DEPOSITIONAL AND GEOCHEMICAL CYCLICITY IN THE … · DEPOSITIONAL AND GEOCHEMICAL CYCLICITY CT&F - Ciencia, Tecnología y Futuro - Vol. 1 Núm. 2 Dic. 1996 5 DEPOSITIONAL AND GEOCHEMICAL

DEPOSITIONAL AND GEOCHEMICAL CYCLICITY

CT&F - Ciencia, Tecnología y Futuro - Vol. 1 Núm. 2 Dic. 1996 5

DEPOSITIONAL ANDGEOCHEMICAL CYCLICITY IN

THE CRETACEOUS FINE-GRAINED STRATA OF COLOMBIA.A MODEL FOR ORGANIC MATTER

CONTENT

Department of Geological Sciences, CB 250, University of Colorado, Boulder, CO 80309-0250, USA.Current address: Advance Exploration Organization OF3084, Conoco Inc., P.O. Box 2197, Houston, TX 77252, USA

pectral analyses of depositional and geochemical time series were conducted on three stratigraphicsections of the Cretaceous Villeta Group, Colombia. Results show statistically-valid cyclicity in distalparasequence stacking patterns and in geochemical variations. Cycles are in the low to high frequency

spectra (500 ky - 20 ky). Timing of cyclicity suggests climatic controls on deposition, probably caused byMilankovitch orbital oscillations. Stratigraphic intervals enriched in total organic matter or organic matterindicators such as barium coincide with condensed intervals that are generally associated with observable high-frequency cyclicity. Very thin stratigraphic cycles (< 0,4 m) contain moderate amounts of organic matter becausecondensation allows organisms to recycle organic carbon. Condensation also allows time for gradual oxidationof organic matter. Thin to moderately thick (0,6-5,0 m) cycles show the highest quantity of organic matter.Sedimentation rates are high enough to inhibit biological destruction and low enough not to dilute organiccarbon within the sediments. Thick cycles are characterized by low organic matter because sedimentation ratedilutes organic material. organic matter content not only depends on sedimentation rate but also onpaleoproductivity and on grain size and shape. The model presented shows how cycle thickness trends may beused as a very simple approach to predict the amount of total organic carbon.

Análisis espectrales de tiempo-frecuencia fueron aplicados a series deposicionales y geoquímicas en tressecciones del Grupo Villeta del Cretácico de Colombia. Los análisis muestran ciclicidad estadísticamentevalida en patrones de apilamiento de parasecuencias en regiones distales de la cuenca y en variabilidadgeoquímicas. Esta ciclicidad varía de baja a alta frecuencia (500 ka-20 ka); los resultados sugieren un controlorbital tipo Miankovitch. Intervalos estratigráficos enriquecidos en carbón orgánico coinciden con seccionescondensadas en las que se observa una ciclicidad estratigráfica de alta frecuencia. Ciclos estratigráficos finos(menores de 0,40 m) contienen cantidades moderadas de materia orgánica, a consecuencia de condensaciónque permite que organismos reciclen el carbón orgánico y al mismo tiempo permite la oxidación gradual de lamateria orgánica. Ciclos estratigráficos finos a medios (0,6 m-5,0 m) muestran la mayor calidad y cantidad demateria orgánica. En estos casos la tasa de sedimentación es suficientemente alta para inhibir la descomposiciónde la materia orgánica por organismos y suficientemente baja como para no diluir la materia orgánica en lamasa de sedimento. Ciclos estratigráficos gruesos (> 5 m) generalmente contienen bajas cantidades de materiaorgánica porque altas tasas de sedimentación la diluyen en su masa. El contenido de materia orgánica dependede la sedimentación, de la paleoproductividad oceánica y del tamaño y forma del grano sedimentario. Semuestra cómo el espesor de los ciclos estratigráficos de alta frecuencia puede ser utilizado como una herra-mienta muy simple para predecir la cantidad de materia orgánica.

Keywords: Depositional and geochemical ciclicity, statigraphy, cretaceous Villeta group

T. VILLAMIL

CT&F - Ciencia, Tecnología y Futuro - Vol. 1 Núm. 2 Dic. 1996

T. VILLAMIL

6

and diagenetically overprinted, making recognition ofprimary signals difficult in some cases but enhancingthem in others. Randomness can be filtered withspectral analyses on a time series. If a stratigraphicrecord, which is composed of many types of signals,shows statistically-valid depositional cyclicity, then thiscyclicity should be observed. However, the fact that acertain stratigraphic section does not show cyclicitydoes not mean that an oscillating control was notinfluencing deposition. Several frequencies ofdepositional and chemical cyclicity were discoveredin the Cretaceous record of Colombia through theapplication of statistical tests to field-derived data andlaboratory analyses of equally-spaced samples. Tocalculate the magnitude of these cycles, a newcomposite macro- and micro-biostratigraphy was used(Villamil, 1994) and was graphically-correlated to theradiometric ages established for the sections in theUnited States Western Interior by Obradovich (1993)and the calculated time scale of Kauffman et al.(1993).

Stratigraphy and location of sections.The Villeta Group and equivalents are composed

primarily of fine-grained facies such as calcareousshales and hemipelagic limestones. The depositionalhistory of the Villeta was highly dynamic and influ-enced by a combination of climatic, sedimentologic,sea level and tectonic factors (Villamil, 1994). Stratafrom the Hiló Formation to the chert-dominated OliniGroup will be discussed. The Hiló Formation consistsof rhythmically bedded shales, cherty shales, and highlycalcareous shales and scattered hemipelagic lime-stones. An overlainly Cenomanian unnamed unit, com-posed primarily of claystones, calcareous shales,bioclastic limestones and calcareous sandstones. Thisunit represents an overall regression. The La FronteraFormation has similar facies to the Hiló Formation buthas a higher calcareous pelagic component. The OliniGroup (latest Coniacian-Campanian) consists of rela-tively shallow-water (~20m-~150m) cherts interbeddedwith shales and some hemipelagic limestones; this unitwas also studied for cyclicity.

