Cyclicity in the Middle Eocene Çayraz Carbonate Formation, Haymana Basin, Central Anatolia, Turkey

Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313 329

ELSEVIER

Cyclicity in the Middle Eocene ( ayraz Carbonate Formation, Haymana Basin, Central Anatolia, Turkey Atilla (~iner a, Max Deynoux b, St6phanie Ricou b, Erdal Kosun ° a Hacettepe University, Geological Engineering Department, Beytepe 06532, Ankara, Turkey b Centre CNRS de Gkochimie de la Surface, Institut de Gkologie, 1 rue Blessig, 67084, Strasbourg, France c General Directorate of Mineral Research and Exploration (MTA), Balgat 06520, Ankara, Turkey Received 10 December 1993; revised and accepted 5 September 1995

Abstract

The Haymana Basin in central Anatolia, Turkey, developed on a fore-arc accretionary wedge during the Late Cretaceous and the Middle Eocene. Turbidite deposits fill the center of the basin and grade northwestward into shallow marine clastics and carbonates, which in turn pass into lacustrine and fluvial deposits toward the margin. The Middle Eocene (~ayraz Formation represents the youngest paleotectonic unit of the Haymana Basin. It is composed of two shelf systems (SS), each formed by a package of nummulitic banks and intervening calcareous mudstones, 150 250 m thick, overlying 70-100 m thick monotonous mudstones. Within each of the two shelf systems, two orders of coarsening-upward and shallowing-upward sequences are recognized. The smaller order sequences, 3-10 m thick, are termed basic units (BU). Four types of basic units have been distinguished. These are made up of either calcareous mudstone overlain by a nummulitic bank, or silty mudstone overlain by sandy to conglomeratic beds. Each basic unit represents a specific depositional environment ranging from quiet lagoon or open marine to more energetic bank or sandy bar. Basic units are vertically associated to form three types of 15 35 m thick basic sequences (BS), which represent the time and space evolution of the shelf systems. Owing to their small scale, their lateral association, and the condensed nature of their bounding surfaces, basic units are considered to be allocyclic transgressive-regressive cycles tuned to Milankovitch bands. At the scale of basic sequences (BS) and shelf systems (SS), the lateral and vertical distribution of the sequences, and the presence of onlap structures strongly suggest sea-level changes. The sequential organization, thicknesses of basic units and basic sequences, and an estimated sedimentation rate favour sea-level changes controlled by Milankovitch climatic cycles. Amalgamation or condensation of basic units in basic sequences, and the existence of locally accentuated onlaps suggest a coupled effect of discontinuous subsidence.

1. Introduction

Cyclicity o f different magnitudes is c o m m o n l y observed in shallow marine c a r b o n a t e deposits (e.g., Fisher, 1964; G o o d w i n and Anderson, 1985; James, 1984; Ricken, 1985; Wright, 1986; Grotzinger, 1986; Bush and West, 1987; 0031-0182/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved SSDI 0031-0182(95)00087-9

G o l d h a m m e r et al., 1987, 1990; Strasser, 1988). The allocyclic a n d / o r autocyclic causes o f such cyclicity remain controversial. While autocyclic models (Ginsburg, 1971) are generally envisioned for small-scale sequences (e.g., James, 1984), allocyclic models depend u p o n factors such as eustatic sea-level fluctuations, variations in tectonic subsi-

314

A. ~iner et al./Palaeogeography, Palaeoclimatology, Palaeoecoh)gy 121 (1996) 313-329

dence, and climate (see examples in Crevello et al., 1989). The aim of this paper is to describe coarsening-upward and shallowing-upward nummulitebearing carbonate cycles of different magnitudes that characterize the Lutetian (~ayraz Formation of the Haymana Basin in Central Anatolia (Onalan et al., 1976). The autocyclic or allocyclic origin of these cycles will be discussed in the light of their nature and spatial relationships.

(Fourquin, 1975; Seng6r and Yilmaz, 1981; G6rfir et al., 1984, 1989; Ko~yigit et al., 1988; Koqyigit, 1991) (Figs. 1, 2). After the closure of the north Neo-Tethys Ocean, deformation continued until the Late Pliocene. As stated by Ko~yigit (1991), one of the characteristics of the Haymana Basin in the Ankara region is its highly deformed sedimentary fill. This fill is more than 5000 m thick in the center of the Haymana Basin and is largely of turbiditic origin. Different depositional conditions existed toward the N W margin of the basin, where shallow marine clastics, reef carbonates, lacustrine and fluvial deposits were deposited (Yfiksel, 1970; 121nalan et al., 1976; (~iner, 1992). During Early to Middle Eocene, the Haymana Basin was characterized by a shallow transgressive sea in which widespread sandy and nummulite-bearing limestones were deposited on the shelf, but deep marine sedimentation continued to the SE, within the basin interior (121nalan et al., 1976; Ko~yigit and Lfinel, 1987; Ko~yigit, 1991). Because of extensive tectonic deformation, it is not possible to trace the passage from turbiditic strata into shelf and continental facies in outcrop ((~iner, 1992).

2. Regional setting The Haymana Basin (presently 60 x 60 km in size) is located about 70 km SW of Ankara in Central Anatolia (Fig. 1). It is a forearc basin that formed on an accretionary wedge (cf. Dickinson and Selly, 1979). This wedge formed during the Late Cretaceous to Late Eocene on the oceanic crust of the northern branch of NeoTethys (Izmir-Ankara Suture Zone) by the convergence and collision of the Eurasian continent to the north, the Gondwana continent to the south, and the intervening Sakarya Continent

[stanbul

[)lack Sea

~

.....

L~I]~._

~

Mediterranean

I z m i r Ankara Intra Pontide

.%uture

Metamorphic of K ir",;ehi r

block

NAt

; Neo SutureJ ]ethys

North A n a t o l i a n Fault

~

~ ~

EAI

13itli~.~ Suture

f

Faults

Oalata i'.;land arc

Y

trust

fast

I

I

m ]

A n a t o l i a n Fault

Fig. 1. Main structural features of Turkey and location map of the Haymana Basin (modifiedafter Ko~yigit, 1991).

