Earliest Triassic microbialites in rk Dag ... - Wiley Online Library

Sedimentology (2011) 58, 739–755

doi: 10.1111/j.1365-3091.2010.01181.x

Earliest Triassic microbialites in C ¸ u¨ru¨k Dag, southern Turkey: composition, sequences and controls on formation STEVE KERSHAW*, SYLV IE CRASQUIN, MARIE-BE´ ATRICE FOREL, CARINE RANDON, PIERRE-YVES COLLIN, ERDAL KOSUN§, SYLVAIN RICHOZ– and AYMON BAUD** *Institute for the Environment, Halsbury Building, Brunel University, Kingston Lane, Uxbridge, Middlesex UB8 3PH, UK (E-mail: [email protected]) CNRS-UMR 7207, CR2P ‘‘Centre de Recherche sur la Paleóbiodiversite´ et les Paleóenvironnements’’, Universite´ Pierre et Marie Curie–Paris 6, T. 46–56, E. 5, Case 104, 4 Place Jussieu, 75252 Paris Cedex 05, France CNRS–UMR 7072, IsTEP (Institute of Earth Sciences of Paris), Universite´ Pierre et Marie Curie–Paris 6 University, T. 56–66, E. 5, Case 117, 4 Place Jussieu, F-75252 Paris Cedex 05, France §Department of Geology, Akdeniz University, 07058 Campus Antalya, Turkey –Commission for the Palaeontological and Stratigraphical Research of Austria, Austrian Academy of Sciences c/o Institute of Earth Sciences, University of Graz, Heinrichstrabe 26, 8010 Graz, Austria **BGC, Rouvraie 28, CH-1018 Lausanne, Switzerland Associate Editor – Daniel Ariztegui ABSTRACT

The Permian–Triassic Boundary sequence at C ¸ u¨ru¨k Dag, near Antalya, Turkey, begins with a major erosion surface interpreted as being the Late Permian lowstand, on which lies ca 0Æ4 m of grainstone/packstone composed of ooids, peloids and bioclasts. Most ooids are superficial coats on fragments of calcite crystals presumed to be eroded from crystal fans which are no longer present. The erosion surface is smooth and shows no evidence of dissolution; the grainstone/packstone contains intraclasts of the underlying wackestone, proving erosion. Next are 15 m of microbialite comprised of interbedded stromatolites, thrombolites, plus beds of planar limestones with small-scale erosion. The latter comprise a complex interlayering of stromatolitic, thrombolitic and peloidal fabrics and precipitated crystal fans, which form a hybrid of microbialite and inorganic carbonate, together with bioclastic debris and micrite. The C ¸ u¨ru¨k Dag microbialite sequence is repetitious; the lower part is more complex, with abundant stromatolites and hybrid microbialites. Some of the stromatolites are themselves hybrids composed of peloids and crystal fans. In the upper part of the sequence stromatolites are missing and the rock is composed mostly of recrystallized thrombolites that develop upwards from tabular to domal form. The domes form directly below small breaks in microbialite growth where very thin shelly micrites and grainstones/ packstones are deposited. Repetition of facies may be controlled by sea-level change; a deepening-up model is consistent with the evidence. Stromatolites (with abundant crystal fans) dominate in shallower water, deepening through hybrid microbialite and interlayered sediments to thrombolite, probably no more than a few tens of metres deep, followed by breaks and renewal of microbialite growth. An interpretation of open marine fully oxygenated waters for microbialite growth is consistent with ongoing parallel work that has identified Bairdioid ostracods in the microbialite, a group known to be open marine. However, other researchers have proposed low oxygen conditions for 2010 The Authors. Journal compilation 2010 International Association of Sedimentologists

739

740

S. Kershaw et al. Permian–Triassic boundary facies globally, so work continues to confirm whether the C ¸ u¨ru¨k Dag microbialite grew in dysoxic or normally oxygenated conditions. The principal stimulus for post-extinction microbialites is likely to be carbonate supersaturation of the oceans. The microbialite sequence is overlain by a further 25 m of grainstone/packstone (without microbialite), followed by Early Triassic shales. Overall, microbialites form a thin aggradational sequence during an overall relative sea-level rise, consistent with global eustatic rise following the Late Permian lowstand. Keywords Hybrid microbialite, mass extinction, microbialite, stromatolite, thrombolite, Triassic.

INTRODUCTION Microbialites in the Permian–Triassic Boundary (PTB) sequence in C ¸ u¨ru¨k Dag, near Antalya, southern Turkish Taurides (Fig. 1), are one of a number of deposits of microbialites formed worldwide after the end-Permian mass extinc-

A A

PalaeoTethys

10 km

TURKEY

tion. These deposits developed in a facies complex of changing environmental conditions. The microbialites of this study were included together with oolites as ‘calcimicrobial caprock’ by Baud et al. (2005) and illustrated in Crasquin et al., (2009). Although PTB microbialites are wellknown, a continuing problem is precise determi-

