Shallow lacustrine carbonate microfacies document ...projects.knmi.nl/publications/fulltexts/abels_sedimentology_2009.pdf · Shallow lacustrine carbonate microfacies document orbitally

Shallow lacustrine carbonate microfacies document orbitallypaced lake-level history in the Miocene Teruel Basin (North-EastSpain)

HEMMO A. ABELS*, HAYFAA ABDUL AZIZ�1, JOSE P. CALVO� and ERIK TUENTER§2

*Stratigraphy/Paleontology, Dept. of Earth Sciences, Utrecht University, Budapestlaan 4, 3584 CD,Utrecht, The Netherlands (E-mail: [email protected])�Department of Earth and Environmental Sciences, Geophysics LMU Munich, Theresienstrasse 41, 80333Munich, Germany�Instituto Geologico y Minero de Espana (IGME), Rıos Rosas 23, 28003 Madrid, Spain§Royal Netherlands Meteorological Institute (KNMI), P.O. Box 201, 3730 AE, De Bilt, The Netherlands

Associate Editor: Tracy Frank

ABSTRACT

Results are presented of a detailed carbonate petrographic study of an Upper

Miocene lacustrine mixed carbonate–siliciclastic succession in the Teruel

Basin (Spain) with the aim of constraining lake-level variability at different

stratigraphic scales. Regular alternations of red to green mudstone and

lacustrine limestone, termed the ‘basic cycle’, reflect lake-level variations at

the metre-scale. In an earlier study, the basic cycle was shown to be controlled

by the climatic precession cycle. Petrographic analysis made it possible to

distinguish two main carbonate microfacies groups characteristic of very

shallow transient and shallow permanent lake environments, respectively. In

addition to the basic cyclicity, the microfacies analysis reveals lake-level

variations on a larger scale. As a consequence, the astronomical forcing

hypothesis of the cyclicity in the Cascante section is explored further. A

climate modelling study of orbital extremes indicates that high lake levels

could relate to enhanced net winter precipitation and runoff during precession

minima, consistent with Mediterranean geological data. Using this phase

relationship, an astronomical tuning of the cycles is established starting from

astronomical ages of magnetic reversal boundaries. Subsequently, successive

basic cycles are correlated to precession minima. The tuning reveals an

identical number of basic cycles in the Cascante section as precession-related

sapropel cycles in the deep marine succession at Monte dei Corvi (Italy),

corroborating the precessional control of the basic cycles at Cascante. Lake-

level highstands in the large-scale cycle identified by the microfacies analysis

relate to maxima in both the ca 100 and 405 kyr eccentricity cycles, again

consistent with Mediterranean geological data. Subtraction of the identified

astronomically related (lake-level) variations from the palaeoenvironmental

record at Cascante indicates a shift to deeper and more permanent lacustrine

environments in the upper half of the section. The cause of this shift remains

unclear, but it may be linked to tectonics, non-astronomical climate, long-

period astronomical cycles or autogenic processes.

1Present address: Deltares/TNO, Geological Survey of The Netherlands, Princetonlaan 6, 3508 TA, Utrecht, TheNetherlands.2Present address: Institute for Marine and Atmospheric Research (IMAU), Utrecht University, Princetonplein 5, 3584CC, Utrecht, The Netherlands.

Sedimentology (2009) 56, 399–419 doi: 10.1111/j.1365-3091.2008.00976.x

� 2008 The Authors. Journal compilation � 2008 International Association of Sedimentologists 399

Keywords Astronomical forcing, carbonate petrography, climate modellingcontinental sediments, cyclostratigraphy, Late Miocene, Spain.

INTRODUCTION

Reconstruction of the evolution of sedimentaryenvironments requires distinction among tec-tonic, climatic and autogenic processes, which isoften difficult to achieve because they may resultin similar palaeoenvironmental shifts and actsimultaneously as well. However, high-resolution(ca 104 yr) age control may provide critical infor-mation about the forcing mechanisms of thepalaeoenvironmental changes. Recent improve-ments of the Cenozoic, and especially the Neo-gene, magnetic polarity timescale (Gradsteinet al., 2004; Lourens et al., 2004) offer an excellentopportunity to correlate lithological and biologicalchanges in sedimentary successions with well-dated tectonic and (global) climatic events(Dupont-Nivet et al., 2007). The improvementsfurther allow the presence of astronomical climateforcing to be examined, which potentially yieldsvaluable information on the response of palaeo-environmental and sedimentological processes toknown climate variations in terms of relativehumidity and seasonality (Weedon, 2003). Inaddition, the identification and subsequentremoval of orbitally induced climate variationsfrom the sedimentary record reveal other varia-tions that may be related to autocyclicity, tectonicsor non-astronomical climate-forcing mechanisms.

Neogene continental successions in Spain offerlong sedimentary records of low-gradient endor-heic basins in which climate and tectonics theo-retically have a great impact on the depositionalenvironment (Sanz et al., 1995; Abdul Aziz et al.,2003b). In the present paper, a continued cyclo-stratigraphic study of the Cascante section in thesouthern Teruel Basin (NE Spain; Fig. 1) ispresented, elaborating on a previous study byAbdul Aziz et al. (2004). The section is charac-terized by a distinct alternation of red to greenmudstone, interpreted to be deposited in a distalalluvial floodplain environment, and shallowlacustrine to palustrine limestone (Fig. 2). Mag-netostratigraphy and small-mammal biostratigra-phy revealed that deposition took place between9Æ4 and 10Æ2 Ma (Abdul Aziz et al., 2004). AbdulAziz et al. (2004) applied spectral analysis onhigh-resolution colour records. Results revealed acyclicity at a 1:2:5 ratio, resembling the ratio of theclimatic precession, obliquity and short eccen-tricity cycles. Combining the statistical and mag-

netostratigraphic results, those authors showedthat the basic, metre-scale mudstone–limestonecycles are related to the precession cycle (AbdulAziz et al., 2004). However, no palaeoenviron-mental significance could be assigned to the otherscales of cyclicity that were tentatively attributedto obliquity and short-eccentricity. Consequently,no astronomical tuning of the Cascante sectionwas established. In this study, a detailed sedimentpetrographic study of all 34 lacustrine and palus-trine limestone beds was therefore performed. Inaddition, simulations with a global climate modelwere carried out as an independent check on thephase relationship between lake-level highstandsand precession as derived from published geolog-ical data. Finally, an astronomical tuning of theCascante succession is presented, allowing aquantification of the influence of climate andtectonic variability on this continental sedimen-tary environment.

GEOLOGICAL SETTING

In the Teruel Basin (Fig. 1A), a 100 km long and15 km wide NNE–SSW trending half-graben,sediment deposition started in the Early Mioceneand lasted until the Late Pliocene (Anadon &Moissenet, 1996). Sedimentation took place in aninternally drained continental setting dominatedby red alluvial terrigenous beds, lacustrine andpalustrine limestones, and gypsum. The sedimen-tary succession is characterized by large-scale (40to 200 m) alternations of red siliciclastic sedimentsand lacustrine carbonate or evaporitic sediments(Anadon et al., 1997; Alonso Zarza & Calvo, 2000).

The studied Cascante section is located in thesouthern part of the basin (Fig. 1B), along theVillel-Cascante del Rio road, in the Teruel-Ade-muz sub-basin. The sub-basin is flanked byMesozoic rocks, predominantly composed of Tri-assic mudstone and evaporite rocks and Jurassiclimestone. The presence of gypsum deposits inthe basin has been related to leaching of Triassicsediments cropping out along the basin margin(Broekman, 1983; Anadon et al., 1992). The Neo-gene infill starts at the basin margin with alluvialfan conglomerates and breccias that grade later-ally, within a few kilometres, into fine-graineddistal alluvial fan sediments and shallow lacus-trine deposits towards the centre of the basin.

400 H. A. Abels et al.

� 2008 The Authors. Journal compilation � 2008 International Association of Sedimentologists, Sedimentology, 56, 399–419

THE CASCANTE SECTION

The studied Upper Miocene section comprisesthe siliciclastic interval of the Loma de CascanteFormation, one of the formation-scale silici-clastic–limestone alternations in the Teruel Basin(Fig. 2). The lithology is characterized by a

distinct regular sedimentary cyclicity of red andgreen, orange mottled, distal alluvial fan mud-stone and white, shallow lacustrine limestone(Abdul Aziz et al., 2004). This basic mudstone–limestone cycle has an average thickness of 2Æ2 m(Abdul Aziz et al., 2004) (Fig. 3). Thirty-fourbasic cycles were observed in the Cascante

Fig. 1. (A) Geological and location map of the region of the NNE–SSW trending Teruel Basin. (B) Detailed geologicalmap and location map of the Cascante section (after Abdul Aziz, 2001). Miocene basinal facies are depicted in thelegend from bottom to top according to their stratigraphic order.

Orbitally paced lake-level history 401


section. A 30 m thick lacustrine limestone unitcharacterized by metre-scale bedding overliesthe mixed carbonate–siliciclastic succession(Fig. 2).

