Integrated Stratigraphic, Geochemical, and Paleontological Late Ediacaran to Early Cambrian Records from Southwestern Mongolia

Home , Cyathus, First appearance datum, Monomorphichnus

Integrated late Ediacaran to early Cambrian records from southwestern Mongolia Integrated stratigraphic, geochemical, and paleontological late Ediacaran to early Cambrian records from southwestern Mongolia

Emily F. Smith†, Francis A. Macdonald, Tanya A. Petach, Uyanga Bold, and Daniel P. Schrag Department of Earth and Planetary Sciences, Harvard University, Cambridge, Massachusetts 02138, USA

ABSTRACT minths, cap-shaped fossils, protoconodonts, limited to a small number of localities with large and Salanacus to just below the positive peak uncertainties in both local and global strati- The Zavkhan terrane in western Mongolia 3p. This interpretation differs from previ- graphic correlations and the lack of absolute preserves thick, fossiliferous, carbonate-rich ous chronostratigraphic placements of first ages directly linked to biostratigraphic and che- strata that span the Ediacaran-Cambrian appearance data of genera between positive mostratigraphic data, particularly with respect transition. Measured stratigraphic sections excursions 1p and 2p. Using this level as the to critical data from Asia. Because the global and geological mapping of these strata, first appearance datum in Mongolia for these data set is biased by just a few localities, refin- which include the Zuun-Arts, Bayangol, genera, we replot global fossil first appearing local correlations and grounding them in Salaagol, and Khairkhan Formations, reveal ances. With this new compilation, there are regional geology are necessary before an accu- large lateral facies changes over short dis- still three distinct pulses of fossil first appear- rate global synthesis of fossil occurrence data tances that necessitate revisions to previous ances, as was suggested in previous compila- can be constructed and interpreted. lithostratigraphic correlations and biostrati- tions; however, we suggest that this pattern is The Ediacaran-Cambrian boundary is defined graphic range charts. Here, we integrate new controlled largely by regional sedimentation at the global bondary stratotype section and geological mapping and measured strati- and taphonomy rather than the rate of taxo- point (GSSP) as the first appearance datum graphic sections across the Zavkhan terrane nomic origination. (FAD) of the bilaterian trace fossil Treptich with high-resolution carbon isotope (δ13C) nus pedum (Landing, 1994), but the boundary chemostratigraphy, sequence stratigraphy, INTRODUCTION has also been associated with the FAD of small and biostratigraphy. Using these data, we shelly fossils and a large negative carbon iso- revise correlations within the Zavkhan ter- Since Charles Darwin’s observation of the tope excursion (Narbonne et al., 1994; Corsetti rane and place small shelly fossil and ichno apparently rapid appearance of fossils in what is and Hagadorn, 2000). These occurrences in the fossil horizons in a refined temporal and spa- now known as Cambrian strata (Darwin, 1859), rock record are generally assumed to be broadly tial framework. Finally, we integrate these the Precambrian-Cambrian boundary has been coincident, but both the precise and relative tim- data from Mongolia with absolute ages and widely regarded as one of the evolutionary pivot ing of these evolutionary milestones and per- chemostratig raphy from sections in Morocco points in the history of life. Despite the persist- turbations to the carbon cycle are not directly and Oman. Our revised age model suggests ing interest on this topic, the causes, triggers, constrained. The subsequent earliest Cambrian that deposition of the Bayangol Formation and tempo of change remain controversial. Fol- period is marked by FADs of small shelly fos- and the lower part of the Salaagol Formation lowing the extinction of Ediacaran biota and the sils in many clades and high-frequency oscilla- was limited to the Nemakit-Daldynian and calcifying metazoan Cloudina, the early Cam- tions in the carbon cycle (Maloof et al., 2010a), Tommotian Stages. brian (or Terreneuvian Series) encompasses an but again, there are few absolute ages tied to With the new correlations and age model, evolutionary interval characterized by increases these FADs or excursions. Because there is no we place the small shelly fossil first appear- in diversity and disparity (Marshall, 2006) that one section globally in which it is possible to ance datum in the basal Bayangol Formation coincide with multiple carbon isotope excur- integrate the ichnofossil record, body fossil instead of the basal Zuun-Arts Formation, sions with amplitudes of <8‰ (e.g., Zhu et al., record, carbon isotope chemostratigraphy, and which moves this horizon hundreds of meters 2006; Porter, 2007; Maloof et al., 2010a; Erwin absolute ages, the calibration of this evolution- higher in the stratigraphy, above the large et al., 2011) over a geologically brief inter- arily important transition remains piecemeal, negative excursion in the Zuun-Arts Forma- val in Earth history (Valentine, 2002). Most resulting in much uncertainty in determining tion. The first appearance datum of Treptich- recently, Maloof et al. (2010a) suggested that rates of origination and geochemical change. nus pedum is ~275 m above the large negative the Cambrian fossil first appearances occurred With a dearth of absolute ages tied to the fossil excursion in the Zuun-Arts Formation and in three discrete pulses associated with rapid record, carbon isotope (d13C) chemostratigra- ~250 m above the first appearance datum reorganizations in the carbon cycle. To test this phy has proven to be a powerful tool for global of small shelly fossils, highlighting the rarity hypothesis and others about mechanistic links correlation of Ediacaran-Cambrian successions and facies dependence of its preservation. between environmental change and evolution- (Brasier et al., 1990, 1993; Kirschvink et al., We shift the first appearance datums of ary milestones across the Ediacaran-Cambrian 1991; Knoll et al., 1995a; Zhu et al., 2006, 2007; tommotiids, orthothecimorphs, hyolithel- transition, it is necessary to integrate records Maloof et al., 2010b). around the globe. However, previous global The Zavkhan (Dzabkhan) terrane in western †E-mail: [email protected] syntheses (e.g., Maloof et al., 2010a) have been Mongolia hosts thick, well-exposed, low-grade,

GSA Bulletin; Month/Month 2015; v. 1xx; no. X/X; p. 1–27; doi: 10.1130/B31248.1; 13 figures; 1 table; Data Repository item 2015265.; published online XX Month 2015.

GeologicalFor Society permission of to America copy, contact Bulletin [email protected], v. 1XX, no. XX/XX 1 © 2015 Geological Society of America Smith et al. carbonate-rich, fossiliferous successions of Tarim (Lehmann et al., 2010; Edel et al., 2014). Cryogenian to Early Ediacaran late Ediacaran through early Cambrian strata, Locally, on the Zavkhan terrane, this late Paleo- Stratigraphy providing an opportunity to more precisely zoic collision is manifested in dextral strike-slip calibrate evolutionary changes in a chemo- and wrench structures (Figs. 1A and 1C) and wide- Along the Zavkhan River, the oldest exposed sequence-stratigraphic context. This succession spread Permian plutonism (Jahn et al., 2009). rocks are felsic volcanic rocks of the Zavkhan hosts dozens of well-exposed sections with sev- The Neoproterozoic through Cambrian stra Formation (Figs. 1B and 1C; Bold et al., 2013), eral horizons of small shelly fossils, archaeo tigraphy of the Zavkhan terrane was first which have been dated using laser ablation cyathid reefs, and ichnofossils (Goldring and described by Bezzubtsev (1963) in the Bayan inductively coupled plasma mass spectrometry Jensen, 1996; Khomentovsky and Gibsher, Gorge. Before the early 1990s, research in (LA-ICP-MS) U/Pb dating on zircon to between 1996; Kruse et al., 1996), many of which are Mongolia was largely restricted to Mongolian 773.5 ± 3.6 and 803.4 ± 8.0 Ma (Levashova being reported and documented here for the and Russian researchers who worked primarily et al., 2010). Granites that intrude the Zavkhan first time. However, because Ediacaran to early on the Ediacaran through early Cambrian strata Formation have been dated using thermal ion- Cambrian strata in the Zavkhan terrane were in the Zavkhan terrane (Zhegallo and Zhuravlev, ization mass spectrometry (TIMS) bulk multi- deposited in a foreland basin (Macdonald et al., 1991). Shortly after Mongolia opened back up grain U/Pb on zircon at 755 ± 3 Ma (Yarmolyuk 2009), topography generated during tectonism to Western researchers, an international research et al., 2008). The Zavkhan Formation is overlain resulted in large lateral facies changes, which team participated in International Geoscience by the synrift siliciclastic deposits informally confounded previous lithostratigraphic correla- Programme (IGCP) Project 303, “Precambrian- referred to as the Khasagt Suite (Ruzhentsev tions and biostratigraphic compilations (Brasier Cambrian Event Stratigraphy,” in 1993, repre- and Burashnikov, 1996) and passive-margin et al., 1996b; Khomentovsky and Gibsher, senting the first effort to bring together specialists deposits of the Tsagaan-Olom Group (Macdon- 1996). In this study, we present new geological from different subdisciplines of earth sciences ald et al., 2009; Bold et al., 2013). mapping and small shelly fossil biostratigraphy for an integrated approach to addressing the Pre- The Tsagaan-Olom Group (note recent for the Nemakit-Daldynian through Tommotian cambrian-Cambrian transition in western Mon- changes in spelling and formalization of stra- strata in the necessary context of a high-reso- golia. During this international excursion, they tigraphy; Bold et al., 2013) contains two Cryo- lution d13C chemostratigraphic age model and visited five previously documented sections to genian glacial diamictites and cap carbonates basinwide sequence stratigraphy. study them in more detail. The Geological Mag (Macdonald et al., 2009). The oldest formation Because there are only a handful of carbon- azine publications that resulted from this proj- in this group is the Maikhan-Uul Formation, a ate-rich fossiliferous sections globally that span ect were the first time these key sections were Sturtian-aged diamictite (Lindsay et al., 1996b; the Ediacaran-Cambrian transition, namely, sec- described in English and placed into a regional Macdonald et al., 2009; Macdonald, 2011), tions in Siberia, China, Kazakhstan, SW United and global context (Brasier et al., 1996a, 1996b; which has been constrained globally by correla- States, NW Canada, and Mongolia, calibrating Goldring and Jensen, 1996; Khomentovsky and tion to between ca. 717 and 660 Ma (Macdonald the timing and tempo of the radiation is heavily Gibsher, 1996; Kruse et al., 1996; Lindsay et al., et al., 2010; Rooney et al., 2014). Overlying the influenced by each record. This contribution of 1996a, 1996b). Maikhan-Uul Formation, there is the carbon- an age model, geological mapping, and revised Voronin et al. (1982), Gibsher and Kho- ate-dominated Taishir Formation, composed biostratigraphy change known global rates of mentovsky (1990), Gibsher et al. (1991), and of the Sturtian cap carbonate and interglacial early Cambrian evolution and provide the nec- Khomentovsky and Gibsher (1996) described strata (Shields et al., 2002; Macdonald et al., essary geological and stratigraphic framework the stratigraphy and published preliminary, 2009; Macdonald, 2011; Johnston et al., 2012; for integrating geochemical, paleontological, schematic, hand-drawn maps of the south- Shields-Zhou et al., 2012). The overlying Khon- and geochronological data from southwest ern and northern blocks of the Bayan Gorge, gor and Ol Formations have been correlated to Mongolia into the global record of the Edia- Tsagaan Gorge, Salaa Gorge, and Taishir. In Marinoan glacial deposits and the ca. 635 Ma caran-Cambrian transition. addition, biostratigraphy is documented from basal Ediacaran cap carbonate, respectively Orolgo (Orolchayan) Gorge (Endonzhamts and (Condon et al., 2005; Macdonald et al., 2009; GEOLOGIC BACKGROUND AND Lkhasuren, 1988) and Kvetee Tsakhir Nuruu Calver et al., 2013). The Ol Formation is over- TECTONOSTRATIGRAPHIC (Zhegallo and Zhuravlev, 1991; Dorjnamjaa lain by the early Ediacaran Shuurgat Formation FRAMEWORK et al., 1993; Esakova and Zhegallo, 1996). (previously referred to informally as the Ulaan Detailed paleontological work was conducted Bulaagyn member of the Tsagaan-Olom Forma- Previous Studies primarily by Russian paleontologists; how- tion), which consists of 100–500 m of carbonate ever, much of this work lacked a stratigraphic and is unconformably overlain by the latest Edi- The Zavkhan terrane is a Precambrian crustal context. The series of 1996 publications was acaran Zuun-Arts Formation (Macdonald et al., fragment embedded in the core of the Central the first attempt at placing the small shelly fos- 2009; Bold et al., 2013). Asian orogenic belt (Fig. 1). After late Edia- sil first appearances in a lithostratigraphic and caran to Ordovician accretion of arc terranes chemostratigraphic framework (Brasier et al., Latest Ediacaran to Early Cambrian to the south, the Late Ordovician to Silurian 1996b; Khomentovsky and Gibsher, 1996). Stratigraphy, Biostratigraphy, and record of Mongolia is marked by extensional Brasier et al. (1996b) presented the first carbon, Chemostratigraphy magmatism and basin formation (Kröner et al., oxygen, and strontium chemostratigraphy for 2010; Gibson et al., 2013). During this time, the the Neoproterozoic through Cambrian carbon- During the latest Ediacaran, the Zavkhan ter- Zavkhan terrane was part of a composite rib- ates in the Zavkhan terrane. More recently, the rane began to subduct to the southwest beneath bon continent that was adjacent to Siberia (Wil- age and depositional environment of the Cryo the Khantaishir-Dariv arc, creating the foreland hem et al., 2012). From the Devonian to Early genian strata have been more tightly constrained basin that accommodated the subsequent thick Permian, the Zavkhan terrane was oroclinally by Macdonald et al. (2009), Macdonald (2011), sequence of latest Ediacaran to earliest Cam- buckled during the arrival of North China and and Bold et al. (2013). brian strata (Macdonald et al., 2009). These

2 Geological Society of America Bulletin, v. 1XX, no. XX/XX

Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

″N ′0 °0 47 N KT 4 0 6C C (‰ VPDB ) 13 –4 δ r ul t wn ve Fa Ri To Area of detailed map 6A

Fm Fm ts Fm Group

han Fm

Olom han-Uul Fm han Fm

ISHIR

sagaan- T TA Maik Khasagt Salaagol Bayangol Zuun-Ar Zavk Khairk

. Ediac C yo Cr early

635 Ma 660 Ma 717 Ma

~541 Ma

C Neoproterozoic C 6B han Fm ne ne oterozoic intrusives Zo Zo ry ~750 Ma granit Khasagt Zavk Pr and metasediments Russia cretiona

d e ambrian Suture e 0 50 100 km s rran

rentiate leozoic Ac

Pa n e rg oup ul t Go

Muru Undi e Ediacaran -C Fa Late

e Ophiolite

agatai Fm

n Bulaagy n

rv Gorg Bayan Ta Ulaa

e r Ophiolit

sgul

han-Uul Fm MONGOLIA

rran e han River rg hongo Olom Gr one Te rrane Khuv Zavk y Z Te Baidrag aishir Fm

Bayank T Shuurgat Ol Fm Maik

"E ″E

e Go Khunkher '0 ′0 Uliastai e cretionar °0 °0 rrane rran n sagaan- Ac Te 96 96 Te T Russia e

han e Altai Gorg Uliastay assive Margin taishir leozoic Bulnai Fault rran rous Basi gure 1D Te fe Pa Fi oterozoic Zavk oterozoic Altai Pr Khan Ophiolit Pr Allochthon

e rboni Fm Ca

Fm Ophiolit e ts Fm v e

Dari Gorg sagaan han Fm

rran onian- Lake T Zo Te tochthonous

Au Allochthonous Dev Ediacaran to Early Ordovician-Silurian P T Gobi-Altai Gorg Salaa China B Khairk Salaagol Bayangol Zuun-Ar

JARGALAN e

a 6D 6E ISHI R

ass

h-Dava TA

s Orolgo Gorg Orolgo al. (2009) and this study plotted against generalized early Cambrian through Neoproterozoic stratigraphy Neoproterozoic Macdonald et al. (2009) and this study plotted against generalized early Cambrian through from C profile 13 km

Khuk 0 t TSAKHIR y and volcanic rocks SALAA VCHIG 15 KHA aleozoic intrusives KHUNKHER ORLOG O Silurian/Ordovician sedimentar P Ultrama c rocks Cambrian cher TSAGAA N AN-UUL Y AA V

BA D A SOUT H KHUKH DA

″N ′0 °0 47 in the Zavkhan terrane. The age constraints are dated by correlation. The basal Maikhan-Uul Formation has been correlated with the basal Sturtian glacial deposit in The basal Maikhan-Uul Formation has been correlated dated by correlation. The age constraints are in the Zavkhan terrane. with the Marinoan cap, which The Ol cap carbonate has been correlated Canada at 717.4 ± 0.2 Ma (Macdonald et al., 2010) and 662.4 3.9 (Rooney 2014). NW Peedee belemnite. (D) Geologic map of the Zavkhan terrane. Red VPDB—Vienna et al., 2004, 2013; Condon 2005). has been dated globally at ca. 635 Ma (Calver Nuruu. Tsakhir 6. KTN—Kvetee boxes mark spatial extent of detailed maps in Figure Figure 1. (A) Shaded outcrop of Zavkhan terrane showing major fault blocks. (B) Terrane map of Mongolia. Area of Figure 1D is boxed in dark blue. Map adapted of Figure Area map of Mongolia. Terrane fault blocks. (B) of Zavkhan terrane showing major 1. (A) Shaded outcrop Figure Bucholz et al. (2014). (C) δ from

Geological Society of America Bulletin, v. 1XX, no. XX/XX 3 Smith et al. strata, along with kilometers of Neoproterozoic that preserves the columnar stromatolite The FAD of anabaritids, the earliest small strata, are exposed in NE-vergent thrust sheets Boxonia grumulosa (Fig. 3A; Markova et al., shelly fossil, has been reported near the base in the Zavkhan terrane. 1972). The distinct stromatolitic horizon is of the Zuun-Arts Formation (Endonzhamts From oldest to youngest, the late Ediacaran overlain by variably phosphatized green shale and Lkhasuren, 1988; Dorjnamjaa et al., 1993; and early Cambrian strata on the Zavkhan ter- with nodular carbonate, marking an abrupt Brasier et al., 1996b; Khomentovsky and Gib- rane are the carbonate-dominated Zuun-Arts flooding surface. In some localities, a lag sher, 1996); however, we refine the position of Formation, the mixed carbonate-siliciclastic deposit of coarse, immature sandstone is pres- this FAD horizon herein. Simple bed planar Bayangol Formation, the carbonate-domi- ent between the stromatolite and phosphatic trace fossils are also present in this formation nated Salaagol Formation, and the siliciclastic shale. Overlying the phosphatic shale, there (Brasier et al., 1996b). Neither cloudinid nor Khairkhan Formation (Fig. 2). Together, these is ~150–400 m of thinly bedded, medium- the Shuram carbon isotope excursion has been strata range from ~800 to 2700 m in thickness. gray limestone (Fig. 3B) with occasional documented in the Zuun-Arts or Shuurgat For- The Khairkhan Formation consists of allochtho- intraclast breccia. The top 5–200 m section of mations. Carbon isotope (d13C) values of the nous mélange, flysch, and molasse, marking the the Zuun-Arts Formation consists of several Shuurgat Formation range between –0.5‰ and closure of the foredeep and the end of deposi- 5–20-m-scale parasequences culminating with +8.5‰, values that more closely resemble the tion in this basin (Macdonald et al., 2009). a distinctive white and black cross-bedded carbon isotopic profile of the lower Ediacaran ooid grainstone (Fig. 3C). The top contact of than the upper Ediacaran (Macdonald et al., Zuun-Arts Formation the Zuun-Arts Formation is defined at a sharp 2009), which would imply a substantial hiatus Strata above the uppermost karst surface in contact between the ooid grainstone and a sec- at the karst surface between the Shuurgat and the underlying Shuurgat Formation are marked ond horizon of microcrystalline phosphatic Zuun-Arts Formations (Macdonald et al., 2009; by buff- to pink-colored microbial dolostone shale (Fig. 3D; Lindsay et al., 1996a). Bold et al., 2013).

