042.7531.6439.4935.1638.5439.6036.02

Note: Compiled from shipboard measurements (Haq, von Rad, O'Connell, etal., 1990).

Table 5. A summary on some important differences between the lightand dark beds.

ing the Milankovitch theory. The obliquity, precession, and eccen-tricity signals are all found in this cyclic sequence, though the last isdetected in only one of the time series. In the study byGolovchenko et al. (this volume) on the same interval, the histo-gram from thickness measurements in the core reveals that themost frequently occurring cycles are correlative to the obliquity andprecession cycles. However, their analysis using the gamma-raylog, whose vertical resolution is 0.3 m and not able to detect manyof the color alternations, did not reveal lower frequencies whosefrequency ratios and time duration are correlative to the eccentric-ity cycles. One possible reason for this may be that during the lateCampanian to early Maestrichtian the eccentricity played a weakand unsteady role in causing cyclic sedimentation. In comparisonto the other two orbital elements, eccentricity periodicities are lessregular.

MODEL FOR THE ORIGIN OF CYCLICITY

With the establishment of Milankovitch-like cycles in Hole762 C and with the available sedimentological evidence, wenow attempt to describe the origin of the sediment cyclicity.

-1

Code

0 +1

621

623

625

.α

Average thicknessand variance (cm)

CaCO3 content andvariance (%)

Clay content andcommon type

Quartz anddolomite

NannofossilsSponge spicules

and ostracodesPhysical properties

Light beds

18 ± 14

86.99 ± 3.17

Low, more smectite thankoalite and illite

Less

MoreLess

Higher water content,higher porosity

Dark beds

22 ± 18

74.46 ± 7.03

High, more koalite andillite than smectite

More

LessMore

Lower water content,lower porosity

szQ.CD

627

629

-

_

-

1

Figure 8. An example of the time series generated from the coreinspection record (time series 2). The value +1 stands for light beds and-1 for dark beds. Sections 122-762G-50X-1 through 122-762C-50X-6.

We suggest that orbital variation may have controlled climaticchange and thus induced the observed cyclic sedimentation onthe Exmouth Plateau during the Late Cretaceous. Unfortu-nately, there are no calculated orbital and insolation changesin the Late Cretaceous. Previous studies have documentedthat the Late Cretaceous was globally warmer than present

270


0.10 -

Time base= 19.58mN=2048a/b=4.36a/c 5.18b/c»1.19

0.05 -

CD

O

0.10 -

o.oo

0.20 -

0.10 -

40.8cm

Time base=9.78N=1024

BW

0.00

Time base=9.78mN=1024

BW

42.5cm

33.7cm

Cycles/m

Figure 9. Power spectra for time series 1. The resolution is indicated by the bandwidth (BW). "N" is thenumber of sampling points on the time series. The time base is the length of the time series being analyzed.The horizontal line is the 90% confidence level (c.l.) from Nowroozi (1967). The wavelength of the stablefrequency peaks is marked in centimeters. Wavelengths in time (k.y.) from the average sedimentation rateof 1.82 cm/k.y. are shown in parentheses. A. Whole-section spectrum for the entire time series 1. B.Subsection spectrum for the upper half of time series 1. C. Subsection spectrum for the lower half of timeseries 1.

271


Table 6. Ratios of modern periods of the precession, obliquity, andeccentricity cycles to each other.

Periods (103 years)

Ratio of precession (PI) to otherorbital elements

Ratio of precession (P2) to otherorbital elements

Ratio of obliquity (O) to otherorbital elements

Ratio of eccentricity (El) to otherorbital elements

Ratio of eccentricity (E2) to otherorbital elements

PI(19.0)

1.00

1.25

2.15

4.99

6.49

P2(23.7)

0.80

1.00

1.73

4.07

5.20

O(41.0)

0.46

0.58

1.00

2.31

3.01

El(94.9)

0.20

0.25

0.43

1.00

1.30

E2(123.3)

0.15

0.19

0.33

0.77

1.00

Note: Orbital periods are according to Berger (1977).

and free of large-scale continental glaciation (Frakes, 1979;Barron, 1983). According to the Cretaceous plate tectonicreconstructions by Barron (1987), during Maestrichtian timethe Australian continent, stretching west to east, was between45°S to 65°S. Ogg (pers. comm., 1990) gives an estimate of53°S for the paleolatitude of the Exmouth region. In a latitu-dinal basin such as the Exmouth Plateau and with cyclicchange in insolation caused by the obliquity and precessionvariations, we suggest a cyclic alternation of two differentclimatic regimes.