B

S Armutcuk

U

ANTALYA

Uçoluk

K

S

Panthalassa

*

U

U

Pangaea

U

CD

U

S

Oyacuk

Kemer

S

U = upwelling S = sinking

NeoTethys

Gedelma

C

D

Fig. 1. (A) Map of Turkey, showing location of C ¸ u¨ru¨k Dag south-west of Antalya, near Kemer ‘K’; see inset for access ¨c to site (asterisk) via a steep unmade road east of the village of U ¸oluk. Detailed geology is provided by Baud et al. (2005) and not repeated here. (B) Palaeogeographic position of C ¸ u¨ru¨k Dag ‘CD’ in western Tethys (map based on Crasquin-Soleau et al., 2001 and Kershaw et al., 2007), with modelled oceanographic characteristics (Kidder & Worsley, 2004). (C) Field photograph of C ¸ u¨ru¨k Dag, highlighting the location of the Permian–Triassic Boundary; the cliff is ca 400 m high. (D) View of Permian–Triassic Boundary transition facies on C ¸ u¨ru¨k Dag, conformable limestones [note person for scale (ca 1Æ8 m tall) in lower left hand corner, circled]. 2010 The Authors. Journal compilation 2010 International Association of Sedimentologists, Sedimentology, 58, 739–755

Earliest Triassic microbialites in southern Turkey nation of the conditions of growth of any individual deposit; PTB microbialite sequences vary from place to place regionally, even within one nappe in southern Turkey, and variation is clear globally. Thus, post-mass extinction microbialites were under the influence of both global and local environmental conditions at any one site (see also Kershaw et al., 2007). Therefore, it is important to attempt to account for microbialite formation in detail, on a site-by-site basis, to understand the processes operating on shallow shelves. A highresolution approach is probably the best way to test models for shallow ocean change in relation to both local and global effects. This study describes and interprets the detailed changes that occurred through the microbialite sequence at C ¸ u¨ru¨k Dag, which is a candidate for the most complex PTB microbialite sequence

worldwide. The aim is to determine the local environmental changes that took place through the development of these post-extinction microbialites, against the backdrop of global environmental changes.

METHODOLOGY This study was undertaken by high-resolution field logging (taking into account palaeontological content and sedimentary structures) and sampling in May 2008 and September 2009. Examination of about 50 polished hand specimens and 120 thin sections through the sequence, from the Late Permian erosion surface through to the top of the microbialite, provided data for facies analysis.

A

B

Shelly micrite with intraclasts

Permian-Triassic boundary

160

microbialite sequence at

150

Çürük Dag, Antalya, Turkey Thrombolite heads and breaks

15

14 6

5

3 Basal microbialite (thrombolite)

K O K A R K U Y U Fm

K O K A R K U Y U Fm

Domal and wavy stromatolite

13

4

P. Fm

70 60 50 40 30

9

20

8

10 cm 0

1 Late Permian wackestone with erosion surface

90

11

2

m

110

80 12

10

Oolite-bioclastic grainstone

130

100

Mini-stromatolite domes

Hybrid microbialite & interbedded carbonates

140

120

7

Tabular bedded thrombolite

741

Bioclastic Late Permian wackestone Intraclasts erosion surface

0

Fig. 2. (A) Log of Permian–Triassic Boundary microbialites at C ¸ u¨ru¨k Dag, from the major Late Permian erosion surface to microbialite top; P. Fm refers to Pamuçak Formation. (B) Detail of lower portion of sequence ca 200 m from the main log location, highlighting the four major facies (grainstone/packstones, stromatolite – two types, thrombolite and hybrid microbialite with interbedded carbonate sediment). 2010 The Authors. Journal compilation 2010 International Association of Sedimentologists, Sedimentology, 58, 739–755

742

S. Kershaw et al.

GEOLOGICAL SETTING The microbialites crop out in the Taurides tectonic complex and the reader is referred to Baud et al. (1997, 2005) and Crasquin et al. (2009) for details, including a geological map. In C ¸ u¨ru¨k Dag (translated as ‘rotten mountain’ due to erosion of shales in its lower part) Permian and Triassic limestones dip to the south at ca 50, and the PTB is accessed easily from the flat top of the mountain and lies at GPS: 3641¢324¢¢ N, 03027¢40.1¢¢ E. Rocks directly below the microbialite are the topmost part of the Pamuçak Formation, wackestones and grainstones/packstones of Late Permian age. The microbialite base marks the beginning of the Kokarkuyu Formation (see Baud et al., 2005 for details), 40 m thick at C ¸ u¨ru¨k Dag. Conodont biostratigraphy is not well-developed for C ¸ u¨ru¨k Dag; Isarcicella staeschei has been reported from the base of the Korkarkuyu Formation; 55 cm higher, Hindeodus parvus was found (Richoz, 2006). The occurrence of H. parvus above I. staeschei suggests that the former is near the top of the range of H. parvus, since H. parvus first appears in the Meishan Global Boundary

A

Stratotype Section and Point, thereby defining the base of the Triassic (Erwin, 2006). From this argument, the base of the microbialite (Kokarkuyu Formation) is already in the Triassic, but further work is needed to define the base of the Triassic in C ¸ u¨ru¨k Dag. Note that Richoz (2006) showed that the prominent negative excursion in the carbon isotope curve begins at the base of the oolite below the microbialite. In contrast, at some other sites, such as in Iran (Wang et al., 2007) (the nearest PTB microbialites to Turkey during that time), the negative change occurs within the microbialite, suggesting diachroneity of the beginning of microbialite growth (assuming that the negative excursion occurred synchronously in Turkey and Iran).