Bayangol Formation The Bayangol Formation, along with the Carbonate grainstone Breccia/conglomerate Stromatolite/thromb Med-coarse sandstone underlying Zuun-Arts and Shuurgat Forma- Calcimicrobial Vf-f sandstone tions, was previously included in the Tsagaan- Ribbonite Siltstone/shale Olom Formation and has been characterized and Rhythmite Phosphorite Olistostrome Unconformity subdivided as numbered units by Voronin et al. (1982), Gibsher and Khomentovsky (1990), and The Khairkhan Fm is mixed conglomerate, sandstone, silt, Gibsher et al. (1991). The base of the Bayan- and shale. In some localities, there are km scale olistoliths of the Salaagol Fm. This unit also contains clasts of gol Formation is marked by phosphatic shale s ultrama c rocks. This unit is the ysch/molasse of the basin.

m with nodular carbonate. The Bayangol Forma-

Khairkhan

n tion consists of mixed carbonate and siliciclas-

g The Salaagol Fm is a microbial, archaeocyath reef with

a a

h minor interbedded conglomerate, sandstone, and siltstone. tic strata, and it ranges from ~500 to 1250 m in

l r

a This formation ranges from 0-400 m in thickness. It is grey,

i thickness. In most sections in the basin, in the

S a SG white, pink, and red in color. Ooid grainstones and h lower to middle part of the formation, there K glauconitic sandstone are present. SSF horizons are also found in this formation. are at least three thick limestone units capping 6 interbedded siliciclastic and limestone units. The Bayangol Fm is a low grade mixed siliciclastic-carbon- Figure 4 shows examples of a few of these mas- 5 ate unit with abundant trace fossils and SSFs. The base is

G sive limestone units and how they mark some

B marked by a phosphatic shale with nodular limestone. of the large-scale sequence boundaries in the The formation is divided into 5 informal members, each strata. In most sections in the basin, above the

capped by a massive limestone unit, for mapping and 4

m sequence stratigraphic purposes. BG1 is equivalent to the massive third limestone marker bed, there are

B Zuun-Arts Fm. BG6 is a variably present white to red hundreds of meters of predominantly fine to o

–2700 meters –2700 arenitic sandstone to pebble conglomerate that is capped

g 0

n by the Salaagol Fm. coarse sandstone.

a 8

y The Bayangol Formation hosts diverse ~ a The Bayangol Fm is highly variable both in terms of

B thickness and lithology. The facies changes occur over very and well-preserved stromatolites, thrombo- 3

G short distances. The massive limestone units represent the lites, and other bioconstructions with a wide B transition from transgressive systems tract in the highstand systems tract. range in size and morphology (Kruse et al.,

1996). Important for this study, there is also 2

G a rich assemblage of small shelly fossil and B abundant trace fossils that have been previ- The base of the Zuun-Arts Fm is marked by the columnar m ously documented in the Bayan, Salaa, and

F stromatolite, Boxonia grumulosa, which is capped by a

ZA phosphatic shale. The rest of the formation is thinnly-bed-

t Tsagaan Gorges, Taishir, and Kvetee Tsakhir r

Zuun- ded limestone that gradually shallows up to a trough A cross-bedded ooid grainstone.. Nuruu (Fig. 1; Korobov and Missarzhevsky, Sh. The top of the Shuurgat Fm is marked by a karst surface. 1977; Voronin et al., 1982; Endonzhamts and Lkhasuren, 1988; Goldring and Jensen, 1996; Figure 2. Generalized late Ediacaran to early Cambrian stratigra- Khomentovsky and Gibsher, 1996). To create a phy of the Zavkhan terrane. Note the recent changes in spelling and composite biostratigraphic range chart for the formalization of stratigraphy in the Zavkhan terrane by Bold et al. Bayangol Formation, these sections were pre- (2013). SSF—Small shelly fossils. viously correlated primarily using lithostrati-

4 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

A B C

D E F

G H I

J K L

Figure 3. Photomontage of the Zuun-Arts and Bayangol Formations. (A) Stromatolite Boxonia grumulosa from the base of the Zuun-Arts Formation. (B) Thinly bedded, medium-gray limestone of lower to middle Zuun-Arts Formation. (C) Cross-bedded ooid grainstone of upper Zuun-Arts Formation. (D) Phosphatic shale with nodular limestone of lower Bayangol Formation. (E) Intraclast conglomerate in lower Bayangol Formation. (F) Thrombolites in upper part of Bayangol Formation in Bayan Gorge. (G) Phosphatized small shelly fossils, grains, and ooids in lag deposit in basal part of Bayangol Formation (137 m above basal contact) in Orolgo Gorge. (H) Anabaritids from basal Bayangol Formation in Orolgo Gorge. (I–K) Early arthropod trace fossils from the middle and upper Bayangol Formation (BG3– BG5) in Khunkher Gorge. (L) Bed-planar trace fossil from the upper Bayangol Formation (BG5) in S. Bayan Gorge. graphic correlations. However, as discussed ful in identifying specific beds within a single been described in detail from Salaa Gorge and later herein, these lithostratigraphic correla- gorge, they cannot be used to construct an Zuun-Arts Ridge (Voronin et al., 1982; Wood tions were compromised by large lateral facies accurate framework for basinwide correlations. et al., 1993; Kruse et al., 1996). It has variably changes. These lithostratigraphic correlations thick red silts and conglomerate interbeds. The utilized the numbered units of Voronin et al. Salaagol Formation formation also contains other bioclastic debris (1982) and were followed by the authors in the The Salaagol Formation is a <400-m-thick that includes chancellorid sclerites, sponge spic- series of 1996 Geological Magazine publica- white, gray, pink, and red microbial and archaeo- ules, coralomorphs, and hyoliths (Kruse et al., tions. Although these numbered units are use- cyathan reef (Figs. 5A and 5B) that has only 1996). The archaeocyathan assemblage at Salaa

Geological Society of America Bulletin, v. 1XX, no. XX/XX 5 Smith et al.

mation at Orolgo Gorge (Figs. 1 and 6E), where A it is overthrust by a klippe of ultramafic cumu- HST lates and serpentinite. METHODS

TST Geological Mapping

BG4 Geological mapping was conducted over LST the course of three summers between 2011 and 2013. The detailed geologic mapping presented here provides the necessary context for HST the lithostratigraphy, chemostratigraphy, and biostratigraphy presented in this study. Higher-

BG3 resolution stratigraphic studies of previously described and undescribed sections would not have been possible without new geological mapping. We returned to the five key areas that were described in the 1996 Geological Magazine pub- B lications, Bayan Gorge, Kvetee Tsakhir Nuruu, Orolgo Gorge, Salaa and Tsagaan Gorges, and Taishir to remap those areas in greater detail (Figs. 6A, 6B, 6C, and 6E). Additionally, we mapped many other parts of the Zavkhan ter- HST rane, targeting areas with good exposure and BG3 continuous sections of the Zuun-Arts, Bayan- gol, and Salaagol Formations (Figs. 6A, 6C, 6D, and 6E). Geological mapping enabled measure- ment of stratigraphic sections in the new, pre- TST viously undescribed areas with the goal of put- BG2 ting together a more complete basinwide facies model and to construct a chemostratigraphic age LST model. Geologic mapping was conducted using HST LANDSAT satellite imagery, which offered the highest-resolution imagery available at the time. Figure 4. (A) Annotated photo looking to the NW in section E1211, NE of Khukh-Davaa For mapping purposes and sequence strati- Pass. The major sequence stratigraphic systems tracts of members BG3 and BG4 of the graphic correlation, we informally divided the Bayangol Formation that are shown in Figure 7 are depicted here (the solid black line marks Bayangol Formation into five different mem- the maximum flooding surface, and the dashed line marks the boundary between BG3 and bers (BG2–BG6), of which the lower three BG4). The dominant facies associations are also shown; however, it is not possible to mark members are capped by massive carbonate all the facies associations at this scale. Human circled for scale. (B) Annotated photo looking units that are clearly visible in satellite imagery to the west in the southern block of the Bayan Gorge. Major sequence stratigraphic systems (Fig. 4). When we initially divided units in the tracts and facies associations of the lower two members of the Bayangol Formation are de- late Ediacaran to early Cambrian strata, we used picted here (the solid white line marks the maximum flooding surface, while the dashed line BG1 interchangeably with Zuun-Arts because marks the top of BG2). Minor facies associations are not shown. HST—highstand systems the Zuun-Arts Formation marks the onset of tract; TST—transgressive systems tract; LST—lowstand systems tract. basin formation and is genetically related to the Bayangol Formation (Bold et al., 2013). Here, we use Zuun-Arts Formation in place of BG1. As the result of our detailed stratigraphic and Gorge is typified byArchaeolynthus , Dokidocy and measured sections across the terrane dem- mapping work, we have defined eight struc- athus, Nochoroicyathus, and Rotundocyathus, onstrate that the Khairkhan Formation is locally tural blocks (Taishir, Salaa, Tsakhir, Khavchig, and based on archaeocyathan biostratigraphy, much thicker and consists of mixed siltstone, Tsagaan, Orlogo, Khunkher, and South Khukh Kruse et al. (1996) assigned it a Tommotian to sandstone, and conglomerate, often containing Davaa) that are characterized by distinct stratig- Botomian age. large boulders of carbonate of the Salaagol For- raphy and are bounded by major faults (Fig. 1A). mation and chert and basalts clasts of unknown Khairkhan Formation provenance. The Khairkhan Formation is the Facies Associations and Sequence The Khairkhan Formation has only been youngest Cambrian unit in the Zavkhan terrane, Stratigraphy described in Salaa Gorge as a >200 m siliciclas- and has been interpreted as the mélange, flysch, tic succession that unconformably overlies older and molasse of the foredeep (Macdonald et al., Nine facies associations were identified units (Voronin et al., 1982; Dorjnamjaa and Bat- 2009). This interpretation is further supported in this succession (Table 1), and for ease of Ireedui, 1991; Kruse et al., 1996). Our mapping by the structural position of the Khairkhan For- discussion, they are grouped into five broad

6 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

A B C

D E F

G H I

Figure 5. Photomontage of the Salaagol and Khairkhan Formations. (A) Interbedded microbial limestone and red silt near the base of the Salaagol Formation in Orolgo Gorge. (B) The upper Salaagol Formation (E1109) in Salaa Gorge with a goatherd for scale. (C) Rounded quartzite cobbles on top bedding surface of archaeocyath reef south of Khukh-Davaa Pass. Camera lens cap for scale. (D) Phosphatized grains, small shelly fossils, and archaeocyaths at top of Salaagol Formation south of Khukh-Davaa Pass. (E) Very poorly sorted, matrix- supported Khairkhan Formation conglomerate in Orolgo Gorge. Limestone clasts contain archaeocyaths. (F) Building-size olistolith of Salaagol Formation in Khairkhan Formation west of Orolgo Gorge. Goatherd for scale. (G) Black to brown silt of upper Khairkhan For- mation south of Khukh-Davaa Pass. (H) Very poorly sorted, matrix-supported Khairkhan Formation conglomerate NW of Orolgo Gorge. (I) Partially silicified red- to buff-colored microbial dolostone of unknown age. categories: (1) carbonate-dominated facies The late Ediacaran to early Cambrian suc- and bypass channels, so to pick out the major associations that were deposited on an outer- cession on the Zavkhan terrane contains sequence boundaries, we use the LSTs, which shelf to slope environment below storm wave multiple sequences that are appropriate for are marked by an influx of medium to coarse base, (2) siliciclastic-dominated associations high-resolution sequence stratigraphy. In our sandstone and conglomerate across the basin that were deposited on an outer-shelf to slope analysis, we use the four commonly recog- (Vail et al., 1977; Mullins et al., 1987; Spence environment below storm wave base, (3) car- nized system tracts: lowstand systems tract and Tucker, 1997). bonate-dominated facies associations that were (LST), transgressive systems tract (TST), high- deposited in shelf to shoreface environments, stand systems tract (HST), and falling-stage δ13C Chemostratigraphy (4) siliciclastic-dominated facies associations systems tract (FSST), following Van Wagoner that were deposited in shelf to shoreface envi- et al. (1990). We interpret carbonate deposi- To test regional and global correlations, we ronments, and (5) flysch and mélange facies tion to represent TSTs and HSTs of relative sea sampled limestone spanning the Zuun-Arts For- associations that were deposited during the level in the basin (Vail et al., 1977). For the mation through the Salaagol Formation from terminal closure of early Cambrian orogenesis. carbonate-dominated slope facies in the lower across the Zavkhan terrane. Limestone rocks These facies associations were used to identify to middle Zuun-Arts and lowermost Bayangol were sampled at 0.5–4 m resolution from 14 the major stratigraphic sequences boundaries Formations, we utilize the TSTs and HSTs for composite stratigraphic sections, many of which in the Zuun-Arts through Salaagol Formation correlation across the basin. The rest of the are composite sections (exact locations of mea- succession. sequence contains abundant parasequences sured sections are indicated in Fig. 6). Clean

Geological Society of America Bulletin, v. 1XX, no. XX/XX 7 Smith et al.

B LEGEND Shuurgat Fm 021kms 6 1262 Ol Fm E11 Silurian/Ordovician E1 O Group Taishir Fm sedimentary and volcanic T- rocks Maikhan-Uul Fm BG2 Paleozoic intrusives Khasagt Fm Ultrama c rocks ~750 Ma granite Cambrian chert Zavkhan Fm Khairkhan Fm 29 Proterozoic intrusives E11291 BG2 E1 1 Salaagol Fm and metasediments 12 BG3 F11201 River BG4 ,F Bayangol Fm 20 F1121 1 Fault F1 BG3 Zuun-Arts Fm E1216 Section BG5 0 BG4 A TAISHIR E113130 ^ E1114/3/7 D706 E1 BG4 Z KHAN RIV BG3 EE1115111 AV E R E1212 E E1211 D 1 21 1 0 3k6 ms

E1123

BG3 BG2 BG4 E1125 7 BG5 EE1207120 E1336 01k0 ms E1209

C E1E1216216 Z ER AVKHAN RIV

E13413 0

E1328, E1329, E1331

BG2 21 1 E1215 BG3 E 063kms

20 DD7207 E1E1110099

8 74 1

E D7 F874F8 E1223 12 13311 F1F1 1 23

BG5 E1220 BG4 BG3 BG2 042kms

Figure 6. Detailed geological maps from across the Zavkhan terrane. Locations of maps are indicated in Figure 1. (A) Geological map of area SE of Taishir. (B) Geological maps of the southern and northern fault blocks in Bayan Gorge. (C) Geological maps of Kvetee Tsakhir Nuruu. (D) Geological map of Khukh-Davaa Pass (NW of town of Jargalant). (E) Geological maps of Orolgo, Salaa, and Tsagaan Gorges.

8 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

limestone samples without siliciclastic compo- nents, secondary veining, or cleavage were tar- geted while sampling. Carbon (d13C) and oxygen (d18O) isotopic measurements were obtained on 2763 samples. Samples were microdrilled along individual laminations, where visible, to obtain 5–20 mg of powder. Veins, fractures, and siliciclastic-rich turbidity currents in submarine channels; molasse and ﬂysch deposits slope via dilute turbidity currents slope environments during marine transgressions slope by debris ﬂow setting; deposited by turbidity currents on outer shelf to lower continental slope; deposited below storm wave base turbidity currents on outer shelf to lower continental slope; deposited below storm wave base areas were avoided. Carbonate d13C and d18O Deposited by high-density High-energy shoreface Deposited on carbonate shel f Deposited on continental High-energy carbonate shoal Deposited in continental Deposited on a continental Pelagic to hemipelagic Pelagic setting; deposited by Origin and depositional environmen t data were acquired simultaneously on a VG Optima dual-inlet mass spectrometer at the Har- vard University Laboratory for Geochemical Oceanography. Carbonate samples were reacted with orthophosphoric acid using a VG Isocarb preparation device, which includes a common acid bath with a magnetic stirrer. Approximately 1 mg of microdrilled samples was reacted in

the bath at 90 °C. Evolved CO2 was collected bases erosional; laterally discontinuous; building-size olistolith s bedding; parallel laminations; ripple cross-lamination; bioturbation common thrombolites, archaeocyaths, nonphosphatized small shelly fossils rarely preserved; rare vugs structures phosphatic hardgrounds with phosphatic grains, ooids, and small shelly fossils possible Ediacaran fossils and bed-planar trace fossils; rare stromatolite s possible hold fast imprints and bed-planar trace fossils; rare stromatolite s rough cross-bedding High-angle cross-bedding; unit High-angle trough cross- T Bioturbation, stromatolites, Normally graded beds; load Limestone nodulates; rare Clast to matrix supported Rare slump folding; rare Rare slump folding; rare Sedimentary structures cryogenically and analyzed using an in-house reference gas. Potential memory effect resulting from the common acid-bath system was minimized by increasing the reaction time for dolomite samples. Memory effect is estimated Y CAMBRIAN STR ATA at <0.1‰ based on variability of standards run

EARL after dolomite samples. Standard deviation (1s)

TO from standards was better than ±0.1‰ for both m thick; laterally m thick d13C and d18O. Carbonate d13C and d18O isotopic ry thin to thin; 1–10 cm bedded discontinuous thick

ery thin to thin; 1–5 cm thick results are reported in per mil (‰) notation Medium to very thickly Thin to medium Thin to very thickly bedded Thin to very thickly bedded Thin to medium 1–20 0.1–5 Ve Bed thickness relative to VPDB (Vienna Peedee belemnite) by using an in-house Carrara Marble standard

TE EDIACARAN that was calibrated against several NBS carbonate standards and cross-calibrated with other laboratories.