Given that the proportion of ocean and land areas isconstant in a region, in higher insolation phases both theocean and land received more energy from the sun. Becauseof the difference in the thermal inertia of water and land, thetemperature increase in the ocean is much smaller than on theland. In this period, land-sea thermal contrast is great insummer, resulting in higher precipitation when more hot airrises over the continent to be replaced by moist, cooler marineair, but the summer-winter temperature difference is relativelysmall because the ocean serves as an efficient thermal reser-voir and decreases the difference. In this phase, the climatehas smaller seasonal contrast in temperature and is rainy. An

"equable" warm and rainy climate favors the weathering ofrocks, especially chemical weathering, and more clay miner-als are formed on land. The high runoff because of higherprecipitation increases summer terrestrial output of clay min-erals (kaolinite and illite) and siliciclastic silt to the ocean. Therainy climate and high runoff may also lead to more precipi-tation than evaporation, which produces lower salinity surfacewater which may show seasonal variation. Warm surfacewater with varying and decreased salinity may not be suitablefor most planktonic species to "bloom." Excessive sedimentsin suspension in the upper water column may also affect thedevelopment of a planktonic community. Total bioproductiv-ity may be reduced in this phase for reasons of nutrientsupply. The reduced equator-pole temperature contrasts dur-ing increased solar insolation (which affects higher latitudesmore) may cause a decrease in wind strength, up welling, anddeep circulation. The nutrient supply may not be sufficient formarine plankton blooming. In such a climate, we consider thatthe dark beds with more clay were deposited at Site 762.

In the low-insolation phase, the total energy from the sun isdecreased. There is a smaller land-sea thermal contrast in thesummer, making it less rainy in comparison with the summerin high-insolation phases. The seasonal contrasts (winter,summer) in temperature become pronounced as the functionof the ocean as a thermal reservoir is reduced. The winter alsobecomes drier. Consequently, climate in the low-insolationphase is colder and drier. A colder, drier climate hinders theweathering of rock, especially chemical weathering, and fewerclay minerals are formed on land. The dilution of terrestrialinput is reduced because of low precipitation and resulting lowrunoff. Planktonic species increase in abundance becausefreshwater influx causes surface waters to be less stratified,and the plankton experience greater nutrient replenishmentfrom deeper levels in a period of increased wind strength andprobably more up welling. Under such a climate, the light darkbeds were deposited at Site 762.

The climatic mechanism considered to produce cyclicsedimentation at Site 762 is summarized in Figure 12, and is inaccordance with the climatic modeling results of the Creta-

0.30

0.20 -

io

Q_0.10

0.00

Time base=8.39mN=1024a/b=2.20a/c=2.40b/c=1.09

BW

84.0cm(46.2ky)

38.2cm(21 .Oky)

35.0cm(19.2ky)

V Λ 7

l/b c 90% c.l.

2 3

Cycles/m

Figure 10. Power spectrum for time series 2. The resolution is indicated by the bandwidth (BW). " N " is thenumber of sampling points on the time series. The time base is the length of the time series. The horizontalline is the 90% confidence level (c.l.) from Nowroozi (1967). The wavelength of the stable frequency peaksis marked in centimeters. Wavelengths in time (k.y.) from the average sedimentation rate of 1.82 cm/k.y.are shown in parentheses.

272


0.15

0.10-

0.05-

0.00

CD

Oü_

0.10-

0.00

0.20-

0.10-

Timebase=i 8.43mN=2048a/c=2.61b/c=2.31

70.9cm80.1cm (38.9ky)(44.0ky)

30.7cm(16.9ky)

c

Time base=9.21 mN=1024

30.7cm27.9cm

Time base=9.21 mN=1024

0.00i

2 3 4

Cycles/m

Figure 11. Power spectra for time series 3. The resolution is indicated by the bandwidth (BW). "N" is thenumber of sampling points on the time series. The time base is the length of the time series being analyzed.The horizontal line is the 90% confidence level (c.l.) from Nowroozi (1967). The wavelength of the stablefrequency peaks is marked in centimeters. Wavelengths in time (k.y.) from the average sedimentation rateof 1.82 cm/k.y. are shown in the parentheses. A. Whole-section spectrum for the entire time series 3. B.Subsection spectrum for the upper half of time series 3. C. Subsection spectrum for the lower half of timeseries 3.