A. ~iner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313-329

Izmir-Ankara branch of the Neo-Tethys

~ayraz I Accretionary Haymana I wedge

\

315

Galata Island Arc i~

~asm

Lithosphere

Oceanic c r u s t ~

Fig. 2. Schematic cross section (not to scale) showing the structural setting of the Haymana Basin during Campanian to Lutetian time (modified after Koqyigit, 1991).

3. Definition and description of cyclic units

3.1. Definition The Lutetian (Middle Eocene) Cayraz Carbonate Formation is the youngest paleotectonic unit in the Haymana Basin. It overlies a thick turbidite complex dating back to the Maastrichtian (Figs. 3, 4). Upon cursory observation (Fig. 4), the (~ayraz Formation appears roughly to be composed of two levels of stacked carbonate beds, each about 200 m thick, enclosed within mudstones several tens to a hundred meters thick. Each of these carbonate-rich levels is made up of a repetitive alternation of calcareous mudstone and nummulitic limestone. These alternations contrast with the monotonous character of the enclosing mudstones. The Cayraz Formation is separated by an angular unconformity from the overlying Neogene volcanolacustrine deposits (Figs. 4, 5). A detailed study of the calcareous mudstones and carbonate beds within the carbonate-rich levels led to the distinction of two orders of embedded cycles (Fig. 5). These have been termed "basic units" (BU) and "basic sequences" (BS). Both cycles consist of calcareous or silty mudstone

overlain by numulitic limestone or calcareous sandstone and, hence, are coarsening-upward. A basic unit is here defined as the smaller repetitive accretion event ("fundamental building block") that we recognized in the succession. Basic units are 3-10 m thick, and are composed of either calcareous mudstone and nummulitic limestone, or silty mudstone and calcareous sandstone beds. Basic units are stacked vertically to form thicker (15-35 m) basic sequences. Four types of basic units (BU1, BU2, BU3 and BU4) embedded in three types of basic sequences (BS 1, BS2 and BS3), each representing a specific depositional environment, were identified (Fig. 6). A shelf system (SS) subdivision is also proposed for the (~ayraz Formation. The two shelf systems are defined by the association of basic sequences that form the nummulite-bearing carbonate packages and the underlying (nearly nummulite-free) mudstones (Fig. 5). The lower shelf system passes downward into the mudstones of the Cuisian Eskipolatli Formation. The upper one is overlain by about 50 m of mudstone (SS ? in Fig. 5), which may represent the lower part of a third shelf system, as suggested by a few nummulitic limestones locally preserved beneath the erosive Neogene unconformity (Fig. 8).

316

A. ~iner et al /Palaeogeographv Palaeoclimatolo~y, Palaeoecology 121 (1996) 313 329

-~ Kirkkavakformation I'htLilctitn f~

YesilyL,rt formation Havmana forination

l)aniz~n

[]

Quaternary,deposifs

~

Neogenedeposits

Nl~tln[liuhti:tu [ ]

Derekovformation I ('let~tccous ~ (ophiolite a m,~lange) ~ Mollaresul formation lli;is'~i~.

•

~'a.~raz['ormation

l

0

~,n I.ult,'li~ll]

J

2kin. measured sections Vaults

E+kipolathformation (kd,tat] , . ~ llgmhkderefortriatinn [l]~:tdian

faults

Fig. 3. Geologicalmap and stratigraphic subdivision of the (~ayraz region, and location of measured sections.

3.2. Basic units Basic unit 1 (BU1) Basic unit 1 is a thickening- and coarseningupward unit 3-7 m thick. The unit consists of two parts (Fig. 6). The lower part ( 1 - 4 m thick) is a homogeneous silty mudstone with fine-grained sandstone beds that are a few centimetres thick and pinch-out over tens of metres. Ostracodes, such as Neocyprideis sp. and Cytherella sp. (M. Duru, pers. comm., 1991) were identified in these mudstones. The upper part (3-5 m thick) consists of fine- to coarse-grained calcareous sandstone beds 1 0 - 3 0 c m thick. These are composed of

poorly rounded quartz grains, volcanic rock fragments, and a minor amount of bioclastic fragments. In thin sections, the siliciclastic fragments occur within a carbonate matrix containing nummulite and echinoderm fragments, with voids filled by a sparry calcite cement. The internal structure of the beds is commonly homogenized by bioturbation. Apart from few centimeter-scale groove casts at the base, the upper and lower bed surfaces show a sharp planar contact with intervening mudstone beds. Channel structures several metres across and filled by micro-conglomerates are typically present within this sandy upper part. in some cases, limestone and andesite pebbles (5-10 cm in diameter)

A. ~iner et a[./Palaeogeography,Palaeoclimatology,PalaeoecologyI21 (1996) 313-329

317

Fig. 4. General view of the ~ayraz Formation and associated formations. A =carbonate (corals) slump deposits of the Yesilyurt Formation. B= submarine channel-fills of the Ilginlikdere Formation. C= basinal mudstones of the Eskipolatli Formation (250 m thick). D =carbonate package of the lower shelf system of the ~ayraz Formation. E=carbonate package of the upper shelf system of the ~ayraz Formation. F= angular unconformityseparating the ~ayraz Formation from the Neogenevolcano-lacustrinesediments. are found within these channels. Cross-bedding and microslumps with no internal deformation are also commonly observed, especially toward the top of the unit. Basic unit 2 (EU2) Basic unit 2 is a coarsening-upward unit with a thickness varying from 3 to 6 m (Figs. 6-9). The unit is composed o f calcareous mudstones (wackestones), containing small nummulites (A-form) and a few ostracods, overlain in sharp contact by a 1-4 m thick nummulitic bed (packstone to grainstone) made-up of larger nummulites (B-form) (Fig. 10). These and other macrofossils identified include Nummulites laevigatus, N. irregularis, N. lamancki, N. helveticus, N. purfoldi, Miliolidae, Microaster sp., Discoyclinasella sp., Alveolina fayrazi, Operculina sp., Asilina exponens, and Assilina spira (Dizer, 1964, 1968; Sirel and Gandaz, 1976). Scour structures filled with larger nummulites than those found in adjacent beds have been locally observed. Toward the top of the unit, wave and current ripple strata, contact and isolated imbrications (according to Laming, 1966, nomenclature) of large nummulites, and crossbedding are typically seen. The upper portions of the nummulitic beds are dark brown and strongly