Stratigraphic sections were measured in the Cre-taceous outcrop belt of Central Colombia (Figure 1).Distalmost facies were measured in the UpperMagdalena Valley near the towns of Yaguará andChaparral. The Yaguará section (2° 53' 30,81''N, 5°

INTRODUCTION

The Cretaceous Villeta Group and equivalent unitsof Colombia offer an excellent opportunity for the studyof depositional and chemical cycles. Cycle analysescan be conducted on stratigraphic thickness variationsand geochemical signals such as oscillations of organiccarbon and major elements. In this paper, beddingfeatures and chemical changes were evaluated forcyclicity with statistical techniques. Depositional andchemical oscillations were analyzed using time seriesMaximum Entropy Spectral Analyses (MESA;Goldhammer et al., 1990 and references therein).Results show that the Villeta and equivalent unitsrecord depositional and geochemical cyclicity that hassignificant peaks at ~20 ky, ~31 ky, ~ 40 ky, ~100 ky,and ~500 ky. The purpose of this paper is to illustratenew geochemical analyses for the Cretaceous ofColombia and emphasize the importance of cyclicityin the record. In addition, a hypothesis for therelationship between cycle thickness and the contentof organic matter is proposed.

Background.The first work related to the study of depositional

cyclicity of the Cretaceous of Colombia was by Bürgl(1961), who claimed that the Cretaceous systemshowed a 6 million-year depositional cyclicityrepresented by stage boundaries. Macellari (1988)suggested 5 low-frequency transgressive-regressivedepositional cycles. Most recently, Föllmi et al. (1992)described higher-frequency depositional rhythmicity inthe Cretaceous of Colombia. Villamil (in press)conducted studies on the sequence stratigraphiccyclicity of distal facies of the Villeta Group ofColombia, utilizing Fischer plots. However, to date,spectral analyses have not been used to statisticallydemonstrate cyclicity for these facies.

It has been hypothesized that cyclicity of thestratigraphic record responds to oscillations of theEarth�s climate-ocean system, the well-knownMilankovitch cycles (see Einsele and Ricken, 1991 andreferences therein). The stratigraphic record isinfluenced by many factors including noise and chaos.Einsele and Ricken (1991) illustrated the point: primarycyclic signals are hidden within depositional backgroundnoise, primarily in basinal facies. Strata are compacted



Maracaibo

Basin

Barin

asfo

rede

ep

basin

Pla

tobasin

SanJorge

basin

Sin

œb

asi

n

San

Jacin

to

Belt

Rom

era

l fault

San

taM

arta

fau

lt

R’

oM

ag

dale

na

Santa Marta massif

Sie

rra

de

Per

ija

Oca fault

Bocono fault

Santa

nd

er

massif

Arauca archBUCARAMANGA

Mid

dle

Mag

dal

ena

Bas

in

BOGOTA

Llan

osfo

rede

epba

sin

EASTE

RN

CO

RD

ILLE

RA

Vaupes arch

Gar

zon

mas

sif

Putum

ayo

fore

deep

basin

NS

BG

SB

Serrania

de

Baud

ó Atr

ato

fault

Atr

ato

basin

Rom

eral

fault

2¼

4¼

6¼

8¼

76¼ 74¼ 72¼

N

0 100 200 300 400 600 Km

YAGUARA SECTION

MESIT AS SECTION

CE

NTR

AL

CO

RD

ILLE

RA

Cau

cath

rust

bel

t

WE

STE

RN

CO

RD

ILLE

RA

AC

CR

ETIO

NA

RY

OLINI REFERENCE SECTION

PA

CIF

ICO

CE

AN

Figure 1. Map of Colombia with the location of stratigraphic sections discussed in text, map modified from Schamel's (1991).

CARIBBEAN


T. VILLAMIL

8

It is not coincidental that stratigraphic intervalsassociated with high magnitude peak transgressions,e.g., second-order maximum flooding surfaces, arecharacterized by high-frequency hemipelagic(carbonate/silica)-clastic cycles because these are thetimes of maximum sediment starvation and ofdiminished autocyclic noise in the record. Examplesof maximum flooding high frequency cyclically-beddedunits are the Hiló and the La Frontera Formations.The fact that these maximum flooding intervals arecharacterized by rhythmic hemipelagic beddingsupports the idea of a climatic control in paleopro-ductivity but does not rule out control in clasticdeposition because climatic changes control the densityand salinity of the water, threby affecting sea level.The idea of considering a dual control on cyclicity, withchanges in sedimentation rate and paleoproductivity,may work better than considering a single controllingfactor. The model proposed by Eicher and Diner (1989,1991) explains a dual control in the development ofcycles. This model, although centered on paleopro-ductivity cycles, also shows a climatic effect on relativesea level (Eicher and Diner, 1989). Eicher and Diner(1989, 1991) associated warm-dry cycles with highproductivity and a slight rise in sea level. An increasein paleoproductivity would enhance deposition ofcalcareous organisms and the rise in sea level wouldtrap sediments shoreward, producing a positivefeedback in the development of hemipelagic limestonecycles.

Changes in sedimentation rate, and oscillations inoceanic chemistry and productivity control patterns andquantities of element concentration. High concentra-tions of elements such as barium, phosphorous andothers have been proposed to be related to oceanicproductivity and therefore can be statistically-analyzedto depict cyclicity. The work of Dymond et al. (1992),Turekian and Tausch (1964), and Bolter et al. (1964)have shown the importance of Barium measurementsin determining the amount of paleoproductivity in astratigraphic interval. Barium measurements may bebetter than TOC analyses in outcrop studies becauseBa is less sensitive to weathering processes. Althoughfluctuations in paleoproductivity cause changes inbarium and phosphorous in the stratigraphic record,changes in sedimentation rate also dramatically affectelement concentrations.

40' 32,43''W) was measured in the Yaguará River nearthe confluence with the Iquira River. The Olini refer-ence section (3° 40' 32,43''N, 75° 32' 25,95''W) wasmeasured in Olini Creek, half way between Chaparraland Coyaima (Department of Tolima) (Figure 1). TheMesitas del Colegio section was measured on thewestern flank of the Eastern Cordillera (4° 32' 25,95''N,74° 26' 45,41''W) at km 28 on the road leading fromBogotá to Mesitas del Colegio along an unnamed creek.