273


ceous by Oglesby and Park (1989). Although in their study anuncoupled climate model is employed, and only insolationvariations caused by the precession cycle are considered, theresults support our explanation for the climate-induced cyclicsedimentation on the Exmouth Plateau. Their results in Southproto-Asia show that the January-July difference in landsurface temperature is greater in a low-insolation phase thanin a high-insolation phase, and in both January and July thespecific humidity and precipitation is higher in a high-insola-tion phase than in a low-insolation phase.

In our model we have integrated the factors causing thevariation in terrestrial material input and the change in bio-productivity. However, during the Late Cretaceous, under thecontrol of orbital elements, the changes on land may haveplayed a leading role in the cyclic sedimentation on theExmouth Plateau. Changes in the ocean are secondary, as theplanktonic microfauna are not only responding to watertemperature changes that may not be very substantial owingto great thermal inertia, but also to input of terrestrial materialand the surface salinity. Nutrient supply is another importantfactor for bioproductivity. As both dark and light beds arebioturbated in both climatic phases, the bottom water is welloxygenated. The oxygen levels in the bottom water may havebeen varying, but apparently never diminished enough toeliminate all bottom life. The fertility of the surface water maybe mainly affected by influx of freshwater but not by upwellingor deep circulation, according to the paleo-water depth (550m) estimated by Gradstein and von Rad (pers. comm., 1990)and the fact that the Exmouth Plateau is not far from thehinterland (terrestrial source).

The development of cycles in Pleistocene sediments in-volved changes in ice volume controlled by orbital variation(Pisias and Moore, 1981). Investigators of Upper Cretaceouscyclic sequences (R.O.C.C. Group, 1986) suggested that thesedimentary response to climatic forcing is variable. Thisstudy on the cyclic sediments recovered from Hole 762C notonly supports the theory of an orbital origin of cyclicity inpelagic sediment, but also furnishes one possible mechanismfor orbital variation controlling climatic change and hencecyclic sedimentation.

DISCUSSIONIn the interval between 630.5 and 649.5 mbsf (time series 3)

there are strong peaks most likely related to the obliquity andone less-strong peak possibly associated with precession (Fig.11 A). Between 621.0 and 630.0 mbsf (time series 2), the peakrelated to the obliquity is still strong (Fig. 10). In the upperinterval 600.0 to 621 mbsf (time series 1), there are peakspossibly related to the eccentricity and precession cycles (Fig.9A). However, according to both wavelength ratios and timeduration from the average sedimentation rate, there is no peakin the whole-section spectrum (Fig. 9A), which can possiblybe related to the obliquity cycle. This implies a transition fromobliquity-controlled climatic variations to eccentricity-preces-sion-controlled climatic variations. This also suggests that theresponse of the pelagic sedimentation system to the control-ling effects of the long-term orbital variations was changing inthe Late Cretaceous. Berger and Pestiaux (1984) suggest thatinsolation curves of lower latitudes mainly reflect the fluctu-ation in eccentricity and precession but in the curves of higherlatitudes the obliquity cycle is clearly expressed; if so, thistransform may be explained by a northward movement of theIndian plate in the late Campanian to early Maestrichtian. Thesequence from the bottom of time series 3 to the top of timeseries 1 represents a time span of about 3 m.y., according toan average sedimentation rate of 1.82 cm/k.y. Using a spread-ing rate of 2.0 cm/yr, the Indian Ocean plate could have

moved 60 km north in this period. This seems insignificant forthe Exmouth Plateau region to move into an eccentricity-precession-controlled low-latitude climate region from a obliq-uity-controlled high-latitude climate region if the cyclic sedi-mentation response to orbital variation were purely latitudedependent. Furthermore, according to Watts (1982), the sep-aration of the Indian plate from the Antarctic plate and itslarge-scale northward motion occurred around the late Pa-leocene, not in the Campanian and Maestrichtian. Anotherexplanation may be that the falling sea level during the end ofthe late Campanian (Haq et al., 1987) caused some changes inthe pattern of climatic variation. It appears that the responseof the pelagic sedimentation system to the controlling effectsof long-term orbital variations is complex and does not adhereto any simple relationship in the geological past. The study byHagelberg and Pisias (1990) indicates that the climatic re-sponse to orbital forcing during the Pliocene is totally differentfrom that during the Pleistocene.