THE PERMIAN–TRIASSIC BOUNDARY ¨ RU ¨ K DAG SEQUENCE AT Ç U In C ¸ u¨ru¨k Dag, PTB facies are visible in several places on the west-facing side of the mountain, and show variations laterally and vertically. Several stages of development of the microbialites (Fig. 2) are described below, from bottom to top.

C

B

Fig. 3. (A) and (B) Field views of the Late Permian erosion surface below the microbialite sequence, showing its smooth nature. The large divisions on the scale are centimetres. (C) Vertical thin section view of the erosion surface, showing eroded shelly wackestones and overlying grainstone/packstones, and wackestone intraclast (arrowed) in the grainstone/packstone. The smooth erosion surface is consistent with physical erosion, without clear evidence of dissolution. Scale bar is 5 mm. 2010 The Authors. Journal compilation 2010 International Association of Sedimentologists, Sedimentology, 58, 739–755

Earliest Triassic microbialites in southern Turkey A

743

B

C

C

D

Fig. 4. Details of grainstone-packstone below the microbialite. (A) and (B) Many superficial ooids have nuclei of rounded fragments of single crystals. (C) Packstone fabric in parts of the unit. (D) Peloidal sediment and shell (gastropod, arrowed) in parts of the unit. Scale bar for (A), (C) and (D) is 1 mm; for (B) it is 0Æ5 mm.

The latest Permian limestones contain a prominent smooth erosion surface (Figs 2 and 3), probably the Late Permian sea-level lowstand. There is no evidence of dissolution of this surface, upon which was laid a 0Æ35 to 0Æ45 m thick grainstone/packstone unit composed of a mixture of shell fragments, peloids, spherical calcitic grains with superficial ooid coats (Figs 3 and 4) and uncommon fully formed ooids. In most cases, the calcitic grains are single crystals and cannot be regarded simply as recrystallized ooids; they are interpreted here as reworked fragments of calcite crystals precipitated after the mass extinction. No in-place calcite crystals have been found in this deposit, or on the Late Permian erosion surface, consistent with erosion of the sea floor in a turbulent shallow water environment. The grainstone/packstone is overlain, with sharp contact, by microbialites. Baud et al. (2005) described a thin (13 cm thick) laterally

impersistent thrombolite, bound above and below by distinct breaks that are likely to be erosion surfaces. However, these surfaces are affected by stylolites directly above the grainstone/packstone. In the field, this unit was found only in one small section of the boundary facies in C ¸ u¨ru¨k Dag, which is not figured here except in Fig. 2. The first major microbialite is stromatolite, developed on a presumed erosion surface comprising the thin thrombolite and grainstone/packstone described above. The stromatolite has varying gross morphology as masses with wavy laminae, prominent domes and narrow columns (Figs 2 and 5). Much of the stromatolite is layered micritic material indicative of sediment-trapping by cyanobacterial activity as in many modern stromatolites (Reid et al., 2000) (Fig. 6A to C); some spheroidal fabrics are present (Fig. 6D), but a common additional fabric is composed of layers of sparite (Fig. 7A to C), commonly interlayered with micritic stromatolite. Thus, much of the

2010 The Authors. Journal compilation 2010 International Association of Sedimentologists, Sedimentology, 58, 739–755

744

S. Kershaw et al. B

A

C

D

Fig. 5. Field photographs of lowermost stromatolites, showing: (A) and (B) domal and wavy and (C) columnar forms. In (A) and (B), the scale bar is 85 mm long; in (C), the large divisions on the scale are 10 mm. (D) Thin section view showing stromatolitic lamination. Scale bar is 10 mm.

stromatolite fabric is a mixture of micritic and peloidal (presumed to be cyanobacterially mediated) and cloudy precipitated cements (probably inorganic but possibly microbially mediated). Directly overlying the lowest stromatolite is the first of several beds of mostly planar laminae (Fig. 8). This deposit is comprised of a variety of fabrics, including stromatolite, thrombolite, crystal precipitates and peloidal sediment (Fig. 9). Stylolitic contact with underlying stromatolites prevents determination of the nature of the depositional contact. Riding (2008) introduced the concept of hybrid microbialites consisting of microbial and inorganically precipitated components within stromatolites; thus hybrid microbialite is an appropriate term for these PTB microbialites in C ¸ u¨ru¨k Dag. In the field, small-scale erosion surfaces are visible (Fig. 8C). In one place the lamination is curved, resembling,

but not proving, hummocky cross-stratification (HCS) (Figs 2 and 8D). However, in some thin sections, erosion surfaces are difficult to prove; part of the fabric appears to be eroded (Fig. 9B), but these features may instead be due to uneven microbial micrite. Shelly grainstone/packstone containing well-preserved crinoid columnals occurs at 2Æ65 m (Fig. 9C and D), but is not found anywhere else in the section. Sedimentary structures suggest that this facies represents a normally low energy environment subject to episodic higher energy, such as storms. Between ca 1Æ1 to 1Æ18 m is a unit of small stromatolite columns occurring at only one horizon (Figs 2 and 10); this was found in two sites 200 m apart (compare Fig. 2A and B) and therefore is probably continuous across at least that distance. Individual stromatolite columns are up to 3 cm diameter and 10 cm tall. This layer