Geologic setting and discussion.The Upper Magdalena Valley and Eastern Cordi-

llera of Colombia were part of a single large basinduring the Albian to Santonian. The paleolatitudinalposition of NW South America favored vigorousupwelling onto the upper water column; high paleoproductivity characterized Cretaceous times ofnorthern South America. Stratigraphic evidence forhigh productivity includes: (a) The presence ofsiliceous-rich rocks such as cherts and siliceous shales,considered as evidence of high paleoproductivity(Berger, 1974, 1976). Phosphates are also consideredas evidence for intense upwelling (e.g., Föllmi, 1991).(b) Rich radiolarian assemblages, found in the LateAlbian-Early Ceno-manian of the Upper MagdalenaValley by Arango (1995) and Villamil and Arango (inpress). (c) Abundant barium, cadmium, phosphorous,and organic carbon values of up to 16% that alsosupports vigorous upwelling and oceanicpaleoproductivity in the Colombian Cretaceous margin(Villamil, 1994 for details).

Stratigraphic units analyzed for chemical anddepositional cyclicity were mostly deposited in distaloffshore environments (~200-300 m water depth). Thedistal setting records subtle as well as marked changesin deposition but it poses severe problems forinterpretations of sedimentation rates and thedevelopment of stratigraphic cycles. The distal offshoresetting is, in general, lithostratigraphically condensed;however, in terms of time, it records much more thannearshore settings. Even when no clastic sedimentreaches this portion of the basin, the pelagic rain oforganisms has the potential to register time of non-deposition of clastics. The development of depositionalsequences and parasequences, however, may be limitedby low sediment supply and submarine erosion.



PITFALLS OF CYCLE ANALYSIS

The biggest obstacle for evaluating cyclicity in atime-dependent basis with the Cretaceous stratigraphicrecord is the relatively poor time control at a continuousscale. Average rock accumulation rate, as a calibrationfor a time scale, has been used by multiplying theaverage rate by the stratigraphic thickness. This pro-cedure poses many critical problems becausesedimentation rate is assumed to be constant andtransferred to «time», and this requires constantsedimentation rate.

The models of Pratt (1985) and Barron et al. (1985)explain the occurrence of shales as fine clastic inputduring wet portions of a climatic cycle, assumingrelatively constant rain of hemipelagic organisms, i.e.,constant calcareous input and variable clasticsedimentation rate. Productivity cycles (Eicher andDiner, 1989, 1991) create lithological differences as afunction of climatic changes that affect paleopro-ductivity. Higher productivity enhances abundance ofcalcareous organisms, and sedimentation rate oflimestones therefore overwhelms clastic input. Anadditional factor that influences the assumption ofconstant rock accumulation rates is that differences ingrain size and shape also affect rock accumulationrates. Compaction differences between coarse,spherical grains and fine, flat particles are marked.An evaluation of relative compaction and sedimentationrate should be conducted before applying spectralanalyses to time series in «real» time domain.

METHODOLOGY

Data for analyzing depositional and chemicalcyclicity were obtained in the field in a careful manner.Stratigraphic sections were measured at differentscales of resolution. Sections were measured with aJacob staff, and individual beds at the cm-scaleresolution with a ruler. Stratigraphic intervals of interestfor cycle analyses such as condensed sections andstage boundaries were measured at the mm-scale.Samples for geochemistry were collected every 0,5 mthroughout the sections studied, and in the intervals ofspecial interest they were collected every 0,05 m.Stratigraphic data were analyzed in terms ofdepositional cycles in the following manner. Afundamental cycle, which is a distal fine-grained facies

equivalent of a nearshore parasequence was proposed.The base of the cycle starts at the base of a hemipelagiclimestone bed, the top of the cycle is the base of thenext hemipelagic limestone.

Geochemical analyses were conducted on discreteoutcrop intervals. Therefore, stable geochemicalmarkers have to be selected for the study of cyclicity.Barium is not easily weathered or altered and mayrecord paleoproductivity in an accurate manner.Evidence for this comes from the present oceans(Bolter et al., 1964 and Schmitz, 1987) and from thedeep sea record (Turekian and Tausch, 1964) amongothers. Three hundred samples were selected andanalyzed with ICP. The entire Villeta Group of theUpper Magdalena Valley was analyzed at a larger scaleto compare to the Cenomanian-Turonian (C-T)boundary interval throughout the same region.

Statistical methods to evaluate cyclicity.Spectral analyses used are based on algorithms of

simplified Fourier series. Data were plotted as valuesversus stratigraphic thickness or cycle number. Spectralanalysis replicates the shape of the data plot usingdifferent frequencies of sine wave curves and outputsthe relative frequency of the dominant waves. Fourieranalysis decomposes a time series into a fundamentalwave and additional harmonics. The transform is ofthe general mathematical form:

f(t)= X1+X2sin(t) + X3cos(t) + X4sin(2t) + X5cos(2t) + ...

Maximum Entropy Spectral Analysis (MESA) withthe fast algorithm developed by Dunn et al. (1986)was applied to all data.

RESULTS

Low-frequency cycles.The Cretaceous of Colombia is composed of several

second-order (sensu Haq et al., 1987) sequences withan average frequency of 6 Ma. This cyclicity wasoriginally recognized and discussed by Bürgl (1961).The spacing of second-order maximum floodingintervals is expressed in the geological record asisotopic facies generally associated with concretionsand black shales. A cyclicity of approximately 500 kyin the Cretaceous of Colombia is apparent when


T. VILLAMIL

10

running a maximum entropy spectral analysis onequally-spaced geochemical data, particularly bariumand phosphorous (Figures 2 and 3, respectively). These~500 ky cycles may have been the result of relativelylong-term variations in productivity, the result of high-frequency, high-order maximum flooding or condensed

sections or a combination of both. Figures 2 and 3 showconcentrations of Ba and P of samples collected andanalyzed every meter in the Yaguará section. The inserton both figures, as in all the following ones, is the resultof spectral analyses on the time series. Frequency isin the x-axis and power, the correlation value of

1000

0

10

CE

NO

MA

NIA

NT

UR

ON

IAN

CO

NIA

CIA

N

La

Fro

nte

raF

orm

atio

n

SA

NT

ON

IAN

Olin

ig

r.