indurated, and sharp contacts occur between these and the overlying calcareous mudstone. The uppermost nummulitic bed at the top of the lower shelf system (Fig. 7) in sections (~1-~5, consists of channel fills several tens of metres wide, with well-rounded pebbles o f nummulitic limestone similar in composition to the nummulitic bed. Some of these channel fills also contain pyritized macrofossils (especially gastropods) and bioturbated horizons. Basic unit 3 (BU3) Basic unit 3 is composed mainly of yellow mudstone (2-6 m thick) overlain by a very thin (10-50 cm) and poorly developed nummulitic bed (wackestone to packstone), which is laterally very continuous (Figs. 6, 12, 13). The contacts between the yellow mudstones and the thin nummulitic beds are generally sharp. Nummulitic beds are yellow and, while the nummulites range in size in these beds, no other macrofossils are present. No structures have been observed in these beds. Basic unit 4 (BU4) Basic unit 4 shows a coarsening-up tendency and is composed of calcareous mudstones (mud-

3l8

A. Ciner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313-329

Neo e:. 6001_ -

4 0 0 I.-~

: -

~ __~

~ ~

Iq O

] Yellow a n d Black m u d s t o n e with or w i t h o u t n u m m u l i t e s Yellow c a l c a r e o u s m u d s t o n e with n u m m u l i t e s Nummulitic limestone F i n e to coarse s a n d s t o n e

O

--\

Basic Unit

(BL0

Basic Sequence

03S) 0m E s k i p o latli [ Formation I I

I

Fig. 5. Subdivision of the Cayraz Formation into lower and upper shelf systems, and proposed model for the sequential subdivision of each shelf system.

A. ~iner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313 329

319

Basic Sequence n°l Basic unit n°l Fine to c o a r s e sandstone, siltstone

calcareous conglomerate.

mudstone

B a s i c s e q u e n c e n°2 l~asic u n i t n°2 Storlll deposits Ioward Ibe top {withor w i t h o u t hardground) Nummulitic bank (packslone to grainslone)

k~

i

~

~wac

lasic u n i t n°4 Storm deposits

and hardground Nunmmllt ic bank (packstone to gramstone)

B a s i c s e q u e n c e n°3

[

-----.----- ,

~

~

Transitional

lower part

muddv

I~ /Calcareous m u d s t o n e I~.~[mudslone to wackeslone) r-

B a s i c u n i t n°3 ~_

Poorly developped { w a c k e s l o n e Io

paclc.slonel

~

5 ~

I ),lltdsl(:llC I

I

_J Fig. 6. Representative logs of the different types of basic units (BU1-BU4) embedded in basic sequences (BS1-BS3), which form the carbonate packages of the upper part of the shelf systems of the (~ayraz Formation. stone to w a c k e s t o n e ) at the base, overlain by a thick ( 1 - 4 m) n u m m u l i t i c bed ( p a c k s t o n e to grainstone) (Figs. 6, 12). Total thickness varies

between 5 and l0 m. The calcareous m u d s t o n e s at the base o f the unit occasionally contain thin h o r i z o n s o f small nummulites, and the passage to

320

A, ~iner et aL/Palaeogeography, Palaeoclimatology, Palaeaecology 121 (1996) 313 329

.=_

..............

~

~

o,~ ~

©

~o

.=_

8~

Hllll IHUlPlP' i

] IMp, I~[ ~'

'~

m~

E

c-

321

A. ~iner et aL/Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313-329

-~ ================================

and

Yellow black mudstone with or without nummulites

L~

Cal. . . . . . . . . with

. . . . .

Nttrnmulttic calcareous Conglomeratic

~mm~Htic

- - _ - - - _- _- - - - _- _ - - _- :

dstone

[B~

lites

limestonemudstone

intercalations

channels limestone

Key bed (nummulitic indicating onlap

°"~'~ :I';-."titt~

~!:!!:!!i ~

•

._-2-5]-:]_2:--]2

~]

',,

~

,,,

Hrnestone)

~

_-::::-!!::--:!!i ------!--! ~~-------~i

.....

. ......

with

.....

~:~'itttiiilli

T

_-

. . . . . . . . . .

~' . . . . .

:'.'.'I::':'I

,','t'l',

~

~ ::::!:!:!:i:i:

iiiiii!!!i!i!iiii ,,

:::::::::::::::::::::

I:!:LI;I;:;I;L!:!:!:LLLL!: ~;LLI;I;LL;;!;L!;!;!:L!:L L~;LLLLLi:!;!;;;LL:;;;I;

?

::::::::::::::::::::::

~

.................

?

.........

I

..........

i Si i i i i i i ;i i i i i i i!V

5 5 5 5 A~5;~;~5~

? ._22_][2-[..........

W ? ................

?

i]ii!!ili

. 311

I} m

~e===

I e~5 I00

20(I

c~

~a

Fig. 8. Subdivision of the upper shelf system of the ~ayraz Formation into basic sequences and their correlations. The thick line in the lower part of the figure indicates a thin nummulitic bank contained in mudstone, and which onlaps westward the last basic sequence of the lower shelf system.

nummulitic bed is always gradual (Fig. 14). At the base of the nummulitic beds, lime mud is present but tends to disappear upward as the number and

size of the nummulites increase. The beds become nearly clay-free upward, and current ripple strata and large-scale wave ripple strata made up of

322

A. ~'iner et al./Palaeogeography, Pulaeoclimatology, Palaeoecology 121 (1996) 313 329

Fig. 9. View of the lower part of the carbonate package of the lower shelf system of the (~ayraz Formation, between sections ~3 and C4. a = amalgamation of Basic units 2. b = 6 m thick Basic unit 2.