C

745

B

D

Fig. 6. Thin section views of lowermost stromatolites. (A) to (C) Lamination with interruptions in growth. (B) Stromatolite with possible filaments. (D) Possible spheroids in stromatolite. Scale bar for (A) is 5 mm; for (B), (C) and (D) it is 1 mm.

indicates unique conditions for a short period in the development of the microbialite sequence. In detail the stromatolite columns are composed of a mixture of peloidal stromatolite with precipitated fans in alternating layers and therefore are hybrid stromatolite (Fig. 10C and D). At several horizons there are tabular thrombolites which develop into thrombolitic heads in some cases (Fig. 11). Thrombolites consist of an open framework of presumed microbial growth (Figs 12 and 13), subsequently infilled with cements and sediments, some of which are geopetal. Thrombolites are all recrystallized, but the margins of the structure are well-defined, so the meso-architecture and micro-architecture of the clotted fabric is shown clearly (Figs 12 and 13). Thin deposits of non-microbial shelly and intraclastic wackestones, packstones and grainstones/ packstones occur directly above three of the thrombolite units (Figs 2 and 14).

Examination of Fig. 2 reveals that the microbialites in C ¸ u¨ru¨k Dag exhibit a repetitive character, with more complex changes in the lower part of the sequence, the details of which are affected by stylolites in some places. Vertical changes in the lower part terminate sharply at the top of the lowest thrombolite head, but no major erosion has been identified. In places the mesoclotted appearance of thrombolites grades into a columnar form of thrombolite, similar in external form to the digitate dendrolites described by Kershaw et al. (2007) in Sichuan, south China. Thus, the architecture of the thrombolites in C ¸ u¨ru¨k Dag is complex in detail. The upper part, above 6 m on the log (Fig. 2), consists mostly of thrombolites with some thin planar hybrid microbialite and deposited sediment. At two further levels there are shelly and intraclastic micrites deposited on the thrombolite head top surfaces (Figs 2 and 14). The total


746

S. Kershaw et al.

A

B

C

D

Fig. 7. (A) to (C) Thin section views of precipitated cements; all these were found in stromatolitic layers and form interlayered cements with stromatolites. Thus stromatolites are hybrid structures containing both cement fans and apparent microbial fabrics. (D) Enlargement of Fig. 6A, showing cement growth on stromatolite laminae, before micrite deposition, indicating a hiatus between cessation of stromatolite growth and sediment deposition, possibly in a deepening setting, discussed in the text. Scale bar for (A) is 5 mm; for (B), (C) and (D) it is 1 mm.

thickness of stromatolites and thrombolites is ca 15 m. The uppermost thrombolite is overlain by ca 25 m of grainstone/packstone, described by Baud et al. (2005) as oolite; however, the actual thickness of this may be affected by faults visible on the cliff face, but which are difficult to detect in logged section. No faults were found cutting the microbialite.

DISCUSSION

Oolite as a form of microbialite? Previous work (Baud et al., 2005) interpreted earliest Triassic microbialite limestones at C ¸ u¨ru¨k Dag to comprise 40 m of limestone, which is the thickest post-extinction microbialite so far described. However, Baud et al. (2005) grouped

grainstones/packstones (which they described as oolite) with microbialites of the earliest Triassic of C ¸ u¨ru¨k Dag as ‘calcimicrobial caprock’ on Late Permian limestones, thereby taking into account the possible microbial contribution to ooid formation (see Flu¨gel, 2004 for discussion). Although ooids may form under at least partial microbial control, perhaps by microbial mediation of carbonate deposition, there are two problems with the above interpretation in C ¸ u¨ru¨k Dag. Firstly, most ooids are superficial ooids, with concentric ooids being uncommon; and much of the rock is composed of bioclasts, so these deposits are not simply classified as oolite. Secondly, if ooids are included with microbialites, without distinction from stromatolites and thrombolites (in C ¸ u¨ru¨k Dag), then there are problems in discriminating facies and making appropriate interpretations of environmental controls. There



C

747

B

D

Fig. 8. Field photographs of hybrid microbialite with interbedded carbonate sediments. (A) and (C) Lowermost stromatolite (55 to 110 cm of log, see Fig. 2). (B) Loose specimen showing cross-lamination in the sediment, indicating energetic deposition in at least part of the fabric. (D) In-place example at 7Æ7 m (Fig. 2) showing possible HCS. Scale bar is 85 mm long.

are also numerous well-known large oolite deposits throughout the rock record which are unrelated to mass extinctions, yet they are not considered to be microbialite. In addition, some other postextinction microbialites are not interbedded with oolites, for example in Sichuan and Guizhou, southern China (Kershaw et al., 1999, 2007). Thus, it is argued that there is no automatic link between oolites and mass extinction, or between oolites and microbialites, and the (oolitic) grainstone/packstone is considered to be separate from microbialite in these post-extinction facies. Nevertheless, such grainstones/packstones are potentially important as indicators of increased carbonate saturation during mass extinction episodes, when calcification of metazoans is suppressed due to a lack of shelly organisms (Groves & Calner, 2004), and are important in facies analysis of the sequence. Thus, maintaining a distinction from microbialites allows facies complexes containing both types of limestones to

be more fully assessed. Therefore, in C ¸ u¨ru¨k Dag, there is only ca 15 m of microbialite, assuming that the microbialite is not faulted.