SEDIMENTOLOGY – OUTCROP SCALE

The outcrop data presented in this paper are acompilation of new and published observationsand interpretations (Broekman, 1983; Broekmanet al., 1983; Kiefer, 1988; Abdul Aziz, 2001;Abdul Aziz et al., 2004). The specific outcropcharacteristics of all 34 basic cycles are consistentlaterally over a distance of 300 m to the west ofthe studied section. Some unique mudstonecolour patterns in outcrops between 1 and 2 km

to the east were proven to be time-equivalentusing detailed magnetostratigraphy (Abdul Azizet al., 2004).

Red and green mudstone

DescriptionThe average thickness of the mudstone part of abasic cycle is 1Æ8 m (ranging from 1Æ4 to 3Æ3 m).The mudstone commonly starts with a dark redcolour at the base, gradually changing to orangewith upward increasing greyish-green mottling,followed by an often sharp transition to greenishmudstone showing abundant yellow mottling(Fig. 3). The mudstone usually is massive andcommonly contains some randomly distributedgypsum aggregates with arbitrarily oriented

Fig. 2. View to the east of the siliciclastic Cascante section with the limestone unit on top of cycle number 34. Red/green mudstone–white limestone cycles are indicated by numbers in black (except for cycle 25, shown in greybecause it is not well visible in the picture). The scale bar is approximately valid for the middle of the photograph (foractual thicknesses, see Fig. 4).



well-developed crystal structure and few milli-metre-grained spherical gypsum nodules. Sparse,millimetre-sized to centimetre-sized carbonatelithoclasts are found locally floating in the mud-stone. Illite and smectite are the dominant clayminerals in these mudstones (Abdul Aziz et al.,2004). Freshwater gastropods, bivalves andsmall-mammal fossil remains are scarce and arespecifically concentrated in dark grey, organic-rich centimetre-thick beds (Broekman, 1983;Abdul Aziz et al., 2004). The organic matter ismillimetre-grained and originates from plantremains. The organic-rich layers are unconsoli-dated and occasionally show millimetre-scalelamination. These layers occur beneath, on topof, and locally within the limestone beds; theyalso contain fragmented charophyte remains andmillimetre-scale to centimetre-scale sphericalgypsum nodules.

InterpretationThe mudstone deposits are interpreted to haveaccumulated out of suspension following floodevents in distal mudflat areas of alluvial fans onan usually dry and exposed basin floor, inaccordance with the interpretation given byKiefer (1988). Lack of evidence for stream actionin the mudstone strata suggests that the supply of

mud occurred through flood events that rapidlywashed the mud into the basin without causingany apparent erosion. However, the massiveinternal structure of these deposits only displaysindications of root bioturbation by mottlingobsuring any sedimentary clues of a transportmechanism. The massive structure could havebeen caused by extensive mud-cracking andbioturbation because of root activity after deposi-tion. Wright & Marriott (2007) report that thesetwo mechanisms can, together with compaction,easily remove sedimentary structures after depo-sition. In addition to the removal of all primarysedimentary structures, prolonged sub-aerialexposure caused pervasive oxidation within themudstone. The green mudstone on top of the redmudstone might be indicative of initially redsediments that were subjected to reduction andleaching after deposition (Eardley et al., 1973;Broekman, 1983; Abdul Aziz et al., 2004). Similarreduction features have also been reported fromthe Upper Kizilirmak Basin in Turkey (Turkmen& Kerey, 2000) and from the Calatayud Basin inSpain (Abdul Aziz et al., 2003b).

As mentioned above, gypsum is present in themudstone as aggregates and millimetre-sizedspherical nodules, which occur especially in theorganic-rich layers just above and below the

Fig. 3. Basic mudstone–carbonate cycle recognized in the Cascante section with an indication of the main lithofaciesfeatures and their sedimentological interpretation.



limestone beds. The aggregates are interpreted assecondary deposits because of their well-devel-oped crystal structures, random orientation andoccurrence. The millimetre-sized spherical nod-ules, that are found mainly in the organic-richlayers, are interpreted as transported andre-deposited gypsum. An alternative origin isinterstitial nodular gypsum growth. The absenceof (other) clearly primary deposited subaqeousgypsum is remarkable. Warren (1999) reports thatdeposition of gypsum in continental settingsmainly depends on three critical factors: (i)abundant water to evaporate and introduce sol-utes; (ii) sufficient solutes to precipitate; and (iii)time. Taking into account the ample time formudstone deposition and the abundance of gyp-sum sources, i.e. Triassic formations, at the basinmargin (Broekman, 1983), the limiting factor forgypsum deposition in the Cascante region iswater. Thus, the lack of well-developed gypsumbeds could indicate that during mudstone depo-sition phreatic groundwater levels in the basinwere not high enough to reach the solute concen-tration necessary to precipitate gypsum (Rosen,1994). This observation suggests that the mud-stone deposits accumulated throughout pro-longed dry periods during which the basin wasflooded only occasionally. The rare small carbon-ate pebbles are interpreted as relicts of sheetfloods that transported larger clasts than average,derived from the Jurassic rocks in the basinmargin.

Limestone

DescriptionThe average thickness of limestone beds in thebasic cycle is 0Æ5 m (0Æ2 to 0Æ9 m). The limestonebeds are poorly laminated or massive, whereastheir top parts commonly show vertical tubes andhorizontal cracks (Fig. 3). Vertical tubes are a fewmillimetres in width and usually up to severalcentimetres in length, but they can reach up to20 cm in massive limestones. Individual tubesshow upward and downward divergent patterns,so that many tubes appear to be interconnected.Millimetre-thick horizontal cracks can be presentubiquitously, especially towards the top of thelimestone beds. The cracks and vertical tubeslocally produce a fragmented limestone, a featurethat is especially visible in the top part ofpronounced massive limestones. Fossil remainsof complete charophyte stems and gyrogonites,broken and complete gastropods, ostracods andsome undifferentiated green algae are found.

Rare, very fine sand-sized quartz grains are pres-ent. Gypsum is present usually as nodules or asfillings of horizontal cracks and vertical struc-tures (Abdul Aziz, 2001). Fine-grained, organic-rich laminae (see description in the Red andGreen Mudstone section) are present locallywithin the limestone beds. The degree of consol-idation of the limestone beds depends on theamount of clay, gypsum and organic matter, withhigh gypsum contents resulting in very con-solidated limestones. In the upper part of thesection, some limestone beds show undulatingstructures, with locally centimetre-scale cross-stratification, which are present mainly at thebase and in the lower half of the limestone beds.

InterpretationThe limestone beds are interpreted as lacustrinecarbonate deposits that accumulated in a shallowlake. The vertical tubes are interpreted as rootbioturbation from a plant cover that developed ontop of the limestone mainly after deposition. Thehorizontal cracks are interpreted as desiccationcracks that developed during short and/or repe-titive periods of subaerial exposure and finallythe desiccation of the lake. The root bioturbationand desiccation cracks lead to the formation ofpalustrine limestone. The cross-stratification andundulating character, mainly in the lower part ofthe limestone beds, are indicative of flowingwater. In contrast, the abundant presence ofcomplete charophyte stems is indicative of alow-energy environment and shallow water depth(Cohen & Thouin, 1987; Platt & Wright, 1991;Garcıa, 1994). The establishment of a shallow lakeon the distal alluvial plain of a fan system couldaccount for the reduced conditions leading togreen-coloured clayey sediment in the top part ofthe mudstone in a basic cycle (Eardley et al.,1973; Broekman, 1983; Abdul Aziz et al., 2004).Alternatively, the green colour could be the resultof reducing conditions during the deposition ofthe mudstone (Sagri et al., 1989). According toBroekman (1983), the fine-grained organic matteris of allochtonous origin and was deposited onthe lake shoreline during the transition fromfloodplain to lake depositional environment andvice versa. This interpretation is followed here,because the occasionally found horizontal lami-nations might indicate transportation of plantmaterial into the basin. The organic materialpossibly was derived from the vegetation coveralong the lake margin, which was washed into thebasin during transgressive and regressive periods.Wetter climate conditions may have caused



higher lake levels during limestone depositionand this may have enhanced vegetation growththat, in turn, reduced the inflow of clastic mate-rial further into the basin (Broekman, 1983; Platt& Wright, 1991).

SEDIMENTOLOGY – MICROFACIESANALYSIS

A detailed petrographic analysis of the limestonebeds was performed to reconstruct duration andstability of lake expansions and water depth.Outcrop-scale description of the carbonate faciesprovides a rough estimate of these lake variablesand the subsequent use of microfacies permitsanalysis of larger scale variations within thesedynamic lake features. Carbonate petrographywas carried out on a total of 145 thin sectionsfrom all 34 limestone beds in the Cascante section(Fig. 4). The limestone beds were sampled at aresolution of at least one thin section per 10 cm.To increase the visibility of carbonate particles,especially bioclast remains, the thin sectionswere partly stained using the red staining methodof Lindholm & Finkelman (1972).