Fig. 10. Close-up view of the large nummulites (B-form) that form the nummulitic banks of Basic units 2 in the lower shelf system of the Cayraz Formation. Pencil for scale.

carbonate and nummulitic fragments occur. In places these upper beds pass into decimeter-scale hummocky cross-stratifications (HCS). Parallel lamination caused by the alignment of large nummulites is also common in these beds. Nummulites and alveolines constitute nearly the whole rock but other macrofossils (echinoderm and gastropod fragments) are also present. Imbrication of the large nummulites is typical. The upper 2 5 cm of the nummulitic beds are dark and highly indurated, forming very continuous horizons that can be traced for several hundred metres (Fig. 12), thereby forming excellent marker beds. On the upper surfaces, large gastropods such as Campanile giganteum and Velatus schmideli (M. Duru, pers. comm., 1991) reach 20 cm in length, and are frequently observed in life position. Examination of thin sections shows that the borings cut through the drusy calcite cement and nummulite tests, and were then filled by calcareous mudstone. In places, the nummulitic beds have a nodular (concretionary) character. The nodules (10-50 cm in diameter) are generally round to ellipsoidal, and are commonly separated by a thin film of lime mud. Basic unit 4 appears to be fairly similar to Basic unit 2. However, it can be distinguished by the more argilaceous nature, and smaller amount of nummulites in the muddy lower part, the transitional character of the contact between the mudstone and nummulitic bed, and the presence of a well-developed indurated horizon capping the unit. 3.3. Basic sequences

Fig. l l . View of the upper part of the carbonate package of the lower shelf system of the (~ayraz Formation, between sections (~2 and (~3. The Basic sequence 2 shown between arrows is 19 m thick, h = hardground that forms the onlapped surface between sections (~2 and (~4 (see Fig. 7). Unlabelled arrows indicate the upper limits of Basic sequences 2.

Basic sequence 1 (BS1) Basic sequence 1 is composed of 2 4 Basic units 1 overlying 5-15 m thick silty mudstones (Fig. 6). Total thickness varies between 18 and 35 m. Question marks shown on Fig. 7 (sections (~6 and (~7) indicate the uncertainties in distinguishing the limits of Basic sequence 1 because of poor exposures at some localities. A Basic unit 2 may form the upper part of a Basic sequence 1. Basic sequence 2 (BS2) Basic sequence 2 is composed of 2-4 (most often 3) Basic units 2 overlying a calcareous mudstone

A. ~iner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313 329

323

Fig. 12. General view of the upper shelf system of the ~ayraz Formation, between sections ~1 and C3 (see Fig. 8). a=poorly developed nummulitic bank forming the upper part of a Basic unit 3. b =nummulitic bank forming the upper part of a Basic unit 4. Unlabelled arrows indicate the upper limits of each Basic sequence 3. The Basic sequence 3 shown with brackets is 23 m thick. h = hardground.

Fig. 13. Calcareous mudstone and thin nummulitic bank of a Basic unit 3 in the upper shelf system of the ~ayraz Formation. Person for scale. (wackestone) unit a b o u t 10 m t h i c k ( F i g s . 6, 11). In this c a l c a r e o u s m u d d y unit, in c o n t r a s t to the n u m m u l i t i c beds, s m a l l e r A - f o r m s ( m a c r o s p h e r ical g e n e r a t i o n , A i g n e r , 1983) largely d o m i n a t e over B - f o r m s ( m i c r o s p h e r i c a l g e n e r a t i o n ) . X - r a y diffraction analysis o f the clay m i n e r a l s in the c a l c a r e o u s m u d s t o n e s indicate t h a t they are m o s t l y c o m p o s e d o f corrensite a n d smectite with m i n o r a m o u n t s o f chlorite a n d illite. T h e thickness o f Basic sequence 2 varies b e t w e e n 15 a n d 30 m with a general c o a r s e n i n g - u p w a r d tendency. H o w e v e r , a m a l g a m a t i o n o f n u m m u l i t i c beds o f Basic units 2 ( t h a t is, the absence o f their lower c a l c a r e o u s m u d s t o n e s ) is also f r e q u e n t l y o b s e r v e d ( F i g . 9). In this case the distinction o f i n d i v i d u a l n u m m u l i t i c units b e c o m e s difficult.

Fig. 14. Detail of the gradual passage from mudstone to nummulitic bank of a Basic unit 4. Upper shelf system of the ~ayraz Formation. a = calcareous mudstones with some small nummulites, b = wavy laminations and decrease in mud content. c = parallel lamination marked by alignment of large nummulites. d= large-scale wave ripples. Arrow = pencil for scale. Basic sequence 3 (BS3) Basic sequence 3 is c o m p o s e d o f two types o f Basic units with a t o t a l thickness between 15 a n d 26 m (Figs. 6 a n d 12). T h e l o w e r p a r t c o n t a i n s 2 - 5 Basic units 3 a n d is o v e r l a i n by 1 - 4 (mostly 3) Basic units 4. T h e u p p e r shelf system is exclusively c o m p o s e d o f Basic sequence 3 ( F i g . 8). H o w e v e r , in section C6, because o f p o o r o u t c r o p , it was n o t possible to limit a n d to differentiate clearly the type o f basic sequence. S o m e large (tens o f metres wide) c h a n n e l structures occur, a n d these are filled b y q u a r t z a n d micritic c a l c a r e o u s p e b b l e s

324

A. Ciner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313-329

(5-10 cm in size) within nummulitic beds. Only a few Basic sequences 3 were identified in the lower shelf system (sections I~1 and ~2 in Fig. 7).