In Figures 9, 10, and 11, the time durations converted fromthe average sedimentation rate for the peaks discussed arevery near to the periodicities (modern values, compare withTable 6) of the orbital elements to which they are interpretedto be related. This may imply that a total 6-m.y. duration forthe late Campanian to early Maestrichtian used for estimatingthe sedimentation rate is near the actual duration. However,there are some differences. For example, in the spectrum fromtime series 2 (Fig. 10) the peak associated with obliquity islonger (46.2 k.y.) than the obliquity periodicity (41.0 k.y.).This difference can be best explained by long-term variation insedimentation rate. The actual sedimentation rate for thissection (time series 2) could be higher than 1.82 cm/k.y.(about 2.05 cm/k.y.). The actual sedimentation rate for theother two sections (see Figs. 9 and 11) may not deviate muchfrom the estimated average one (1.82 cm/k.y.). However,there are sufficient indications of sedimentation rate change inthe intervals. For example, in Figure 11A there are two peakswhose time durations (44.0 k.y. and 38.9 k.y.) are both nearobliquity periodicity (41 k.y.). They are the product of onedominant obliquity cycle (41 k.y.). Two sedimentation rates(1.95 cm/k.y. and 1.73 cm/k.y.) are derived when the wave-lengths of these peaks are divided with the obliquity periodic-ity (41 k.y.). These two values may represent the fluctuatingrange of sedimentation rate in this interval (time series 3). Inboth Figures 9A and 10, two closely spaced peaks of about 20k.y. in duration may be respectively related to the 23.7-k.y.and 19.0-k.y. precession periodicities, but more likely theycould be the result of just one of the two precession periods,separated by change in sedimentation rate.

The brown to green change in background hue at 603.5 mbsfis believed to have been caused by a relative sea-level change(Haq, von Rad, 0'Connell, et al., 1990) from highstand tolowstand. According to this explanation, the actively mixedocean and the oxygenated water during the highstand periodresults in a brownish hue. This is not in accordance with theavailable evidence. In the whole interval we have studied, thereis neither an obvious difference in the degree of bioturbation norin the TOC content (see Table 2) in the sediments of differenthue, indicating little change in the oxygen level in bottom waterduring the period when the sequence was deposited. In our view,the change in hue was more directly caused by the decreasingintensity of weathering on land. When the brownish-hued inter-val was deposited the global climate was warmer and morehumid and the weathering products (clayey minerals, etc.) werebrownish because of complete oxidation. When the greenish-hued interval was deposited, the background climate becamecolder or at least less humid, and the weathering products(clayey minerals, etc.) were not fully oxidized and appeared

274


Orbital cycles:obliquity;precessioneccentricity, j

Cyclic variation of insolation ina latitudinal zone with a certainland-sea area proportion.

High insolation phases±

Low insolation phases

High land-sea thermal contrast in summer;small seasonal temperature difference;equable, warm and humid climate.

Low land-sea thermal contrast in summer;large seasonal temperature difference;dry and cold climate.

Strong weathering process on the land;more terrestrial input to the ocean. Bio-productivity is decreased.

Weak weathering process on the land;less terrestrial input to the ocean. Bio-productivity is increased.

Dark clayey nannofossil chalk Light nannofossil chalk

Cyclic1 sediments

Figure 12. Suggested controlling mechanism of cyclic orbital variation for both the change of climate and theformation of cyclic sediments in the Late Cretaceous Indian Ocean (see text for detailed explanation).

greenish. It is also possible that the terrestrial input of ironoccurred in a reduced state. This explanation is not in contra-diction with the correlation between the hue change and therelative sea-level change, nor with the occurrence of pyrite in thegreenish-hued interval. The relative sea-level change may haveaccompanied the overall hue change, but may not have neces-sarily been the final cause.