The microfacies analysis resulted in a groupingbased on petrographic features and textures. Twomain microfacies groups and eight sub-groupswere differentiated (Table 1). The first groupincludes limestone beds showing microfaciestypical of a shallow permanent lake environment,whereas the second group includes carbonatedeposits showing features characteristic of a veryshallow, transient (sensu Currey, 1990) lakeenvironment. Currey (1990) defines transientlakes as having an inundation to desiccation timeratio of around 1:1, whereas permanent lakeshave a ratio of over 1000:1. Letter labels formicrofacies sub-groups were arbitrarily chosen.Thin sections that show characteristics inbetween two sub-groups were labelled as inter-mediate between the two sub-groups. Limestone

beds consequently were labelled according to themicrofacies code indicative of the deepest andmost permanent lake environment recordedwithin the bed.

Fig. 4. Lithologic column and magnetostratigraphy ofthe Cascante section. Grey shading in the polarityrecord represents uncertainty intervals between sam-ples with certain polarity. Grey shades in the litholog-ical column represent colour differences, see legend.Basic mudstone–carbonate cycles are numbered to theleft of the column. Location of limestone samples col-lected for thin sections is also indicated. The graph iscompleted by a* (red) and L* (blue) colour reflectancerecords and their 2Æ38 and 2Æ39 m filters (dotted lines)(modified from Abdul Aziz et al., 2004).



Table

1.

Sp

ecifi

cati

on

of

petr

ogra

ph

icd

esc

rip

tion

,m

atr

ix,an

dbio

cla

stch

ara

cte

rist

ics

of

the

eig

ht

carb

on

ate

mic

rofa

cie

ssu

b-g

rou

ps

recogn

ized

inth

eC

asc

an

tese

cti

on

.

Gen

eral

cha

ract

eris

tics

(cla

ssifi

catio

n ac

cord

ing

to D

unha

m,

1962

, and

Fol

k, 1

959)

Micrite Concentration

Lamination

Primary gypsum

Burrows/roots

Sorting

Diversity

Orientation

Fragmentation

Concentration

Net

wat

er b

udge

tdu

ring

dep

ositi

on(in

terp

reta

tion)

low

high

Group

Pack

ston

e, b

iom

icrit

e or

bio

mic

rudi

te.

Cha

roph

ytes

and

gyr

ogon

ites,

few

ga

stro

pods

, biv

alve

s and

thin

ostra

cods

.

Wac

kest

one

and

biom

icrit

e.B

iocl

asts

sam

e as

in A

.

Wac

kest

one

and

biom

icrit

e.G

astro

pods

, som

e ch

arop

hyte

stem

s,gy

rogo

nite

s and

som

e os

traco

ds.

Wac

kest

one

and

biom

icrit

e to

bi

omic

rudi

te.

Som

e qu

artz

gra

ins a

nd o

rgan

ic m

atte

r.

Mud

to w

acke

ston

e an

d bi

omic

rite.

Ost

raco

ds a

nd so

me

gast

ropo

ds a

nd

biva

lves

. Org

anic

mat

ter a

nd p

eloi

ds.

Intra

clas

t cha

ract

er. S

ome

quar

tz g

rain

s.

Pack

ston

e an

d bi

omic

rite.

Cha

roph

yte

stem

s, gy

rogo

nite

s, os

traco

ds

and

biva

lves

.

Wac

kest

one

and

biom

icrit

e.O

stra

cods

, gas

tropo

ds, c

haro

phyt

e st

ems,

gyro

goni

tes a

nd b

ival

ves.

Som

e cl

otte

d m

icrit

e +

fene

stra

l fea

ture

s.

Wac

kest

one

and

biom

icru

dite

. O

stra

cods

and

cha

roph

ytes

.C

lotte

d m

icrit

e an

d fe

nest

ral f

abric

s.

A B C D

Very Shallow transient lake microfacies

ES R

Shallow permanent lake microfacies

Q

++- -

++++

- -- -

- -++

- -

+/-

++

+/-

-- -

-

+/-

+/-

+/-

+/-

+/-

-

-++

- -- -

+++

++++

++

+-

- -+

+/-

++

++

+++/

-+/

-- -

++/

-+/

--

- -- -

- --

+++/

-+

++

+/-

- --

+/-

++

++/

-+

Sub-group

No

sign

ind

icate

sla

ck

of

specifi

cbeh

avio

ur,

+/)

ind

icate

savera

ge

pro

pert

yw

ith

resp

ect

toth

eoth

er

mic

rofa

cie

s,an

d+

())

stan

ds

for

hig

her

(low

er)

than

avera

ge.

To

the

righ

ta

sch

em

ati

cin

terp

reta

tion

isgiv

en

inte

rms

of

net

wate

ravail

abil

ity,

see

text

for

exp

lan

ati

on

.



Shallow permanent lake microfacies

DescriptionThe carbonate microfacies representative ofshallow, permanent lake conditions include

sub-groups Q, R and S (Figs 5A,B and 6A). Thesesub-groups show a dense packing of carbonatemicrite and bioclast fragments and, fromsub-group Q to R to S, an increase in sub-millimetre-scale lamination of micrite and

200 µ

m200 µ

m200 µ

m

200 µ

m200 µ

m200 µ

m

20

0 µ

m2

00

µm

20

0 µ

m

20

0 µ

m2

00

µm

20

0 µ

m

20

0 µ

m2

00

µm

20

0 µ

m

200 µ

m200 µ

m200 µ

mA B

D

C

E F

Fig. 5. (A) Example of permanent lake microfacies sub-group R showing lamination of micrite and broken remains ofcharophyte stems and gyrogonites that were transported to (slightly) deeper lake environments, normal polarizedlight. (B) Example of permanent lake microfacies sub-group Q showing lamination of dense micrite and both brokenand complete charophyte remains indicative for only short transportation, red stained section, normal polarizedlight, scale as on other photomicrographs. (C) Example of transient lake microfacies sub-group A a highconcentration of charophyte stems that shows a slight orientation, some gastropod remnants are present, the com-pleteness of delicate charophyte remains indicative for low-energy environments without much transportation beforedeposition, section red stained, normal polarized light. (D) Example of transient lake microfacies sub-group Cshowing a relatively high concentration of broken and complete bioclast remains of mainly gastropods, red stainedsection, normal polarized light. (E) Example of transient lake microfacies sub-group D showing broken gastropod andostracod shell remains, (primary) lenticular gypsum occurs as filling of horizontal desiccation cracks, red stainedsection, cross-polarized light. (F) Example of transient lake microfacies sub-group E showing microbreccia with hostrock and filling of cracks with micrite and gypsum, palustrine limestone, upper half red-stained, cross-polarizedlight.



bioclast fragments. Bioclast diversity and frag-mentation, as well as packing of the bioclastfragments, increase from sub-group Q to S (‘con-centration’ in Table 1). Bioclasts consist mainlyof fragmented charophyte stems and gyrogonitesand minor ostracod and gastropod shells, whichare present chiefly in sub-group Q (Fig. 5B). Noneof the three sub-groups contains siliciclastic mate-rial coarser than the mud fraction, whereas claymaterial is present in minor amounts. No cracksand burrows have been observed, nor gypsumcrystals. On outcrop scale, the limestone beds ofthis microfacies are indurated and have a well-defined base and top. Beds are internally lami-nated to massive; small-scale cross-stratification

and undulating structures are common, especiallyin the basal part of the beds. Bed thickness rangesfrom 0Æ4 to 0Æ9 m.

InterpretationThe dense packing of micrite and bioclast frag-ments and the sub-millimetre-scale laminationsare interpreted as the result of deposition inrelatively low-energy water environments. Len-ticular gypsum, that is interpreted as primarygypsum by Anadon et al. (1992), is presentabundantly in the very shallow transient micro-facies. The absence of cracks, burrows and lenti-cular gypsum in the shallow permanent lakemicrofacies indicates stable lacustrine conditions

200 µm200 µm200 µm200 µm200 µm200 µm

200 µm200 µm200 µm200 µm200 µm200 µm

A B

C D

Fig. 6. (A) Example of permanent lake microfacies sub-group S lamination of micrite and broken remains ofcharophyte stems and gyrogonites transported to (slightly) deeper lake environments, normal polarized light. (B)Example of transient lake microfacies sub-group B showing abundant and mostly complete charophyte remainsindicative of low energy conditions, no preferred orientation of micrite or bioclasts, some signs of sub-aerial exposureand/or root action in upper part of photograph, section red stained, normal polarized light. (C) Example of transientlake microfacies sub-group C/D showing gastropod and unidentified shell remains at intermediate densities withoutclear orientation and organic matter, section red stained, normal polarized light. (D) Root burrow or striotubule filledwith lenticular gypsum, host rock around burrow shows primary lenticular gypsum as well, not red stained, normalpolarized light.



that prevailed for relatively long time periods(Currey, 1990). The abundance of fragmentedcharophyte stems and gyrogonites suggests thatthey were transported over a short distance fromvery shallow lake areas, where charophytesflourished, towards slightly deeper waters (Platt,1989; Perez et al., 2002). Platt (1989) and Perezet al. (2002) state that water depth in theseenvironments probably does not exceed a fewmetres. Stable lacustrine conditions prevail whenevaporation and groundwater discharge from thelake system are lower than precipitation andgroundwater influx. These circumstances may berelated to relatively continuous water inflow,together with low evaporation, and/or lack ofextremely dry climatic conditions for a certainperiod of time. The increase in horizontallyoriented bioclasts, clast fragmentation, andmicrite and bioclast lamination from sub-groupQ to S is interpreted as slightly higher energy anddeeper conditions towards sub-group S. Nomicrofacies are observed that indicate the deeperand less energetic settings that are expected whenlake levels rise even more; this means that in thecase of Cascante the ‘deepest’ observed lake isalso the relatively highest energy environmentprobably indicating a shallow lake environmentwith a flat bottom.