Up

Up

Lo

wLo

wU

pp

er

Un

nam

ed

un

it

Mid

Mid

20

30

40

50

60

70

80

90

100

m

110

200 300

Ba, ppm

400 500

0,50,40,30,20,1

10.000

30.000

20.000

Figure 2. Barium data and spectral analysis of a barium time series of the Yaguará section in the Uper Magdalena Valley ofColombia. Shaded area represents values below average and white area values above average. The power / frequency insetare the results of the spectral analysis utilizing Maximum Entropy Spectral Analysis.The central peak represents -500 ky cyclicity

Frequency

Po

we

r

»500 ky



frequencie is in the y-axis. ~500 ky oscillations inconcentration are detected in all sections measured asvariations in different elements. The origin of thesecycles is uncertain but ~500 ky cycles may be sufficientlyclose to the 400 ky Milankovitch ellipticity cycles.

Vertical variations in Ba content can be interpretedin terms of paleoproductivity because it has strongbiogenic associations (Chow and Goldberg, 1960 andDymond et al., 1992). Bishop (1988) showed thatbarium is taken up and transported by sinking biogenic

Figure 3. Geochemical data and spectral analysis of phosphorous time series of the Yaguará section in the Upper MagdalenaValley of Colombia. Shaded area represents values below average and white area values above average. The power / frequency

inset are the results of the spectral analysis utilizing Maximum entropy Spectral Analysis. Peaks of the inset represent- 500 ky and - 100 ky cyclicity.

CE

NO

MA

NIA

NT

UR

ON

IAN

CO

NIA

CIA

N

La

Fro

nte

raF

orm

atio

n

SA

NT

ON

IAN

Olin

ig

r.

Up

Up

Lo

wLo

wU

pp

er

Un

nam

ed

un

it

Mid

Mid

0

0 1.000 2.000 3.000 4.000

P. ppm.

Po

we

r

Be

low

ave

rag

e

5.000 6.000 7.000

10

20

30

0,00,00e+0

1,00e+6

100 ky

500 ky2,00e+6

0,1 0,2 0,3 0,4 0,5

40

50

60

70

80

90

100

m

110

Frequency


T. VILLAMIL

12

particles in the ocean. Spectral analysis, in the inset ofFigure 2, shows two peaks at high and low frequencies,probably reflecting combined cycles. ~500 ky cyclicitycan also be detected in Ba of the Olini reference section(Figure 4). The most pronounced peak in Figure 4 is at~500 ky and represents the same type of cyclesdiscussed above. Cycle analysis for the entire Olinisection produces low power results because data fromvarious systems tracts and facies are mixed. However,even with limited reliability, time-series analyses stillare sensitive to the ~500 ky signal.

High-frequency cycles.Depiction of 100 ky cycles has to be evaluated with

closely-spaced samples. Figure 5 shows thicknesstrends of the Olini reference section. The stratigraphicsection was divided into systems tracts. Running timeseries on different systems tracts diminishes thepotential error caused by differences in sedimentationrate and measuring across unconformities. Figure 5shows the Olini reference section and the data basecomposed of cycle thickness vs. cycle number (bottominset). Four time series results are shown, one for allthe data and three for different segments of thestratigraphic section. The top left diagram showsanalyses for all cycles. The result is quite noisysuggesting blending and interference of signals,probably by differences in sedimentation rate. Thereare two peaks, however, near frequencies 0,3-0,4; thepeak to the left is quite strong and rises above noiselevel suggesting that it represents some sort ofstatistically-valid peak in cyclicity. The value of thispeak is approximately 100 ky. The other three graphsare results of spectral analyses on individual systemstracts. The upper right graph (A) is the resultingspectral analysis of the lower Villeta Group of theUpper Magdalena Valley. This portion of the Villetawas measured from the top of the Caballos Formation(lower Albian) to the latest portion of the highstandtract (Albian-Cenomanian boundary), and it representsa short transgressive systems tract, a maximumflooding interval, and a condensed highstand. Timeseries analysis shows a peak at ~100 ky, reflectingbed repetition in the lower Villeta Group. Stratigraphicequivalents to this portion of the Villeta in other regionsof Colombia also show a strong imprint of relativelyhigh-frequency cyclicity.

The middle left graph of Figure 5 (late HST B)represents time series results from the Albian-Cenomanian boundary to the base of the La FronteraFormation. Results show a strong peak at ~100 ky;this cycle is reflected stratigraphically as parasequencerepetitions and as bundling of shale beds. This intervalrepresents gradual progradation from the Cenomanian-Albian boundary interval to the latest Cenomanian. Thethird spectral analysis was conducted on the LaFrontera highstand (Figure 5, HST C). The La Fronterahighstand shows two peaks in depositional cyclicity.The right peak represents ~100 ky, reflected strati-graphically as relatively thick limestones with minorbundles inside.

The Yaguará stratigraphic section also showsmarked 100 ky depositional cyclicity (Figure 6). Thissection was divided into discrete intervals to avoidcomparing differences in sedimentation rates. Thelower time series analysis represents the stratalrepetitions of the condensed section interval. Thisanalysis shows three peaks; the highest one is at ~100 ky,followed by ~40 and <20 ky peaks. The time series inthe middle of Figure 6 shows cycle analysis for theentire La Frontera Formation and part of the underlyingunnamed unit. The remaining time-series plot does notshow cyclicity in any frequency. This may reflectmixing sediments representing very low and relativelyhigh sedimentation rates. The graph at the top of Figu-re 6 shows the second-order early highstand La Fron-tera Formation. This plot shows a peak at ~100 and asmaller-scale peak at ~40 ky.

The Mesitas del Colegio stratigraphic section (Figure7) is located in a more proximal region of the basin.Figure 7 shows the stratigraphic section to the left, thedata of cycle thickness versus cycle number on thebottom, and the time series conducted on this data set.Time series analysis shows a peak at ~100 ky and asmaller one at a higher frequency. The statistical validityof these cycles cannot be precisely evaluated; however,peaks suggest that a control with a frequency on theorder of 100 ky affected sedimentation patterns.

Geochemical analyses depict ~100 ky cyclicity butnot as strongly as ~500 ky cyclicity. Figures 3 and 4show higher frequency cycles that are close to 100 ky.Phosphorous content in the Yaguará section (Figure3) shows a strong ~100 ky signal attributed tohighfrequency changes in sedimentation rates, paleo-



?

?

?

200 400 600 800

Ba, PPM

0

0,50,40,30,20,10,0

0

10.000

20.000

30.000

40.000

Po

we

r

Frequency

≈

»100 ky

»500 ky

ALB

IAN

SA

NT

Olin

ig

r..C

EN

OM

AN

IAN

TU

R.

