Paleomagnetic and palynologic analyses of Albian to Santonian strata at Bayn Shireh, Burkhant, and Khuren Dukh, eastern Gobi Desert, Mongolia

Cretaceous Research (1999) 20, 829–850 Article No. cres.1999.0188, available online at http://www.idealibrary.com on Paleomagnetic and palynologic analyses of Albian to Santonian strata at Bayn Shireh, Burkhant, and Khuren Dukh, eastern Gobi Desert, Mongolia *J. F. Hicks, †D. L. Brinkman, ‡D. J. Nichols and §M. Watabe *Department of Earth and Space Sciences, Denver Museum of Natural History, 2001 Colorado Boulevard, Denver, CO 80205-5798, USA †Department of Geology and Geophysics, Yale University, New Haven, CT 06520-8109, USA ‡US Geological Survey, MS 939, Box 25046, Denver, CO 80225-0046, USA §Hayashibara Museum of Natural Sciences, Shimoishii 1-2-3, Okayama 700, Japan Revised manuscript accepted 23 September 1999 Cretaceous terrestrial sediments deposited in a series of intracratonic basins across the Gobi Desert region of southern Mongolia and northern China contain a unique and diverse vertebrate fauna. In 1996 an expedition jointly sponsored by the Mongolian Paleontological Center and the Hayashibara Museum of Natural Sciences revisited a number of famous vertebrate fossil localities in the eastern Gobi region of Mongolia and, as part of a broad geological and paleontological study, collected a series of paleomagnetic samples from measured sections at Bayn Shireh, Burkhant and Khuren Dukh, as well as from an unmeasured locality adjacent to Khuren Dukh. Expedition members also collected palynologic samples from Khuren Dukh and the adjacent locality. Paleomagnetic analysis shows that all the sites from which samples were collected display detrital remnant magnetization that is consistently normal in polarity. The measured Cretaceous magnetic directions are oriented to the east or northeast of the present day expected direction (declination 356.2, inclination 65.2), and they are wholly concordant with that expected for a mid-latitude Northern Hemisphere sampling locality, and with the directions for this period reported by other workers. These results, when considered in tandem with the known biostratigraphy, strongly suggest that the sedimentary deposits at all four localities in the eastern Gobi correlate to the normal polarity chron 34 (the Cretaceous Long Normal), which ranges in age from approximately 121 to 83.5 million years. Previous vertebrate, invertebrate and palynological data from Khuren Dukh suggest that the lower and middle parts of the stratigraphic interval exposed there (which have been assigned to the Shinekhudag Formation) are ‘Khukhtekian’ in age and correspond to the Aptian–Albian interval that can be broadly correlated to the older, Early Cretaceous part of the Cretaceous Long Normal, C34n. New palynologic data presented here indicate that these strata are no older than middle to late Albian. The rocks at Bayn Shireh (the Bayn Shireh Formation) have been assigned a ‘Baynshirenian’ biostratigraphic age that may range from Cenomanian to early Campanian. The magnetostratigraphy results presented here indicate that the strata at both the Bayn Shireh and Burkhant localities do not cross the Santonian/Campanian Stage boundary, however, as this is believed to lie at, or very near, the C34n/C33r reversal boundary. Thus, the Bayn Shireh Formation was most likely deposited near the end of the Cretaceous Long Normal Interval, no later than the latest Santonian.  1999 Academic Press K W: Early Cretaceous; Late Cretaceous; Mongolia; Gobi Desert; paleomagnetism; magnetostratigraphy; palynostratigraphy; Cretaceous Long Normal; C34n; Bayn Shireh1; Dzun Bayan2; Khuren Dukh3; Shinekhudag. 1–3 These names are transliterated into English from the original Mongolian in a variety of different ways. 1There is no fundamentally correct spelling: ‘‘Bayn Shireh’’ by Gradzinski et al. (1968, ‘‘Glossary of Mongolian Place-Names with English Equivalents’’); ‘‘Bayn Shire’’ by Jerzykiewicz & Russell (1991); ‘‘Bayn Shiren’’ by Jerzykiewicz et al. (1993; e.g., p. 2193); ‘‘Bayan Shire’’ by Samoilov & Benjamini (1996); ‘‘Bainshire’’ by Sochava (1975); 2‘‘Dzun Bayan’’ and ‘‘Dzunbayan’’ (Minzhin, 1994); ‘‘Dzunbayn’’ (Shuvalov, 1974). 3Generally consistent translation in the literature, after Barsbold et al. (1971), but alternative spellings have been noted: ‘‘Khoren Dukh’’, ‘‘Khyren Dukh’’, ‘‘Huren Duh’’ reported by Jerzykiewicz & Russell (1991, p. 364); ‘‘Huren Dukh’’ by Narmandakh (1997). 