Very shallow transient lake microfacies

DescriptionLimestones of this microfacies consist of massive,uniform carbonate micrite lacking lamination(microfacies A to E) and scarcely containing veryfine sand-sized quartz grains (see also Freytet &Verrecchia, 2002). The distinction between sub-group Q (a microfacies typical of deposition inshallow permanent lake conditions; Fig. 5B) andsub-group A (Fig. 5C) is based primarily on theabsence of lamination and reduced packing ofmicrite in sub-group A. From sub-group A to E, adecrease in diversity and packing of bioclasts isobserved (Figs 5C to F and 6B,C; Table 1). Charo-phyte stems and gyrogonites prevail in sub-groups A and B (Figs 5C and 6B), while manycomplete and some broken ostracod and gastro-pod shells are more common in sub-groups D andE (Fig. 5E and F). In sub-group C (Fig. 5D), allfour types of, dominantly broken, bioclasts arepresent, although they are often not well-mixed.An increase in the amount of sub-millimetre-sized particles of organic matter, most probablyderived from plant material, can be observedfrom sub-group C to D (Figs 5D, E and 6C).

Sub-millimetre-sized lenticular gypsum crystalsoccur often floating in the carbonate micrite, withtheir presence increasing from sub-group B to E.

Limestone beds showing this microfacies groupusually display mainly horizontal and somevertical cracks (Fig. 5E and F), and vertical bur-row (striotubules) structures (Fig. 6D). The bur-rows become significantly more abundant fromsub-group A to E. Horizontal cracking and bio-turbation features are virtually absent in sub-group A, while in sub-group E brecciation isintense (Fig. 5F). Cracks commonly are filled witha mosaic of lens-shaped sub-millimetre-scalegypsum crystals but, in some cases, they arefilled partially with silt-sized to clay-sized sili-ciclastic material or fragmented host carbonaterock, the latter especially in desiccation breccia(sensu Normati & Salomon, 1989) of sub-group E.The sub-groups B, C and D show gradationalmicrofacies features between the two end-mem-ber sub-groups A and E.

In outcrop, the carbonate beds displaying veryshallow transient lake microfacies usually aremassive and/or poorly laminated and frequentlyintercalate with thin green clay and dark greyorganic-rich mudstone beds that have a thicknessbetween 1 and 10 cm.

InterpretationThe lack of lamination in all sub-groups, minoramounts of gypsum and abundant well-preservedcharophyte cysts and stems, and gyrogonites inthis microfacies indicates a very shallow, low-energy freshwater lake environment. The delicatecharophyte stems grow in shallow lake marginsand usually break apart when agitated and trans-ported, as observed in modern lake environments,where meadows of charophytes flourish at depthsof less than 5 m (Platt, 1989; Garcıa, 1994; Perezet al., 2002).

The decrease in abundance and diversity ofbioclasts and the increased organic matter andamount of root burrows in the limestone, asobserved from sub-group A to E, indicate agradual change from a low energy, shallow lakeenvironment (sub-group A; Fig. 5C) towards ashallow lake environment that is more swampy.This interpretation is supported by the increasein lenticular gypsum floating in the carbonatemicrite from sub-group A to E. Anadon et al.(1992) relate similar lenticular gypsum crystals toa primary evaporate phase in extremely shallowlake environments, even though Albesa et al.(1997) report that the salinity is not necessarilyhigh. The higher amount of horizontal cracks, that



are interpreted as desiccation cracks, and rootburrows (striotubules) observed from sub-group Ato E, suggests an increase in frequency andduration of sub-aerial exposure (sub-group E inFig. 5F). Alonso Zarza (2003) states that theabsolute exposure time and frequency needed toform palustrine features in freshwater lake car-bonate deposits are not known and may berelatively short, even a season. The primaryexposure features can form under fluctuating lakewater level, which is accompanied by depositionof organic plant material and clay laminae inter-calated in the limestone bed. In terms of inunda-tion versus exposure time, carbonate microfaciessub-groups A and B characterize a ‘persistent’lake environment, suggesting an inundation ver-sus desiccation ratio of 10 to 1000:1 (Currey,1990), while microfacies sub-groups C to E rep-resent a ‘transient’ lake environment, with a ratioof 1 to 10:1. Limestone beds of the very shallowtransient lake microfacies sub-groups generallyshow their relatively deepest lake phase justabove the middle of the bed.

PALAEOENVIRONMENTAL CYCLES

Sedimentary model for the basic cycle

The remarkable lateral continuity of lithologicfeatures of both mudstone and limestone beds inthe Cascante succession suggests that they weredeposited in a low-energy, endorheic ramp-typebasin. In this setting, small-scale lake-level fluc-tuations resulted in the submergence or desicca-tion of large areas of the basin (Platt & Wright,1991). The description and interpretation of thebasic mudstone–limestone sedimentary cycle areschematically depicted in Fig. 3. It should benoted that the mudstone part of the cycle is onaverage three to four times thicker than thelimestone interval. Within the basic cycle, redmudstone is interpreted as a low-energy sheetflow deposit that underwent sub-aerial exposureafter deposition in a distal alluvial plain setting.The transition from a red to a greenish colourtowards the top of the mudstone deposit mostprobably is caused by reduction and leaching ofthe underlying muddy sediment when lake levels

rose following mudstone deposition. Organic-richlayers below and on top of the lacustrine lime-stones are interpreted as deposited during thelake transgression and regression. The freshwaterlimestone is associated with lake-level highstandsresulting in a shallow lake undergoing varyingscales of fluctuating lake levels. In some cases,low lake levels resulted in episodic exposure ofthe carbonate mud and the formation of palus-trine features. Lack of significant siliciclasticinput into the carbonate lake can be related tohigh lake levels, resulting in a rise of the baselevel of erosion trapping the sediment in higherareas (Picard & High, 1981), and/or to the devel-opment of a dense vegetation cover along the lakemargin that kept sediments in place and pro-moted chemical instead of physical weathering(Platt & Wright, 1991).

Larger-scale environmental variations

The codes of the different microfacies sub-groupsfor the 34 limestone beds in the Cascante sectionhave been plotted in Fig. 7. Carbonate microfaciescharacteristics of very shallow transient lakeenvironments prevail in the lower part of thesection below cycle 26. Two intervals with threebasic cycles (10 to 12 and 15 to 17) and oneinterval with two cycles (20 and 21) containlimestones representative of more ‘persistent’ (seeabove) lake environments (sub-groups A and B).Except for cycle 4, all other limestones in thelower part of the section show more ‘transient’lake environments (i.e. palustrine limestones ofsub-groups C, D and E) representing shallowerand more variable lake conditions. Below cycle26, no limestone bed with microfacies indicativeof shallow permanent lake conditions is present.Above that cycle, the limestone beds comprisemicrofacies sub-groups that indicate the morepersistent environments of the very shallow tran-sient lake microfacies (sub-groups A and B) andthe permanent environments of the shallow per-manent lake microfacies (sub-groups Q, R and S),except for cycle 27. Thus, lake levels increase andlake environments become more stable from cycle26 onwards, probably indicating the transition tothe thick lacustrine limestone unit that overliesthe studied section. This uppermost interval is

Fig. 7. Microfacies analysis results of all limestone beds in the Cascante section and their correlation to successiveprecession minima and insolation maxima according to the magnetostratigraphy of Husing et al. (2007). Grey in thepolarity record indicates age uncertainty intervals as given by these authors. Colours and different line characteristicsindicate microfacies sub-groups (see legend). Asterisks indicate thick mudstone intervals. For details on the litho-logical column see Fig. 4.





characterized by alternation of relatively deeperand shallower lakes in every other bed. Thispattern shifts one cycle from 28 to 29. Somewhatdeeper lakes are found in cycles 26, 28, 29, 31 and33, whereas shallower lakes are found in cycles27, 30, 32 and 34.