CONCLUSIONS1. The cyclic color variation in the upper Campanian to

lower Maestrichtian pelagic sediments from Site 762 is adepositional feature rather than a diagenetic result. Diage-netic processes at most enhanced the appearance of thecyclicity.

2. Spectral analysis has revealed that the predominantwavelength of the color cycles are 34-41 cm and 71-84 cm.With an average sedimentation rate of 1.82 cm/k.y., the time

durations of the color cycles are around 41 k.y. and 21 k.y.,correlative to the obliquity and precession periods andstrongly suggesting an orbital origin for the cycles. A weaklow-frequency signal correlative to the eccentricity cycle isfound in the upper part of the sequence (time series 1).Therefore, the cyclic variations in the Upper Cretaceousstrata from Hole 762C resulted from cyclic variations inorbital elements and we advocate that long-term orbital vari-ations have been at work in the geological past.

3. During the Late Cretaceous, when there was no conti-nental glaciation, cyclic sedimentation on the Exmouth Pla-teau of the eastern Indian Ocean was primarily caused bycyclic changes on land (soil formation and runoff) in responseto orbitally-controlled insolation variations. Bioproductivityvaried in phase with changes in terrestrial input, but nostagnant, oxygen-depleted bottom water was involved in theprocess.

275


4. Spectral analysis indicates that the way in which orbitalvariations control cyclic sedimentation may vary with time. Inthe older part of the sequence (time series 2 and 3), there arestrong obliquity signals with precession signals, but in theupper parts of the sequence (time series 1) there are onlyprecession and eccentricity signals. A transition from obliqui-ty-controlled climatic variations to eccentricity-precession-controlled climatic variations is implied. We suggest that thesedimentary response to the ultimate controlling force oforbital variations is neither necessarily constant through geo-logical time in one place nor globally uniform at one time.Many factors can constrain the sedimentary response. Thechange from brownish hue to greenish hue in the sequencemay indicate a major shift in the background climatic pattern,from very humid to less humid.

5. This investigation illustrates that building discrete timeseries on the basis of lithological inspection and calculatingpower spectra from such time series with the Walsh spectralanalysis is an efficient, economical, and practical procedurefor studying cyclic sediments. If subsurface materials are tobe investigated, the constraints on this procedure are mainlythe length of the interval and core preservation. Our expe-riences show that in examining and interpreting the powerspectra it is more helpful to use both wavelength ratios andtime durations from the estimated average sedimentationrate.

ACKNOWLEDGMENTS

We thank the Shipboard Scientific Party of Leg 122 for thecooperative scientific effort that made this study possible, G.Hampt for calcium carbonate measurements, and J. Tribblefor XRD analysis. SEM examination was assisted by F.Thomas. This work benefited from discussions with Dr. R.Wilkens. Our manuscript was improved by comments andsuggestions from Drs. F. M. Gradstein and J. G. Ogg. We aregrateful to Drs. T. Herbert and T. Hagelberg for their criticalreview. Financial support for Z.H. and R.B. was provided bya Canadian NSERC Collaborative Special Project Grant and aUniversity Research Grant from Imperial Oil (Canada), andfor S.O. by USSAC.

REFERENCES

Arthur, M. A., Dean, W. E., Bottjer, D., and Scholle, P. A., 1984.Rhythmic bedding in Mesozoic-Cenozoic pelagic carbonate se-quences: the primary and diagenetic origin of Milankovitch-likecycles. In Berger, A., Imbrie, J., Hays, J., Kukla, G., andSaltzman, B. (Eds.), Milankovitch and Climate (Pt. 1): Dordrecht(D. Reidel), 191-222.

Barron, E. J., 1983. A warm, equable Cretaceous: the nature of theproblem. Earth Sci. Rev., 19:305-338.

Barron, E. J., 1987. Cretaceous plate tectonic reconstructions. Palae-ogeogr., Palaeoclimatol., Palaeoecol., 59:3-29.

Barron, E. J., Arthur, M. A., and Kauffman, E. G., 1985. Cretaceousrhythmic bedding sequences: a plausible link between orbitalvariations and climate. Earth Planet. Sci. Lett., 72:327-340.