4. Depositional environment The overall shallow marine and transgressive nature of the (~ayraz Formation has previously been documented in geodynamic reconstructions (Unalan et al., 1976; Ko~yigit and Lt~nel, 1987; Koqyigit, 1991). The presence and nature of in situ macrofossils, cross-stratification, and strong bioturbation confirms this shelf character. The conglomeratic channels cutting into silty mudstones, as well as the presence of nummulite and echinoderm fragments, indicate local or periodic energetic environments with reworking processes. The benthic foraminifera N u m m u l i t e s are believed to have lived in warm (greater than 20°C) marine waters, at depths of 10-80 m, with normal salinity (Nemkov, 1960; Blondeau, 1972). Ostracodes found in the silty mudstones are thought to live in shallow more-or-less brackish water. Siliciclastics, which form the Basic units and Basic sequences 1, represent continental input to the shelf. The nummulitic beds are interpreted as nummulitic banks, with " b a n k " defined as a non-reef mound-like buildup, in this case formed by in situ accumulation of primarily nummulites (Arni, 1965). In the (~ayraz Formation, the nummulitic banks can be traced laterally over hundreds of metres to a few kilometres. Nummulite reefs (Nemkov, 1962), nummulitic banks (Arni, 1965), nummulite bioherms (Bignot, 1972), reefal nummulitic banks (Arni and Lanterno, 1976) or parautochthonous coquinas (Aigner, 1983) are the terms proposed in the literature for these types of banks. Fusulinid shoals in the Pennsylvanian of North America (e.g., Wilson, 1975, p. 209) and nummulite banks around the Tethyan Sea (Aigner, 1983) provide good examples. However, as Aigner (1983) pointed out, these kind of banks cannot be attributed to true reefs (characterized solely by framebuilding organisms) or to shoals (characterized solely by physical processes). Because nummulites were limited to the Paleogene, the nummulitic banks do not have modern counterparts (Aigner,

1983). We think that the nummulitic banks in ~ayraz occupied considerably wider areas than modern reefs. They probably covered most of the shelf, thinning towards shelf lagoons and the platform margin. The dark, indurated upper horizons with borings and macrofossils in life positions developed on Basic units 4, and less so on Basic units 2, are interpreted as hardgrounds. Because the biogenic boring walls are very sharp in nature and were later filled by calcareous mud from the overlying unit, the hardgrounds are thought to be synsedimentary in origin. The hardgrounds suggest at least a decrease or even a cessation in the rate of carbonate sedimentation. Vertical and lateral changes in the strata reflect a range in shelf conditions. The basal parts of Basic units 2 and 4 represent an overall low-energy shelf environment where lime mud accumulated with small nummulites (A-form), with occasional storms, as evidenced by nummulite (A-form) concentrations thought to represent storm-lag units. The nummulitic banks represent a shallower and more energetic environment. Cross-bedding, imbrications, large wave ripples, or even hummocky cross-stratification, show that storms and currents were continuously washing out mud-size particles, leaving behind a concentration of large nummulites (B-form). Based on the A/B forms ratio, Aigner (1983) suggested that most of the nummulitic tests are reworked in situ without lateral transport. This would imply that the specific distribution of the A- and B-forms in basic units reflects particular environmental conditions. Aggradation of nummulitic banks or water shallowing, therefore, is implied by the upward transition of calcareous mudstones (wackestone) to nummulitic bank (packstone to grainstone). The upper part of the lower shelf system (Fig. 7) shows a strong tendency for the emergence of the last nummulitic bank. Here, storms and waves reworked the nummulitic bank, forming intraformational conglomerates in a similar way to the formation of oolites in an offshore bar. A deeper shelf environment is suggested for Basic unit 3, characterized by the near absence of nummulites in the yellow mudstones and the poor development of nummulitic banks. This unit is

A. ~iner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313-329

thought to represent the distal extension of the well developed numulitic banks of Basic units 2 and/or 4. The more argilaceous nature and small amount of nummulites in the muddy lower part of Basic units 4, and the occurrence of other macrofossils beside nummulites may also reflect the passage from nummulitic banks of Basic unit 2 to a deeper environment. Such transitions (schematized in Fig. 15) can be observed in the lower shelf system. The lower shelf system shows from W to E the passage from Basic sequence 1, with a significant terrigenous sand component, to Basic sequences 2 and 3, to a yellow and black shelf mudstones (Fig. 7). The limited lateral extent (1-3 km) and the two dimensional W - E orientation of the exposure do not allow extensive paleogeographic reconstructions. According to broader paleogeographic and geodynamic reconstructions that locate the continent to the NW and the basin to the SE, the W-E orientation of the exposure is slightly oblique to the general paleoslope. This is also suggested in the lower shelf system by the N W - SE orientation

325

of channel axis. The occurrence of siliciclastic deposits in the western part of the exposure (Figs. 7, 8), and the passage eastward from nummulitebearing mudstones and carbonates to nearly nummulite-free yellow and black mudstones (Fig. 7) seems to confirm this continent-to-basin orientation. Also because of the limited exposure, we are hesitant to identify clearly various environments such as those proposed by Wilson (1975) for carbonate shelves. Even the distinction between carbonate ramp or platform or open shelf and lagoon remains difficult. However, we can conclude that the ~ayraz Formation represents a shelf environment upon which laterally extensive nummulitic banks developed. The nummulitic limestones that are associated with the nummulite banks within basic units or basic sequences may represent shelf lagoonal deposits partly contaminated by clastics brought from the continent and from nummulitic reefs by washover effects during storms. On the other hand, the yellow and black mudstones, which enclose the shelf deposits

A

B

R

m

Hardground

O~

"~

Toward the numrnulite bank ~

Numrnulite b a n k

Toward the lagoon or open shelf IOnl

Calcareous m u d s t o n e with n u m m u l i t e s -~ Mudstone

f

! 0

lkm

Fig. 15. Idealised correlated sections showing the possible relationships between basic units and basic sequences in the platform systems of the ~ayraz Formation. See text for discussion.

326

A. ~'iner et al./Palaeogeograph3, Palaeoclimatology, Palueoecology 121 (1996) 313 329

(Fig. 5) and are represented in the eastern part of the lower shelf system (Fig. 7), may correspond to open shelf sedimentation.