Constructing an Age Model TIONS FOR LA

Because global stages 2–4 of the early Cam- ry ﬁne to very coarse sand, pebble to boulder size clasts; olistoliths present locally dominantly ﬁne to medium; locally contains dolomitic matrix sandston e minor amounts of phosphate with lime matrix partings are of shale to siltstone

ASSOCIA brian have not been defined yet, and previous Poorly sorted sand matrix with Ve Packstone to grainstone Boundstone to bindstone Siltstone to medium-grained Microcrystalline shale with Packstone; limestone clasts Lime mudstone to wackestone; Lime mudstone to wackeston eV Grain size global correlations of early Cambrian data use

ACIES stratigraphic terminology from Siberia and Kazakhstan, for the sake of consistency, we also use the stage system for the Lower Cambrian on As) 1 and

ABLE 1. F the Siberian Platform in this study. The Lower T Cambrian on the Siberian Platform is divided into the Nemakit-Daldynian, Tommotian, Atda- banian, Botomian, and Toyonian (Repina and Rozanov, 1992; Rozanov et al., 2008). Gener- ally, the Nemakit-Daldynian and Tommotian ry poorly sorted, thickly bedded conglomerate oncoids; thickly bedded stromatolite, thrombolite, and archaeocyath reef. Light-gray limestone occasionally with minor sandstone to siltstone beds sandstone with isolated limestone beds minor nodular limestone; amounts of phosphate breccia; clasts derived from facies associations (F 2; laterally discontinuous limestone and mudstone dark-gray limestone; sometimes interbedded with very minor shale Ve Bedded sandstone Limestone containing ooids and Massive microbialaminite, Interbedded siltstone and Phosphatic shale to siltstone with Intraformational limestone tabular Interbedded medium-gray Clean, well-bedded medium- to General character are thought to be correlative with the Terre- neuvian Series. The Atdabanian and Botomian are thought to be correlative with the undefined Series 2 of the early Cambrian. Similar to the way in which previous workers created composite d13C curves and correlated sections globally, a combination of positive and negative excursions was used as tie points in this study. These d13C excursions were first labeled

urbiditic siltstone and in Siberia (e.g., Kouchinsky et al., 2007) and T Thinly and well-bedded 9: Massive to graded 8: Cross-bedded to swaley conglomerate bedded sandstone sandston e breccia limeston e limeston e have been used in several global compilations A6: Oolitic to oncolitic limestone A7: Platformal limestone A5: A4: Phosphatic shale to siltstone A3: Slope-derived carbonate A2: Parted to ribbon-bedded A1: FA FA F F F F F F F Association since. With these data, 1n, the most negative

Geological Society of America Bulletin, v. 1XX, no. XX/XX 9 Smith et al. d13C value at the Ediacaran-Cambrian bound- Small shelly fossil taxon names from previ- ing to depositional paleodepth (Fig. 7). Next, ary, and five positived 13C excursions (2p–6p) ous small shelly fossil horizons were taken from we organize the results from more proximal to were identified in each section. A composite the compilation done by Brasier et al. (1996b) more distal. d13C chemostratigraphic curve was constructed and Khomentovsky and Gibsher (1996) and the by stretching the d13C values between each of publication by Esakova and Zhegallo (1996). Facies Associations these tie points, assuming constant sedimenta- For generic and higher group assignments for tion rates in each section. Calibration points for individual taxa, we followed the supplementary The facies associations (FAs) described next, positive excursions are the same as those used in material of Maloof et al. (2010a, and references grouped according to paleo–water depth, and Maloof et al. (2010a). therein). the descriptions and environmental interpretations are summarized in Table 1. Paleontology RESULTS Slope- to Outer Shelf–Deposited Carbonate Body and trace fossils were sampled and Reconstructions of the Zavkhan Terrane Facies Associations photographed while measuring stratigraphic Three carbonate-dominated FAs show evi- sections. This study places previously described Based on geological mapping, we identified dence of deposition in deeper-water slope to small shelly fossil horizons into the necessary major faults in the Zavkhan terrane (Fig. 1A) outer-shelf settings. These facies are characteris- stratigraphic context and a refined age model. and grouped measured sections that are on the tic of the lower to middle Zuun-Arts Formation. Revised mapping and correlation among sec- same fault block together. The southern thrust Facies association 1. FA 1 is character- tions have resulted in different interpretations sheets preserve more distal sections, and the ized by thinly bedded, normally graded pack- of the stratigraphic position of the small shelly northern ones are more proximal. The measured stone, wackestone, or lime mudstone with very fossil horizons. sections presented here are ordered accord- minor amounts of siliciclastic material. Typi-

A B

Figure 7. Integrated lithostratigraphy, sequence stratigraphy, and δ13C chemostratigraphy of the Zuun-Arts and Bayangol Formations. Measured sections are numbered according to proximity to the paleoshoreline, with more distal sections on the right and more proximal sections on the left. Locations of the individual measured sections are shown on the map in the lower-left side of the figure. Small shelly fossils (SSFs) horizons from this study are marked with red stars, and those from previous studies are marked with green stars. Purple circles mark ichnofossil horizons. Previous paleontological data from Orolgo (Orolchayn) Gorge is revised from Endonzhamts and Lkhasuren (1988). Paleontological data from Salaa and Tsagaan Gorges are from Voronin et al. (1982), Esakova and Zhegallo (1996), and Khomen- tovsky and Gibsher (1996). The paleontological data from Bayan Gorge are from Khomentovsky and Gibsher (1996). Each of the previous small shelly fossil horizons is labeled according to the labeling schemes of the different collections. Shaded colored boxes show correlation between sections. VPDB—Vienna Peedee belemnite; HST—highstand systems tract; TST—transgressive systems tract; LST—lowstand systems tract.

10 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia cally, these beds are dark to light gray in color. quartz grains. This facies is found in the upper- almost entirely brittle deformation and, gener- The limestone beds have sharp bases and tops most Zuun-Arts Formation, throughout the Bay- ally, is dominated by a conjugate set of right- and are normally graded. Occasionally, soft angol Formation, and in parts of the Salaagol lateral strike-slip faults and oblique en echelon sedimentary slumping is present. FA 1 is the Formation. thrust faults. dominant facies in the lower to middle part of Facies association 7. FA 7 is anything rang- the Zuun-Arts Formation and is interpreted to ing from microbialaminites, stromatolites, and Taishir (E1114, E1113, E1115, represent outer-shelf to upper-slope deposition thrombolites to an archaeocyath reef. Different E1117, E1325) below storm wave base. microbial and thrombolitic textures are com- Most previous studies at this locality focused Facies association 2. FA 2 is a ribbon-bed- mon throughout the Bayangol Formation and on the underlying Tsagaan-Olom Group (Gold- ded limestone, similar to FA 1 in that is a thinly parts of the Salaagol Formation. The Salaagol ring and Jensen, 1996; Khomentovsky and Gib- bedded, carbonate-dominated facies. Typically, Formation also contains beautifully preserved sher, 1996; Macdonald et al., 2009; Macdon- these beds are medium to light gray in color. archaeocyath and hyolith reefs that range in ald, 2011) and the fossils, bioherms, and patch FA 2 differs from FA 1 in that it contains shale size from laterally discontinuous patch reefs to reefs of the Bayangol Formation (Dorjnamjaa or siltstone partings. This facies association is hundreds of meters thick and for more than a and Bat-Ireedui, 1991; Goldring and Jensen, common in the middle part of Zuun-Arts For- kilometer along strike. 1996; Kruse et al., 1996). Detailed mapping of mation as well as portions of the Bayangol For- the Zuun-Arts and Bayangol Formations just mation, and we interpret it to represent outer- Shelf to Shoreface Siliciclastic to the southeast of the town of Taishir allowed shelf to lower-slope deposition below storm Facies Associations us to measure a composite section of the two wave base. Facies association 8. FA 8 is a cross-bedded formations at this locality and to place previ- Facies association 3. FA 3 is a slope-derived to swaley-bedded sandstone. It is characterized ously poorly contextualized fossil horizons into carbonate breccia that consists of clast- to by finely to thickly bedded, very fine- to very a coherent stratigraphic framework. Additional matrix-supported intraformational limestone coarse-grained quartzarenite to calcarenite. geologic mapping to the SSW and SW of the conglomerate. Clasts are tabular with compo- Sedimentary structures include high-angle cross- town of Taishir led to the discovery of more sition similar to the underlying and overlying stratification, parallel bedding, current ripples, previously unreported, well-exposed sections. units and indicative of derivation from sur- and normally graded beds. Bioturbation is com- Sections of the Zuun-Arts and Bayangol For- rounding slope deposits. This type of sedimen- mon is this FA. FA 8 is found throughout the mations are well-exposed to the SSW in a low tary breccia is present in some sections of the Bayangol Formation. creek bed (Fig. 6A). Additionally, just south lower Zuun-Arts Formation and lower Bayan- Facies association 9. FA 9 is an immature, of the large Permian intrusion, sections of the gol Formation. Some of the sedimentary brec- poorly sorted pebble to boulder conglomer- Upper Bayangol through Khairkhan Formations cias in the Zuun-Arts Formation are found in ate. Conglomerates are found in the Bayangol are well exposed. All of these sections described association with slump folding, suggesting through Khairkhan Formation but differ in tex- here are on the same fault block as the Bayan that these breccias were deposited during slope ture, composition, and thickness. The Bayangol Gorge sections. failure. Other breccias could represent storm- Formation contains centimeter-thick lenticular One of the measured sections of the Zuun- dominated facies. pebble conglomerates. Often, these conglomer- Arts Formation, E1114, was measured near the ates are red-colored and feldspathic, and they town of Taishir. Section E1114 was measured Slope-Deposited Siliciclastic can contain phosphatized grains and small just south of the Zavkhan River and east of the Facies Associations shelly fossils. In the Salaagol Formation, there town of Taishir. The total measured thickness Facies association 4. FA 4 is present at the is a wide range of conglomerates that include of E1114 is 176 m. Only the top surface of the base of the Zuun-Arts and Bayangol Formations centimeter-thick pebble conglomerates with stromatolite Boxonia grumulosa is exposed in and is characterized by laminated phosphatic angular clasts to cobble conglomerates that a gully. Brown and black, cherty, phosphatic shale to siltstone. Locally, this FA is present at are tens of meters thick with rounded clasts of shale overlies the buff- to pink-colored Boxonia the contact between the Salaagol and Khairkhan underlying strata. These conglomerates occur grumulosa. Similar to the Bayan Gorge sections, Formations. Tan carbonate nodules in the shale between archaeocyath reefs. Additionally, this the basal portion of this section is dominated by and siltstone are locally present. Volumetrically, facies is found in the shallow-water facies of the very thinly bedded limestone with nodular and FA 4 is not a major component of the succes- Khairkhan Formation, and it is interpreted as bedded chert and abundant interclast breccia sion; however, it is a distinct and important the molasse of the foreland. beds that pinch out laterally. The top 24 m por- stratigraphic marker. tion of the Zuun-Arts Formation is light-gray Facies association 5. FA 5 is characterized Key Locations: Geologic Mapping, cross-bedded ooid grainstone. by thinly bedded, finely laminated siltstone, tur- Sedimentology, and Stratigraphy For the Bayangol Formation, Kruse et al. biditic sandstones, and pebble conglomerates. (1996) presented a 30 m section, which they Fining-upward sequences are common. Bouma Next, we present the results of geological referred to as “Tayshir I” (not to be confused units A–C (and D–E in some sections) are com- mapping (Figs. 1 and 6), sedimentology, and with the Tayshir I and II sections of the Maikhan monly preserved. FA 5 is present in the middle stratigraphy that have enabled the refinement Uul and Taishir Formations in Khomentovsky to upper part of the Khairkhan Formation. of the small shelly fossil biostratigraphy in the and Gibsher [1996]), which we mapped as Zavkhan Basin. These results are synthesized members BG2 and BG3. They also described Shelf to Shoreface Carbonate for each locality. In some cases, the geological two 7 m sections that they called “Discovery Facies Associations mapping has greatly changed the interpretation site” and “Ovoo site,” and which we interpret Facies association 6. FA 6 is cross-bedded of the stratigraphy, including the small shelly as being a portion of BG4 in our section E1115. to swaley-bedded oolitic to oncolitic limestone. fossil biostratigraphy and the chemostratigra- However, without much context, it is difficult to Occasionally, the oolitic limestone contains phy. The deformation in the Zavkhan terrane is know exactly where these previously measured

Geological Society of America Bulletin, v. 1XX, no. XX/XX 11 Smith et al. sections are. The fossil horizons from Taishir Kvetee Tsakhir Nuruu (E1216, E1215, described already, and E1219 is a measured are not included in previous small shelly fossil E1217, E1219) section of member BG3, exposed in the limb biostratigraphic compilations (i.e., Khomen- This ridge is located ~30 km southeast of the of a syncline. tovsky and Gibsher, 1996; Maloof et al., 2010a). town of Taishir on the south side of the river (Fig. Here, we present a composite measured section 6C). The Zuun-Arts Formation and the basal SE of Bayan Gorge (E1123) of the Bayangol Formation (E1113, E1115, and Bayangol Formation are exposed on the south- South of the town of Taishir and SE of Bayan E1117) that enables us to put new and previous western side of the ridge. Additional mapping Gorge, there is a previously undescribed, well- fossil finds into the correct stratigraphic context. northeast of the river resulted in the discovery of exposed section of the Bayangol Formation Members BG2 through lower BG4 sit con- a well-exposed section of the Zuun-Arts Forma- (E1123). Because this section is on the same formably above the Zuun-Arts Formation in a tion (E1216) and a well-exposed but faulted and fault block as the southern Bayan Gorge section, well-exposed section to the south and west of folded section of the lower Bayangol Formation it is similar in facies and has similar thickness to a bend in the Zavkhan River. The rest of the (E1215, E1217, E1219). the composite section there. Bayangol section (very upper BG3 through Phosphatic shale and subcrop of thinly bed- The basal portion of the Zuun-Arts Formation BG5) is exposed in a broad syncline to the west ded limestone mark the base of the Zuun-Arts is not well exposed at this locality, so only the of a left-lateral strike-slip fault. There is a poorly Formation at this locality. Boxonia grumulosa upper 106 m portion was measured. The beds in sorted red conglomerate overlain by elongate is present in other sections at Kvetee Tsakhir this section are shallowly dipping, and the expo- stromatolites at the BG3-BG4 contact that were Nuruu but is not exposed at the exact place in sure of this section is not as good as the expo- used to tie the sections together. Member BG5 which this section was measured. The basal sures in Bayan Gorge or Taishir. Nonetheless, is very shallowly dipping near the top of the sec- 50 m portion of this measured section is thinly the section offers another opportunity to mea- tion, and the upper contact of the formation is bedded limestone with black, phosphatic shale sure and sample the middle to upper part of the not exposed. and minor intraclastic breccia. The black shale Zuun-Arts Formation. The measured thickness Members BG4 and BG5 of this section here was thicker and more abundant than in any of BG2–BG4 here is 357 m, and sedimentologi- (E1115) have some exceptional features, other section. The middle part of the section cally, it is very similar to the measured section including abundant and diverse trace fossils is dominated by thinly bedded limestone with in Bayan Gorge. The upper part of the formation and small shelly fossil assemblages. The elon- more intraclastic breccia and slump folding with (BG5) was not measured in detail at this local- gation of the stromatolites at the base of BG4 vergence to the southeast (n = 41). The top of ity, because, due to a major fault, there is no top is oriented at 210°, an indicator that the paleo- the section contains more thinly bedded lime- contact with the Salaagol Formation. shoreline was oriented perpendicular to the stone and debris flows but also abundant sand- direction of elongation along a NW-SE tran- stone and siltstone interbeds. This is the only Bayan Gorge (E1126, E1129, F1120, sect in modern coordinates (Hoffman, 1967). section in which sandstone, siltstone, and shale F1121, E1130, D706) The upper part of BG5 is notable because of its were observed at the top of the section instead of The Bayan Gorge is on the south side of the >100 m of siltstone to sandstone beds that con- cross-bedded ooid grainstone. Zavkhan River, ~15 km west of Taishir (Fig. tain few sedimentary features. Some of the beds The measured section of the Bayangol For- 6B). It hosts a northern and southern fault block are slightly graded, but many are structureless mation (E1215) is the same section as the pre- of very well-exposed sections of the Zuun-Arts medium to coarse sandstone units that lack bed- viously described section at Kvetee Tsakhir and Bayangol Formations. Despite abundant, ding surfaces. Nuruu, south of the Zavkhan River (Zhegallo but minor, conjugate fault sets in the gorge, the Just SE of Taishir, there is an excellent sec- and Zhuravlev, 1991; Dorjnamjaa et al., 1993; mapping is tractable. We present revised maps tion of the Salaagol Formation called E1325. Brasier et al., 1996b). At this locality, the of Bayan Gorge and used this new mapping to The basal contact of the Salaagol Formation is Bayangol Formation is relatively thin, and only construct a composite section of the Zuun-Arts not exposed at this section, but laterally, it rests the basal part is exposed. The base of the sec- (E1129 from the southern block) through Bay- conformably on a white to pink quartz pebble tion is marked by thinly bedded limestone that angol Formations (F1120, F1121, E1130, and conglomerate. The Salaagol Formation is 177 m grades up into massively weathering, partially D706 from the northern block). Previous workers thick at this locality. The basal ~38 m section of recrystallized limestone that looks similar to suggested that in the southern fault block, there the formation consists of microbialaminite and the upper part of the Zuun-Arts Formation. was repetition of the lower Bayangol Formation massive recrystallized limestone. The recrystal- This limestone is capped by shale, an obvious (their units 17 and 18, our BG2) due to thrust lization is patchy and is most likely the result lenticular pebble conglomerate that contains faults (Gibsher et al., 1991; Khomentovsky and of the contact metamorphism from nearby intru- phosphatized ooids and small shelly fossils, and Gibsher, 1996); however, we observed only sion. At 38 m above the base of the formation, bioturbated limestone. small offsets due to the conjugate sets of faults. well-preserved archaeocyaths appear, and at More strata of the Bayangol Formation are Importantly, this revised mapping changed the 84 m, they are the dominant reef former. exposed north of the Zavkhan River, where stratigraphy and placement of small shelly fossil The Khairkhan Formation rests conformably the basal contact with the Zuun-Arts Forma- horizons into the composite small shelly fossil on top of the Salaagol Formation at this local- tion is clearly exposed. However, there, much range chart. Instead of repeating strata (units 17 ity. Just to the south of E1325, this formation is of the Bayangol Formation strata are faulted, and 18, our BG2), on the west side of the gorge, better exposed as a poorly sorted, very coarse folded, relatively flat-lying, and massively we observed continuous stratigraphy that is sandstone to pebble conglomerate. The top con- weathering, preventing us from measuring a much thicker than previously reported. tact of the Khairkhan Formation is not exposed complete section of the formation. Sections Two sections of the Zuun-Arts Formation, anywhere in this area. There is ~10–20 m of E1217 and E1219 (not plotted in Fig. 7) com- one on the southern fault block (E1129) and the nonexposure between the top contact of the prise a composite section of BG2 and BG3 other on the northern fault block (E1126), were Khairkhan Formation and a red to black chert strata. E1217 sits conformably on E1216, the measured in Bayan Gorge; however, the data in of unknown age. measured section of the Zuun-Arts Formation Figure 7 are from E1129. The two sections are