Beauchamp, K. G., 1975. Walsh Functions and Their Applications:New York (Academic Press).

Berger, A. L., 1977. Support for the astronomical theory of climaticchange. Nature, 269:44-45.

Berger, A. L., and Loutre, M. F., 1989. Pre-Quaternary Milankovitchfrequencies. Nature, 342:133.

Berger, A. L., and Pestiaux, P., 1984. Accuracy and stability of theQuaternary terrestrial insolation. In Berger, A., Imbrie, J., Hays,J., Kukla, G., and Saltzman, B. (Eds.), Milankovitch and Climate(Pt. 1): Dordrecht (D. Reidel), 83-111.

Bottjer, D. J., Arthur, M. A., Dean, W. E., Hattin, D. E., and Savrda,C. E., 1986. Rhythmic bedding produced in Cretaceous pelagiccarbonate environments: sensitive recorders of climatic cycles.Paleoceanography, 1:467-481.

Cotillon, P., and Rio, M., 1984. Cyclic sedimentation in the Creta-ceous of Deep Sea Drilling Project Sites 535 and 540 (Gulf ofMexico), 534 (Central Atlantic) and in the Vocontian Basin(France). In Buffler, R. T., Schlager, W., et al, Init. Repts. DSDP,77: Washington (U.S. Govt. Printing Office), 339-376.

Darmedru, C , Cotillon, P., and Rio, M., 1982. Climatic and biologicalrhythms in marine pelagic realm: their relationships in Cretaceousalternating beds from the Vocontian Basin (SE France). Bull. Soc.Geol. Fr., 7:672-640.

Frakes, L., 1979. Climate through Geological Time: New York(Elsevier).

Hagelberg, T. K., and Pisias, N., 1990. Nonlinear response ofPliocene climate to orbital forcing: evidence from the EasternEquatorial Pacific. Paleoceanography, 5:595-617.

Hallam, A., 1986. Origin of minor limestone-shale cycles: climaticallyinduced or diagenetic? Geology, 14:609-612.

Haq, B. U., Hardenbol, J., and Vail, P. R., 1987. Chronology offluctuating sea levels since the Triassic. Science, 235:1156-1167.

Haq, B. U., von Rad, U., O'Connell, S., et al., 1990. Proc. ODP, Init.Repts., 122: College Station, TX (Ocean Drilling Program).

Hart, M. B., 1987. Orbitally induced cycles in the chalk facies of theUnited Kingdom. Cretaceous Res., 8:335-348.

Hays, J. D., Imbrie, J., and Shackleton, N. J., 1976. Variations in theearth's orbit: pacemaker of the ice ages. Science, 194:1121-1132.

Herbert, T. D., and D'Hondt, S. L., 1990. Precession climate cyclic-ity in Late Cretaceous-Early Tertiary marine sediments: a highresolution chronometer of Cretaceous-Tertiary boundary events.Earth Planet. Sci. Lett., 99:263-275.

Herbert, T. D., and Fischer, A. G., 1986. Milankovitch climatic originof Mid-Cretaceous black shale rhythms in Central Italy. Nature,321:739-742.

Huffman, E.W.D., 1977. Performance of a new automatic carbondioxide coulometer. Microchem. J., 22:567-573.

Imbrie, J., and Imbrie, J. Z., 1980. Modeling the climatic response toorbital variations. Science, 207:943-953.

Jarrard, R. D., and Arthur, M. A., 1989. Milankovitch Paleoceano-graphic cycles in geophysical logs from ODP Leg 105, LabradorSea and Baffin Bay. In Srivastava, S. P., Arthur, M., Clement, B.,et al., Proc. ODP, Sci. Results, 105: College Station, TX (OceanDrilling Program), 757-772.

Kent, D. V., and Gradstein, F. M., 1985. A Cretaceous and Jurassicgeochronology. Geol. Soc. Am. Bull, 96:1419-1427.

King, D. T., Jr., 1990. Upper Cretaceous marl-limestone sequences ofAlabama: possible products of sea-level change, not climateforcing. Geology, 18:19-22.

Negi, J. G., and Tiwari, R. K., 1983. Matching long term periodicitiesof geomagnetic reversals and galactic motions of the solar system.Geophys. Res. Lett., 10:713-716.