5. Origin of cyclicity The (~ayraz Formation is characterized by a repetition of four types of basic units embedded in three types of basic sequences. Regardless of its type or its position, each basic unit or basic sequence shows the superposition of two distinct facies (e.g., calcareous mudstones and nummulite banks), which implies a deepening-shallowing (or trangressive-regressive) character. These deepeni n , s h a l l o w i n g units may be regarded as autocyclic or allocyclic depending upon the correlation of individual units and upon the nature of the bounding surfaces (Anderson et al., 1984; Bush and West, 1987). From field observations, the upper part of Basic sequences 2 and 3 consists of generally three basic units (2 and 4), while the lower part of these sequences consists of only one Basic unit 2 and three Basic units 3 (Fig. 15). Correlation of the sections as shown in Fig. 15 argues that the three Basic units 3 in section B are represented in section A by their amalgamation with removal of the muddy intervals. The allocyclic hypothesis is also supported by the locally observed continuity of a nummulitic bank from Basic unit 3 to Basic units 4 and 2, and by the presence of hardgrounds capping Basic units 2 and 4. During this hardground period, a variety of organisms developed freely, proliferated, and bored the lithified surface. A rapid rise in sealevel with starvation of sediment supply is the usual cause of such hardground formations, which are better developed on Basic units 4 than on their lateral shallower equivalent Basic units 2. According to an allocyclic hypothesis, Fig. 15 shows how the lateral association of the various kinds of basic units in a basic sequence reflect the response of a nummulitic bank to an overall sealevel rise. Each group of basic units are time equivalent. There is no evidence for landward or basinward migration of the basic units, but instead an in-situ aggradational pattern. The frequent amalgamation of some basic units (for example

BU2 in Fig. 9) in different parts of the platform indicates that the nummulitic banks were in a high energy environment where most calcareous mud was winnowed away. This means that the high accumulation rates of nummulite tests in these banks can easily compensate for sea-level rise and subsidence. As Busch and Rollins (1984) pointed out, this might also be explained by a model of allocyclic dominance with some concomitant autocyclic influence. The nummulitic banks in Basic unit 4 and especially in Basic unit 3 (Fig. 15, section B), compared to the nummulitic banks observed in Basic unit 2, are thinner and muddier in character and overlie calcareous and yellow mudstones. This indicates that in this part of the shelf, accommodation space was still available for the accumulation of sediments. It can be concluded that in a nummulitic shelf, sea-level changes do not necessarily result in environmental shifts. The specific lithofacies assemblage described above for each basic unit and their vertical and lateral evolution within each basic sequence reflects a high degree of environmental dependence, which partly conceals the allocyclic parameters. At the scale of the basic sequences, the vertical and lateral distribution of different types of basic sequences also reflect sea-level variations. In the lower shelf system (Fig. 7) Basic sequences l, which form the lower half of the sections between (~3 and C7, are replaced progressively by Basic sequences 2, and finally Basic sequences 3 predominate in the eastern (~1 section. Because Basic sequence 2 and Basic sequence 3 represent a deeper shelf environment compared to the more proximal terrigenous environment represented by Basic sequence 1, one can identify a westward transgression. However, the presence of wide channels filled with intraformational conglomerates on the uppermost part of the sections between C1 and (75, and the large extension of Basic sequence 2, which overlies Basic sequences 3 in the eastern C1 (72 sections (Fig. 7), indicate the end of the general transgressive tendency. The geometric evolution of basic sequences also confirms this transgressive pattern. In the middle of the first shelf system, yellow mudstones of a Basic sequence 3 become progressively thicker from W to E. Thin nummulitic banks (shown by a heavy line in Fig. 7, sections CI (~4) in the

A. Ciner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313-329

middle of the yellow mudstones, are easily followed to the west in an onlap geometry that ends at section (~4 (Fig. 16). This onlap geometry results from a sharp change in paleo-slope between sections (~4 and (~1, probably created by a synsedimentary fault system present between sections (~4 and (~5. The accomodation space thus created was quickly filled by yellow mudstones, and the nummulitic banks of Basic sequence 2 were subsequently deposited across the region. Onlap also occurs at the base of the thick mudstone unit that forms the lower part of the upper shelf system (Fig. 8). This onlap is again evidenced by a thin (1 m) nummulitic bank. This bank overlies yellow mudstones, which become progressively thinner toward the west (45 m in section (~1 and only 10 m in section (~6). These locally accentuated onlap geometries and the unexpected channels filled with conglomerates on top of the lower shelf system suggest that local tectonics played a role in the organization of basic sequences. Unlike the lower shelf system, the upper shelf system is characterized by the regularity of its basic sequences. This is probably because of more uniform subsidence of the shelf during the deposition of the upper shelf system. The result is a general aggradational pattern represented only by Basic sequence 3. Conceivable mechanisms for eustatic sea-level changes in shallow marine sediments are activity of ice caps, cyclic evaporation of isolated oceanic basins, thermal expension of ocean water, and

327

changes in the geoid (cf. Strasser, 1988). However, glacio-eustatic sea-level oscillations controlled by Milankovitch climatic rhythms is the best tailored model for generating small-scale carbonate cycles (e.g. Fischer, 1964; Hays et al., 1976; Kendall and Schlager, 1981; Grotzinger, 1986; Goldhammer et al., 1987, 1990; Strasser, 1988). Stacks of such small-scale carbonate sequences are known from the Precambrian to the present, and most prePleistocene examples, which may include the Cayraz carbonate-platform, show that Milankovitch-related oscillation in sea-level took place at a time for which no direct evidence of glaciation has been found. This raises the unsolved question of the role of waxing and waning of alpine or limited polar ice caps during non glacial periods (Fischer, 1986; Grotzinger, 1986; Goldhammer et al., 1987). Considering a glacio-eustatic Milankovitch control, the thickness (3-7 m in average) of basic units and their association in basic sequences (16-30 m thick in average) would be more or less consistent, respectively with precession cycles (20 ka) and eccentricity cycles (100 ka). This assuption is based on a 25 cm/103 yr sedimentation rate (Grotzinger, 1986), which is a reasonable (even low) value considering the generally prolific development of a nummulitic community (Blondeau, 1972). However, the ratio of 2-6 basic units per basic sequence do not match a perfect Milankovitch rhythmic pattern. Fluctuations in amplitude of sealevel oscillation leading to amalgamations, condensations, or missed beats may be caused by discontin-

Fig. 16. Onlap geometry in the lower shelf system of the (~ayraz Formation (section (~4). The nummulite bank, on which the onlap arrow is placed, is 2 m thick.