12 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia both very well exposed with similar thickness Geologic mapping was focused on the eastern BG3 at this locality is carbonate dominated in and stratigraphy. side of the NW-SE Late Ordovician to Silu- the middle part of the member, as well as the Beginning at the top contact of the stromato- rian graben. There is a major fault separating upper part. It has thick, bioturbated limestone lite Boxonia grumulosa, the measured section the NE section from the SE one, so the results beds, oncolites, and oolites that are interbedded on the southern block (E1129) is 221 m thick. from these two sections are discussed separately with siltstone and fine to coarse sandstones in The base of the section is marked by green next. The SE section is interpreted to have been the lower to middle part of the member. phosphatic shale with bedded chert and nodu- deposited on the same fault block as the section The base of section E1336 is marked by the lar limestone that is overlain by thinly bedded from Salaa and Tsagaan Gorges (Figs. 1A and resistant thick limestone unit that is just south limestone with bedded and nodular chert. The 1D), whereas the NE Khukh-Davaa section is of the left-lateral fault marking Khukh-Davaa lower part of the Zuun-Arts Formation is pre- interpreted to be on the same fault block as the Pass. This section is composed of predomi- dominantly thinly bedded (1–2 cm), dark-gray section from Khunkher Gorge. Based on map- nantly green-colored, fine- to medium-grained limestone (Fig. 3B) with 0.5–1.5 m beds of ping of these two fault blocks, we interpret the sandstone. Here, 1–3 m microbial patch reefs limestone interclastic breccia. In the middle to thrust sheet with the sections from SE Khukh- and small centimeter-scale microbial balls with upper part of the Zuun-Arts Formation, there are Davaa and Salaa and Tsagaan Gorges to be the diffuse edges are present throughout the sec- several 5–15 m parasequences. The bases of the more proximal of the two. tion. Additionally, coarse white quartzarenites parasequence are marked by planar, thinly bed- become increasingly more dominant and coarser ded limestone interbedded with pink calcisiltite. SE Khukh-Davaa (E1209, E1207, E1336, up section. In the 20 m below the first archaeo- Up section, the beds become thicker, and the E1328, E1329, E1331, E1339, E1340) cyath reef, the sandstone is white, medium to tops of some of the parasequences are capped The section of the Zuun-Arts Formation at very coarse, and quartz dominated. We call this by cross-bedded ooid grainstone. this locality (E1209) is exceptional in several upper 20 m of the Bayangol Formation “mem- Beginning at the top contact of the stromato- regards, including unusual sedimentary fea- ber BG6”; however, the contact between BG5 lite Boxonia grumulosa, the measured section tures and expanded portions of different parts and BG6 is gradational. in the north block (E1126) is 203 m thick. This of the formation. The total thickness of this The Salaagol and Khairkhan Formations are section is very similar sedimentologically to measured section is 263 m. The base of the sec- well exposed in several localities south of the the section E1129. Well-preserved simple bed- tion is marked by a silicified karst surface fol- Khukh-Davaa Pass (Fig. 6D). Both formations planar trace fossils were found at 169 m. The lowed by thinly bedded, graded dolosiltite and are highly variable along strike, and two sec- upper contact with the Bayangol Formation is a 10 m horizon of Boxonia grumulosa. Capping tions that capture that variability are presented a sharp transgression between massive ooid the stromatolites are phosphatic shales inter- here (Fig. 8). The Salaagol Formation ranges grainstone of the Zuun-Arts Formation and bedded with mechanically bedded light-gray in thickness from 0 to 175 m. In many sections green phosphatic shale interbedded with lime- dolostone. Overlying the basal interbedded in this area, the limestone of the Salaagol For- stone of the basal Bayangol Formation. phosphatic shale and dolostone, there is dark- mation interfingers with siltstone, sandstone, The section of the Bayangol Formation on gray to black, sulfidic limestone and dolostone and conglomerate that are lithologically simi- the southern block of Bayan Gorge is the most- with rare domal stromatolites. The top 134 m lar to the overlying Khairkhan Formation. We studied section and the type locality of the Bay- portion is dominated by cross-bedded limestone define the top of the Salaagol Formation as the angol Formation. Two sets of unit numbering, ooid and mudchip grainstone. Paleoflow of the uppermost limestone unit. In some sections, the painted onto the cliff faces, have been used by cross-beds is to the southeast. The cross-bedded Salaagol Formation is composed of meter-scale Zhegallo and Zhuravlev (1991) and Gibsher ooid grainstone shoal at the top of this section is patch reefs that pinch out laterally. In other sec- et al. (1991), which were subsequently used by much thicker here than it is in any other section tions, it is composed of ~10-m-scale archaeo- Khomentovsky and Gibsher (1996). However, measured. cyath reefs with ~50 m cobble conglomerates due to the mismapping of the gorge, we do not The section of the Bayangol Formation is a between (Fig. 5C). In still other sections, the use either of the numbering schemes. The three composite section in which BG2–BG4 of the Salaagol Formation is entirely absent, with a thick carbonates that cap members BG2–BG4 Bayangol Formation (E1207) were measured cobble conglomerate resting unconformably on are clearly visible in satellite imagery and are just north of Khukh-Davaa Pass, and the upper very coarse quartz-rich sands of the upper Bay- present in many other localities. This is an part (BG5) of the formation (E1336) was mea- angol Formation. important distinction because previous writers sured just south of the pass in a broad anticline. In the thickest measured section of the have interpreted what we refer to as BG3 (units The two sections are separated by a left-lateral Salaagol Formation (E1328 and E1329), the 18 and 19 in Gibsher et al., 1991) as being a strike-slip fault; however, the offset of the fault base of the Salaagol Formation is composed of thrust repetition of BG2. As a result, our mea- is not major. thinly bedded, medium-gray limestone grain- sured section is thicker and contains additional The base of E1207 is marked by an unusually stone with ooids. At ~50 m, there is a gradational strata that are missing in the middle of previous thick section of red and green shale sharply over- contact with ooid grainstone and thrombolites measured sections. lying cross-bedded ooid grainstone of the upper with occasional quartz pebble conglomerates. Zuun-Arts Formation. There are many small, Further up section, thrombolitic and hyolith Khukh-Davaa insidious faults and few distinct marker beds in reefs dominate, with occasional sandstones and “Khukh-Davaa” is the name of the pass along the lowermost part of the Bayangol Formation conglomerates. Dispersed archaeocyaths first a left-lateral, E-W–oriented, strike-slip fault in in this section, and it is possible that the thick- appear at 111 m on the edges of stromatolites the northwestern Khevtee Tsakhir Range (Fig. ness in Figure 7 is exaggerated due to structural and become the dominant reef builders at 124 m. 6D). New mapping in this area resulted in the repetition of the shale beds. The top of BG2 is The upper 8 m portion of the section is thinly discovery of many new sections of the late marked by a few small shelly fossil horizons bedded limestone with glauconitic sandstone Ediacaran through early Cambrian stratigraphy interbedded with black shale and a phosphatic and phosphatic hardgrounds. Phosphatized and both north and south of the Khukh-Davaa Pass. hardground. Similar to other sections, member partially phosphatized hyoliths, chancelorids,

Geological Society of America Bulletin, v. 1XX, no. XX/XX 13

Smith et al.

LST

LST LST LST LST

TST

TST TST LST TST

HST

HST HST

TST

TST T HS T HS TST HST HST

LST HST 04 Dista l 9 –4 C (‰ VPDB) E1220

BG5 SG BG3 2 BG 4 BG OROLGO GORGE 04 –4 8 C (‰ VPDB) 13

J801, E113 8

BG4 BG2 ZA BG3 BG5 KHUNKHER GORGE 04 –4 7 C (‰ VPDB)

BG4 BG5 BG3b BG3a BG2 ZA E1211, E1212 NE KHUKH-DAVAA I 0 04 C (‰ VPDB) –4 13 6 δ

ALAA AND

BG5 ZA BG2 BG3 874, D718, D72 0 F TSAGAAN GORGES 04 C (‰ VPDB) –4 5 13 δ

E1336

4 BG BG2 BG5 BG3 ZA E1209, E1207, SE KHUHKH-DAVAA Limestone grainstone Microbialaminte/stromatolite/thrombolite Ribbonite limestone Rhythmite limestone Limestone debris ow breccia Conglomerate Medium-very coarse sandstone Very ne- ne sandstone Siltstone/shale Phosphatic shale 0 0 04 C chemostratigraphy of the uppermost Bayangol, Salaagol, and Khairkhan Formations. The five C chemostratigraphy of the uppermost Bayangol, Salaagol, and Khairkhan Formations. –4 13 4 C (‰ VPDB) 13 δ

E1130, D706

BG5 BG4 ZA BG2 3 BG S. BAYAN GORGE E1126, F1120, F1121, 1 04 3 –4 3 C (‰ VPDB) E1123

13 4

BG5 ZA BG4 BG2 3 BG SE of BAYAN GORG E 8 04 –4 C (‰ VPDB) 2 13 δ 6 KTN

Ooid s Archaeocyaths Oncoids Boxonia grumulosa Fault

BG3 BG2 ZA E1215, E1216 9 5 D 7 04 C (‰ VPDB) –4

revious 1 δ his study

3 BG BG2 BG4 ZA BG5 TAISHI R E1115, E1117 E1114, E1113, SSF s—p Incorrect SSF FA Ichnofossils Not exposed SSF s—t Proximal 250 500 0 (m) measured sections are numbered according to proximity to the paleoshoreline, with more distal sections on the right and more proximal sections on the left. Locations of proximal distal sections on the right and more with more to the paleoshoreline, according to proximity numbered sections are measured The sections 6.) indicated on Figure sections are locations of measured (Precise side of the figure. shown on the map in lower-left sections are the individual measured HST—highstand belemnite; Peedee VPDB—Vienna 8. Figure used in those same as the are Lithologies sequence. of the tract systems lowstand using the correlated are systems tract; LST—lowstand SSFs—small shelly fossils. TST—transgressive systems tract; Figure 8. Integrated lithostratigraphy, sequence stratigraphy, and δ stratigraphy, sequence lithostratigraphy, 8. Integrated Figure

14 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia and archaeocyaths are present in the sandy lime- facies, they are absent. In the most southern part etition in the lower part of the Bayangol Forma- stone at the contact with the Khairkhan Forma- of Figure 6D, the Khairkhan Formation contains tion; however, we saw no evidence for a major tion (Fig. 5D). Phosphatic hardgrounds and lag building-size limestone olistoliths that are visi thrust repetition and instead mapped this area as deposits with coated phosphate grains, small ble in satellite imagery. The upper part of the having a number of right-lateral faults. Unfor- shelly fossils, and archaeocyaths are also pres- Khairkhan Formation consists of light-green, tunately, there is no exposure between Tsagaan ent in other sections nearby in this area. The top gray, brown, and black finely laminated siltstone Gorge and the exposure to the NW, close to of the section is marked by a transgression into (Figs. 5G and 9C). The thickness of the forma- where section D720 was measured, which hundreds of meters of thinly bedded sandstone tion is difficult to determine due to faulting, obscures some of the fault traces in the lower and brown siltstone (Figs. 5G and 5H). folding, lateral facies change, and poor exposure Bayangol Formation. Because this area was In the other measured section in this area of the upper Khairkhan Formation, but in this more structurally complex than was previously (E1340), archaeocyath reefs are interbedded area, it is hundreds of meters thick. suggested (Khomentovsky and Gibsher, 1996) with tens of meters of cobble to boulder con- Overlying the Khairkhan Formation in the and there was uncertainty in the nature and off- glomerate. The conglomerate beds are well area just south of the Khukh-Davaa Pass (Figs. set of some of the faults, we did not measure or sorted with rounded clasts of sandstone and 1D and 6D), there is a buff- to red-colored, vari- sample the middle Bayangol at high resolution. granite to very poorly sorted with angular car- ably silicified microbial dolostone of unknown The Tsagaan Gorge section of the Zuun-Arts bonate and sandstone clasts. Clasts of almost all age (Fig. 5I). This unit is also present in the sec- Formation was previously measured, described, of the underlying Neoproterozoic to Cambrian tion southeast of Taishir and near Kvetee Tsakhir and sampled by Voronin et al. (1982), Gib- units were found, and we suggest these con- Nuruu. The contact between the Khairkhan For- sher and Khomentovsky (1990), Brasier et al. glomerate beds mark a regional unconformity. mation and this chert unit was not exposed in (1996b), and Khomentovsky and Gibsher The archaeocyath reefs in the section are 1–3 m any of the localities included in this study. (1996). These studies failed to note that the in scale and pinch out to the east. This section upper-middle part of the section is faulted and does not have a top contact because the section Salaa Gorge and Tsagaan Gorge (F874, recrystallized, obscuring bedding and sedimen- is exposed in a shallowly dipping syncline. D718, D720, F1131, E1109) tary features. This portion of the strata was not In the area south of the Khukh-Davaa Pass, These two gorges are ~4 km away from measured or sampled, and as a result, the thick- many facies of the Khairkhan Formation are each other on the northern flank of the Khevtee ness of the Zuun-Arts Formation at this locality present, ranging from siltstone to sandstone to Tsakhir Range, with Tsagaan Gorge being the is uncertain. The unmeasured portion is marked conglomerate. In some conglomeratic facies, eastern of the two (Fig. 6E). They are ~20 km by a fault in Figure 7. The basal Bayangol For- limestone archaeocyath-rich clasts of the southwest of the village of Jargalant. Khomen- mation was measured between Tsagaan and Salaagol Formation are dominant, but in other tovsky and Gibsher (1996) mapped a thrust rep- Salaa Gorges.

Proximal Distal 2 3 1100 1 4 SE KHUKH-DAVAA SALAA E1328, E1329, AREA GORGE OROLGO E1340 E1331

SE TAISHIR F1131, E1109 GORGE h

E1325 h E1220, E1223

Khairkhan Fm Khairkhan

K K K 750 TST Gl

HST

SG SG

SG TST

6 BG –4 –2 024 –4 –2 0 13 500

13 BG6 HST δ C (‰ VPDB) δ C (‰ VPDB) Fm Salaagol –4 –2 0

δ13C (‰ VPDB) Salaagol Fm Salaagol –4 –2 024 δ13C (‰ VPDB) 250 TST 2 SSFs—this study Ooids Exposure surface LST 3 Archaeocyaths 4 Gl Glauconite Oncoids HST 1 Cross bedding

Break in section Fm Bayangol TST Olistostromes 0 m Slump folding –4 –2 024 δ13C (‰ VPDB) Figure 9. (A) Autochthonous to allochthonous thrust sheets of different facies of the Khairkhan Formation and cumulate ultramafics and serpentinite. Clasts of ultramafic rock have been found in the Khairkhan Formation. Together, these stratigraphic and sedimentological results support the claim that the Zavkhan Basin was a foreland basin (Macdonald et al., 2009). (B) Photo of the cumulative ultramafics in thrust contact with the Salaagol Formation in northern Orolgo Gorge. (C) Photo of the parautochthonous facies of the Khairkhan Formation from SE Khukh-Davaa area.

Geological Society of America Bulletin, v. 1XX, no. XX/XX 15 Smith et al.

The sedimentology, paleontology, and paleo major structural complications of the Zuun-Arts fault, and this section restores to the east. The environment of the Salaagol Formation at Salaa and Bayangol Formations at Khunkher Gorge, exposure in the creek is excellent, but abundant Gorge have been described in detail by Kruse so no detailed map is presented herein, but the minor conjugate fault sets had to be mapped in et al. (1996). We remeasured this section to use location is shown in Figure 1D. detail before measuring this section (Fig. 6E). the detailed sedimentology and paleontology of The Zuun-Arts Formation was thickest at this The section has both a basal contact with the previous workers as a reference point, and also locality. Sedimentologically, it resembles the Zuun-Arts Formation and a top contact with to sample the entire formation at a higher reso- Zuun-Arts sections in SE Khukh-Davaa. The the Salaagol Formation. lution (~4 m) for d13C chemostratigraphy. The Bayangol Formation has a basal contact with Similar to other sections, the base of the total measured thickness of the composite sec- the Zuun-Arts Formation ooid grainstone and an Bayangol Formation at this locality is marked tion at this locality is 380 m. A red to yellow upper contact with the Salaagol Formation. The by cross-bedded ooid grainstone sharply over- carbonaceous silt was used to tie sections F1131 upper contact is not shown in Figure 7 because lain by phosphatic green shale with nodular and E1109 together. Figure 5B shows the upper there could be minor faulting at the base of the limestone. The basal ~155 m of this section are half of the measured section (E1109) overlain Salaagol Formation. Similar to other sections, sedimentologically different from other mea- by the Khairkhan Formation. members BG2 and BG3 are siliciclastic succes- sured sections. It is characterized by very thinly sions capped by thick limestone units. However, bedded (0.5–5 cm) gray to tan limestone that is NE Khukh-Davaa (E1211, E1212) member BG4 is either much thicker or is much often interbedded with black shale. Horizons A nearly complete, well-exposed section of more siliciclastic-dominated in this section. The containing small shelly fossils are abundant. At the Bayangol Formation was found north of upper part of the section has abundant beds with 137.5 m above the basal contact, there is a len- Khukh-Davaa Pass and northeast of a major ooids, coated grains, and oncoids. The thick ticular phosphatic hardground overlain by 10 cm NW-SE–trending fault, which faults the Edi- oncolite bed near the top of the section looks of phosphatized ooids (Fig. 3G). Above this dis- acaran-Cambrian strata against the Zavkhan very similar to the oncolite bed in Salaa Gorge. tinct marker bed, the limestone beds are thicker, Formation (Fig. 6D). The total thickness of the and no additional small shelly fossil horizons Bayangol Formation at this locality is 765 m but Orolgo Gorge (E1220, E1223) were found for hundreds of meters. Above the lacks a top contact with the Salaagol Formation. Insidious minor faulting is present through- thick, resistant limestone unit in the middle of The base of the section is marked by the contact out this section; however, the predictability of BG2, there is ~70 m of nonexposure in the creek. between a cross-bedded ooid grainstone of the the conjugate fault sets, the excellent exposure, Above this covered interval, there is a green bio- Zuun-Arts Formation and red to green phos- and the numerous distinct marker beds allow turbated siltstone and fine-grained sandstone that phatic shale of the basal Bayangol Formation. for tractable geological mapping, a detailed contains slump folds and some fault repetitions. In present-day coordinates, this section is measured section, and high-resolution sam- Although BG2 is expanded at this locality, closest to E1207, a section just north of Khukh- pling. Our mapping in Orolgo Gorge shows BG3 is condensed. However, it is similar to Davaa Pass; however, these sections are sepa- that the uppermost Zuun-Arts Formation is other sections in that there are thick limestone rated by a major fault and, as a result, have faulted against the Taishir Formation (Fig. 6E); units in the lower-middle part of the member. different facies. The overall thicknesses for the nowhere in this gorge is the lower to middle One distinctive feature of BG4 is the basal por- different members of the Bayangol Formation Zuun-Arts Formation exposed. This finding is tion, which is characterized by red and green are relatively similar, but the carbonate content important for small shelly fossil biostratigraphy shale interbedded with pink thrombolites and and rock type within each of the members dif- because the FAD of Anabarites trisulcatus and limestone microbialaminites. A thick micro- fer. For example, BG3 is expanded in both of Cambrotubulus decurvatus is reported from the bial limestone ridge-former caps member BG4. these sections; however, it is more carbonate- base of the Zuun-Arts Formation at this locality, Member BG5 is perhaps the most unusual part dominated in section E1207 than it is in this most likely because the phosphatic shale at the of this measured section because it has far more one. At this locality, on the other hand, BG4 base of the Bayangol Formation was mistaken limestone beds than any other measured section is more carbonate-dominated and thicker. The as the phosphatic shale at the base of the Zuun- in the basin. The base of BG5 is characterized by limestone capping this member was difficult to Arts Formation (Endonzhamts and Lkhasuren, fine to very coarse, planar to cross-bedded sand- measure and sample because the terrain is very 1988). As a result of our revised geologic map, stone. There is a distinct microbialaminite and steep and massive weathering, making it dif- we place this small shelly fossil horizon 200– microbial balls in the middle of the siliciclastics. ficult to see bedding and sedimentary textures. 300 m stratigraphically higher, in the base of the The top of BG5 is characterized by abundant, Member BG5, consisting of poorly sorted fine Bayangol Formation. channelized cross-bedded calcarenites, oolites, to very coarse sandstone, is only 87 m thick Small shelly fossil horizons had been previ- and oncolites interbedded with cross-bedded here, which is relatively thin compared to other ously described at Orolgo Gorge, but no mea- sandstone. Above the carbonate-dominated localities. The top of the section is marked by sured section or chemostratigraphy has been portion of BG5, there are more coarse to very what appear to be massively recrystallized white reported from this locality. The Bayangol, coarse cross-bedded quartzarenites and calcar- to pink limestone stromatolites that are likely Salaagol, and Khairkhan Formations are partic- enites. The top contact with the Salaagol For- still part of the upper Bayangol Formation. The ularly well exposed, expanded, and carbonate- mation is marked by red to gray silty limestone top of this measured section is faulted against a dominated in this section, and geologic map- microbial reefs with dispersed archaeocyaths. Paleozoic intrusion. ping allowed for detailed lithostratigraphic and The section of the Salaagol Formation mea- chemostratigraphic studies in this gorge. sured here (top of E1220) is a continuation of Khunkher Gorge (E1138) The measured section in Orolgo Gorge is the the section of the Bayangol Formation from this This previously undescribed section of the thickest, most carbonate-dominated section that locality. The upper part of the Bayangol Forma- Zuun-Arts Formation through Bayangol Forma- was measured of the Bayangol Formation. In tion has thick limestone units that are not pres- tion was measured in Khunkher Gorge and on modern coordinates, it is close to Salaa Gorge; ent in other sections. As a result, the boundary the ridge to the west of the gorge. There are no however, the two are separated by a major between the Bayangol and Salaagol Formations