Repetition of microbialites and models of growth This study reveals repetitious growth of PTB microbialites in the study area. All types of microbialite calcification are probably driven by high levels of carbonate saturation (Riding, 2005), which has also been cited as the reason for abundant oolites in earliest Triassic sediments (Groves & Calner, 2004; Calner, 2005). Changes in ocean saturation may be expected to take place over longer time scales than the presumed rapid facies changes observed at this site. Therefore, because of the dominance of microbialites throughout the basal Triassic in C ¸ u¨ru¨k Dag, it is very unlikely that variations in saturation levels caused the small-scale microbialite repetitions.


748

S. Kershaw et al.

A

C

B

D

Fig. 9. (A) and (B) Typical appearance of hybrid microbialite in thin section, at 1Æ4 m in the log in Fig. 2. Detailed labelling in (B). Note the contrast between the (interpreted) eroded peloidal sediments, stromatolitic peloidal fabrics, cement fans and thrombolite. (C) Shelly grainstone/packstone at 2Æ65 m (Fig. 2). (D) Detail of shelly grainstone/ packstone at 2Æ65 m containing several crinoid columnals (with syntaxial cements); note the well-preserved character of the columnals. Scale bar for (A) is 5 mm, for (B) and (C) it is 1 mm; and for (D) it is 0Æ5 mm.

Also, the microbialites are underlain and overlain by grainstone/packstone, so it is likely that the entire microbialite sequence is associated with shallow water processes, and sea-level change is the simplest explanation for the cause of these repetitions. The outcrops currently sit in a nappe sequence and, although tectonic activity may have affected the region where the microbialites grew, sea-level rise was rapid during the Early Triassic and was likely to mask any tectonic vertical motion. Modern microbialites thrive in the brackish conditions of Lake Clifton, the hypersaline conditions of Hamelin Pool in Shark Bay, and the

open marine environments of the Bahamas. Microbialites in general apparently are influenced little by salinity but instead are stimulated by raised carbonate saturation, a point emphasized by Burne & Moore (1987) and Riding (2000) in their descriptions of microbialites from different environments. McNamara (2009) specifically noted that salinity is unimportant in microbialite growth. Therefore, it is difficult to use modern analogues to provide the precise information necessary to interpret the PTB microbialites. However, preliminary evidence from ostracods in work in progress by Forel et al. shows that the depositional environments were all open marine.



C

749

B

D

Fig. 10. Small stromatolite domes at 110 cm. (A) Field view; scale divisions 10 mm. (B) Polished sample; large scale divisions are 10 mm. (C) and (D) Thin section views showing that most of these domes are composed of cloudy precipitated cement fans and imply a hybrid organic–inorganic fabric. Scale bars are 5 mm.

Bairdioidea forms, well-known as open marine ostracods, are abundant; there is no indication of restricted facies based on an ostracod study. Thus, the vertical changes in the microbialites of C ¸ u¨ru¨k Dag are interpreted as due to environmental change in open marine conditions; oxygen levels are discussed below. However, whether the repeated microbialites represent shallowing-up or deepening-up changes is difficult to prove from the current evidence. Post-extinction stromatolites in the Early Triassic are known from open marine carbonate platform systems (for example, in Iran; Baud et al., 1997; Wang et al., 2007) and in Hungary there is a strong argument for an outer ramp setting (Hips & Haas, 2006), therefore being in relatively deeper water. However, stromatolites in general are also common components of lagoonal environments well-known from such sites as Shark Bay, Australia (McNamara, 2009). Lake Clifton in Western Australia is well-known for its large domal thrombolites in very shallow water (Burne & Moore, 1987), in contrast to

thrombolites in deeper water in other lakes (Lake Van, Turkey, Kempe et al., 1991; Pavilion Lake, British Columbia, Laval et al., 2000). By analogy, in C ¸ u¨ru¨k Dag, in a deepening-up model, stromatolites may have grown in very shallow water, and the thrombolites in deeper water; or the reverse scenario prevailed in a shallowing-up model. Below is an interpretive outline of these two alternative pathways, followed by a discussion which favours only one of them.

Pathway 1: Shallowing-up model Shallow-marine grainstone/packstone bars developed on top of the Late Permian erosion surface, with a thin thrombolite in shallow water. Rapid deepening stimulated stromatolite growth in open marine conditions (however, this does not account for the small-scale, sharply bounded changes in stromatolite morphology; see Fig. 2). Shallowing led to tabular and then domal thrombolites, with termination at the tops of domal thrombolite heads (but note that no erosion of these heads has been observed). A second episode


750

S. Kershaw et al.

A

B

C

Fig. 11. Field photographs of thrombolites. (A) Tabular thrombolite changes upwards to domal head at 8Æ5 m in the log in Fig. 2. The length of the yellow scale is 0Æ5 m. (B) Typical thrombolitic mesoclot fabric. (C) Transverse section of tabular thrombolite showing some organization into a columnar structure. Scale in (B) and (C) is 85 mm long.

of rapid deepening stimulated renewed stromatolite growth for the lower part of the sequence, while in the upper part the lack of stromatolites suggests that the deepening was less. The microbialite sequence was terminated by a return to the shallow waters of a grainstone/packstone formation that is maintained for ca 25 m above the microbialites, suggesting steady subsidence or sea-level rise, keeping shallow water for these overlying deposits.