CO

NIA

CIA

N

Lo

we

rL

ML

MU

Mid

dle

Ville

tag

rou

pLa

Fro

nte

ra

Up

pe

rLo

we

rM

idU

pp

er

300

50

100

150

200

250

0

380

450

500

550

600

Figure 4. Geochemical data and spectral analysis of the Olini reference section barium time series (see Figure 1 for location).Several peaks in cyclicity can be observed in this 600 meter-thick section; the strongest peak is represented in the spectral

analysis --inset-- by the highest peak, 500 ky in duration. Additional peak to the right represent higher relative frequencies 100.


T. VILLAMIL

14

Figure 5. Different spectral analyses stratigraphic segments of the Oliní reference section. The inset on the bottom of the figurerepresents the thickness of cycles plotted versus the cycle number. Data has been separated in systems tracts for spectral

computations. The upper left spectral analyses represents all measured beds, the upper right inset is the spectral analysis for onlythe lower portions of the Villeta Group (A) --the Hiló Formation--. The central left spectral analysis was conducted on the late

highstand portion of the Villeta, upper Albian to uppermost Cenomanian. The central right inset represents the spectral analisysfor the condensed section and early highstand HST (C).

Frequency Frequency

Po

wer

Po

wer

All villetagroup

Late HST

300

380

450

500

550

600

50

100

150

200

250

0

0,50,40,30,20,1

10

20

30

40

0,0

0,50,40,30,20,10,0

20

40

60

80

100

0,50,40,30,20,10,00

2

4

6

8

10

La FronteraHST (C)

200150100500

0

10

Cicle number

Th

ickn

ess

(m)

18115387Condensed sectionand early highstand

(A)

Late highstand

(B) (C)

Co

nd

en

sed

sectio

na

nd

ea

rly

hig

hsta

nd

(A)

Hig

hsta

nd

(C)

La

teh

igh

sta

nd

(B)

100-40 ky

Frequency Frequency

»500 ky

»100 ky

Lower villetagroup (a)

4

Po

wer

0,50,40,30,20,10,02

6

8

10

»100 ky

»100 ky

»40 ky

»20 ky

»100 ky

Highstand

Cab

allo

sV

ille

tag

rou

pLa

Fro

nte

raO

lin

ig

r.

ALB

IAN

Mid

dle

Up

pe

rLo

we

r

SA

NT. .

Lo

we

rM

id.

?

?

Up

pe

r

TU

R.

CE

NO

MA

NIA

N

LM

CO

NIA

CIA

N

LM

?

U



Up

pe

rLo

wM

idU

p

Un

nam

ed

un

itLa

Fro

nte

raF

orm

atio

n

110

100

90

80

70

60

50

40

30

20

10

Olin

ig

r.

SA

NT

OR

INIA

NC

ON

IAC

IAN

TU

RO

NIA

NC

EN

OM

AN

IAN

Lo

wM

idU

p

m

0

0,60,50,40,30,20,1

0,0

0,1

0,2

0,3

0,4

Po

we

r

Early HST

0,60,50,40,30,20,1

0

1

2

3

4

Frequency

Frequency

Po

we

rCen-base of

santonian

0,60,50,40,30,20,10,00

0,02

0,04

0,06

0,08

0,10

Frequency

Po

we

r

Condensed

»100 ky

»100 ky

»40 ky

»40 ky

Figure 6. Spectral analyses for different systems tracts of the Yaguará section.The stratigraphic section was divided into three segments to avoid marked differences in sedimentation rate, insets are correlated

to stratigraphic segments


T. VILLAMIL

16

Up

pe

r

La

Fro

nte

raF

orm

atio

n

Lo

we

rM

idd

le

TU

RO

NIA

NC

ON

IAC

IAN

CE

NO

MA

NIA

N

72

74

76

78

80

82

84

86

88

90

92

94

96

98

100

102

104

106

108

38

40

42

44

46

48

50

52

54

56

58

60

62

64

0,50,40,30,20,10,0

0,0

0,2

0,4

0,6

0,8

1,0

Frequency

Po

we

r

50403020100

0

1

2

3

4

Limestone-shale

cycle number

Cycle

thic

kn

ess

Figure 7. Spectral analysis on parasequence thickness trends of the Mesitas del Colegio stratigraphic section, Eastern Cordillera.

»40 ky

»100 ky



and 20 ky cycles (Figure 9). 20 ky cyclicity can beobserved with spectral analyses on 0,5 m time series(Figure 9 a, b). Figure 8 shows a restricted and high-powered peak at or near 20 ky. There are at least twopossibilities for explaining why the Olini stronglyrecords 20 ky cycles. The first possibility is that slightvariations in paleoproductivity or sediment supply, whenupwelling is very strong, affect the stratigraphic recorddramatically compared to similar variations in a lessvigorous upwelling regime. The second possibility isthat chert deposition is more sensitive to variations insediment supply and paleoproductivity than otherlithologies.

INFLUENCE OF CYCLICITY IN THECONTENT OF ORGANIC MATTER.

The percentage of organic carbon or organic matterstored in a rock depends on: production of organiccarbon, transport to the bottom and storage insediments, and preservation of organic carbon in thesediment and rock. Studies with emphasis onpaleoproductivity control on organic carbon depositionhave been conducted by Hallam (1967) and Eicherand Diner (1991). Other studies suggest that organiccarbon variations within stratigraphic cycles have adiagenetic origin (Ricken, 1986). Studies that attributechanges in organic carbon within stratigraphic cyclesas a result of a combination of effects, i.e., variationsin sediment supply, paleoproductivity and diagenesis,have been conducted by Arthur and Dean (1991).Oceanic paleoproductivity, differential sedimentation,and the type of sediment, determine the total contentand quality of organic carbon stored in a rock.