16 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia was difficult to determine. Here, we define the Small Shelly Fossil Biostratigraphy small shelly fossil horizon in the upper part of base of the Salaagol Formation as the base of the member BG3 (see Fig. 7). last major carbonate unit that is stratigraphically We place previously reported small shelly New small shelly fossil horizons are reported below the Khairkhan Formation. The base of fossil horizons (Korobov and Missarzhevsky, here, but the phosphatized small shelly fossils this section is marked by a distinct red siltstone 1977; Voronin et al., 1982; Endonzhamts and within them have not yet been dissolved, identi- and silty microbial limestone (Fig. 5A). Lkhasuren, 1988; Zhegallo and Zhuravlev, 1991; fied, or imaged. The small shelly fossil horizons The Salaagol Formation in this gorge is over Goldring and Jensen, 1996; Khomentovsky are almost always preserved as lenticular lag 343 m thick. The top of the formation is a talus and Gibsher, 1996) into the new stratigraphic deposits that pinch in and out laterally (Fig. 3G). slope and is not exposed on either side of the and d13C chemostratigraphic context, and we They are often found in association with phos- gorge. The basal to middle part of the formation document new small shelly fossil horizons in phatized coated grains and ooids. One exception is predominantly gray to pink to green limestone multiple sections from across the Zavkhan ter- to this is the type of preservation seen in the microbialites with siltstone interbeds. The first rane. Here, we present changes in the results of small shelly fossil horizons in the basal part of archaeocyaths appear 70 m above the basal con- stratigraphic placement of previously described Orolgo Gorge, where the small shelly fossils are tact. The middle to upper portion of the section small shelly fossil horizons and the resulting large, unbroken, and not found in association continues to be dominated by microbialites, but revised FADs as a result of new geologic map- with other grains (Fig. 3H). Additionally, they thinly bedded nodular limestone beds and onco- ping and refined correlations. are found on dozens of bedding planes in the lites become increasingly more common in the The lowermost small shelly fossil finds on the basal ~200 m of measured section E1220. More upper part of the section. Zavkhan terrane are reported from the Zuun-Arts generally, the small shelly fossil horizons are The overlying Khairkhan Formation is folded Formation in three localities: Kvetee Tsakhir not discrete beds that are continuous markers in a syncline at this locality. The formation was Nuruu (Dorjnamjaa et al., 1993), Orolgo Gorge, across the basin, but they rather occur at differ- not measured in detail but was instead just gener- and Salaa Gorge (Endonzhamts and Lkhasuren, ent stratigraphic horizons in different sections, ally characterized on an overturned limb on the 1988). Dorjnamjaa et al. (1993) reported consistent with the rapid lateral facies changes north side of the fold (Fig. 6E). The estimated Anabarites trisulcatus from siliceous phospho- across the basin. thickness of the formation here is ~550 m. The rite at the base of the Zuun-Arts Formation; how- basal part of the formation contains limestone ever, other workers were not able to confirm this Ichnofossil Biostratigraphy boulders and olistoliths that are concordant to finding (Brasier et al., 1996b). After extensive the surrounding matrix of shale and silt, and searching near the base of the Zuun-Arts Forma- Abundant and diverse trace fossils are pres- slumping. The rest of the formation is dominated tion at this locality, we were not able to find any ent throughout the Bayangol Formation, and as by sequences of very poorly sorted, matrix- small shelly fossil horizons, and we suspect that Goldring and Jensen (1996) pointed out, west- supported conglomerate with angular clasts of the siliceous phosphatic horizon at the base of the ern Mongolia is one of the few places globally black chert, quartz, ultramafics, and limestone Bayangol Formation was confused for the one at in which they are interbedded with skeletal fos- (Figs. 5E and 5H) that fine up to sandstone and the base of the Zuun-Arts Formation. sil assemblages, providing the opportunity to siltstone. Many of conglomerates and coarse Similarly, the FADs of Anabarites trisulcatus calibrate the relative timing of these two types sandstone are channelized. Some of the con- and Cambrotubulus decurvatus from the base of of fossil records. Here, we present a few more glomerates contain larger boulders and olisto the Zuun-Arts Formation in Orolgo and Salaa notable finds to add to the well-documented liths of archaeocyath-rich limestone. To the Gorges could not be confirmed (Endonzhamts ichnotaxa that have been previously described north and west of this section, there are build- and Lkhasuren, 1988). In fact, in Orolgo Gorge, and adjust their stratigraphic placement. ing-size olistoliths of the Salaagol Formation in the Zuun-Arts Formation is faulted against the Given that the Precambrian-Cambrian bound- the Khairkhan Formation (Fig. 5F). The contact Taishir Formation, and only the uppermost part ary is defined as occurring at the FAD of the between the Salaagol and Khairkhan Formations of the former is present (Fig. 6E). Again, we trace fossil T. pedum at the Fortune Head section is not well exposed in this gorge; however, sam- suspect that confusion between the two phos- (Landing, 1994), ichnologists working in the ples with phosphatized archaeocyaths and small phatic horizons resulted in the misplacement of Zavkhan terrane have devoted much time and shelly fossils (similar to those in Fig. 5A) were this important FAD. These revisions are shown energy trying to document the FAD of this trace discovered in a talus slope near the contact. in Figure 7. fossil. The FAD of this fossil, according to Gold- The Khairkhan and Salaagol Formations are Using the locations of previously published ring and Jensen (1996), is in unit 20 of the Bay- in fault contact with ultramafic cumulates and small shelly fossil horizons from Bayan Gorge angol Formation. With our revised geological serpentinite (Fig. 7B), and in the area to the (Khomentovsky and Gibsher, 1996) and our mapping and stratigraphy (see previous results north and west of Orolgo Gorge, the Khairkhan revised mapping, some of the small shelly fossil sections “Geological Mapping” and “Descrip- Formation contains clasts of the ultramafic rocks horizons (R-IV through R-X; see Fig. 7) have tion of Key Sections”), we place this unit at the (Fig. 6E). Northwest of the ultramafic rocks in been moved stratigraphically up into mem- contact between members BG3 and BG4 (see Orolgo Gorge and on the northeastern edge of ber BG3. The previous interpretation was the Fig. 7). Other ichnogenera from unit 20 include the large Paleozoic intrusion, there are hundreds result of incorrect mapping of BG3 as a thrust Monomorphichnus isp., Hormosiroidea isp., of meters of very finely laminated green to repetition of BG2 (Khomentovsky and Gib- and Rusophycus cf. avalonensis (Goldring and white to brown siltstone of the Khairkhan For- sher, 1996). Jensen, 1996). Just below, in unit 18, or what mation (Figs. 5G, 9A, and 9C). Due to the poor The small shelly fossil horizons from the we interpret as the lower part of member BG3, exposure and the nearby intrusion, the nature Taishir section had not been put into strati- Helminthoida, Spatangopsis, and Nemiana were of the underlying contact is not clear; however, graphic context because the area had neither a found. The trace fossil assemblage from member the map relation with the underlying olistolith detailed map nor a composite measured section. BG4 at Taishir includes Oldhamia, Plagiogmus, facies of the Khairkhan Formation suggests that With our revised mapping and composite mea- and possible Zoophycus (Fig. 7; Goldring and it is a fault contact (Fig. 9A). sured section, we place the previously reported Jensen, 1996).

Geological Society of America Bulletin, v. 1XX, no. XX/XX 17 Smith et al.

We emphasize that the preservation of T. The d13C values from nine measured sections archaeocyaths, and coated grains. In many sec- pedum and the other trace fossils is highly facies of the Bayangol Formation are presented here. tions, this contact is faulted, perhaps because dependent in the Zavkhan terrane. Although we Brasier et al. (1996b) identified positive excur- the contact represents a sharp rheologic bound- were not extensively searching for T. pedum, we sions B–F in the Bayangol Formation, and data ary between a massive limestone reef and more did not find it at any other stratigraphic horizon presented here confirm the existence of at least fissile mixed siliciclastic units. Nonetheless, or in any other measured section. We agree with five positive excursions. All measured sections there are many sections that are well preserved previous workers that placing the Precambrian- presented here show high-frequency oscilla- and not compromised by faulting for which the Cambrian boundary at the FAD of T. pedum tions in d13C values that range between ~–6‰ transgression can be used as the datum for the on the Zavkhan terrane contradicts other lines and +6‰. Interestingly, the positive excursions sections in Figure 8. of evidence that suggest that the Precambrian- are sharp and pronounced, whereas the negative The tie lines in Figures 7 and 8 show how the Cambrian boundary is actually much lower in excursions have less-pronounced minimum val- measured sections in the Zavkhan terrane corre- the stratigraphy. Although some workers have ues. The limestone beds in BG5 from Orolgo, late to one another using major sequence bound- shown that there is a wide environmental toler- Bayan, and Khunkher Gorges, the three gorges aries within the Bayangol and Salaagol Forma- ance for T. pedum and have used this to argue that have exposure of carbonates in the upper tions. Portions of the major sequences within for further support of the GSSP type section of Bayangol Formation, have unusually high d13C the Bayangol Formation are also shown and the Precambrian-Cambrian boundary (Buatois values of <+6‰. annotated in photographs from portions of the et al., 2013), many other workers have shown The d13C profiles for the Salaagol Forma- sections in N Khukh-Davaa Pass and S Bayan trace fossil production and preservation of tion were surprisingly different from the previ- Gorge in Figure 4. Generally, the massive car- T. pedum to be strongly lithofacies controlled ously reported d13C values from Salaa Gorge bonate reefs that cap each informal member of (Geyer and Uchman, 1995; MacNaughton and and Zuun-Arts Ridge (Brasier et al., 1996b). the Bayangol Formation represent the HST of Narbonne, 1999; Gehling et al., 2001; Geyer, In Orolgo Gorge, the d13C values are as high the sequences, with the maximum flooding sur- 2005). As a result, using T. pedum as a global as 7‰. In Salaa Gorge and in SE Taishir, the face (MFS) near the base of each of these. These chronostratigraphic marker has been problem- d13C values in the lower-middle part of the are in turn overlain by medium to coarse sands atic on most Cambrian paleocontinents, nota- section reach ~+4.5‰. The d13C values in the that represent the LST of the sequences. The bly in Siberia, China, Mongolia, and Kazakh- upper part of the measured sections at Salaa thicknesses of the LSTs and TSTs vary between stan (Babcock et al., 2011, 2014; Landing Gorge, Khukh-Davaa, and SE Taishir drop to sections, depending on a section’s proximity to et al., 2013). ~–2‰ to –1‰. Brasier et al. (1996b) reported the shoreline. The more distal sections tend to Although T. pedum was extremely diffi- 23 d13C values that range between –0.6‰ and preserve thinner LSTs and thicker TSTs than the cult to find in the Zavkhan terrane, there were –1.6‰ from 44 m of strata of the Salaagol For- more proximal ones. many sandstone horizons that yielded abun- mation at the Zuun-Arts locality and d13C values Complicating the identification of major dant and diverse trace fossils. Figures 3I, 3J, that range between –0.2‰ and –1.2‰ from 22 sequence system tracts are minor shallowing- 3K, and 3L display a few of the exceptional samples from the upper 117 m of the Salaagol up sequences within the large sequences as well ichnofossils that were found in the Khunkher Formation at Salaa Gorge. The discrepancy as bypass channel sands. Obvious examples and Bayan Gorges. The early arthropod trace between the range of d13C values reported here of bypass channel sands are in members BG4 fossils in Figures 3I and 3J are from middle and previously reported values is the result of and BG5 of the Khunkher Gorge section (Fig. BG3 and that in Figure 3K is from BG5 in earlier workers only sampling the upper por- 10). The nearby sections contain ~150-m-thick the Khunkher section (for exact horizons, see tions of the Salaagol Formation. limestone reefs capping BG4 and BG5, whereas Fig. 7). The trace fossil in Figure 3L is from the section in Khunkher Gorge instead has rela- BG5 in Bayan Gorge. DISCUSSION tively thin limestone units with abundant oncolite, ooid, and calcarenite beds at the same strati- Carbon Isotope Chemostratigraphy Facies Model and Sequence Stratigraphy graphic horizon. The d13C chemostratigraphic curves provide independent confirmation that Carbon isotope (d13C) chemostratigraphy Within the Zuun-Arts through Khairkhan the lithologically different units of BG4 and for individual sections of the Zuun-Arts and succession, there are several major sequence BG5 in the Orolgo and Khunkher Gorges were Bayangol Formations and for the Salaagol boundaries that can be used as stratigraphic being deposited synchronously. Formation for each locality is presented in tie lines across the basin. The phosphatic shale One interesting feature of the integrated Figures 7 and 8, respectively. The d13C values units at the bases of the Zuun-Arts and Bayan- chemostratigraphy and sequence stratigraphy from seven measured sections of the Zuun-Arts gol Formations vary in thickness and color but is that the large d13C excursions of the Bayan- Formation are presented here. The top of the are easily recognizable stratigraphic markers of gol Formation generally seem to occur at the Zuun-Arts Formation is marked by at least one transgressions in the sequence. The phosphatic same frequency as the major sequence bound- large negative excursion that has a minimum of shale at the base of the Bayangol Formation is aries. With sedimentological, stratigraphic, and ~–7‰. This excursion was recognized by previ- a distinct stratigraphic marker that we use as chemostratigraphic data from across the basin, ous workers (referred to as anomaly “W”), and the datum upon which to align the sections in mechanistic hypotheses about the relationships it has been correlated globally with the negative Figure 7. The sequence boundaries with the between feedbacks in sea level, associated nutri- d13C excursion that, despite some uncertainty in Bayangol Formation are discussed in detail. ent input, and the carbon cycle can begin to be precise correlations (or summary, see Babcock The top contact of the Salaagol Formation with tested with data from the rock record. The nega- et al., 2014 f), has been found in association the Khairkhan Formation is marked by a LST tive d13C excursions rarely coincide with major with the FAD of T. pedum (Narbonne et al., overlain by a sharp transgression that, in some sequence boundaries, but the pronounced posi- 1994; Brasier et al., 1996b; Corsetti and Haga- sections, is marked by a phosphatic hardground tive d13C excursions are often found preserved dorn, 2000). overlain by phosphatized small shelly fossils, between the TST and the HST of the sequences.

18 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

Siberia (Kouchinsky et al., 2007), the sections Limestone reefs at of the Bayangol Formation preserve additional top of BG4 minor excursions (Fig. 7). Bypass sand channel The magnitude of the positive excursions as well as the thicknesses over which they occur vary between the sections. It is also ZAVKHAN TERRANE clear from comparing both the lithostratigra- SUBDUCTING TO SW KHANTAISHIR- phy and chemostratigraphy between measured DARIV ARC sections that the facies transitions are diachro nous across the basin. This interpretation is supported by comparing the sections from Figure 10. Schematic depositional model for member BG4 of the Khunkher and Orolgo Gorges (Fig. 7); the Bayangol Formation. Evidence for bypass channel sands exists in chemostratigraphic profiles for the upper parts the measured section of the upper Bayangol Formation in Khunkher of the Bayangol Formation are almost identi- Gorge (Fig. 7). cal, but the lithostratigraphy is very different. There is also obvious diachroneity between the two measured sections near Khukh-Davaa Pass, Using the age model presented herein and the points between sections in the Zavkhan terrane with the chemostratigraphic profiles closely resulting estimates for sedimentation rates, we are in the supplementary material.1 matching each other but with noticeable dif- calculated the lag time between transgressions Although high-resolution d13C chemostrati- ferences in the lithostratigraphy. These are just and positive d13C excursions. graphic correlation within the Zuun-Arts For- two of many obvious examples of the lithologi- An important tectonic result of the new facies mation is complicated by high-frequency vari- cal variability of the sections of the Bayangol model for the Zavkhan terrane is the documen- ability in the lower half of the formation, there Formation, highlighting the importance of the tation of the deep-water, allochthonous facies is one large negative excursion that is <9.5‰ integrated approach used here. of the Khairkhan Formation in northern Orolgo in magnitude preserved near the top of all mea- One interesting chemostratigraphic feature Gorge (Fig. 9A). These include poorly sorted sured sections of the Zuun-Arts Formation. The is the positive excursion (~7‰) at the base of conglomeratic facies of the Khairkhan For- more proximal sections (Taishir and Kvetee the measured section at Orolgo Gorge. Sedi- mation that contain ultramafic clasts. Nearby, Tsakhir Nuruu) preserve a negative d13C excur- mentologically, the basal part of this section the Khairkhan and Salaagol Formations are sion in the middle of the formation. The lateral looks very different from any other section in thrust contact with a large ultramafic body variability within the Zuun-Arts Formation is because it contains a much thicker section of (Figs. 6E and 9B). The second deep-water the result of diachronous deposition. The two phosphatic shale, black shale, and interbedded facies of the Khairkhan Formation is hundreds phosphatic shales that bookend the formation limestone. We have interpreted this portion of meters of very finely laminated siltstone in mark two transgressions that represent two of the stratigraphy as an expanded basal part the area just to the north of Orolgo Gorge (Figs. synchronous sequence boundaries across the of the Bayangol Formation. It is important to 9A and 9C). We suggest that the deeper-water Zavkhan terrane. Using this sedimentologi- note, however, that the sharp drop from posi- facies of the Khairkhan Formation are alloch- cal and sequence stratigraphic argument, the tive to negative d13C values occurs across a thonous to the Zavkhan terrane, bolstering the upper part of the Zuun-Arts Formation in the sharp sedimentological contact that contains stratigraphic argument that the late Ediacaran proximal sections cannot be correlative with a phosphatic hardground, phosphatized ooids, to early Cambrian strata were deposited in a the lower Bayangol Formation in the more dis- and pyrite pseudomorphs (Fig. 3G). We have foreland basin (Macdonald et al., 2009). Thus, tal sections. interpreted this as a small unconformity in the although small-scale transgressions associated Correlations between the nine measured succession. with positive d13C excursions may be global sections of the Bayangol Formation are not Compared to the correlations within the and eustatic, the larger-scale sequences and as straightforward as those in the Zuun-Arts lower formations, sequence and chemostrati- basin architecture are undoubtedly the product and Salaagol Formations for three reasons: graphic correlations between sections of the of local tectonism. (1) Unlike the other two formations, there is Salaagol Formation (Fig. 8) are relatively not continuous carbonate deposition in the straightforward. Comparison of the sedimen- Constructing a Composite Carbon Isotope Bayangol Formation, (2) there are large lateral tology and chemostratigraphy between sec- Curve for the Zavkhan Terrane changes in facies and bed thickness across the tions demonstrates that there is diachronous basin that are further complicated by sandstone- deposition of this formation and large facies The d13C chemostratigraphic data are gener- filled bypass channels, and (3) the carbon cycle changes across the basin. This interpretation is ally consistent with the sequence stratigraphic rapidly oscillated during the earliest Cambrian supported by comparing parts of the formation correlations used in Figures 7 and 8. In Figure 11, with several excursions that, taken alone, are between the sections at Orolgo (E1220) and the composite curve is presented next to the d13C nondistinct. To further complicate correlations, Salaa Gorge (F1131 and E1109). The section curve from Orolgo Gorge, the most expanded, compared to the very smooth d13C profile with at Orolgo Gorge is much more expanded with carbonate-dominated section in the Zavkhan five clear positive excursions from Sukharikha, abundant oncolite beds. Also of note are d13C terrane in which 2p–6p are preserved. Details values, which jump from positive to negative regarding the specific tie points used in each of 1GSA Data Repository item 2015265, strati- across a sequence boundary in the Salaagol graphic, chemostratigraphic, biostratigraphic, and the late Ediacaran to early Cambrian Mongolian chronostratigraphic data presented in this paper, is Formation. The details and age model implica- 13 formations are discussed in stratigraphic order. available at http://www.geosociety.org/pubs/ft2015 tions of this change in d C values are discussed The specificd 13C values that were used as tie .htm or by request to editing@geosociety .org. later herein.