CYCLOSTRATIGRAPHY

Astronomical forcing

The stratigraphic regularity of the basic cycle inthe Cascante section strongly suggests an allo-cyclic forcing mechanism. Abdul Aziz et al.(2004) therefore applied Blackman–Tukey (BT)and maximum entropy (ME) spectral analysis oncolour reflectance records (Fig. 4) in the depthdomain. The dominant peak in the resultingpower spectra could be related to the basic,2Æ2 m cycle. Additional power was found at lowerfrequencies, with a thickness ratio of about 1:2:5.This ratio closely corresponds to the astronomicalperiod ratio of climatic precession, obliquity andshort (ca 100 kyr) eccentricity (Abdul Aziz et al.,2004). Detailed age control in the section wasderived from the magnetostratigraphy (Fig. 4)calibrated to the geomagnetic polarity timescaleCK95 using mammal biostratigraphic data (AbdulAziz et al., 2004). Application of this age modelrevealed a period of 20 to 26 kyr for the basiccycle. This period is well within the range of theduration of the climatic precession cycle. Theother peaks identified by the spectral analysisindicate a period of ca 41 kyr for the cycle with athickness twice that of the basic cycle and around100 to 150 kyr for the cycle with a thickness offive times the basic cycles. These cycles aretentatively related to obliquity and the ca100 kyr period of the short eccentricity cycle.Abdul Aziz et al. (2004) concluded that theseresults strongly indicate that the sedimentarycyclicity in the Cascante section is related toastronomical forcing. The use of the recentlyimproved astronomical tuned timescale (Lourenset al., 2004) and the astronomical ages resultingfrom the tuning of the deep marine Monte deiCorvi section (Husing et al., 2007) does notchange the outcome of the statistical analysis.

The potential astronomical control requiresfurther exploration. In order to elaborate on this,the sedimentary cycles need to be interpreted interms of astronomical climate forcing. Once thephase relationships have been established, anastronomical tuning of the sedimentary cycles by

correlation of individual cycles to calculatedastronomical target curves can be constructedstarting from the magnetostratigraphic age cali-bration. Finally, the environmental interpreta-tions derived from the carbonate microfaciesanalysis can be compared with the target curvesand the working hypothesis can be examined.This cyclostratigraphic exploration does notimply that alternative interpretations of thecyclicity are excluded. However, the data stronglysuggest orbital forcing and this option has beenelaborated to (finally) prove astronomical forcingusing an integrated approach.

Climatic interpretation of the sedimentarycycles

The alternation of shallow lake and distal allu-vial plain facies indicates constant shifts at104 year timescales between an annual or sea-sonal net positive and negative water budget,when the basic cycles are interpreted as forcedby climate variations. From the sedimentologicaldata alone, it is impossible to deduce whetherthese variations were caused by changes inprecipitation, evaporation, seasonality, rain sea-son and/or possibly other climatic factors. Hori-zontal desiccation cracks and root traces, whichincrease towards the top of the carbonate beds,indicate that lake levels often were variable andhigh lake levels only prevailed for a short periodof time. Lake-level fluctuations resulted in rapiddrowning or desiccation of the depositional site.The weak development of palustrine featuresindicates that sedimentation resumed not longafter the ending of lake conditions, as palustrinefeatures can form rapidly in ca 101 to 103 years(Alonso Zarza, 2003). The amount of time repre-sented by the limestone versus mudstone part ofthe basic cycle is difficult to specify. The lack ofwell-developed pedogenic features in the mud-stones suggests that sedimentation kept pacewith soil formation processes within the mud-stones (Buurman, 1980). Sedimentation wasapparently relatively continuous during boththe deposition of the mudstone and limestonepart of the basic cycle. However, major differ-ences in the sedimentation rate between the twoend-member lithologies may exist. In conclusion,the observed regular and repetitive developmentof full lacustrine conditions in a prevailing sub-aerially exposed distal floodplain requires acyclic forcing mechanism to yield a positivewater budget during specific and relatively shorttime periods.



Phase relationships – geological data

Correlation of individual mudstone and carbon-ate cycles with calculated astronomical targetcurves requires a knowledge of the phase rela-tionship of the two basic cycle end-memberlithologies with the orbital parameters. Conse-quently, the climatic influence in terms of netwater budget of orbital extremes and, additio-nally, the climatic conditions required for develo-pment of the depositional environments ofmudstone and limestone have to be unravelled.This rationale is termed phase relationshipbetween lithology and orbital parameters.

In the northern hemisphere, climatic preces-sion minima (summers in perihelion) cause highsummer and low winter insolation at the top ofthe atmosphere resulting in enhanced seasonality.Numerous studies have demonstrated the influ-ence of precession on circum-Mediterraneanclimate as recorded by Neogene marine (Rossig-nol-Strick, 1983; Hilgen et al., 1999; Foucault &Melieres, 2000) and continental sediments(Steenbrink et al., 1999; Abdul Aziz et al., 2000;Magri & Tzedakis, 2000). Increased clay contentsin cyclic Pliocene marl sequences depositedoffshore Gulf of Cadiz suggest higher annualrunoff from a southern Spanish drainage area attimes of precession minima (Sierro et al., 2000).Similarly, indications for humid conditions dur-ing precession minima have been implied formarl sequences of the Abad Formation in theSorbas Basin in southern Spain (Sierro et al.,1999; Vazquez et al., 2000). Sapropel formation inthe eastern and western Mediterranean basins hasbeen related to increased winter precipitationfrom southern, and possibly northern, Mediterra-nean borderlands during precession minima(Lourens, 2004). Magri & Tzedakis (2000)reported temperate-stage forest expansions dur-ing boreal summer insolation maxima (duringprecession minima) for the last 200 kyr in centralItaly, which they relate to higher annual netprecipitation. Maximum forest coverage occursduring autumn insolation maxima, which Magri& Tzedakis linked to the absence of extremesummer evaporation and, consequently, to max-imum net water budget. On the other hand, asignificant contraction of the forest coverage isfound during winter insolation maxima (i.e.summer insolation minima during precessionmaxima) and March insolation maxima.

Reduced clastic supply that resulted in marl-rich intervals in the Pliocene Gulf of Cadiz isobserved during short and long eccentricity min-

ima (Sierro et al., 2000). Moreover, intervalsdevoid of sapropels in the Mediterranean occurduring eccentricity minima and are related tolower precipitation in the southern, and possiblynorthern, borderlands (Lourens, 2004). Short (ca100 kyr) and long (405 kyr) eccentricity minimamost probably result in prolonged periods ofrelatively dry climate conditions that can lead tolake-level lowstands in the Spanish mainland.

Phase relationships – climate modelling

In order to determine the impact of climaticprecession extremes on climate in Spain, modelsimulations were performed by using an atmo-spheric global climate model of intermediatecomplexity, SPEEDY, coupled to a slab oceanand a simple land model. SPEEDY is a primitiveequation model with simplified physics parame-terizations (Molteni, 2003) that has a horizontalresolution of about 3Æ75� · 3Æ75� (T30) and sevenvertical levels (Hazeleger et al., 2003). The landmodel consists of a simple bucket model, whichimplies that the soil moisture cannot exceed15 cm per grid box. Redundant soil moisture isdefined as runoff. Three 100 year simulationsusing the SPEEDY climate model were carriedout: (i) a control run with present-day orbitalparameters; (ii) a minimum precession run(eccentricity = 0Æ056, precession = )0Æ055); and(iii) a maximum precession run (eccentricity =0Æ058, precession = 0Æ058). A precession maxi-mum (minimum) means winter (summer) solsticein perihelion. Precession is defined as esin(p + x) with e the eccentricity of the orbit ofthe Earth and x the angle between the vernalequinox and perihelion (measured counter-clock-wise). The used values for precession are theextreme values occurring in the last one millionyears (Berger, 1978). For the precession simula-tions, obliquity was kept fixed at 22Æ08, i.e. theminimum value for the last one million years. Afixed obliquity was used to focus on the preces-sion signal without the possible interference ofobliquity. For all three runs, boundary conditionssuch as orography, concentration of trace gases,vegetation and ice sheets were kept to present-dayvalues. A more extensive description of theexperimental set-up can be found in Tuenteret al. (2003). The model results are shown asaverages over the last 10 years of 100 year simu-lations. Ten years are used to average out annualvariability, while a longer averaging period is notused to ensure that the model is in equilibriumafter 90 years.



The model shows that during winter, netprecipitation (precipitation minus evaporation)in precession minimum is higher than in preces-sion maximum (Fig. 8C) while, for summer, thereis not much difference between the precessionextremes. The large difference in net winterprecipitation is caused by stronger precipitationduring minimum precession, whereas winterevaporation is equal for both minimum andmaximum precession (Fig. 8B). In the model,the reason for enhanced winter precipitationduring precession minima is higher Mediterra-nean sea surface temperatures both in summerand in winter. These higher temperatures are dueto increased summer insolation which, in thewinter, results in higher evaporation and forma-tion of clouds over the Mediterranean. Insummer, higher precipitation occurs duringmaximum precession (Fig. 8A), although summer

net precipitation is similar for both precessionextremes (Fig. 8C) because of an increase inevaporation (Fig. 8B). Annually, higher net pre-cipitation occurs during minimum precession(Table 2).