0195–6671/99/020829+22 $30.00/0  1999 Academic Press 830 J. F. Hicks et al. 1. Introduction tuffs and volcaniclastics occurred. Many of the Creta- ceous formations of the Gobi region are essentially Many of the Mesozoic vertebrate localities known undisturbed, and few are very steeply tilted, except from Asia are concentrated in a relatively few areas in close to the margins of faulted mountain ranges. the Gobi Desert region of southern Mongolia and There are large unconformities at critical levels adjacent northern China. In comparison to strata of (e.g., at the Cretaceous/Tertiary boundary) as much similar age in North America and Europe, the paleon- of the sedimentation appears to have been controlled tological importance of the Gobi Desert and sur- by episodic block faulting, with significant hiatuses rounding areas was discovered late, only within the of non-accumulation and gentle erosion separating last 75 years, and until the sweeping political changes periods of relatively rapid fluviolacustrine and/or of the last ten years, circumstances ensured that it aeolian deposition. Fossils are locally abundant (for an remained closed to all but a handful of western overview of localities, see Sochava, 1975), but al- scientists. Cretaceous and Paleocene rocks are widely though some of the invertebrates can be correlated to distributed across this region, but good exposures bordering marine intervals that are datable, almost all suitable for geological research are often quite limited of the vertebrates from the Gobi are endemic, with the and access to them is hindered by the huge distances genera usually confined to Asia, though there are a few to be covered, lack of roads and difficult logistics. notable exceptions among dinosaurs and mammals. Comparisons are often made between the two major The Cretaceous sequences in the Gobi Desert make dinosaur provinces of the Cretaceous, the Western up a series of widely separated, poorly dated, and Interior of North America and the Gobi of Asia relatively unstudied sedimentary sections that are (Jerzykiewicz, 1996). Although each province con- bounded by large unconformities and have been only tains a unique and important paleontological record of broadly correlated to the geologic time scale using the Cretaceous period, the two provinces were fossil plant, invertebrate and very rare vertebrate as- deposited under very different tectonic and physio- sociations. The Cretaceous sequences in the Gobi graphic regimes. The Western Interior Basin is a show similarities in regional climate change that may geographically broad, readily accessible, well-studied be correlative to the North American sequence. The and well-dated foreland basin sequence of consider- remoteness and harsh climate of the region has meant able economic and paleontologic importance; a sedi- that the area has mainly been accessible only to large, mentary sequence that is thousands of meters thick in well-equipped, and highly organized expeditions able the deepest, western part of the asymmetrical basin, to access and stay at the widely-separated localities and one that can be correlated at a number of strati- for any appreciable length of time (for example, graphic levels, both chronologically and biostrati- see narratives of the Polish–Mongolian expeditions; graphically, to the type sections of the European stage Kielan-Jaworowska & Dovchin, 1968, 1972). boundaries. The nature of the exposures and distribution of the In contrast, the Gobi region of southern and eastern fossiliferous facies have also meant that most of the Mongolia is remote and relatively inaccessible, and in research carried out to date has focused on the exten- comparison to the Western Interior, less well-studied. sive badlands located within three subordinate basins The last major marine incursion into Central Asia was in the western Gobi: Nemegt, Shiregin Gashun and during the Permian. Thus, nearly all of the sediments Ulan Nur (Figure 1); areas where Upper Cretaceous that accumulated after that time are entirely terrestrial sediments have yielded spectacular fossils. In contrast, and were deposited in a series of broad and shallow the exposures in the eastern Gobi, which are primarily intracratonic basins that contain only one or two centred around Sayn Shand (the capital of the Eastern thousand meters of sediment, of which only a small Gobi Aymag, Figure 1) are generally older and have fraction is exposed at the surface. These basins are been the focus of considerably less attention. In 1996, also isolated relative to each other, making it difficult an expedition jointly sponsored by the Mongolian to correlate between them. Thin sedimentary cover Paleontological Center and the Hayashibara Museum and relatively low-lying exposures also mean that the of Natural Sciences revisited three known fossil verte- major formational contacts are seldom visible, and brate localities in the eastern Gobi region: Khuren even the