Geological Society of America Bulletin, v. 1XX, no. XX/XX 19 Smith et al.

Orolgo Gorge (BG/SG data) + Composite data from Composite data from

Kh A B C

Bayan Gorge (Z-A data) 6 localities in Mongolia 4 localities in Morocco 010

2 al., et

Maloof Age model from model Age Tommotian 525 Mya

BG5 5p

6p BG4 4p 530 Mya

5p BG3 3p 4p 535 Mya 3p 2p BG2 2p Nemakit-Daldynia n

540 Mya Z-A 1p

1p Sh

–5 05–5 05–5 05diacaran δ13C (‰ VPDB) δ13C (‰ VPDB) δ13C (‰ VPDB) E

Figure 11. Revised δ13C chemostratigraphic age model for the Zuun-Arts through Salaagol Formations in the Zavkhan terrane. (A) δ13C isotopes plotted next to a generalized stratigraphic column for the Zuun-Arts Formation at Bayan Gorge and the Bayangol through Khairkhan Formations at Orolgo Gorge. Here, δ13C values are plotted against measured thickness. These data are plotted because Orolgo Gorge has the thickest, most carbonate-dominated section of the Bayangol and Salaagol Formations. (B) A composite δ13C curve of six sections from the Zavkhan terrane. Generally, sections are correlated using sequence stratigraphy, and here, the composite curve is constructed by using positive δ13C excursions for tie points and stretching the δ13C curves between. The uppermost Salaagol and Khairkhan Formations are Tommotian in age. (C) A composite δ13C curve from four localities in Morocco, a locality where there are absolute age constraints (Maloof et al., 2005, 2010b). Here, δ13C values are plotted against age. U/Pb tie points are marked by yellow (Oman—Bowring et al., 2007) and gray (Maloof et al., 2005, 2010a) circles. The age model on the right side of the figure is from Maloof et al. (2010a). The transparent colored boxes show how these three δ13C chemostratigraphic curves correlate to each other. Following Maloof et al. (2010a), positive excursions in the late Ediacaran through Nemakit-Daldynian are numbered and labeled with white circles. VPDB—Vienna Peedee belemnite; HST—highstand systems tract; TST—transgressive systems tract; LST—lowstand systems tract.

Refined Age Model documented globally near the Ediacaran-Cam- onto the Moroccan d13C record, which is con- brian boundary in localities such as Morocco, strained with absolute U/Pb ages from Morocco Carbon isotope chemostratigraphy is useful China, Siberia, Oman, the western United (Maloof et al., 2005, 2010b) and Oman (Bowring as both a correlation tool and, with the absence States, and northwestern Canada (Narbonne et al., 2007). Although Morocco lacks fossils in of absolute ages, for constructing a more precise et al., 1994; Corsetti and Hagadorn, 2000; the earliest Cambrian, it has a very expanded age model that can be used to correlate between Maloof et al., 2005; Kouchinsky et al., 2007; Li d13C curve from four localities with four chemi- the composite d13C curve from the Zavkhan ter- et al., 2009). This negative excursion at the Edi- cal abrasion–isotope dissolution–thermal ion- rane and other sections globally that do contain acaran-Cambrian boundary, the high-frequency ization mass spectrometry (CA-ID-TIMS) U/Pb absolute ages. Additionally, under the assump- oscillations during the Nemakit-Daldynian, and zircon ages on ash beds of 525.343 ± 0.088 Ma tion that changes in d13C values mark synchro- the two very high excursions at the end of the (Maloof et al., 2005, 2010b), 524.837 ± nous, global markers, chemostratigraphic cor- Nemakit-Daldynian and in the early Tommotian 0.092 Ma (Maloof et al., 2010b), 523.17 ± relation avoids circular arguments that result (Maloof et al., 2005, 2010b; Kouchinsky et al., 0.16 Ma (Maloof et al., 2010b), and 520.93 ± from correlating poorly established, facies- 2007), are all present in the Zavkhan terrane 0.14 Ma (Landing et al., 1998; Compston et al., dependent, and highly regional paleontological strata, increasing confidence in the use ofd 13C 1992; Maloof et al., 2010b). The age from the biozones. chemostratigraphic profiles as a tool for global Ediacaran-Cambrian boundary is a CA-ID- We correlate the largest negative d13C excur- correlation. TIMS U/Pb zircon age on an ash bed from sion at the top of the Zuun-Arts Formation with Here, we map late Ediacaran and early Cam- Oman dated at 541.00 ± 0.29 Ma (Bowring other negative d13C excursions that have been brian records of d13C variability from Mongolia et al., 2007; Peng et al., 2012).

20 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

Although there is some uncertainty in pre- age model is that the Salaagol Formation is Refined Biostratigraphy cise chemostratigraphic correlation between Nemakit-Daldynian to Tommotian in age with the Mongolia and Morocco composite curves, relatively continuous strata. This age model is By placing the previously described fossil large-scale uncertainty is mitigated by the based, in part, on the similarity between Sr iso- horizons in an accurate sequence stratigraphic unambiguous end-Ediacaran negative excur- tope values from Tommotian strata in Siberia and chemostratigraphic framework, we can sion and the end Nemakit-Daldynian positive (Derry et al., 1994; Kaufman et al., 1996) and compile a new composite biostratigraphic range excursion. High-frequency d13C oscillations from the upper Salaagol Formation (Brasier chart for ichnofossils and body fossils for the during the earliest Cambrian are bookended by et al., 1996b). It is also based on the complete Zavkhan terrane (Fig. 12). Additionally, this these two easily identifiable chemostratigraphic lack of any trilobite fossils (whole or disarticu- work establishes the necessary framework for markers that are preserved in both Mongolia and lated) in the upper Salaagol and Khairkhan For- more detailed paleontological studies in new Morocco. Following Kouchinsky et al. (2007) mations, which would be expected to occur if sections from across the basin. and Maloof et al. (2010a), we assigned the posi- these units were Atdabanian or Botomian in age. One important result is that this work helps tive excursions numbers 2 through 6 (1 is the However, this age model is in conflict with the resolve the long-standing problem of under- positive excursion during the end-Ediacaran). archaeocyath biostratigraphy from the Zavkhan standing the relative timing of the FAD of It is interesting that both in Mongolia and terrane, which uses a few archaeocyath species T. pedum, the FAD of small shelly fossils, and Morocco, there appear to be additional minor that are thought to be diagnostic of the Atdaba- the negative excursion that is associated with excursion between 1p and 2p, 3p and 4p, and nian and Botomian Stages. Most of the archaeo- the Ediacaran-Cambrian boundary. By adhering 4p and 5p, suggesting that the carbon cycle was cyaths collected by Kruse et al. (1996) from the strictly to the GSSP definition of the boundary oscillating even more rapidly than previously Salaagol Formation range from the Tommotian as the horizon that coincides with the FAD of thought during the earliest Cambrian. to Botomian in age; however, there are a few T. pedum (Landing, 1994), the Ediacaran-Cam- This refined age model results in a couple of species, such as Tabulacyathellus bidzhaensis, brian boundary in the Zavkhan terrane would be important new interpretations about the age Archaeopharetra marginata, Couberticyathus placed between members BG3 and BG4 (Fig. of the strata in the Zavkhan terrane. The first is lepidus, Robustocyathellus abundans, Orbi 12), contradicting other lines of evidence that that the Bayangol Formation is entirely Nemakit- cyathus mongolicus, and Ladaecyathus mel suggest the boundary is much lower. These lines Daldynian in age, rather than extending into the nikovae, that are thought to be restricted just to of evidence include (1) the FAD of T. pedum is Tommotian as was previously suggested. The the Atdabanian and/or Botomian (Paleobiology ~275 m above the large negative carbon isotope evidence supporting this interpretation is the Database). We emphasize that the age ranges excursion, which has been commonly used as very high d13C values of ~7‰ near the top of for these specimens were calibrated in just Sibe- a secondary chronostratigraphic marker (Grotz- the measured sections of the Bayangol Forma- ria, and thus it is possible that these specimens inger et al., 1995; Brasier et al., 1996b; Maloof tion in Bayan, Khunkher, and Orolgo Gorges. have a longer stratigraphic extent globally. The et al., 2005, 2010a; Zhu et al., 2006; Peng et al., The only two times during the early Cambrian second age model for the Salaagol Formation, 2012), in the Zuun-Arts Formation, (2) the FAD that the d13C values were this high was during which is based on the robustness of using these of T. pedum is ~250 m above the FAD of small the middle to late Nemakit-Daldynian. A second few archaeocyth species for biostratigraphy, shelly fossils, and (3) the FAD of T. pedum is important point is that there is a second very high requires there to be a 6–10 m.y. unconformity in stratigraphically above the FAD of Rusophycus, d13C excursion that begins in the uppermost Bay- the medium sandstone to pebble conglomerate which inverts the normal ichnofossil biostra- angol Formation and reaches values of 8‰ in the beds of the middle-upper Salaagol Formation tigraphy (Crimes, 1987, 1992; MacNaughton lower Salaagol Formation. The level of peak val- (Fig. 8). and Narbonne, 1999). The facies dependence ues is most clearly documented at Orolgo Gorge, The second age model necessitates a multi and rarity of T. pedum in the Zavkhan terrane but it is reproduced in other measured sections of stage subsidence history in which the mar- provide further support for reassessing the Edi- the Salaagol Formation as well (Figs. 7 and 8). gin subsides for ~15 m.y., there is a hiatus for acaran-Cambrian boundary with a marker that This second very high d13C peak coincides 6–10 m.y., and then it subsides again for a few is not as facies dependent as the first evolution- with the FAD of archaeocyaths in the Zavkhan million years during deposition of the upper ary appearance of the behavior represented by terrane. Correlating this d13C excursion with the Salaagol and Khairkhan Formations. The T. pedum. excursion at the Nemakit-Daldynian to Tommo- 6–10 m.y. hiatus in subsidence could reflect We were not able to reproduce any of the tian boundary distinguishes the archaeocyaths either slab breakoff or the passing of a periph- previous claims that the FAD of small shelly in the lower to middle Salaagol Formation as eral forebulge before the final obduction of the fossils is in the basal Zuun-Arts Formation, and among the earliest documented archaeocyaths allochthonous flysch of the Khairkhan Forma- we attribute the previous placement to misiden- globally. The d13C values between the two high tion. The subsidence history for the first age tification of the phosphatic shales at the basal peaks in Mongolia are not as low as the –2‰ model is much simpler, with a margin that is Bayangol Formation as those at the basal Zuun- to –1‰ values in Morocco, which we inter- subsiding for 16–20 m.y. before the final colli- Arts Formation. With the FAD of small shelly pret as representing a small unconformity in sion of the Khantaishir-Dariv arc. With a lack of fossils shifted stratigraphically up into the the conglomerate beds just below the base of the trilobites, Tommotian-like Sr values, and a sim- basal Bayangol Formation, globally, the FAD Salaagol Formation. pler subsidence mechanism, we prefer the first of small shelly fossils and T. pedum only occur Unfortunately, the d13C values of –2‰ to age model for the upper Salaagol and Khairkhan stratigraphically above the large negative carbon 0‰ of the upper Salaagol Formation are not Formations (Fig. 11). We suggest that the few isotope excursion at the top of the Ediacaran- diagnostic of d13C values of any one member archaeocyath species from the upper Salaagol Cambrian boundary (Brasier et al., 1990; Knoll in the early Cambrian, and they are consistent Formation that are thought to be diagnostic of et al., 1995b; Zhou et al., 1997; Kouchinsky with a Tommotian, Atdabanian, or a Botomian the Atdabanian or Botomian were either mis- et al., 2001, 2007). The one section globally that age. Here, we discuss two possible age mod- identified or have longer stratigraphic ranges is a possible exception to this is the section at els for the upper Salaagol Formation. The first than was previously thought. Khorbusuonka River (Knoll et al., 1995a).

Geological Society of America Bulletin, v. 1XX, no. XX/XX 21 Smith et al. BODY FOSSILS Anabaratid s Protoconodonts Cap-shaped fossils Salanacus Hyolithelminthes Coeloscleritophoran s Tommotiids Orthothecimorphs Molluscs Calcareous brachiopods Archaeocyaths ICHNOFOSSILS Simple bed planar traces Helminthoida c. miocenica Spatangopsis sp. Rusophycus Treptichnus pedum Oldhamia Zoophycus Plagiogmu s

Salaagol/Khairkhan fm s ~526 Mya 6p Tommotian

BG6 ?

5p BG5 Nemakit-Daldynian

P-C Boundary, BG4 as de ned at GSSP 4p BG3

3p BG2 Transition

s 2p

~541 Mya Precambrian-Cambrian e Z-A Fm 8323 (Orolgiin Gorge) D,E (Salaa Gorge) D,E (Salaa Gorge) D,E (Salaa Gorge) D,E (Salaa Gorge) 8326 (Orolgiin Gorge) D,E (Salaa Gorge) D,E (Salaa Gorge) L,M (Salaa Gorge) P,O (Salaa Gorge) Orolgiin + Salaa gorge Unit 18, S. BG Unit 18, S. BG Khunkher Gorg Unit 20, S. BG Taishir Taishir Taishir S. BG –5 0 5 Sh. δ13C (‰ VPDB)

Figure 12. Composite small shelly fossil and ichnofossil biostratigraphic range chart with revised δ13C chemostratigraphic age model. First appearance datums (FADs) in the fossil record are from previous workers (Korobov and Missarzhevsky, 1977; Voronin et al., 1982; Endonzhamts and Lkhasuren, 1988; Brasier et al., 1996b; Esakova and Zhegallo, 1996; Khomentovsky and Gibsher, 1996). Marker beds and section locations are listed below each circled FAD. Bold red line marks a depositional hiatus. Generic and higher group assignments used here are consistent with those from Maloof et al. (2010a, and references therein). VPDB—Vienna Peedee belemnite; GSSP—global stratotype section and point.

22 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

Many of the small shelly fossil FAD horizons 40 Siberia in the Zavkhan terrane (tommotiids, orthotheci morphs, hyolithelminths, cap-shaped fossils, 35 China protoconodonts, Salanacus) are from beds “D” a and “E” in the Salaa Gorge (Voronin et al., 1982; 30 Esakova and Zhegallo, 1996). With the revised Mongolia 25 correlations and age model presented here, we place these between positive excursions 2p and 20 3p (Figs. 7 and 12) instead of between 1p and 2p (Maloof et al., 2010a). Other important fossils 15 for calibrating biostratigraphic ranges globally Number of new gener 10 are the molluscs Watsonella and Aldanella (Parkhaev, 2007; Steiner et al., 2007; Rozanov 5 et al., 2008; Li et al., 2011; Parkhaev et al., 2011); however, more recently, the FADs of both taxa 0 have been shown to be diachronous, both locally and globally (Landing et al., 2013). Watsonella 544–542 542–540 540–538 538–536 536–534 534–532 532–530 530–528 528–526 526–524 524–522 522–520 has only been documented in one small shelly 2 million year age bins (ages in millions of years) fossil horizon in the upper Bayangol Formation Figure 13. Number of new genera appearing in 2 m.y. time bins with revised age at Salaa Gorge, whereas Aldanella has not been model from Mongolia. Data from China and Siberia are modified from Maloof found in the Zavkhan terrane. et al. (2010a). Global Origination Rates during the Earliest Cambrian entirely fine- to medium-grained sandstone— pilation, most of the Dahai Member is missing With the revised mapping, stratigraphy, and not the appropriate facies to preserve phospha- due to nondeposition or later erosion. As small age model, we have adjusted the FADs of small tized small shelly fossils. This lack of appropri- shelly fossil FADs from new sections in south shelly fossils in Mongolia. Previously, early ate facies is most likely due to the fact that these China continue to be added to this compilation, Cambrian origination had been assessed with units were being deposited during the closure of and as regional and global correlations continue global FADs of small shelly fossils (Maloof the foreland basin rather than due to the closure to be refined, the pulsed pattern of global FADs et al., 2010a); however, because these com- of the phosphatization window at the end of the will most certainly change. pilations have been limited to three localities, early Cambrian (Porter, 2004). Finally, in many of the fossiliferous Siberian namely, Siberia, China, and Mongolia, updating The second pulse of fossil FAD is due almost sections, there is a sub-Tommotian unconformity the age model or occurrence data for any one entirely to a huge number of FADs from China. (Khomentovsky and Karlova, 1993; Knoll et al., region causes adjustments to apparent global These fossils come from the phosphatic lime- 1995b; Zhu et al., 2001). We suggest that the origination rates. Figure 13 combines the new stone of the Zhongyicun Member. The apparent apparent pulses of fossil first appearances are data from Mongolia with data from China and large spike in FADs in the Zhongyicun Member the result of intervals of nondeposition in the Siberia to show the number of small shelly fos- is due in part to three major disconformities in sections included in this compilation and do not sil genera that first appear globally in each time the strata: one at the Ediacaran-Cambrian bound- represent global evolutionary patterns; FADs bin of 2 m.y. ary, one in the lower Dahai Member, and one at will not be found during periods in which sedi- The global compilation of fossil occurrence the top contact of the Dahai Member (Steiner ment is not deposited. Charles Darwin (1859) data still shows pulses of the first appearances et al., 2007). An erosional surface at the top of suggested that the apparently rapid appearance of new genera during the earliest Cambrian. the Dengying Formation has been observed in of fossils found in Cambrian strata was a prod- One possibility is that these peaks of fossil the Meishucun section (He et al., 1988; Qian uct of the incompleteness in the stratigraphic first appearances reflect peaks of diversifica- et al., 1996). The result is that, particularly in record—at a smaller scale, this indeed may be tion. Another possibility is that these fossil the southern sections of Yunnan Province, the the case. first appearance peaks instead reflect regional Daibu Member and portions of the Zhongyicun or global sedimentation patterns and/or tapho- Member are missing (Steiner et al., 2007). In the Implications for Early Cambrian nomic biases. The regional distribution of FADs overlying Nemakit-Daldynian to Tommotian Carbon Isotope Variability for each peak as shown in Figure 13 demon- Upper Zhongyicun and Lower Dahai members, strates that each of the three peaks is dominated there are also several disconformities, erosional The large-amplitude, high-frequency excur- by one region. This pattern suggests that peri- surfaces, and phosphatic hardgrounds, includ- sions of inorganic d13C values that accompany ods of deposition and the facies dependence of ing a sub-Tommotian one in the sequence (Qian the Cambrian radiation have long perplexed phosphatized fossils are the primary controllers and Bengtson, 1989; Steiner et al., 2007; Yang geologists. Brasier and Lindsay (2001) sug- of the FAD peaks in Figure 13. For example, in et al., 2014), which help to explain the dearth of gested that the d13C excursions are modulated Mongolia, nearly all of the small shelly fossils FADs during the late Nemakit-Daldynian. There by sea-level change, which, in effect, changes are in the lower to middle Bayangol Forma- is another major disconformity at the boundary rates of organic carbon burial. They suggested tion, which contains phosphatic limestone beds, between the Dahai Member and the overlying that the positive excursions are the result of whereas small shelly fossils are rare in the upper Shiyantou Formation. In the Meishucun sec- marine transgressions and the increase in the Bayangol Formation, which consists almost tion, one of the key fossil localities in this com- burial of organic matter, and the negative excur-