In the model, the annual net precipitation isdominated by winter precipitation for both mini-mum and maximum precession, suggesting thatwinter precipitation plays an important role inrunoff and groundwater supply that togetherdetermine lake levels. Runoff acts accordinglywith higher winter and annual runoff duringprecession minima (Fig. 8D; Table 2). Therefore,it is probable that high lake levels and carbonatedeposition in the Cascante section occurred dur-ing precession minima.

The phase relationship of high lake levelsduring precession minima derived from theSPEEDY model are in line with the sedimento-

A B

C D

Fig. 8. (A) to (D) Monthly areaaveraged values for ‘Spain’ (10Æ5 to2� W; 37 to 44Æ5� N, four gridboxes)of: (A) precipitation; (B) evapora-tion; (C) precipitation minus evap-oration; and (D) runoff for thecontrol run (dotted black line),minimum (climatic) precession run(solid line) and the maximum pre-cession run (dashed line). All val-ues are in mm day)1 and averages ofthe last 10 years of 100 year simu-lations (see text for details).

Table 2. Annually averaged valuesfor ‘Spain’ (10Æ5 to 2� W; 37 to44Æ5� N, four gridboxes) of precipi-tation, evaporation, precipitationminus evaporation (net precipita-tion), and runoff for the controlrun, minimum precession, andmaximum precession.

Simulatedannual values

Precipitation(mm day)1)

Evaporation(mm day)1)

Precip.–Evap.(mm day)1)

Runoff(mm day)1)

Control run 2Æ22 1Æ61 0Æ61 0Æ66Precession min. 2Æ47 1Æ47 1Æ00 1Æ10Precession max. 2Æ37 1Æ88 0Æ49 0Æ54

All values are averages of the last 10 years of 100 year simulations.



logical findings in the Gulf of Cadiz (Sierro et al.,2000) and the Sorbas Basin (Vazquez et al., 2000)as well as those from sapropelic layers in theMediterranean (Lourens, 2004). Moreover, theresults agree with forest expansions during pre-cession minima (Magri & Tzedakis, 2000). Con-sidering the geological and modelling data, thephase relationship between high lake level (andlimestone formation) and enhanced net winterprecipitation during precession minima isfavoured.

Astronomical Tuning

An astronomical tuning is constructed for theCascante section using the astronomical ages forthe reversal boundaries of Husing et al. (2007)and the inferred phase relationships between thecarbonate beds to minimum precession andmudstone beds to maximum precession, as dis-cussed above. The astronomical tuning is estab-lished by correlating the basic cycles to theprecession target curve of Laskar et al. (2004)starting from all the reversal boundaries upwardsand downwards (Fig. 7). Except for the base ofchron C5n.1r, which is ca 10 kyr younger inCascante, the tuning produces similar astronom-ical ages for chron boundaries as calculated byHusing et al. (2007). This observation impliesthat exactly the same number of precession-related cycles is present in the Cascante sectionas in the time-equivalent interval in the Montedei Corvi section. This outcome confirms theastronomical forcing and dominant precessioncontrol of the basic cyclicity at Cascante.

As described above, the microfacies analysisrevealed larger scale lake-level variations com-prising several basic cycles. Lowstands therein,with beds showing features indicative of theshallowest transient lake environment (micro-facies sub-groups C, D and E; cycles 1 to 3, 5 to 6,7 to 9, 13 to 14, 18 to 19 and 22 to 25), correlate tominimum eccentricity related to the ca 100 kyrcycle (Fig. 7). Highstands with beds that formedin the relatively deeper transient lake environ-ment (microfacies sub-groups A and B; cycles 4,10 to 12, 15 to 17 and 20 to 21) were depositedduring maxima of the short eccentricity cycle.The prolonged lowstands (cycles 1 to 9 and 22 to25) correspond to long (405 kyr) eccentricityminima and prolonged highstands (cycles 10 to21 and 26 to 34) to long eccentricity maxima(Fig. 7). This link between lake level, as derivedfrom microfacies analysis, and eccentricity indi-cates that the eccentricity modulation of the

precession amplitude played an important rolein determining longer-term variations in lakelevel. The observed lake-level lowstands duringshort and long eccentricity minima are in linewith the modulating effect of eccentricity andwith Mediterranean marine geological data(Sierro et al., 2000; Lourens, 2004).

In the interval above cycle 25, successivelimestone beds display an alternation of deeperand shallower lake carbonate deposits (Fig. 7).Moreover, the mudstone beds reveal a regularalternation in this interval; successive mudstonesshow a thick and thin alternation while redcolouring is present in the thicker beds (shownwith an ‘asterisk’ in Fig. 7) and absent in thethinner beds (cycles 27 to 31 in Fig. 7). Thesekinds of alternations in basic precession con-trolled cycles are found in the marine realm aswell, where thick pronounced and thin, vaguesapropels alternate (Husing et al., 2007). Suchalternating patterns that occur superimposed on abasic precession-forced cyclicity are called pre-cession–obliquity interference patterns. Thesepatterns originate from interference of these twoorbital parameters, which typically are foundduring times of low eccentricity when the pre-cession amplitude is reduced. In particular, thesepatterns occur at times of minimum eccentricityrelated to the long-period 2Æ4 Myr eccentricitycycle (Abdul Aziz et al., 2003a; Laskar et al.,2004; Husing et al., 2007). Precession–obliquityinterference patterns have also been recognized incontinental sediments (Abdul Aziz et al., 2003a).The interference pattern in the Cascante section(cycles 26 to 34) indeed occurs during a minimumof the 2Æ4 Myr eccentricity cycle. This cyclereaches its absolute minimum at 9Æ52 Ma, whichis the minimum in the 2Æ4 Myr bandpass filter ofthe eccentricity time series (Laskar et al., 2004).

In the marine realm, the interference patterns insapropel successions fit with those in the insola-tion target curves (Husing et al., 2007). Such a fitcan be used to confirm the astronomical tuning inthese sensitive deep marine settings. In theCascante section, the details in the interferencepattern also fit with the combined precession–obliquity target curve (P–0.5T in Fig. 7). Deeper,more permanent lakes occur during extremesummer insolation maxima (P–0.5T minima)and thicker, more intense red mudstones duringextreme summer insolation minima, whichresults from the additional influence of obliquitymaxima and minima, respectively. This fit isconsistent with the astronomical tuning of theCascante section. In the lower part of the section,



interference patterns also are present, especiallybelow cycle 9 (see Fig. 7). However, this is notfurther discussed in detail because the patternsare much less pronounced in this interval and anindependent quantitative record is needed toprove the correlation. Nevertheless, there is agood fit with the interference patterns in P–0.5Tin this interval as well.

In conclusion, the astronomical tuning of theCascante cycles to the precession time series doesnot reveal any significant difference in the astro-nomical age of the magnetic reversal boundariesas compared with the astronomically tunedsapropel succession at Monte dei Corvi (Husinget al., 2007). The larger scale lake-level variations,revealed by microfacies analysis, fit well with theeccentricity target curve as expected from Medi-terranean marine geological data. Additionally,precession–obliquity interference patterns in theupper part of the section reveal a good fit with theinsolation target curve. These results confirmthe astronomical forcing working hypothesisof the sedimentary cycles in the continentalCascante section.

DISCUSSION

The integrated cyclostratigraphic study was car-ried out to explore further the astronomical forcingof sedimentary cyclicity in the Cascante section, aspreviously suggested by the results of spectralanalysis of colour records (Abdul Aziz et al., 2004).All data are now consistent with the astronomicalforcing hypothesis, indicating that astronomicalforcing is indeed the most probable mechanism.

Outcrop sedimentology, microfacies analysis,magnetostratigraphy and modelling results showthat (all) palaeoenvironmental variations at 104 to105 year timescales could be related to astronom-ically forced climate variations. The basic cycle isforced by the climatic precession cycle, whereaslonger-term lake-level variations are related to theshort and long eccentricity cycles. Obliquityinfluence is observed during periods of loweccentricity.

The dominance of astronomically forced envir-onmental variations throughout the Cascantesection implies that orbitally induced changesin insolation had a prominent imprint on climatein Spain during the Late Miocene. These varia-tions are recorded, especially in low-gradientclosed basin settings such as Cascante. Lake-levelfluctuations could have been forced by variousmechanisms, mainly tectonics, climate and

autogenic processes. Based on the combinedcyclostratigraphy and microfacies study of theCascante section it is argued that tectonic, auto-genic and local climate processes had no signi-ficant effect on the Cascante palaeoenvironmenton 104 to 105 year timescales. From a tectonicpoint of view, basin subsidence and/or marginuplift apparently occurred at constant rates or atleast the difference between the two factorsremained similar on the 104 to 105 year time-scales, such that recorded palaeoenvironmentalchanges at these timescales are exclusivelyrelated to orbital climate variations.