Oceanic paleoproductivity controls the amount oforganic matter that reaches the sea bottom. NorthwestSouth America was located during most of theCretaceous in a position that favored high productivityand organic matter rain (Villamil, 1995) that reached amaximum during the Santonian with chert-dominatedfine-grained, sediment-depleted settings. Santonian-Maastrichtian fine-grained strata are generally highlysiliceous shales or pure cherts (e.g., the Olini Group,Plaeners Fm.). The latest Cretaceous Guadalupe Groupis composed of medium-to coarse-grained sandstonesinterbedded with cherts. While sandstones were beingdeposited in one region of the basin, shallow-watercherts were being deposited in the areas of lowsediment supply.

productivity oscillations, or both. Stratigraphically, ~100ky cyclicity in phosphorous peaks reflects landwardshifts of facies belts; e.g., phosphatic-rich intervals areassociated with transgressive pulses or distalparasequence boundaries. These phosphatic lags re-flect sediment starvation and condensation (e.g., Föllmiet al.,1991) and have been used extensively to depictcondensed sections. Geochemical peaks in phos-phorous content coincide with the occurrence of phos-phatic lags. Evidence for sedimentary concen-trationof phosphatic lags comes from the shape and size ofphosphatic pebbles and sedimentary structures. Thesepebbles vary in size from large (up to 0,03 m) to verysmall (< 0,001 m). Additional geochemical indicatorsto inger 100 ky cyclicity used in the present work arephosphorous and barium (Figures 2-4).

40 ky cyclicity can be interpreted from verticalvariations in the stratigraphic thickness of distalparasequences (Figure 5, late HST B). 40 ky cyclicityof condensed intervals such as the lower portion ofthe Hiló and the base of the La Frontera Formationsare easily observed in the field. ~40 ky cyclicity has astrong signal in cherts and siliceous shales of the OliniGroup and the La Luna Formation (Figure 8). TheOlini Group is composed primarily of black chertsinterbedded with siliceous shales deposited undervigorous upwelling conditions. The Olini Group wasmeasured in the Yaguará River in the southern portionof the Upper Magdalena Valley. Dating the OliniGroup has been difficult because of very rare occu-rrence of ammonites. The amount of chert (biogenic)input apparently was controlled by climatic cyclicity.Strata of the Lower Chert of the Olini Group arecyclically-bedded and bundled into larger cycles. Figu-re 8 shows spectral analyses of the Lower and UpperChert intervals of the Olini Group measured in the UpperMagdalena Valley.

Geochemically, ~40 ky cyclicity can be detectedby spectral analyses of calcium, barium, phosphorousand other element time series. Figure 9 shows resultsof spectral analyses conducted on elementalconcentrations on closely-spaced samples (0,5 m).Spectral analyses on time series of geochemicalinformation every 0,5 m were only conducted for thebase of the La Frontera Formation. In many instances,analyses of high-frequency cyclicity of the Cretaceousof Colombia yield 31 ky peaks in the spectral analyses;these 31 ky peaks may be the result of blending 40


T. VILLAMIL

18

20 40 60 80 100

0,0

0,5

1,0

1,5

Time X 1.000 years

Po

we

r

126

124

122

120

118

116

114

112

110

108

106

104

102

100

98

PERIODICITY DIAGRAM

Un

it1

LO

WE

RS

AN

TO

NIA

N

Olin

ig

rou

p(p

art

)firs

tch

ert

inte

rval

Un

it2

Un

it3

Sh

ale

2040

Figure 8. Spectral analyses of two different intervals of the Santonian-Early Campanian Olini Group in the Upper MagdalenaValley. Note strong peaks at 20 and 40 ky.



The combination of these relationships would hy-pothetically locate the highest organic matter contentin distal toes of parasequences where sedimentationrate allows sufficient organic matter to be trapped inthe sedimentary matrix (Figure 11). Figure 11 showsan extrapolation of this relationship within a sequencestratigraphic model. Thicker cycles with higher sedi-mentation rate contain more organic material becausea higher sedimentation rate favors trapping and pres-ervation of organic matter. Cycles that are too thick,however, are depleted in TOC because organic mat-ter is diluted within the sediment.

STRATIGRAPHIC RESPONSE TO ORBITALCYCLES, SUMMARY AND ANALYSIS

Periodic changes of climate due to astronomicallycontroled variations of the distribution of solar radia-tion over the Earth influence the stratigraphic record

26

25

24

23

TU

RO

NIA

NC

EN

OM

AN

IAN

La

Fro

nte

raF

orm

atio

n

0 2 4 6 8

P, PPM (X 10.000)

0,60,50,40,30,20,10,07.00e+7

8.00e+7

9.00e+7

1.00e+8

1.10e+8

1.20e+8

Frequency Frequency

Po

we

r

Po

we

r

100 ky

40 ky

20 ky

%

10 20 30 40

0,50,40,30,20,10

100

200

300

400

40 ky

100 150 200 250

Ba, PPM

Frequency

0,50,40,30,20,10,0

0

5.000

10.000

Po

we

r

100 ky

40 ky

20 ky

0 50

Figure 9. Geochemical data and high-frequency (100, 40, and 20 ky) cyclicity in a phosphorous (a), barium (b) and calcium (c)time series of the La Frontera Formation measured near the town of Yaguará, Upper Magdalena Valley. Strongest peaks in results

represent ~20 ky Milankovitch cycles.

Sedimentation rate and the type of sediment thatreaches the bottom control the storage of organic mat-ter (Figure 10). Johnson (1982) showed that organicmatter in sediments is quite low with very slow ratesof deposition because microbes have time to attackand decompose organic matter. As sedimentation rateincreases, the content of organic matter increases be-cause time for microbe destruction of organic matterdiminishes. This direct relationship holds true until fastsedimentation rate dilutes organic matter (Figure 10).If sedimentation rate determines the thickness of strati-graphic cycles and sedimentation rate controls theamount of organic carbon, then there should be a rela-tionship between cycle thickness, type of cycle andTOC content. In addition to this relationship, oceanicproductivity also determines the type of sediment (bio-genic) that reaches the bottom, i.e., as productivity in-creases, populations adapted to higher nutrient con-tents will replace previous populations.