Pathway 2: Deepening-up model Shallow water grainstone/packstone bars developed on top of the Late Permian erosion surface. A small sea-level rise in the earliest Triassic led to stromatolite growth in shallow environments. The lowest stromatolite unit consists of several different architectural sub-units which could be interpreted as a lagoonal setting, subject to small scale, rapid sea-level fluctuations. Thus, the stromatolites may be formed with little sea-level change in an open marine environment behind

shoals of mobile grainstone/packstone bars, or just seaward of the bars in shallow sub-tidal conditions. However, in this model the thin impersistent thrombolite (Fig. 2), described by Baud et al. (2005), may represent a relic of a largely eroded microbialite that formed before the lowest stromatolite, thereby making the sequence of events at this site more complex. The lowest stromatolites may have formed on a hardground of pre-existing limestones and, therefore, may not have been near any active ooid shoals, potentially accounting for the lack of ooids and bioclasts between stromatolite columns. Continued sealevel rise led to open water where the hybrid microbialites and interbedded sediments developed, subject to storm action creating low-angle cross-stratification, including possible HCS. Further deepening stimulated thrombolite growth in more open marine and, presumably, quieter conditions. At first, thrombolite growth was tabular, then, as accommodation space increased, water deepened and light levels fell, these developed



C

751

B

D

Fig. 12. (A) Vertical section of a layer of thrombolite showing minor erosion at the top and overlain intraclast-rich micrite. (B) Vertical section of thrombolite with open fabric. (C) Thin section view of thrombolite frame mostly infilled with micrite. (D) Transverse section of mesoclot structure. Scale bar for (C) and (D) is 5 mm.

into domal thrombolites. Thrombolite growth was terminated either by storm action and deposition of sediment on the thrombolites or, alternatively, by deepening water that lowered light levels sufficiently to impede thrombolite growth, after which sediments accumulated. Rapid shallowing and re-establishment of stromatolites in the lower part of the sequence followed. As the upper part of the sequence lacks stromatolites, perhaps the later sea-level fluctuations did not always result in the shallowest water depositional environments. If this is true, then there was an overall deepening trend up-section through the microbialite. The microbialite sequence is terminated by a return to the shallow waters of grainstone/ packstone formation that is maintained for ca 25 m above the microbialites, suggesting continued shallow water through steady subsidence or sea-level rise. The end of microbialite growth may be due to a fall in carbonate saturation of the water, higher energy or by grazing pressure from herbivores.

The deepening-up model is thought to be more consistent with the evidence for the following reasons. Firstly, the lowest part of the microbialite contains the most complex components, dominated by stromatolites, so it could be expected to be a consequence of small changes of sea-level in a very shallow setting. The lower part also has abundant precipitated calcite fans associated with the stromatolites. These fans are compatible with extreme carbonate saturation in the shallowest waters of the presumably warm, low-latitude setting of the Taurides in the PTB transition. Furthermore, the formation of stromatolites with micritic sediments and no ooids is consistent with a rising base level where the ooid shoals are likely to have been migrating landwards and, therefore, are less likely to be available to supply ooids to the area of stromatolite formation. Nevertheless, the lowest stromatolites may have grown on lithified and eroded grainstones/packstones, as discussed earlier. Whatever the precise conditions at the base of the first


752

S. Kershaw et al.

A

B

C

D

Fig. 13. (A), (B) and (D) Thin section details of thrombolites, showing primary frame, partially infilled by micrite. (C) Microgastropod encased in thrombolite. Scale bars for (A), (B) and (C) are 1 mm; for (D) the scale bar is 0Æ5 mm.

stromatolite, all the stromatolite deposits in the sequence are associated with micrites, reinforcing the view that they developed in low energy environments, interpreted as being deep (or at least quiet) enough to be out of reach of ooid shoals. Secondly, if the sequence represents a shallowing-up process, then there should be more coarsegrained material associated with the thrombolites. The thin, shelly, intraclastic wackestones deposited above the thrombolites may have been imported by storm action that terminated thrombolite growth. Thirdly, laminated sediments between stromatolite and thrombolite beds contain a hybrid of microbialite and deposited sediments, as well as precipitated crystal fans. The sedimentary structures of small-scale erosion (Fig. 8C) (and possible HCS) are consistent with slightly deeper water of the shoreface which is subject to episodic storms. In this view, the thrombolites would have formed as the deepest deposit.

Fourthly, global sea-level rose after the Late Permian lowstand, consistent with deepeningup; although this is somewhat circumstantial (because local rapid tectonic uplift could theoretically override a global signal), it is clear that the microbialite sequence aggraded to form 15 m of deposit. The sequence must have been laid down either on a subsiding sea floor or under steady sea-level rise, with deposition to fill the accommodation space. The deepening-up model is summarized in Fig. 15.

Oxygenation levels of the microbialite sequence There is growing evidence of lowered oxygen in shallow water facies associated with upwelling of deep ocean waters and elevation of the chemocline (Wignall & Twitchett, 1996; Dolenc et al., 2001; Algeo et al., 2007; Kershaw et al., 2007; and references therein). However, in C ¸ u¨ru¨k Dag, the microbialite sequence contains bivalves, gastro-


Earliest Triassic microbialites in southern Turkey B

A

C

753

D

Fig. 14. Sediment layers between microbialites, outlined with white lines in (A) and (B). (A) Shelly and intraclast micrite on thrombolite head at 8Æ5 m (Fig. 2), overlain by a tabular thrombolite of the next thrombolite unit. (B) Micrite layers between thrombolites at 6Æ8 m (Fig. 2). The yellow rule in (A) and (B) is 0Æ5 m long. (C) Polished sample of shelly and intraclast micrite from sediments shown in (B). The large divisions on the scale bar are 10 mm. (D) Thin section of intraclast wackestone from sample location ‘21’ shown in (B); scale bar is 1 mm.