Gradual change

Different scales of lithological cyclicity in theCascante section are ascribed to astronomicalforcing. In theory, these astronomical signals canbe subtracted from the record in order to examinewhether other unexplained palaeoenvironmentalvariations exist in the Cascante section; however,this can strictly be done only by using quantita-tive records. The present proxies are not indica-tive of any variations on timescales of 104 to105 years, that are not related to astronomicalforcing. On longer timescales, a gradual shift fromdominant transient to mixed transient–perma-nent lake environments is observed after cycle 25(Fig. 7) culminating in the fully lacustrine lime-stone unit on top of cycle 34 (Fig. 2). Clearly, thisshift cannot be explained by the astronomicalcycles discussed before. The causes for this majorshift can be tectonics, non-astronomical climate,astronomical climate forcing because of cycleswith a longer period, or autogenic processes.

Million-year timescale changes, such as theshift to deeper and more permanent lake envi-ronments after cycle 25 in the Cascante section,commonly are related to tectonic processes. In thelacustrine record of the Triassic Newark Basin,large-scale lithological variations are interpretedto reflect long-period astronomical climate varia-tions (Olsen & Kent, 1996). However, a decisiveconclusion about the potential long-period orbitalclimate forcing in the succession studied byOlsen & Kent (1996) cannot be made because ofinsufficient absolute age control and lack of anastronomical target curve for the Triassic. In theNeogene, eccentricity and obliquity compriselong-period cycles of ca 2Æ4 Myr and ca 1Æ2 Myr,respectively (Laskar et al., 2004). Recently, theinfluence of these quasi-cycles on temperatureand precipitation of climate of the Iberian Penin-sula has been suggested by their impact on



mammal turnover rates (Van Dam et al., 2006). Inparticular, long-period eccentricity minima, thatare successively 2Æ0 and 2Æ8 Myr apart, resulted inprolonged absence of extreme summer conditionsleading to an increase in both wet-adapted small-mammal lineages and total diversity (Van Damet al., 2006). In Cascante, the shift above cycle 25from transient lake environments in the wettestphase of the basic cycle to a mix of transient andpermanent lake environments occurs at an eccen-tricity minimum of ca 2Æ4 Myr (Fig. 7), whereasthe environmental effect is consistent with theresults of Van Dam et al. (2006). This coincidenceis tempting, although proof for such a link shouldcome from the accurate dating of a number ofsuch large-scale transitions, ideally in differentbasins. Note that the phase relationship witheccentricity would then be different for theca 2Æ4 Myr cycle than for the much better knownshort and long eccentricity cycles.

CONCLUSIONS

In the Cascante section, high-resolution magneto-stratigraphic age control, together with carbonatepetrography and outcrop sedimentology, as envi-ronmental proxies, reveals that all palaeoenviron-mental signals at 104 and 105 year timescales arerelated to astronomical climate forcing. Model-ling results show, consistent with geological data,that lake-level highstands are related to higher netwinter precipitation during precession minima.The robust astronomical tuning reveals that thelonger-term lake-level highstands, as revealed bythe microfacies analysis, occur during ca 100 kyrand 405 kyr eccentricity maxima, as a conse-quence of the modulation of the precessionamplitude. Subtracting the astronomical signalsfrom the palaeoenvironmental variation in theCascante section revealed a shift towards deeperand more permanent lake environments at timesof precession minima in the upper part of thesection. This shift, which culminates in a thicklacustrine limestone unit on top of the studiedsection, remains unexplained, but may be trig-gered by tectonics, non-astronomical climatechange, astronomical climate forcing because ofvery long period cycles, and autogenic processes.

In summary, the results illustrate that cyclo-stratigraphy is a powerful tool to study palaeo-environmental variations in continentalsuccessions at timescales greater than 103 years.Astronomical climate forced variability can beconstrained very well, if sufficient age control and

palaeoenvironmental proxies are present. Such adetailed study can be especially useful in LateNeogene successions, for which the geomagnetictimescale has been improved considerably by theincorporation of astronomically tuned reversalages (Lourens et al., 2004; Husing et al., 2007).

ACKNOWLEDGEMENTS

Jesus Sanchez-Corral from the Universidad Com-plutense de Madrid and Otto Stiekema fromUtrecht University are thanked for thin sectionpreparation, the Instituto Aragones de GestionAmbiental (INAGA) for sampling licenses, and DrAnne van der Weerd for field assistance. HAacknowledges the financial support of the DutchScience Foundation (NWO-ALW). Dr Frits Hilgenis thanked for useful discussions and improve-ments on earlier versions of the manuscript. Prof.Ildefonso Armenteros, Dr Miguel Garces andDr Wan Yang are acknowledged for their con-structive reviews.

REFERENCES

Abdul Aziz, H. (2001) Astronomical Forcing in ContinentalSediments. PhD thesis, Utrecht University, Utrecht, the

Netherlands.

Abdul Aziz, H., Hilgen, F.J., Krijgsman, W., Sanz, E. and

Calvo, J.P. (2000) Astronomical forcing of sedimentary

cycles in the middle to late Miocene continental Calatayud

Basin (NE Spain). Earth Planet. Sci. Lett., 177, 9–22.

Abdul Aziz, H., Krijgsman, W., Hilgen, F.J., Wilson, D.S. and

Calvo, J.P. (2003a) An astronomical polarity timescale for

the late Middle Miocene based on cyclic continental

sequences. J. Geophys. Res., 108, B3, 2159. doi:10.1029/

2002JB001818.

Abdul Aziz, H., Sanz-Rubio, E., Calvo, J.P., Hilgen, F.J. and

Krijgsman, W. (2003b) Palaeoenvironmental reconstruction

of a middle Miocene alluvial fan to cyclic shallow lacustrine

depositional system in the Calatayud Basin (NE Spain).

Sedimentology, 50, 211–236.

Abdul Aziz, H., Van Dam, J., Hilgen, F.J. and Krijgsman, W.(2004) Astronomical forcing in Upper Miocene continental

sequences: implications for the Geomagnetic Polarity Time

Scale. Earth Planet. Sci. Lett., 222, 243–258.

Albesa, J., Calvo, J.P., Alcala, L. and Alonso Zarza, A.M.(1997) Interpretacion paleoambiental del yacimiento de La

Gloria 4 (Plioceno, Fosa de Teruel) a partir del analisis

de facies y de asociaciones de gasteropodos y de mamıferos.

Cuad. Geol. Iberica, 22, 239–264.

Alonso Zarza, A.M. (2003) Palaeoenvironmental significance

of palustrine carbonates and calcretes in the geological

record. Earth-Sci. Rev., 60, 261–298.

Alonso Zarza, A.M. and Calvo, J.P. (2000) Palustrine sedi-

mentation in an episodically subsiding basin: the Miocene

of the northern Teruel Graben (Spain). Palaeogeogr. Palae-

oclimatol. Palaeoecol., 160, 1–21.



Anadon, P. and Moissenet, E. (1996) Neogene basins in the

eastern Iberian range. In: Tertiary Basins of Spain (Eds

P.F. Friend and C.J. Dabrio), World and Regional Geology

Series, 6, 68–71.

Anadon, P., Rosell, L. and Talbot, M.R. (1992) Carbonate

replacement of lacustrine gypsum deposits in two Neogene

continental basins, eastern Spain. Sed. Geol., 78, 201–216.

Anadon, P., Ortı, F. and Rosell, L. (1997) Unidades evaporıt-

icas de la zona de Libros-Cascante (Mioceno, Cuenca de

Teruel): Caracterısticas estratigraficas y sedimentologicas.

Cuad. Geol. Iberica, 22, 283–304.

Berger, A.L. (1978) Long-term variations of daily insolation

and Quaternary climatic changes. J. Atmos. Sci., 35, 2362–

2367.

Broekman, J.A. (1983) Environments of deposition, sequences

and history of Tertiary continental sedimentation in the

Basin of Teruel-Ademuz (Spain). Proc. K. Ned. Acad. Wet.

Ser. B, 86, 25–37.

Broekman, J.A., Besems, R.E., Daalen, P.v. and Steensma, K.(1983) Lithostratigraphy of Tertiary continental deposits in

the Basin of Teruel-Ademuz (Spain). Proc. K. Ned. Acad.

Wet Ser. B, 86, 1–16.

Buurman, P. (1980) Paleosols in the Reading Beds (Paleocene)

of Alum Bay, Isle of Wight, U.K. Sedimentology, 27,593–606.

Cohen, A.S. and Thouin, C. (1987) Nearshore carbonate

deposits in Lake Tangayika. Geology, 15, 414–418.

Currey, D.R. (1990) Quarternary palaeolakes in the evolution

of semidesert basins, with special emphasis on Lake

Bonneville and the Great Basin, USA. Palaeogeogr. Palaeo-climatol. Palaeoecol., 76, 189–214.

Dunham, R.J. (1962) Classification of carbonates according to

their depositional texture. In: Classification of CarbonateRocks (Ed. W.E. Ham), AAPG Mem., 1, 108–121.

Dupont-Nivet, G., Krijgsman, W., Langereis, C.G., Abels, H.A.,Dai, S. and Fang, X. (2007) Tibetan plateau aridification

linked to global cooling at the Eocene-Oligocene transition.

Nature, 445, 635–638.