Geological Society of America Bulletin, v. 1XX, no. XX/XX 23 Smith et al. sions are the result of regressions and reduced mate the lag time between a maximum flooding orthothecimorphs, hyolithelminths, cap-shaped rates of organic matter burial. Kirschvink and surface and a positive excursion between the fossils, protoconodonts, and Salanacus are Raub (2003) offered an alternative methane TST and HST as being between 200,000 and moved up to just below the positive d13C peak fuse hypothesis in which the d13C excursions 500,000 yr. These rates are highly dependent on 3p. We demonstrate that with the new data from are attributed to the periodic release of large the choice of calibration points for the positive Mongolia, there are still three pulses of fossil volumes of methane gas, which they attributed excursions, and they will continue to be refined first appearances. We suggest that the pulsed to seismic activity, sediment failure, or sedi- as more early Cambrian ash horizons are dated nature of these first appearances is largely the ment erosion. Similarly, Maloof et al. (2005) globally. If there were high-latitude glaciers dur- result of surges in regional sedimentation and invoked Paleocene-Eocene thermal maximum– ing the late Ediacaran and early Cambrian, the limited phosphatization windows in the sections like rapid (<100,000 yr) releases of methane or large-scale sequences could be consistent with in Mongolia, China, and Siberia rather than organic carbon to explain the high-frequency Milankovitch cycles (Baode et al., 1986; Chu- global evolutionary tempos. Rates of global negative excursions in the earliest Cambrian. makov and Elston, 1989; Chumakov, 2011a; origination and diversification during this evolu- Squire et al. (2006) elaborated on this hypoth- Chumakov, 2011b; Etemad-Saeed et al., 2015; tionarily critical transition remains an important esis by suggesting that the periodic release of Hebert et al., 2010; Meert and Van der Voo, problem for paleontologists and geologists, and methane trapped in sedimentary rocks was the 1994; Rice et al., 2011; Stodt et al., 2011; Tors- future global correlations should include confi- consequence of rapid erosion of the Transgond- vik et al., 1996; Zhu and Wang, 2011). Nonethe- dence intervals that take into account regional wanan Supermountain. Maloof et al. (2010a) less, the repeating nature and scale of the oscil- and local sampling biases. modified their earlier hypothesis for the drivers lation, together with evidence of changing redox With integrated sequence stratigraphy and of geochemical change by stating that the rapid- conditions (Schröder and Grotzinger, 2007; high-resolution d13C chemostratigraphy, we ity and magnitude of the early Cambrian d13C Wille et al., 2008), are suggestive of a forcing show that the large d13C excursions and major excursions would require volcanic outgassing intrinsic to the carbon cycle, rather than external sequence boundaries occur at roughly the same up to an order of magnitude larger than mod- forcings such as volcanic or methane release. frequency in the Zavkhan terrane, providing ern fluxes and a 2.25 times increase in seafloor potential stratigraphic evidence that the high-fre- spreading during the early Cambrian. More CONCLUSION quency carbon isotope excursions in the earliest recently, Schrag et al. (2013) suggested that the Cambrian could have been modulated by global burial of authigenic carbonate was an important Wide ranges of causes have been presented sea-level change and corresponding weathering sink of carbon in anoxic oceans, and that the to explain the Cambrian radiation (review by and burial events. Using a refined age model, we advent of bioturbation or variations in surface Marshall, 2006); however, the strength of any of estimate that the lag time between maximum redox could have resulted in fluctuations in the these explanations is limited by a precise cali- flooding surfaces and positived 13C excursions size of this additional carbon sink. bration of rates of geochemical and biological is 200,000–500,000 yr. Combined, these data Coupled sequence stratigraphy and chemo change. Understanding the feedbacks between provide parameters and testable hypotheses that stratigraphy in Mongolia demonstrate that the the radiation of animals and associated geo- can be used to better constrain the relationships negative d13C excursions occur predominantly chemical and environmental changes requires between biological and environmental change in HSTs to LSTs, and the positive d13C excur- a precise age model and accurately calibrated during this major transition in life. sions occur in TSTs to HSTs. None of the fossil record. With new geological mapping, hypotheses described above provides direct sequence stratigraphy, and a chemostratigraphic ACKNOWLEDGMENTS explanations for a mechanistic link between age model, we place previous small shelly fossil We would like to thank A. Bayasgalan, B. Altant- eustatic sea-level change and the isotopic varia- and ichnofossil finds into an accurate geological setseg, S. Oyungerel, D. Oyun, and S. Lkhagva-Ochir tions. If a weathering process is invoked, the and temporal framework. for logistical assistance in Mongolia. We would like isotopic variability indicates differential weath- The new age model refines the age con- to thank our field assistants G. Sarantuya, J. Chuluun baatar, M. Delger, M. Jugder, A. Dorj, O. Dandar, ering of nutrients (i.e., phosphorous) that leads straints on the Bayangol Formation to deposi- U. Khuchitbaatar, A. Rooney, D. Bradley, S. Moon, to organic carbon burial versus silicate weather- tion entirely within the Nemakit-Daldynian. C. Dwyer, R. Anderson, and E. Smith for the tremen- ing that in turn leads to carbonate burial. Alter- The Nemakit-Daldynian–Tommotian boundary dous help in the field and laboratory. We thank J. Crev- natively, nutrient availability due to recycling is in the lower Salaagol Formation, coinciding eling and D. Jones for insightful comments and help with some of the initial field work. We thank S. Porter, on the continental shelf or slope might explain with the FAD of archaeocyaths in the Zavkhan M. Zhu, E. Sperling, and A. Maloof for detailed and carbon isotope changes linked to sea level. It terrane. The lack of trilobite fossils and Sr iso- thoughtful reviews. We thank C. Bucholz for useful is also possible that authigenic carbonate may tope chemostratigraphy suggest that the middle discussions. We thank S. Manley and G. Eischeid for have varied with sea-level changes due to dif- to upper Salaagol Formation is Tommotian in the use of and assistance in laboratories at Harvard ferential organic burial and anaerobic oxida- age rather than Atdabanian or Botomian, as was University. We thank National Aeronautics and Space Administration (NASA) Massachusetts Institute of tion in slope sediments (Schrag et al., 2013). previously suggested based on the identification Technology (MIT) Astrobiology node, NASA Geo- On the other hand, the sequence architecture in of archaeocyath species with poorly constrained biology grant NNH10ZDA001N-EXO, and National the Zavkhan terrane could have been driven by age ranges. Science Foundation Graduate Research Fellowship regional tectonic forces that have no bearing on The FAD of small shelly fossils in the Program to E. Smith for financial support. the global carbon cycle. However, large-scale Zavkhan terrane is moved stratigraphically hun- REFERENCES CITED sequence boundaries in the Zavkhan terrane dreds of meters up into the Bayangol Forma- occur at roughly the same frequency as position. With one possible exception (Knoll et al., Babcock, L.E., Robison, R.A., and Shanchi, P., 2011, Cam- brian stage and series nomenclature of Laurentia and 13 tive d C excursions. Using the same calibra- 1995a), the FAD of small shelly fossils and the the developing global chronostratigraphic scale: Mu- tion points for the positive excursions as Maloof FAD of T. pedum globally occur above the nadir seum of Northern Arizona Bulletin, v. 67, p. 12–26. Babcock, L.E., Peng, S., Zhu, M., Xiao, S., and Ahlberg, P., et al. (2010a) and assuming constant rates of of the negative carbon isotope excursion at the 2014, Proposed reassessment of the Cambrian GSSP: sedimentation between the excursions, we esti- base of the Cambrian. The FADs of tommotiids, Journal of African Earth Sciences, v. 98, p. 3–10.

24 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

Baode, G., Ruitang, W., Hambrey, M. and Wuchen, G., 1986, Corsetti, F.A., and Hagadorn, J.W., 2000, Precambrian- Gibson, T., Myrow, P., Macdonald, F., Minjin, C., and Glacial sediments and erosional pavements near the Cambrian transition: Death Valley, United States: Geol Gehrels, G., 2013, Depositional history, tectonics, and Cambrian—Precambrian boundary in western Henan ogy, v. 28, no. 4, p. 299–302. detrital zircon geochronology of Ordovician and Devo- Province, China: Journal of the Geological Society, Crimes, P.T., 1987, Trace fossils and correlation of late Pre- nian strata in southwestern Mongolia: Geological So- v. 143, no. 2, p. 311–323. cambrian and early Cambrian strata: Geological Maga- ciety of America Bulletin, v. 125, no. 5–6, p. 877–893. Bezzubtsev, V.V., 1963, On the Precambrian-Cambrian zine, v. 124, no. 2, p. 97–119. Goldring, R., and Jensen, S., 1996, Trace fossils and bio stratigraphy of the Dzabkhan River Basin, in Ma- Crimes, T.P., 1992, The record of trace fossils across the fabrics at the Precambrian-Cambrian boundary interval terials on the Geology of MPR: Gostopotekhizdat, Proterozoic-Cambrian boundary, in Lipps, J.H., and in western Mongolia: Geological Magazine, v. 133, p. 29–42. Signor, P.W., eds., Origin and Early Evolution of the no. 4, p. 403–415. Bold, U., Macdonald, F.A., Smith, E.F., Crowley, J.L., Metazoa: New York, Plenum Press, p. 177–202. Grotzinger, J.P., Bowring, S.A., Saylor, B.Z., and Kaufman, and Minjin, C., 2013, Elevating the Neoproterozoic Darwin, C.R., 1859, On the Origin of Species by Means A.J., 1995, Biostratigraphic and geochronologic con- Tsagaan-Olom Formation to a group: Mongolian Geo- of Natural Selection, or the Preservation of Favoured straints on early animal evolution: Science, v. 270, scientist, v. 39, no. 5, p. 89–94. Races in the Struggle for Life: London, Murray, 502 p. no. 5236, p. 598–604. Bowring, S.A., Grotzinger, J.P., Condon, D.J., Ramezani, Derry, L., Brasier, M., Corfield, R.A., Rozanov, A.Y., and He, T., Shen, L., and Yan, J., 1988, A new understanding on J., Newall, M.J., and Allen, P.A., 2007, Geochrono- Zhuravlev, A.Y., 1994, Sr and C isotopes in Lower Meishucun Sinian-Cambrian boundary section in Jin- logic constraints on the chronostratigraphic framework Cambrian carbonates from the Siberian craton: A ning: Yunnan Journal of Chengdu University of Tech- of the Neoproterozoic Huqf Supergroup, Sultanate of paleoenvironmental record during the ‘Cambrian ex- nology (Science & Technology Edition), v. 3, p. 5. Oman: American Journal of Science, v. 307, no. 10, plosion’: Earth and Planetary Science Letters, v. 128, Hebert, C.L., Kaufman, A.J., Penniston-Dorland, S.C., and p. 1097–1145. no. 3, p. 671–681. Martin, A.J., 2010, Radiometric and stratigraphic con- Brasier, M., and Lindsay, J.F., 2001, Did supercontinental Dorjnamjaa, D., and Bat-Ireedui, Y., 1991, The Precambrian straints on terminal Ediacaran (post-Gaskiers) glacia- amalgamation trigger the “Cambrian explosion,” in of Mongolia: Ulaanbaatar, Geological Institute of the tion and metazoan evolution: Precambrian Research, Zhuravlev, A.Yu., and Riding, R., eds., The Ecology of Mongolian Academy of Sciences. v. 182, no. 4, p. 402–412. the Cambrian Radiation: New York, Columbia Univer- Dorjnamjaa, D., Bat-Ireedui, Y.A., Dashdavaa, Z., and Hoffman, P., 1967, Algal stromatolites: Use in stratigraphic sity Press, p. 69–89. Soelmaa, D., 1993, Precambrian-Cambrian Geology of correlation and paleocurrent determination: Science, Brasier, M., Magaritz, M., Corfield, R., Huilin, L., Xiche, the Dzavkhan Zone: Oxford, UK, Earth Sciences De- v. 157, no. 3792, p. 1043–1045. W., Lin, O., Zhiwen, J., Hamdi, B., Tinggui, H., and partment, University of Oxford, 36 p. Jahn, B.-M., Litvinovsky, B., Zanvilevich, A., and Reichow, Fraser, A., 1990, The carbon-and oxygen-isotope rec Edel, J.B., Schulmann, K., Hanžl, P., and Lexa, O., 2014, M., 2009, Peralkaline granitoid magmatism in the ord of the Precambrian-Cambrian boundary interval in Palaeomagnetic and structural constraints on 90° anti- Mongolian–Transbaikalian belt: Evolution, petrogen- China and Iran and their correlation: Geological Maga- clockwise rotation in SW Mongolia during the Permo– esis and tectonic significance: Lithos, v. 113, no. 3, zine, v. 127, no. 4, p. 319–332. Triassic: Implications for Altaid oroclinal bending. p. 521–539. Brasier, M., Khomentovsky, V., and Corfield, R., 1993, Sta- Preliminary palaeomagnetic results: Journal of Asian Johnston, D., Macdonald, F., Gill, B., Hoffman, P., and ble isotopic calibration of the earliest skeletal fossil as- Earth Sciences, v. 94, p. 157–171. Schrag, D., 2012, Uncovering the Neoproterozoic car- semblages in eastern Siberia (Precambrian-Cambrian Endonzhamts, Zh., and Lkhasuren, B., 1988, Stratigrafi ya bon cycle: Nature, v. 483, no. 7389, p. 320–323. boundary): Terra Nova, v. 5, no. 3, p. 225–232. pogranichnykh tolshch dokembriya i kembriya Dzab Kaufman, A.J., Knoll, A.H., Semikhatov, M.A., Grotzinger, Brasier, M., Dorjnamjaa, D., and Lindsay, J., 1996a, The khanskoy zony, in Khomentovsky, V.V., and Shenfi J.P., Jacobsen, S.B., and Adams, W., 1996, Integrated Neoproterozoic to early Cambrian in southwest Mon- l’, V.Yu., eds., Pozdniy dokembriy i ranniy paleozoy chronostratigraphy of Proterozoic-Cambrian boundary golia: An introduction: Geological Magazine, v. 133, Sibiri , Rifey i vend: Novosibirsk, Institut Geologii i beds in the western Anabar region, northern Siberia: no. 4, p. 365–369. Geofi ziki, Sibirskoe Otdelenie, Akademiya Nauk: Geological Magazine, v. 133, no. 5, p. 509–533. Brasier, M., Shields, G., Kuleshov, V., and Zhegallo, E., SSSR, p. 150–162. Khomentovsky, V., and Gibsher, A., 1996, The Neoprotero- 1996b, Integrated chemo- and biostratigraphic calibra- Erwin, D.H., Laflamme, M., Tweedt, S.M., Sperling, E.A., zoic–Lower Cambrian in northern Govi-Altay, west- tion of early animal evolution: Neoproterozoic–early Pisani, D., and Peterson, K.J., 2011, The Cambrian ern Mongolia: Regional setting, lithostratigraphy and Cambrian of southwest Mongolia: Geological Maga- conundrum: Early divergence and later ecological suc- biostratigraphy: Geological Magazine, v. 133, no. 04, zine, v. 133, no. 4, p. 445–485. cess in the early history of animals: Science, v. 334, p. 371–390. Buatois, L.A., Almond, J., and Germs, G.J., 2013, Envi- no. 6059, p. 1091–1097. Khomentovsky, V., and Karlova, G., 1993, Biostratigraphy ronmental tolerance and range offset of Treptich Esakova, N.V., and Zhegallo, E.A., 1996, Stratigrafaya i of the Vendian-Cambrian beds and the Lower Cam- nus pedum: Implications for the recognition of the fauna nizhnego Kembriya Mongolii (Lower Cam- brian boundary in Siberia: Geological Magazine, Ediacaran-Cambrian boundary: Geology, v. 41, no. 4, brian stratigraphy and fauna of Mongolia): Sovmest- v. 130, no. 1, p. 29–45. p. 519–522. naya Sovetsko-Mongol’skaya Paleontologicheskaya Kirschvink, J., and Raub, T., 2003, A methane fuse for the Bucholz, C.E., Jagoutz, O., Schmidt, M.W., and Sambuu, Ekspeditsiya, v. 46, p. 208. Cambrian explosion: True polar wander: Comptes O., 2014, Phlogopite- and clinopyroxene-dominated Etemad-Saeed, N., Hosseini-Barzi, M., Adabi, M.H., Rendus Geoscience, v. 335, p. 71–83. fractional crystallization of an alkaline primitive melt: Miller, N.R., Sadeghi, A., Houshmandzadeh, A., Kirschvink, J.L., Magaritz, M., Ripperdan, R.L., Zhuravlev, Petrology and mineral chemistry of the Dariv igneous and Stockli, D.F., 2015, Evidence for ca. 560 Ma A.Y., and Rozanov, A.Y., 1991, The Precambrian- complex, western Mongolia: Contributions to Mineral- Ediacaran glaciation in the Kahar Formation, central Cambrian boundary: Magnetostratigraphy and carbon ogy and Petrology, v. 167, no. 4, p. 1–28. Alborz Mountains, northern Iran: Gondwana Research isotopes resolve correlation problems between Siberia, Calver, C., Crowley, J., Wingate, M., Evans, D., Raub, (in press). Morocco, and South China: GSA Today, v. 1, no. 4, T., and Schmitz, M., 2013, Globally synchronous Gehling, J.G., Jensen, S., Droser, M.L., Myrow, P.M., and p. 69–91. Marinoan deglaciation indicated by U-Pb geochronol- Narbonne, G.M., 2001, Burrowing below the basal Knoll, A.H., Grotzinger, J.P., Kaufman, A.J., and Kolosov, ogy of the Cottons Breccia, Tasmania, Australia: Geol- Cambrian GSSP, Fortune Head, Newfoundland: Geo- P., 1995a, Integrated approaches to terminal Protero- ogy, v. 41, no. 10, p. 1127–1130. logical Magazine, v. 138, no. 2, p. 213–218. zoic stratigraphy: An example from the Olenek Uplift, Calver, C.R., Black, L.P., Everard, J.L., and Seymour, D.B., Geyer, G., 2005, The Fish River Subgroup in Namibia: northeastern Siberia: Precambrian Research, v. 73, 2004, U-Pb zircon age constraints on late Neoprotero- Stratigraphy, depositional environments and the Pro- no. 1, p. 251–270. zoic glaciation in Tasmania: Geology, v. 32, no. 10, terozoic-Cambrian boundary problem revisited: Geo- Knoll, A.H., Kaufman, A.J., Semikhatov, M.A., Grotzinger, p. 893–896. logical Magazine, v. 142, no. 5, p. 465–498. J.P., and Adams, W., 1995b, Sizing up the sub-Tommo- Chumakov, N., and Elston, D., 1989, The paradox of Late Geyer, G., and Uchman, A., 1995, Ichnofossil assemblages tian unconformity in Siberia: Geology, v. 23, no. 12, Proterozoic glaciations at low latitudes: Episodes, v. 12, from the Nama Group (Neoproterozoic–Lower Cam- p. 1139–1143. no. 2, p. 115. brian) in Namibia and the Proterozoic-Cambrian Korobov, M., and Missarzhevsky, V., 1977, O pogranichnykh Chumakov, N.M., 2011a, Glacial deposits of the Baykonur boundary problem revisited: Beringeria, v. 2, Special sloyakh kembriya i dokembriya Zapadnoy Mongolii Formation, Kazakhstan and Kyrgyzstan: Geological Issue, p. 175–202. (khrebet Khasagt-Khairkhan): Bespozvonochnye paleo- Society, London, Memoirs, v. 36, no. 1, p. 303–307. Gibsher, A., and Khomentovsky, V., 1990, The section of zoya Mongolii: Sovmestnaya Sovestsko-Mongol’skaya Chumakov, N.M., 2011b, Glacial deposits of the Bokson the Tsagaan Oloom and Bayan Gol Formations of the Paleontologicheskaya Ekspeditsiya: Trudy, v. 5, p. 7–9. Group, East Sayan Mountains, Buryatian Republic, Vendian–Lower Cambrian in the Dzabkhan zone of Kouchinsky, A., Bengtson, S., Missarzhevsky, V.V., Russian Federation: Geological Society, London, Mongolia, in The Late Precambrian and Early Paleozoic Pelechaty, S., Torssander, P., and Val’kov, A.K., 2001, Memoirs, v. 36, no. 1, p. 285–288. of Siberia: Novosibirsk, Institut Geologii i Geofiziki, Carbon isotope stratigraphy and the problem of a pre– Compston, W., Williams, I.S., Kirschvink, J.L., Zhang Sibirskoe Otdelenie, Akademiya Nauk SSSR, p. 79–91. Tommotian Stage in Siberia: Geological Magazine, Zichao, and Ma Guogan, 1992, Zircon U-Pb ages from Gibsher, A.S., Bat-Ireedui, Y.A., Balakhonov, I.G., and Efre- v. 138, no. 4, p. 387–396. the Early Cambrian time-scale: Journal of the Geologi- menko, D.E., 1991, The Bayan Gol reference section Kouchinsky, A., Bengtson, S., Pavlov, V., Runnegar, B., cal Society of London, v. 149, p. 171–184, doi:10 .1144 of the Vendian–Lower Cambrian in central Mongolia, Torssander, P., Young, E., and Ziegler, K., 2007, Car- /gsjgs.149.2.0171. in Khomentovsky, V.V., ed., Late Precambrian and bon isotope stratigraphy of the Precambrian–Cambrian Condon, D., Zhu, M.Y., Bowring, S., Wang, W., Yang, A.H., Early Palaeozoic of Siberia. Siberian Platform and its Sukharikha River section, northwestern Siberian plat- and Jin, Y.G., 2005, U-Pb ages from the Neoprotero- Framework: Novosibirsk, Ob’edinennyy Institut Ge- form: Geological Magazine, v. 144, no. 4, p. 609–618. zoic Doushantuo Formation, China: Science, v. 308, ologii, Geofiziki i Mineralogii, Sibierskoe Otdelenie, Kröner, A., Lehmann, J., Schulmann, K., Demoux, A., Lexa, no. 5718, p. 95–98. Akademiya Nauk SSSR, p. 151. O., Tomurhuu, D., Štípská, P., Liu, D., and Wingate,

Geological Society of America Bulletin, v. 1XX, no. XX/XX 25 Smith et al.