328

A. ~iner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313 329

uous subsidence as suggested by the observed locally accentuated onlaps (see above). Cyclic build-up and release of elastic stresses along stick-slip faults bordering the platform have been proposed by Cisne (1986) as a model for the origin of shallowing-upward sequences on a carbonate platform margin. Such an origin would be suggested by the observed accentated onlaps. However, these features are not rhythmic and probably result from synsedimentary faults associated with a discontinuous subsidence.

6. Conclusions The two shelf systems that form the (~ayraz Formation are characterized by the repetitive alternation of calcareous mudstones and nummulitic banks, which covered part of the shelf. The relative abundance of silt and clay, the nature and abundance of fossils, the presence of sandy to conglomeratic beds, and structures such as crossbeddings, imbrications, wave ripples, and HCS permit the distinction of quiet low energy shelf lagoon environment, reef-like nummulitic banks, or open shelf conditions. Although the limited extent of the exposure does not allow a precise environmental reconstruction, these distinctions are in good agreement with the broader paleogeographic reconstructions that place the (~ayraz area on the shelf margin bordering the deepest part (southeast) of the H a y m a n a Basin. The subdivision of the shelf systems into different coarsening- and shallowing-upward types of basic units embedded in several types of coarsening and shallowing upward basic sequences suggest allocyclic parameters for their development. Such a hypothesis is confirmed by the transgressiveregressive character of these units, the presence of non-depositional surfaces atop some of them, and the unit-to-unit correlations that are not environment dependent. The small-scale, transgressive regressive basic units are controlled by allocyclic sea-level rises, probably tuned to Milankovitch bands. Basic sequences also show transgressive regressive tendencies related to sea-level changes. This is confirmed by the time and space relationships of the different types of basic sequences and by the

presence of onlap structures. According to the sequential organization, thickness and estimated sedimentation rate, an origin by eustatic sea-level changes controlled by Milankovitch climatic cycles is favoured, even if no direct evidence of glaciation has been reported during Lutetian time. Such stacks of shallowing-upward sequences on carbonate platforms of various ages are known in the rock record, so their origin appears to be independant of their age and of glacial periods. Discontinuous subsidence, suggested by locally accentuated onlap structures, was probably an important factor on such an active margin basin. It may account for the amalgamation or condensation of basic units in basic sequences, and on a larger scale, for the deepening effect represented by the thick mudstone units that form the lower part of the platform systems.

Acknowledgments This research is part of the results of a thesis by A. (~iner. This work was done within a French Turkish cooperative program, initiated by N Gfindogdu of the Hacettepe University, and funded by the French Foreign Office. We would especially like to thank G. Kocurek and C. Swezey for discussion and correction on several drafts of this paper. Thanks also to P i . De Boer, N. G6rt~r, and G. Kelling for their constructive reviewing comments, and to J.M. Bock for his technical assistance.

References Aigner, T., 1983. Facies and origin of nummulitic build-ups: an example from the Giza Pyramids Plateau (M. Eocene, Egypt). Neues Jahrb. Geol. Pal~iontol. Abh., 166: 347-368. Anderson, E.J., Goodwin, P.W. and Sobieski, T.H., 1984. Episodic accumulation and the origin of formation boundaries in the Helderberg Group, New York State. Geology, 12:120 123. Arni, P., 1965. L'6volution des Nummulilinae en tant que facteur de modification des d6p6ts littoraux. In: Coll, int. Micropal6ontol., Dakar, 1963. Mere. BRGM, 32:7 20. Arni, P. and Lanterno, E., 1976. Observationspaleo6cologiques dans l'Eoc6ne du Gargano (Italie M6ridionale). Arch, Sci. Genbve, 287-311. Bignot. G., 1972. Recherches stratigraphiques sur les calcaires du Cr6tac6 Sup6rieur et de l'Eoc6ne d'Istrie et des r6gions voisines. Essaie de r6vision du Liburnien. Trav. Lab. Micropal6ontol., 2, 353 pp.