Eardley, A.J., Shuey, R.T., Gvosdetsky, V., Nash, W.P., Picard,M.D., Grey, D.C. and Kukla, G.J. (1973) Lake cycles in the

Bonneville Basin, Utah. Geol. Soc. Am. Bull., 84, 211–216.

Folk, R.L. (1959) Practical petrographic classification of

limestones. AAPG Bull., 43, 1–38.

Foucault, A. and Melieres, F. (2000) Palaeoclimatic cyclicity

in central Mediterranean Pliocene sediments: the mineral-

ogical signal. Palaeogeogr. Palaeoclimatol. Palaeoecol., 158,311–323.

Freytet, P. and Verrecchia, E.P. (2002) Lacustrine and palus-

trine carbonate petrography: an overview. J. Paleolimnol.,

27, 221–237.

Garcıa, A. (1994) Charophyta: their use in paleolimnology.

J. Paleolimnol., 10, 43–52.

Gradstein, F.M., Ogg, J.G., Sith, A.G., Bleeker, W. and

Lourens, L.J. (2004) A new Geologic Time Scale, with spe-

cial reference to Precambrian and Neogene. Episodes, 27,83–100.

Hazeleger, W., Molteni, F., Severijns, C., Haarsma, R., Bracco,A. and Kucharski, F. (2003) SPEEDO: a flexible coupled

model for climate studies. Exchanges, 28, 1–3.

Hilgen, F.J., Abdul Aziz, H., Krijgsman, W., Langereis, C.G.,Lourens, L.J., Meulenkamp, J.E., Raffi, I., Steenbrink, J.,Turco, E., Van Vugt, N., Wijbrans, J.R. and Zachariasse,W.J. (1999) Present status of the astronomical (polarity)

time-scale for the Mediterranean Late Neogene. Phil. Trans.

Roy. Soc. London A, 357, 1931–1947.

Husing, S.K., Hilgen, F.J., Abdul Aziz, H. and Krijgsman, W.(2007) Completing the Neogene geological time scale

between 8.5 and 12.5 Ma. Earth Planet. Sci. Lett., 253, 340–

358.

Kiefer, E. (1988) Facies development of a lacustrine tecto-

sedimentary cycle in the Neogene Teruel-Ademuz Graben

(NE Spain). Neues Jb. Geol. Palaontol., 6, 327–360.

Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia,A.C.M. and Levrard, B. (2004) A long term numerical

solution for the insolation quantities of the Earth. Astron.

Astrophys., 428, 261–285.

Lindholm, R.C. and Finkelman, R.B. (1972) Calcite staining:

semi-quantitative determination of ferrous iron. J. Sed.

Petrol., 42, 239–242.

Lourens, L.J. (2004) Revised tuning of Ocean Drilling Program

Site 964 and KC01B (Mediterranean) and implications for

the d18O, tephra, calcareous nannofossil, and geomagnetic

reversal chronologies of the past 1.1 Myr. Paleoceanogra-

phy, 19, PA3010. doi: 10.1029/2003PA000997.

Lourens, L.J., Hilgen, F.J., Shackleton, N.J., Laskar, J. and

Wilson, D.S. (2004) The Neogene Period. In: A Geologic

Time Scale 2004 (Eds F.M. Gradstein, J.G. Ogg and

A. Smith), pp. 405–440. Cambridge University Press,

Cambridge.

Magri, D. and Tzedakis, P.C. (2000) Orbital signatures and

long-term vegetation patterns in the Mediterranean. Quar-tern. Int., 73/74, 69–78.

Molteni, F. (2003) Atmospheric simulations using a GCM with

simplified physical parametrizations; Model climatology

and variability in multi-decadal experiments. ClimateDynam., 20, 175–191.

Normati, M. and Salomon, J. (1989) Reconstruction of

a Berriasian lacustrine paleoenvironment in the Cameros

basin (Spain). Palaeogeogr. Palaeoclimatol. Palaeoecol., 70,215–223.

Olsen, P.E. and Kent, D.V. (1996) Milankovitch climate forcing

in the tropics of Pangaea during the Late Triassic. Palaeo-geogr. Palaeoclimatol. Palaeoecol., 122, 1–26.

Perez, A., Luzon, A., Roc, A.C., Soria, A.R., Mayayo, M.J. and

Sanchez, J.A. (2002) Sedimentary facies distribution and

genesis of a recent carbonate-rich saline lake: Gallocanta

Lake, Iberian Chain, NE Spain. Sed. Geol., 148, 185–202.

Picard, M.D. and High, L.R. (1981) Physical stratigraphy of

ancient lacustrine deposits. SEPM Spec. Publ., 31, 233–259.

Platt, N.H. (1989) Continental sedimentation in an evolving

rift basin: the lower Cretaceous of the western Cameros

Basin (northern Spain). Sed. Geol., 64, 91–109.

Platt, N.H. and Wright, V.P. (1991) Lacustrine carbonates:

facies models, facies distributions and hydrocarbon aspects.

Spec. Publ. Int. Ass. Sediment., 13, 57–74.

Rosen, M.R. (1994) The importance of groundwater in playas:

a review of playa classifications and the sedimentology and

hydrology of playas. In: Paleoclimate and Basin Evolution

of Playa Systems (Ed. M.R. Rosen), GSA Volume Special

Paper, 289, 1–18.

Rossignol-Strick, M. (1983) African monsoons, an immediate

climate response to orbital insolation. Nature, 304, 46–49.

Sagri, M., Abbate, E. and Bruni, P. (1989) Deposits of

ephemeral and perennial lakes in the Tertiary Daban basin

(Northern Somalia). Palaeogeogr. Palaeoclimatol. Palaeo-

ecol., 70, 225–233.

Sanz, M.E., Alonso-Zarza, A.M. and Calvo, J.P. (1995)

Carbonate pond deposits related to semi-arid alluvial

systems: examples from the Tertiary Madrid basin, Spain.

Sedimentology, 42, 437–452.



Sierro, F.J., Flores, J.A., Zamarreno, I., Vazquez, A., Utrilla,R., Frances, G., Hilgen, F.J. and Krijgsman, W. (1999)

Messinian pre-evaporite sapropels and precession-induced

oscillations in western Mediterranean climate. Mar. Geol.,153, 137–146.

Sierro, F.J., Ledesma, S., Flores, J.-A., Torrescusa, S. and

Martınez del Olmo, W. (2000) Sonic and gamma-ray astro-

chronology: cycle to cycle calibration of Atlantic climatic

records to Mediterranean sapropels and astronomical

oscillations. Geology, 28, 695–698.

Steenbrink, J., Van Vugt, N., Hilgen, F.J., Wijbrands, J.R. and

Meulenkamp, J.E. (1999) Sedimentary cycles and volcanic

ash beds in the Lower Pliocene lacustrine succession of

Ptolemais (NW Greece): discrepancy between 40Ar/39Ar and

astronomical ages. Palaeogeogr. Palaeoclimatol. Palaeo-ecol., 152, 283–303.

Tuenter, E., Weber, S.L., Hilgen, F.J. and Lourens, L.J. (2003)

The response of the African summer monsoon to remote and

local forcing due to precession and obliquity. Global Planet.Change, 36, 219–235.

Turkmen, I. and Kerey, I.E. (2000) Alluvial and lacustrine

facies of the Yenicubuk formation (lower-middle Miocene),

Upper Kizilirmak basin, Turkiye (Turkey). In: Lake Basins

through Space and Time (Eds E.H. Gierlowski-Kordesch

and K. Kelts), AAPG Stud. Geol., 46, 449–464.

Van Dam, J., Abdul Aziz, H., Alvarez Sierra, M.A., Hilgen,F.J., Van den Hoek Ostende, L.W., Lourens, L.J., Mein, P.,Van der Meulen, A.J. and Pelaez-Campomanes, P. (2006)

Long-period astronomical forcing of mammal turnover.

Nature, 443, 687–691.

Vazquez, A., Utrilla, R., Zamarreno, I., Sierro, F.J., Flores,J.A., Frances, G. and Barcena, M.A. (2000) Precession-

related sapropelites of the Messinian Sorbas Basin (South

Spain): paleoenvironmental significance. Palaeogeogr.

Palaeoclimatol. Palaeoecol., 158, 353–370.

Warren, J. (1999) Evaporites: Their Evolution and Economics.

Blackwell Science, Oxford, 438 pp.

Weedon, G.P. (2003) Time-series Analysis and Cyclostratig-

raphy – Examining Stratigraphic Records of Environmental

Cycles. Cambridge University Press, Cambridge, 259 pp.

Wright, P.V. and Marriott, S.B. (2007) The dangers of taking

mud for granted: Lessons from Lower Old Red Sandstone

dryland river systems of South Wales. Sed. Geol., 195,91–100.

Manuscript received 6 July 2007; revision accepted 22April 2008



Documents

Shallow lacustrine carbonate microfacies document ...projects.knmi.nl/publications/fulltexts/abels_sedimentology_2009.pdf · Shallow lacustrine carbonate microfacies document orbitally