T. VILLAMIL

20

through variations of climatic and oceanic circulationsystems. This influence controls the distribution of car-bonate and organic matter within the rock (De Boer,1990). Changes in insolation: at low latitudes, the in-fluence of the precession cycle is dominant and causeslatitudinal shifts of the caloric Equator (De Boer, 1991).High levels of insolation may produce the followingeffects: a) Increased oceanic and continental evapo-ration leading to shorter residence time of water masseson land reservoirs and on the surface, and productionof relatively dense saline waters in the ocean (Pratt,1985 and Barron et al., 1985). Shorter residence timeof water over the continents and in the ground waterreservoir may result in changes of precipitation andcontinental runoff and in a decrease of net water trans-port to oceanic basins. The stratigraphic response is adecrease in sedimentation rate and a decrease in thedistance of sediment transport. b) A rise in tempera-ture also causes a double effect on the volume of watermasses. On one side, water masses will increase involume because of the thermal expansion (Eicher andDiner, 1991). On another side, however, water mixing

is favored by an increase in circulation, and densitymay vary with mixing. The stratigraphic response ofan increase in water volume would be an increase inaccommodation space in the nearshore setting and trap-ping of sediment shoreward. Thermal expansion pro-duces a positive effect that adds up to a decrease inresidence time; both factors combined trap sedimentsshoreward by creation of accommodation space anddecrease the net rate of sediment transport. There-fore the warm-dry portion of the cycle would producea parasequence boundary in nearshore settings and aportion of the cycle that is low in sediment supply indistal regions.

In addition, differences in insolation may haveproduced changes in the velocity of water circulationbecause of different degrees of energy in the ocean-atmosphere system. Increase in water circulation ratesresulted in higher paleoproductivity. An increase ofaccommodation space with lowered sediment supplycoupled with an increase in paleoproductivity favorsdeposition of calcareous- or siliceous-rich hemicycles

Increase TOCwith sed. rate

Dilutionof TOC

1 10 100

Sedimentation rate (SR) m/m.y.

1,0

0,1

Black shale

"Cherts"

"Limestones"

Thincycles

Thickcycles

Optimumthickness

Figure 10. Relationship between sedimentation rate and organic matter content in different types of lithologies and cyclethickness (modified from Johnson, 1982).



rather than clastic-rich hemicycles. A decrease ininsolation favors precipitation and runoff and does notfavor vigorous oceanic circulation. The stratigraphiceffect observed during low insolation periods is clastic-rich hemicycles.

A completely different set of factors that shouldalso be considered when proposing stratigraphicresponses to Milankovitch astronomical cycles arechanges in gravitational forces due to different positionsof the planets. These changes in gravitation directlyand markedly affect tides which, in turn, affect netsediment transport and stability of sediment piles. Tidaldifferences caused by astronomical cycles of theMilankovitch frequencies have not been studiedsufficiently but could probably change current viewsof the stratigraphic response to orbital cycles.

CONCLUSIONSDepositional and geochemical cyclicity in low to

Shoreward Basinward

OneParasequence

TO

MIn

cre

ase

Total OM expected

OM, litho-control

OM, Sed. rate control

OM

=High organicmatter values

OM OM

ON

EON

E

Pa

raseq

uen

ce

Pa

raseq

uen

ce

LST

SB

SB

MFS

TS

HST

TST

= High OM (Organic matter) values

Sequence scale

Parasequence scale

Figure 11. Model that shows the hypothetical distribution of organic matter within one depositional cycle or parasequence. Thecircle in the parasequence shows the location of highest organic matter content within the parasequence. Simplified model with

the location of highest organic matter sites at the sequence scale. Note how organic-rich levels are diachronous theiraccumulation begins earlier in distal regions and terminates later also in distal regions.

high frequencies dominated the Cretaceous distal,fine-grained deposits of Colombia and are inagreement with Milankovitch orbital oscillations.Results presented here, however, do not proveMilankovitch cyclicity because analyses wereconducted on the stratigraphic thickness domainrather than in the time domain. Spectral analysesof geochemical and stratigraphic time series do notallow a clear test of paleoproductivity or dilutionmodels. Cretaceous chronology does not allowprecise calculations of sedimentation rate todetermine whether variations in sedimentation orin paleoproductivity produced observed cyclicity.Frequencies of approximately 500, 100, and 40 kywere detected with spectral analyses of geochemi-cal and stratigraphic data. Sedimentation rate, sedi-ment type and paleoproductivity determine the quan-tity and quality of organic matter preserved in arock. Because climatic cyclicity controls sedimen-tation rate and paleoproductivity, it also determines


T. VILLAMIL

22

the position of organic-rich facies within a strati-graphic unit. Very thin and amalgamated cycles arecharacterized by low sediment supply and there-fore may contain low amounts of organic matter.Thicker cycles with higher sedimentation rate con-tain more TOC, but cycles that are too thick aredepleted in TOC because organic matter is dilutedwithin the sediment.

ACKNOWLEDGMENTS

Claudia Arango, Erle Kauffman and William W. Theyprovided helpful comments and improved ideas. MaríaClemencia Serrano and the editorial committee of thisjournal provided helpful suggestions. BHP Mineralsthrough Barry Bolton and Patrick Okita funded mostgeochemical analyses. Juan Carlos Ramón and anonymousreviewer improved the quality of the manuscript.

REFERENCES

Arango, C., 1995. "Foraminifera sensing of paleo-ceanographic changes across the Cenomanian -Turonian(C-T) boundary interval in Colombia and Pueblo,Colorado", MSc. unpublished thesis, University ofColorado, Boulder, 140 p.

Arthur, M. A. and Dean, W. E., 1991. "An holistic geo-chemical approach to cyclomania: examples from Cre-taceous pelagic limestone sequences", in Cycles andevents in stratigraphy, Einsele, G., Seilacher, A. andRicken, W. (eds.), Springer, Berlin: 23 MS p.

Barron, E. J., Arthur, M. A. and Kauffman, E. G., 1985."Cretaceous rhythmic bedding sequences -A plausi-ble link between orbital variations and climate", Earthand Planetary Science Letters, 72: 327 - 340.

Berger, W. H., 1974. "Deep-sea sedimentation", in TheGeology of Continental Margin, Burk, C.A. and Drake,C. D., (eds.): 213 - 242.

Berger, W. H., 1976. "Biogenous deep sea sediments:production, preservation and interpretation", inChemical Oceanography, Riley, J. P. and Chester, R.(eds.), Academic Press NY: 265 - 347.

Bishop, J. K. B., 1988. "The Barite-Opal-Organic carbonassociations in oceanic particulate matter", Nature, 331:341 - 343.

Bolter, E., Turekian, K. K. and Schutz, D. F., 1964 . "Thedistribution of rubidium, cesium and barium in theoceans", Geochimica et cosmochimica Acta, 28: 1459 -1466.