M.T., 2010, Lithostratigraphic and geochronological Markova, N., Korobov, M., and Zhuravleva, Z., 1972, redox-sensitive trace elements and rare earth elements constraints on the evolution of the Central Asian oro- K voprosu o vend-kembriyskikh otlozheniyakh in Oman: Journal of the Geological Society of London, genic belt in SW Mongolia: Early Paleozoic rifting fol- Yugo-Zapadnoy Mongolii (On the problem of the v. 164, no. 1, p. 175–187. lowed by late Paleozoic accretion: American Journal of Vendian-Cambrian deposits of southwestern Mongo- Shields, G.A., Brasier, M.D., Stille, P., and Dorjnamjaa, Science, v. 310, no. 7, p. 523–574. lia): Bulleten’Moskovskogo Obshchestva Ispytateley D.-I., 2002, Factors contributing to high d13C val- Kruse, P.D., Gandin, A., Debrenne, F., and Wood, R., 1996, Prirody, Otdel Geologicheskiy, v. 47, no. 1, p. 57–70. ues in Cryogenian limestones of western Mongolia: Early Cambrian bioconstructions in the Zavkhan Basin Marshall, C.R., 2006, Explaining the Cambrian “explosion” Earth and Planetary Science Letters, v. 196, no. 3, of western Mongolia: Geological Magazine, v. 133, of animals: Annual Review of Earth and Planetary Sci- p. 99–111. no. 4, p. 429–444. ences, v. 34, p. 355–384. Shields-Zhou, G., Hill, A., and Macgabhann, B., 2012, Landing, E., 1988, Lower Cambrian of eastern Massachu- Meert, J.G., and Van der Voo, R., 1994, The Neoproterozoic The Cryogenian Period, in Gradstein, J.G., Ogg, setts: Stratigraphy and small shelly fossils: Journal of (1000–540 Ma) glacial intervals: No more snowball M., Schmitz, and Ogg, G., eds., The Geologic Time Paleontology, v. 62, p. 661–695. Earth?: Earth and Planetary Science Letters, v. 123, Scale: Cambridge, UK, Cambridge University Press, Landing, E., 1994, Precambrian-Cambrian boundary global no. 1, p. 1–13. p. 393–411. stratotype ratified and a new perspective of Cambrian Mullins, H.T., Gardulski, A.F., Wise, S.W., and Applegate, Spence, G.H., and Tucker, M.E., 1997, Genesis of limestone time: Geology, v. 22, no. 2, p. 179–182. J., 1987, Middle Miocene oceanographic event in megabreccias and their significance in carbonate se- Landing, E., Geyer, G., Brasier, M.D., and Bowring, S.A., the eastern Gulf of Mexico: Implications for seismic quence stratigraphic models: A review: Sedimentary 2013, Cambrian evolutionary radiation: Context, cor- stratigraphic succession and Loop Current/Gulf Stream Geology, v. 112, no. 3, p. 163–193. relation, and chronostratigraphy—Overcoming defi- circulation: Geological Society of America Bulletin, Squire, R.J., Campbell, I.H., Allen, C.M., and Wilson, C.J., ciencies of the first appearance datum (FAD) concept: v. 98, no. 6, p. 702–713. 2006, Did the Transgondwanan Supermountain trigger Earth-Science Reviews, v. 123, p. 133–172. Narbonne, G.M., Kaufman, A.J., and Knoll, A.H., 1994, In- the explosive radiation of animals on Earth?: Earth and Lehmann, J., Schulmann, K., Lexa, O., Corsini, M., Kroner, tegrated chemostratigraphy and biostratigraphy of the Planetary Science Letters, v. 250, no. 1, p. 116–133. A., Stipska, P., Tomurhuu, D., and Otgonbator, D., Windermere Supergroup, northwestern Canada: Impli- Steiner, M., Li, G., Qian, Y., Zhu, M., and Erdtmann, B.-D., 2010, Structural constraints on the evolution of the cations for Neoproterozoic correlations and the early 2007, Neoproterozoic to early Cambrian small shelly Central Asian orogenic belt in SW Mongolia: Ameri- evolution of animals: Geological Society of America fossil assemblages and a revised biostratigraphic cor- can Journal of Science, v. 310, no. 7, p. 575–628. Bulletin, v. 106, no. 10, p. 1281–1292. relation of the Yangtze Platform (China): Palaeogeog- Levashova, N.M., Kalugin, V.M., Gibsher, A.S., Yff, J., Parkhaev, P.Y., 2007, Shell chirality in Cambrian gastro- raphy, Palaeoclimatology, Palaeoecology, v. 254, no. 1, Ryabinin, A.B., Meert, J., and Malone, S.J., 2010, pods and sinistral members of the genus Aldanella p. 67–99. The origin of the Baydaric microcontinent, Mongolia: Vostokova, 1962: Paleontological Journal, v. 41, no. 3, Stodt, F., Rice, A.H.N., Björklund, L., Bax, G., Halverson, Constraints from paleomagnetism and geochronology: p. 233–240. G.P., and Pharaoh, T., 2011, Evidence of late Neo Tectonophysics, v. 485, no. 1–4, p. 306–320. Parkhaev, P.Y., Karlova, G., and Rozanov, A.Y., 2011, Tax- proterozoic glaciation in the Caledonides of NW Scan- Li, D., Ling, H.-F., Jiang, S.-Y., Pan, J.-Y., Chen, Y.-Q., Cai, onomy, stratigraphy and biogeography of Aldanella dinavia: Geological Society, London, Memoirs, v. 36, Y.-F., and Feng, H.-Z., 2009, New carbon isotope stra- attleborensis—A possible candidate for defining the no. 1, p. 603–612. tigraphy of the Ediacaran-Cambrian boundary interval base of Cambrian Stage 2: Museum of Northern Ari- Torsvik, T., Smethurst, M., Meert, J.G., Van der Voo, R., from SW China: Implications for global correlation: zona Bulletin, v. 67, p. 298–300. McKerrow, W., Brasier, M., Sturt, B., and Walder- Geological Magazine, v. 146, no. 4, p. 465–484. Peng, S., Babcock, L., and Cooper, R., 2012, The Cambrian haug, H., 1996, Continental break-up and collision in Li, G., Zhao, X., Gubanov, A., Zhu, M., and Na, L., 2011, Period, in Gradstein, J.G., Ogg, M., Schmitz, and Ogg, the Neoproterozoic and Palaeozoic—A tale of Baltica Early Cambrian mollusc Watsonella crosbyi: A poten- G., eds., The Geologic Time Scale, Volume 2: Cam- and Laurentia: Earth-Science Reviews, v. 40, no. 3, tial GSSP index fossil for the base of the Cambrian bridge, UK, Cambridge University Press, p. 437–488. p. 229–258. Stage 2: Acta Geologica Sinica, v. 85, no. 2, p. 309– Porter, S.M., 2004, Closing the phosphatization window: Vail, P.R., Mitchum Jr, R., and Thompson, S., III, 1977, 319 [English edition]. Testing for the influence of taphonomic megabias on Seismic stratigraphy and global changes of sea level: Lindsay, J., Brasier, M., Dorjnamjaa, D., Goldring, R., the pattern of small shelly fossil decline: Palaios, v. 19, Part 4. Global cycles of relative changes of sea level. Kruse, P., and Wood, R., 1996a, Facies and sequence no. 2, p. 178–183. Section 2. Application of seismic reflection configu- controls on the appearance of the Cambrian biota in Porter, S.M., 2007, Seawater chemistry and early carbonate ration to stratigraphic interpretation, in Payton, C.E., southwestern Mongolia: Implications for the Pre- biomineralization: Science, v. 316, no. 5829, p. 1302. Seismic Stratigraphy—Applications to Hydrocarbon cambrian-Cambrian boundary: Geological Magazine, Qian, Y., and Bengtson, S., 1989, Palaeontology and biostra- Exploration: American Association of Petroleum Ge- v. 133, no. 4, p. 417–428. tigraphy of the Early Cambrian Meishucunian Stage in ologists Memoir 26, p. 83–97. Lindsay, J., Brasier, M., Shields, G., Khomentovsky, V., Yunnan Province, South China: Oslo, Norway, Univer- Valentine, J.W., 2002, Prelude to the Cambrian explosion: and Bat-Ireedui, Y., 1996b, Glacial facies associations sitetsforlaget, 156 p. Annual Review of Earth and Planetary Sciences, v. 30, in a Neoproterozoic back-arc setting, Zavkhan Basin, Qian, Y., Zhu, M., He, T., and Jiang, Z., 1996, New inves- no. 1, p. 285–306. western Mongolia: Geological Magazine, v. 133, no. 4, tigation of Precambrian-Cambrian boundary sections Van Wagoner, J.C., Mitchum, R., Campion, K., and Rah- p. 391–402. in eastern Yunnan: Acta Micropalaontologica Sinica, manian, V., 1990, Siliciclastic Sequence Stratigraphy Macdonald, F.A., 2011, The Tsagaan Oloom Formation, v. 13, p. 225–240. in Well Logs, Cores, and Outcrops: Concepts for High- southwestern Mongolia, in Arnaud, E., Halverson, Repina, L., and Rozanov, A., 1992, The Cambrian of Sibe- Resolution Correlation of Time and Facies: American G.P., and Shields-Zhou, G., eds., The Geological Rec ria: Novosibirsk: Trudy Instituta Geologii Geofi ziki, Association of Petroleum Geologists Methods 7, 55 p. ord of Neoproterozoic Glaciations: Geological Society v. 788, p. 135. Voronin, Y.I., Voronova, L.G., Girgor’eva, N.V., Drozdova, of London Memoir 36, p. 331–337. Rice, A.H.N., Edwards, M.B., Hansen, T.A., Arnaud, E., N.A., Zhegallo, E.A., Zhuravlev, A.Y., Ragozina, Macdonald, F.A., Jones, D.S., and Schrag, D.P., 2009, and Halverson, G.P., 2011, Glaciogenic rocks of A.L., Rozanov, A.Y., Sayutina, T.A., Sysoev, V.A., and Stratigraphic and tectonic implications of a new gla- the Neoproterozoic Smalfjord and Mortensnes for- Fonin, V.D., 1982, The Precambrian/Cambrian bound- cial diamictite–cap carbonate couplet in southwestern mations, Vestertana Group, E. Finnmark, Norway: ary in the geosynclinal areas (the reference section of Mongolia: Geology, v. 37, p. 123–126. Geological Society, London, Memoirs, v. 36, no. 1, Salaany-Gol, MPR): Moscow, Akademia Nauk SSSR, Macdonald, F.A., Schmitz, M.D., Crowley, J.L., Roots, C.F., p. 593–602. Trudy Sovmestnoy Sovetsoko-Mongol’skoy Paleonto- Jones, D.S., Maloof, A.C., Strauss, J.V., Cohen, P.A., Rooney, A.D., Macdonald, F.A., Strauss, J.V., Dudás, F.Ö., logicheskoy Ekspeditsii, 150 p. Johnston, D.T., and Schrag, D.P., 2010, Calibrating the Hallmann, C., and Selby, D., 2014, Re-Os geochro- Wilhem, C., Windley, B.F., and Stampfli, G.M., 2012, The Cryogenian: Science, v. 327, no. 5970, p. 1241–1243. nology and coupled Os-Sr isotope constraints on the Altaids of Central Asia: A tectonic and evolutionary in- MacNaughton, R.B., and Narbonne, G.M., 1999, Evolution Sturtian snowball Earth: Proceedings of the National novative review: Earth-Science Reviews, v. 113, no. 3, and ecology of Neoproterozoic–Lower Cambrian trace Academy of Sciences of the United States of America, p. 303–341. fossils, NW Canada: Palaios, v. 14 p. 97–115. v. 111, no. 1, p. 51–56. Wille, M., Nägler, T.F., Lehmann, B., Schröder, S., and Maloof, A.C., Schrag, D.P., Crowley, J.L., and Bowring, Rozanov, A.Y., Khomentovsky, V., Shabanov, Y.Y., Karlova, Kramers, J.D., 2008, Hydrogen sulphide release to S.A., 2005, An expanded record of early Cambrian G., Varlamov, A., Luchinina, V., Demidenko, Y.E., surface waters at the Precambrian/Cambrian boundary: carbon cycling from the Anti-Atlas Margin, Morocco: Parkhaev, P.Y., Korovnikov, I., and Skorlotova, N., Nature, v. 453, no. 7196, p. 767–769. Canadian Journal of Earth Sciences, v. 42, no. 12, 2008, To the problem of stage subdivision of the Lower Wood, R., Zhuravlev, A.Y., and Anaaz, C.T., 1993, The ecol- p. 2195–2216. Cambrian: Stratigraphy and Geological Correlation, ogy of Lower Cambrian buildups from Zuune Arts, Maloof, A.C., Porter, S.M., Moore, J.L., Dudás, F.Ö., Bow- v. 16, no. 1, p. 1–19. Mongolia: Implications for early metazoan reef evolu- ring, S.A., Higgins, J.A., Fike, D.A., and Eddy, M.P., Ruzhentsev, S.V., and Burashnikov, V.V., 1996, Tectonics of tion: Sedimentology, v. 40, no. 5, p. 829–858. 2010a, The earliest Cambrian record of animals and the western Mongolian Salairides: Geotectonics, v. 29, Yang, B., Steiner, M., Li, G., and Keupp, H., 2014, Ter- ocean geochemical change: Geological Society of no. 5, p. 379–394. reneuvian small shelly faunas of East Yunnan (South America Bulletin, v. 122, no. 11–12, p. 1731–1774. Schrag, D.P., Higgins, J.A., Macdonald, F.A., and Johnston, China) and their biostratigraphic implications: Palaeo- Maloof, A.C., Ramezani, J., Bowring, S.A., Fike, D.A., D.T., 2013, Authigenic carbonate and the history of geography, Palaeoclimatology, Palaeoecology, v. 398, Porter, S.M., and Mazouad, M., 2010b, Constraints on the global carbon cycle: Science, v. 339, no. 6119, p. 28–58. early Cambrian carbon cycling from the duration of the p. 540–543. Yarmolyuk, V., Kovalenko, V., Anisimova, I., Sal’nikova, Nemakit-Daldynian–Tommotian boundary d13C shift, Schröder, S., and Grotzinger, J., 2007, Evidence for anoxia E., Kovach, V., Kozakov, I., Kozlovsky, A., Kudrya Morocco: Geology, v. 38, no. 7, p. 623–626. at the Ediacaran-Cambrian boundary: The record of shova, E., Kotov, A., and Plotkina, Y.V., Late Riphean

26 Geological Society of America Bulletin, v. 1XX, no. XX/XX Integrated late Ediacaran to early Cambrian records from southwestern Mongolia

alkali granites of the Zabhan microcontinent: Evidence Zhu, M., and Wang, H., 2011, Neoproterozoic glaciogenic Zhu, M.-Y., Babcock, L.E., and Peng, S.-C., 2006, Advances for the timing of Rodinia breakup and formation of diamictites of the Tarim Block, NW China: Geological in Cambrian stratigraphy and paleontology: Integrat- microcontinents in the Central Asian fold belt, in Pro- Society, London, Memoirs, v. 36, no. 1, p. 367–378. ing correlation techniques, paleobiology, taphonomy ceedings Doklady Earth Sciences 2008, Volume 420: Zhu, M., Li, G., Zhang, J., and Michael, S., 2001, Early and paleoenvironmental reconstruction: Palaeoworld, Springer, p. 583–588. Cambrian stratigraphy of east Yunnan, southwestern v. 15, no. 3, p. 217–222. Zhegallo, L., and Zhuravlev, A.Y., 1991, Guidebook for China: A synthesis, in Zhu, M., Van Iten, H., Peng, S., the International Excursion to the Vendian-Cambrian and Li, G., eds., The Cambrian of South China: Acta Science Editor: Christian Koeberl Deposits of the Dzvkhan Zone of Mongolia: Moscow, Palaeontologica Sinica, v. 40, p. 4–39. Associate Editor: Kenneth G. MacLeod Palaeontological Institute, 24 p. Zhu, M., Strauss, H., and Shields, G.A., 2007, From snow- Zhou, C., Zhang, J., Li, G., and Yu, Z., 1997, Carbon and ball Earth to the Cambrian bioradiation: Calibration Manuscript Received 8 December 2014 Revised Manuscript Received 4 May 2015 oxygen isotopic record of the early Cambrian from the of Ediacaran–Cambrian Earth history in South China: Manuscript Accepted 8 July 2015 Xiaotan Section, Yunnan, South China: Scientia Geo- Palaeogeography, Palaeoclimatology, Palaeoecology, logica Sinica, v. 32, no. 2, p. 201–211. v. 254, no. 1, p. 1–6. Printed in the USA

Geological Society of America Bulletin, v. 1XX, no. XX/XX 27