Fig. 15. Model of microbialite facies at C ¸ u¨ru¨k Dag, showing thrombolite heads in the deepest water. Note that the contact between the ooid shoal and the stromatolites is interpreted as an erosion surface in the field, so that the stromatolites may have initiated on a solid surface. See text for further explanation. 2010 The Authors. Journal compilation 2010 International Association of Sedimentologists, Sedimentology, 58, 739–755

754

S. Kershaw et al.

pods and crinoids, presumed to have been living nearby because many shells appear to be wellpreserved (one gastropod was found encased in thrombolite; Fig. 13C). Beatty et al. (2008) viewed well-developed ichnofossils in early Triassic deposits as indicators of areas where recovery from extinction was more rapid, perhaps due to more favourable conditions, particularly in the high latitude location of their study. Whether or not this argument also applies to the tropical setting of central Tethys (where C ¸ u¨ru¨k Dag is located, see Fig. 1) requires further work, but it is possible that recovery from mass extinction was faster than has been interpreted previously. Thus, the location of C ¸ u¨ru¨k Dag may well have escaped the worst effects of poor oxygenation (as supported by abundant open marine ostracods); reinforcing the view that carbonate saturation is the principal cause of microbialite growth after the mass extinction. Also, recent work on ostracods from China indicates that microbialites in the Great Bank of Guizhou did not grow in low oxygen conditions (Forel et al., 2009), in contrast to those in Sichuan, for which there is evidence of lowered oxygen (Crasquin-Soleau & Kershaw, 2005), so there is a more complex situation which requires more research to resolve. However, in almost all cases of microbialites in C ¸ u¨ru¨k Dag, a thin isopachous cement is present on the surfaces of microbialite carbonate, deposited prior to sedimentary micrite (for example, Figs 7D, 9B and 13A). Thus, there is a time gap between the end of individual microbialite growth and the deposition of sediment. It is possible that the oxygen levels of the shallow shelf fluctuated over a short time scale, so that the microbial structures developed in low oxygen conditions, while the ostracods which are deposited between thrombolite branches and stromatolite heads represent more oxygenated water. More research is required to investigate such small-scale changes, which may originate in climatically driven oceancirculation fluctuation below the resolution of current stratigraphy.

CONCLUSIONS 1 The post-extinction microbialite sequence at C ¸ u¨ru¨k Dag is much more complex than previously recognized. 2 Microbialite form is divided broadly into stromatolite, thrombolite and hybrid microbialite, indicating significant changes in environmental controls as the sequence developed, probably due

to sea-level fluctuations. Some stromatolites are composed of peloidal material and precipitated crystal fans formed in close association, forming hybrid microbialites. 3 The microbialites show repetitious changes which may be interpreted in terms of sea-level change after the Late Permian lowstand, when sea-levels rose globally. Vertical changes in the microbialite facies are more consistent with deepening-up processes, so that stromatolites formed in shallow water, and thrombolite heads were the deepest component, probably a few tens of metres below sea-level. 4 This work has permitted exploration of the controls on post-extinction microbialite facies in greater detail than has previously been attained and emphasizes the need for high-resolution sampling in the interpretation of the very thin microbialite sequences found after the endPermian mass extinction.

ACKNOWLEDGEMENTS We are very grateful to Rainer Brandner and Charles Henderson for stimulating field discussion. This paper is a contribution to IGCP572 ‘‘Restoration of marine ecosystems following the Permian-Triassic mass extinction: lessons for the present’’. We are grateful to Paul Wignall and an anonymous referee for comments on the manuscript.

REFERENCES Algeo, T.J., Ellwood, B., Nguyen, T.K.T., Rowe, H. and Maynard, J.B. (2007) The Permian-Triassic boundary at Nhi Tao, Vietnam: evidence for recurrent influx of sulfidic watermasses to a shallow-marine carbonate platform. Palaeogeogr. Palaoeclimatol. Palaeoecol., 252, 304–327. Baud, A., Cirilli, S. and Marcoux, J. (1997) Biotic response to mass extinction: the Lowermost Triassic microbialites. Facies., 36, 238–242. Baud, A., Richoz, S. and Marcoux, J. (2005) Calcimicrobial cap rocks from the basal Triassic units: western Taurus occurrences (SW Turkey). C R Palevol, 4, 569–582. Beatty, T.W., Zonnefeld, J.-P. and Henderson, C.M. (2008) Anomalously diverse Early Triassic ichnofossil assemblages in northwest Pangea: a case for a shallow-marine habitable zone. Geology, 36, 771–774. Burne, R.V. and Moore, L.S. (1987) Microbialites: organosedimentary deposits of benthic microbial communities. Palaios, 2, 241–254. Calner, M. (2005) A Late Silurian extinction event and anachronistic period. Geology, 33, 305–308. Crasquin, S., Baud, A., Kershaw, S., Richoz, S., Kosun, E. and Forel, M.B. (2009) The Permian-Triassic transition in the