Bürgl, H., 1961. "Sedimentación cíclica en el geosinclinalCretáceo en la Cordillera Oriental de Colombia", BoletínGeológico, 7: 85 - 118.

Chow, T. J. and Goldberg, E. D., 1960. "On the marinegeochemistry of Barium", Geochimica et CosmochimicaActa, 20: 192 - 198

De Boer, P. L., 1990. "Astronomical cycles reflected insediments", Zentralblatt für Geologie und Paläon-tologie, Teil I, H8: 911 - 930.

Dunn, P. A., Goldhammer, R. K. and Hardie, L. A., 1986. "Mr.Sediment- a computer model for carbonate cyclicity",Geological Society of America Abstracts with Programs,18: 590.

Dymond, J., Suess, E. and Lyle, M., 1992. "Barium in deepsea sediments: a geochemical proxy for paleoproduc-tivity", Paleoceanography, 7: 163 - 191.

Eicher, D. L. and Diner, L., 1991. "Environmental factorscontrolling Cretaceous limestone-marlstone rhythms",in Cycles and Events in Stratigraphy, Einsele, G., Ricken,W. and Seilacher, A. (eds.), Springer-Verlag: 79 - 93.

Eicher, D.L. and Diner, R., 1989. "Origin of the CretaceousBridge Creek cycles", in the Western Interior, UnitedStates. Palaeogeog., Palaeoclimat. and Palaeoecol., 74(1/2): 127 - 146.

Einsele, G., and Ricken, W., 1991. "Limestone-marlalternations - an overview�, in Cycles and Events inStratigraphy, Einsele, G., Ricken, W. and Seilacher,A. (eds.), Springer-Verlag: 23 - 47.

Föllmi, K. B., Garrison, R. E. and Grimm, K. A., 1991."Stratification in phosphatic sediments: illustrationsfrom the Neogene of California", in Cycles and Eventsin Stratigraphy, Einsele, G., Ricken, W. and Seilacher,A. (eds.), Springer-Verlag: 492 - 507.

Föllmi, K. B., Garrison, R. E., Ramírez, P. C., Zambrano,F. O., Kennedy, W. J. and Benedikt, L. L., 1992. "Cyclicphosphate-rich successions in the Upper Cretaceousof Colombia", Palaeogeography, Palaeoclimatology,Palaeoecology, 93: 151 - 182.



Goldhammer, R. K., Dunn, P. A. and Hardie, L. A., 1990."Depositional cycles, composite sea level changes, cyclestacking patterns, and the hierarchi of stratigraphic forc-ing: examples from Alpine Triassic platform carbonates",Geological Society of America Bulletin, 102: 535 - 562.

Hallam, A., 1967. "The depth significance of shales withbituminous laminae", Marine Geol., 5: 481 - 493.

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

Johnson, I. L., 1982. "Relationship between sedimentationrate and total organic carbon content in ancientsediments", American Association of Petroleum Geo-logists, Bulletin, 66 (2): 170 - 188.

Kauffman, E. G., Sageman, B. B., Kirkland, J. I., Elder, W. P.,Harries, P. J. and Villamil, T., 1993. "Molluscanbiostratigraphy of the Western Interior Cretaceous Basin,North America", in Cretaceous Evolution of theWestern Interior basin, North America, Caldwell, W. G.E., and Kauffman, E. G. (eds.), Special PublicationGeological Association of Canada (in press).

Macellari, C. E., 1988. "Cretaceous paleogeography anddepositional cycles of Western South America", Journalof South American Earth Sciences, 1: 373 - 418.

Obradovich, J. D., 1993 . "A Cretaceous Time Scale", inCretaceous Evolution of the Western Interior basin,North America, Caldwell W. G. E. and Kauffman E. G.(eds.), Special Publication Geological Association ofCanada: 379 - 397.

Pratt, L., 1985. "Isotopic Studies of Organic Matter andCarbonate in rocks of the Greenhorn Marine Cycle",in Fine Grained Deposits and Biofacies of the CretaceousWestern Interior Seaway: Evidence of Cyclic SedimentaryProcesses, Pratt, L., Kauffman, E. G. and Zelt, F. B. (eds.),SEPM Second Annual Midyear Meeting,: 38 - 48.

Ricken, W., 1986. Diagenetic Bedding: A model for marl-limestone alternations, Springer-Verlag, Berlin: 210 p.

Schmitz, B., 1987. �Barium equatorial high-productivity andthe northward wandering of the Indian Continent�,Paleoceanography, 2: 63 - 78.

Turekian, K. K. and Tausch, E. H., 1964. "Barium in deep-sea sediments of the Atlantic Ocean", Nature, 201: 696- 697.

Villamil, T. (in press). " High resolution biostratigraphy andsea level changes, Albian to Santonian organic-richinterval in Colombia and Venezuela", in Mesozoic-Cenozoic stratigraphy and tectonic evolution of theCaribbean region/Northern South America:Implications for eustasy from exposed sections of aCretaceous-Eocene passive margin, Pindell, J. L. andDrake (eds.), SEPM Special Paper.

Villamil, T. and Arango, C. (in press). "High-resolutionanalyses of the Cenomanian-Turonian Boundary inColombia: evidence for sea level rise, condensation andupwelling", in Mesozoic-Cenozoic stratigraphy andtectonic evolution of the Caribbean Region/NorthernSouth America: Implications for eustasy from exposedsections of a Cretaceous-Eocene passive margin setting,Pindell, J. L. and Drake, C. D. (eds.), SEPM Special Paper.

Villamil, T., 1994. "High-resolution stratigraphy, chronol-ogy and relative sea level of the Albian-Santonian (Cre-taceous) of Colombia", Ph.D. dissertation, Universityof Colorado, Boulder: 462.

Villamil, T., 1995. "Un modelo tectónico-paleoceanográfico yestratigráfico secuencial para el intervalo de roca generadoradel Cretácico del norte de Sur América", Memorias VICongreso Colombiano del Petroleo, 1: 53 - 57.

Documents

DEPOSITIONAL AND GEOCHEMICAL CYCLICITY IN THE … · DEPOSITIONAL AND GEOCHEMICAL CYCLICITY CT&F - Ciencia, Tecnología y Futuro - Vol. 1 Núm. 2 Dic. 1996 5 DEPOSITIONAL AND GEOCHEMICAL