A. ~iner et al./Palaeogeography, Palaeoclimatology, Palaeoecology 121 (1996) 313-329 Blondeau, A., 1972. Les Nummulites. Vuibert, Paris, 245 pp. Busch, R.M. and Rollins, H.B., 1984. Correlation of carboniferous strata using a hierarchy of transgressive-regressive units. Geology, 12: 471-474. Busch, R.M. and West, R.R., 1987. Hierarchal genetic stratigraphy: a framework for paleoceanography. Paleoceanography, 2: 141-164. ~iner, A., 1992. Sedimentologie et stratigraphie sdquentielle du bassin d'Haymana/~ l'Eoc6ne moyen, Turquie. Thesis. Univ. Louis Pasteur, Strasbourg, 190 pp. Cisne, J.L., 1986. Earthquakes recorded stratigraphically on carbonate platforms. Nature, 323: 320-322. Crevello, P.D., Wilson, J.L., Sarg, J.F. and Read, J.F., Editors, 1989. Controls on carbonate platform and basin development. SEPM Spec. Publ., 44, 405 pp. Dickinson, S.R. and Selly, D.R., 1979. Structure and stratigraphy of fore-arc regions. AAPG Bull., 63: 67-94. Dizer, A., 1964. Sur quelques Alv6olines de l'Eoc6ne de Turquie. Rev. Micropal6ontol., 7: 265-279. Dizer, A., 1968. Etude micropal6ontologique du Nummnlitique de Haymana (Turquie). Rev. Micropal6ontol., 1: 13-21. Fischer, A.G., 1964. The Lofer cyclothems of the Alpine Triassic. Kans. State Geol. Surv. Bull., 169: 107-149. Fischer, A.G., 1986. Climatic rhythms recorded in strata. Annu. Rev. Planet. Earth Sci., 14:351 376. Fourquin, C., 1975. L'Anatolie du Nord Ouest, marge m6ridionale du continent Europ~en, histoire pal~og6ographique, tectonique et magmatique durant le Secondaire et Tertiaire. Bull. Soc. G6ol. Fr., (7), 2: 1058-1070. Ginsburg, R.N., 1971. Landward movement of carbonate mud: new model for regressive cycles in carbonates. AAPG Bull., 55, p. 340. Goldhammer, R.K., Dunn, P.A. and Hardie, L.A., 1987. Highfrequency glacio-eustatic sealevel oscillations with Milankovitch characteristics recorded in Middle Triassic platform carbonates in North Italy. Am. J. Sci., 287: 853-892. Goldhammer, R.K., Oswald, E.J. and Dunn, P.A., 1990. Forward modeling of high-frequency depositional sequences: an example from Middle Pensylvanian shelf carbonates of SW Paradox Basin, Honaker Trail, Utah. AAPG Bull., 74, p. 663. Goodwin, P.W. and Anderson, E.J., 1985. Punctuated aggradational cycles: A general hypothesis of episodic stratigraphic accumulation. J. Geol., 93: 515-533. G6rt~r, N., Oktay, F.Y., Seymen, I. and Seng6r, A.M.C., 1984. Paleotectonic evolution of the Tuzg61t~ Basin complex, Central Turkey: sedimentary record of a Neo-Tethyan closure. In: J.E. Dixon and A.H.F. Robertson (Editors), The Geological Evolution of the Eastern Mediterranean. Geol. Soc. London, pp. 467-482. G6rfir, N., Seng6r, A.M.C. and Oktay, F.Y., 1989. Sedimentological evolution of Central Anatolian basins and the assembly of the alpine tectonic collage in Turkey. In: Tethyan Tectonostratigraphic Terrane-models Tested, EUG V. Terra Abstr., 55. Grotzinger, J.P., 1986. Cyclicity and paleoenvironmental

329

dynamics, Rocknest Platform, northwest Canada. Geol. Soc. Am. Bull., 97: 1208-1231. Hays, J.D., Imbrie, J. and Schackleton, J.J., 1976. Variations in the earth's orbit: pacemaker of the Ice Ages. Sciences, 194: 1121-1132. James, N.P., 1984. Shallowing-upward sequences in carbonates. In: R.G. Walker (Editor), Facies Models.Geosci. Can. Repr. Ser., 1 (2nd Ed.): 213 228. Kendall, C.G.St.C. and Schlager, W., 1981. Carbonates and relative changes in sea level. Mar. Geol., 44: 181-212. Ko~yigit, A., 1991. An example of an accretionary fore-arc basin from northern Central Anatolia and its implications for the history of subduction of Neo-Tethys in Turkey. Geol. Soc. Am. Bull., 103: 22-36. Ko~yigit, A. and L~inel, T., 1987. Geology and tectonic setting of Aka region, Ankara. Middle East Tech. Univ. J. Pure Appl. Sci., 20:35 57. Koqyigit, A., IDzkan, S. and Rojay, F.B., 1988. Examples from the fore-arc basin remnants at the active margin of northern Neo-Tethys; development and emplacement ages of the Anatolian Nappe, Turkey. Middle East Tech. Univ. J. Pure Appl. Sci., 21: 183-210. Laming, D.C., 1966. Imbrication, paleocurrents and other sedimentary features in the Lower New Red Sandstone, Devonshire, England. J. Sediment. Petrol., 36:940 959. Nemkov, G.I., 1960. Le diamorphisme chez les Nummulites. Vopr. Micropal. SSSR, 3 : 5 0 - 6 6 (translated by BRGM, Orl6ans). Nemkov, G.I., 1962. Remarques sur la pal6odcologie des Nummulites. Vopr. Micropal. SSSR, 6:64 72 (translated by BRGM, Orl6ans). Ricken, W., 1985. Epicontinental marl-limestone alternation: event deposition and diagenetic bedding (Upper Jurassic southwest Germany). In: U. Bayer and A. Seilacher (Editors), Sedimentary and Evolutionary Cycles. Springer, Berlin, pp. 127-162. Seng6r, A.M.C. and Yilmaz, Y., 1981. Tethyan evolution of Turkey: a plate tectonic approach. Tectonophysics, 75: 181-241. Sirel, E. and Gttnd~iz, H., 1976. Description and stratigraphic distribution of some species of the genera Nummulites, Assilina and Alveolina from the Illerdian, Cuisian and Lutetian of Haymana region (S. Ankara). Bull. Geol. Soc. Turkey, 19: 31-44. Strasser, A., 1988. Shallowing-upward sequences in Purbeckian peritidal carbonates (lowermost Cretaceous, Swiss and French Jura Mountains). Sedimentology, 35: 369-383. Onalan, G., Y~iksel, V., Tekeli, T., G6nen~, O., Seyirt, Z. and Ht~seyin, S., 1976. Upper Cretaceous-Lower Tertiary stratigraphy and paleogeographic evolution of Haymana-Polatli region (SW Ankara). Bull. Geol. Soc. Turkey, 19:159-176 (in Turkish). Wilson, J.L., 1975. Carbonate Facies in Geologic History. Springer, Berlin, 471 pp. Wright, V.P., 1986. Facies sequences on a carbonate ramp: the Carboniferous Limestones of south Wales. Sedimentology, 33:221 241. Yt~ksel, S., 1970. Etude g6ologique de la rdgion d'Haymana (Turquie Centrale). Thesis. Univ. Nancy, 179 pp.