Nowroozi, A. A., 1967. Table for Fischer's test of significance inharmonic analysis. Geophys. J. R. Astron. Soc, 12:512-520.

Ogg, J. G., 1987. Early Cretaceous magnetic polarity time scale andthe magnetostratigraphy of Deep Sea Drilling Project Sites 603 and534, Western central Atlantic. In van Hinte, J. E., Wise, S. W.,Jr., et al., Init. Repts. DSDP, 93: Washington (U.S. Govt. PrintingOffice), 849-880.

Oglesby, R., and Park, J., 1989. The effect of precessional insolationchanges on Cretaceous climate and cyclic sedimentation. J. Geo-phys. Res., 94:14973-14816.

Pisias, N. G., and Mix, A. C , 1988. Aliasing of the geologic recordand the search for long-period Milankovitch cycles. Paleoceanog-raphy, 3:613-619.

Pisias, N. G., and Moore, T. C , Jr., 1981. The evolution of Pleis-tocene climate: a time series approach. Earth Planet. Sci. Lett.,52:450-458.

Research on Cretaceous Cycles (R.O.C.C.) Group (Arthur, M. A.,Bottjer, D. J., Dean, W. E., Fischer, A. G., Hattin, D. E.,Kauffman, E. G., Pratt, L. M., and Scholle, P. A.), 1986.Rhythmic bedding in Upper Cretaceous pelagic carbonate se-quences: varying sedimentary response to climatic forcing. Geol-ogy, 14:153-156.

Schlumberger, 1989. Log Interpretation Principles/Applications:Houston, TX (Schlumberger Educ. Services).

Schwarzacher, W., 1987. The analysis and interpretation of stratifi-cation cycles. Paleoceanography, 2:79-95.

276


Schwarzacher, W., and Fischer, A. G., 1982. Limestone-shale bed-ding and perturbation of the Earth's orbit. In Einsele, G., andSeilacher, A. (Eds.), Cyclic and Event Stratification: New York(Springer-Verlag), 72-95.

ten Kate, W. G., and Sprenger, A., 1989. On the periodicity in acalcilutite-marl succession (SE Spain). Cretaceous Res., 10:1-31.

Watts, A. B., 1982. Tectonic subsidence, flexure and global changesof sea level. Nature, 297:469-474.

Weedon, G. P., 1986. Hemipelagic shelf sedimentation and climaticcycles: the basal Jurassic (Blue Lias) of South Britain. EarthPlanet. Sci. Lett., 76:321-335.

Date of initial receipt: 28 May 1990Date of acceptance: 29 January 1991Ms 122B-148

APPENDIXCalcium Carbonate Measurement With Coulometric Method

Samples were dried in either a 100°C oven or in a freeze drier,powdered, weighed to the nearest milligram, and analyzed for calcium

carbonate using a Coulometrics CO2 Coulometer. The Coulometerhas an accuracy of 0.15% (±0.2 mg).

The Coulometric titration technique measures all of the CO2 that isliberated by acidifying and heating sediment samples in a closedsystem. To do this, the powdered samples are placed in a test tubewhich is attached to the Coulometer, placed in a heat shield, and 2HHC1 is pumped into the test tube. The liberated CO2 is transferredthrough scrubbers (solutions to remove interfering substance) by aCO2-free gas into an absorption cell where it is titrated throughcoulometric means. The absorption cell contains an aqueous mediumof thanolamine and a coulometric indicator. The interaction of theCO2 and the cell solution creates a titratable acid. The coulometerforms a base electrically and titrates to an endpoint determined by theoptical transmission of the indicator (Huffman, 1977). The pulseoutput was scaled and fed to a counter in terms of CO2. A stableCoulometer reading indicates that all of the CO2 has been evolved andtitrated. The reading was recorded and the results were calculated inmicrograms carbon and CaCO3 (wt% CaCO3 = wt% Cinorganjc× 8.334). This method of calculation assumes that all of the CO2 isliberated from calcium carbonate, but some of the CO2 may beliberated from dolomite, siderite, or other carbonate minerals.

277

13. upper cretaceous cyclic sediments from hole 762c

Documents