Earliest Triassic microbialites in southern Turkey Southwestern Taurus Mountains (South Turkey). In: IGCP 572 annual Meeting & Field Workshop in southern Turkey, Antalya, Sept. 2–6, 2009. (Ed. A. Baud), Field Guide Book. 48 pp. Crasquin-Soleau, S. and Kershaw, S. (2005) Ostracod fauna from the Permian-Triassic boundary interval of South China (Huaying Mountains, eastern Sichuan Province): palaeoenvironmental significance. Palaeogeogr. Palaeoclimatol. Palaeoecol., 217, 131–141. Crasquin-Soleau, S., Broutin, J., Besse, J. and Berthelin, M. (2001) Ostracodes and palaeobotany from the Middle Permian of Oman: implications on Pangea reconstruction. Terra Nova, 13, 38–43. Dolenc, T., Lojen, S. and Ramovs, A. (2001) The PermianTriassic boundary in western Slovenia (Idrijca Valley section): magnetostratigraphy, stable isotopes and elemental variations. Chem. Geol., 175, 175–190. Erwin, D.H. (2006) Extinction. Princeton University Press, Princeton and Oxford, 296 pp. Flu¨gel, E. (2004) Microfacies of Carbonate Rocks: Analysis Interpretation and Application. Springer, Berlin, 976 pp. Forel, M.-B., Crasquin, S., Kershaw, S., Feng, Q.L. and Collin, P.-Y. (2009) Ostracods (Crustacea) and water oxygenation in the earliest Triassic of South China: implications for oceanic events at the end-Permian mass extinction. Aust. J. Earth Sci., 56, 815–823. Groves, J.R. and Calner, M. (2004) Lower Triassic oolites in Tethys: a sedimentologic response to the end-Permian mass extinction. Geol. Soc. America Abstracts with Programs, 36, 336. Hips, K. and Haas, J. (2006) Calcimicrobial stromatolites at the Permian-Triassic boundary in a western Tethyan section, Bu¨kk Mountains, Hungary. Sed. Geol., 185, 239–253. Kempe, S., Kazmierczak, J., Landmann, G., Konuk, T., Reimer, A. and Lipp, A. (1991) Largest known microbialites discovered in Lake Van, Turkey. Nature, 349, 605–608. Kershaw, S., Zhang, T. and Lan, G. (1999) A? microbialite crust at the Permian-Triassic boundary in south China, and its palaeoenvironmental significance. Palaeogeogr. Palaeoclimatol. Palaeoecol., 146, 1–18. Kershaw, S., Li, Y., Crasquin-Soleau, S., Feng, Q., Mu, X., Collin, P.-Y., Reynolds, A. and Guo, L. (2007) Earliest Triassic microbialites in the South China block and other areas: controls on their growth and distribution. Facies, 53, 409–425.

755

Kidder, D.L. and Worsley, T.R. (2004) Causes and consequences of extreme Permo-Triassic warming to globally equable climate and relation to the Permo-Triassic extinction and recovery. Palaeogeogr. Palaeoclimatol. Palaeoecol., 203, 207–237. Laval, B., Cady, S.L., Pollack, J.C., McKay, C.P., Bird, J.S., Grotzinger, J.P., Ford, D.C. and Bohm, H.R. (2000) Modern freshwater microbialite analogues for ancient dendritic reef structures. Nature, 407, 626–629. McNamara, K. (2009) Stromatolites. Geoscientist, 19, 16–22. Reid, R.P., Visscher, P.T., Decho, A.W., Stolz, J.F., Bebout, B.M., Dupraz, C., MacIntyre, I.G., Paerl, H.W., Pinckney, J.L., Prufert-Bebout, L., Steppe, T.F. and DesMarais, D.J. (2000) The role of microbes in accretion, lamination, and early lithification of modern marine stromatolites. Nature, 406, 989–992. Richoz, S. (2006) Stratigraphie et variations isotopiques du Carbon dans le Permien supe´rieur et le Trias infe´rieur de la Neóte´thys (Turquie, Oman et Iran). Me´moires de Geólogie (Lausanne), 46, 264. Riding, R. (2000) Microbial carbonates: the geological record of calcified bacterial-algal mats and biofilms. Sedimentology, 47(Suppl. 1), 179–214. Riding, R. (2005) Phanerozoic reefal microbial carbonate abundance: comparisons with metazoan diversity, mass extinction events, and seawater saturation state. Rev. Esp. Micropaleontol., 37, 23–39. Riding, R. (2008) Abiogenic, microbial and hybrid authigenic carbonate crusts: components of Precambrian stromatolites. Geologia Croatica, 61, 73–103. Wang, W., Kano, A., Okumura, T., Ma, Y., Matsumoto, R., Matsuda, N., Ueno, K., Chen, X., Kakuwa, Y., Gharaie, M.H.M. and Ilkhchi, M.R. (2007) Isotopic chemostratigraphy of the microbialite-bearing Permian–Triassic boundary section in the Zagros Mountains, Iran. Chem. Geol., 244, 708–714. Wignall, P.B. and Twitchett, R.J. (1996) Ocean anoxia and the end-Permian mass extinction. Science, 272, 1155–1158.

Manuscript received 21 October 2009; revision accepted 11 June 2010