best exposures contain only part of the Dukh, Bayn Shireh and Burkhant (Figure 1), to sedimentary sequence, with the upper and lower which have been attributed a Khukhtekian through boundaries obscured. There was some volcanism in Baynshirenian age (approximately Aptian through the Gobi during the Cretaceous, but it was limited to early Campanian, see Figure 2). From the analysis of distinct and widely separated localities where thin a suite of paleomagnetic samples collected through the andesitic to basaltic lava flows and some locally thick Cretaceous sequences exposed at these localities we  Paleomagnetic and palynologic analyses of Albian to Santonian strata 831 Figure 1. Map of Mongolia showing the major tectonic elements of the Gobi Basin, discrete sub-basins or grabens, major towns and fossil localities described in the text. The paleomagnetic sample localities of Pruner (1987, 1992) are labeled A and B, and the present day field direction and the paleomagnetic orientation are shown along with the inferred lines of paleolatitude. present the first extensive magnetostratigraphic analy- (Jerzykiewicz & Russell, 1991; Norman, 1996, 1998). sis of any Cretaceous sedimentary sequence in the Likewise, the only vertebrate genera found in Upper Gobi region. Cretaceous deposits of both Asia and North America At present there is no way to correlate the Creta- are a multituberculate (Catopsalis), a hadrosaur (Sau- ceous terrestrial sequences of the Gobi Basin to the rolophus) and a theropod (Velociraptor; Jerzykiewicz & European stage boundaries as seen in the Western Russell, 1991; Burnham et al., 1997). Of these verte- Interior of North America. The sediments are not brates, Saurolophus may provide the best biostrati- interbedded with marine index fossils that can be graphic correlation to date. This hadrosaur was first dated, and there are few interbedded volcanic layers collected by the Mongolian Paleontological Expedi- that can be isotopically dated. Biostratigraphic corre- tion in the 1940s from the Nemegt Formation lations of regional significance have, however, been (Rozhdestvensky, 1952). It has also been reported made on the basis of freshwater mollusks (Barsbold, from the lower Edmonton Formation in Alberta, 1972; Martinson, 1982; Makulbekov & Kurzanov, Canada (Russell & Chamney, 1967), a sequence 1986), plant megafossils (Krassilov, 1982; Krassilov which, on the basis of the associated ammonite fauna, & Makulbekov, 1995), palynomorphs (Bratseva & can be correlated with precision to the very latest Novodvorskaya, 1975), charophytes (Karczewska & Campanian and earliest Maastrichtian, from between Ziembinska-Tworzydlo, 1983) and turtles (Shuvalov 71.5 and 69.5 Ma (Obradovich, 1993). & Chkhikvadze, 1979). In general, the resolution of In the absence of a well-defined biostratigraphic these correlations is poor. Moreover, except for one zonation through all but the very latest Cretaceous, genus each of triconodont mammal (Gobiconodon) and with no isotopic ages available in the critical and ornithopod dinosaur (Iguanodon), no other sedimentary sections, the only method that holds any vertebrate genera from the Lower Cretaceous are great promise for the global correlation of the Gobi known to co-occur in both Asia and North America region is magnetostratigraphy and correlation to the 832 J. F. Hicks et al. Figure 2. Summary diagram of the geomagnetic polarity time scale (GPTS) and Cretaceous stage boundaries (after Harland et al., 1990), with the revised stage boundary ages of Obradovich (1993). To the right labeled A to H and with the published source, are the different biostratigraphic ranges for the sections and formations studied, which are shown with a stippled pattern. The biostratigraphic ages referred to in the text are shown with a diagonal pattern. Geomagnetic Polarity Time Scale (GPTS). Yet this great potential for a direct magnetostratigraphic cor- method is hindered by the unbroken Cretaceous Long relation to the GPTS, if in the future the same Normal interval (C34n) that prevailed for some 35 or methods can be successfully applied to the younger 40 million years through the late Early to early Late Cretaceous sediments of the Gobi Basin. Our results Cretaceous. The paleomagnetic results that we also indicate that magnetostratigraphy may be present here indicate, however, that a measurable calibrated by palynostratigraphy in the Gobi region. magnetic signal that is wholly concordant with the paleolatitude of this region for the Cretaceous can be 2. Geologic setting obtained from fluvial or fluviolacustrine sediments that lie within the Cretaceous Long Normal (Hicks & The Gobi Basin is a broad semi-arid plateau that is Brinkman, 1997). This demonstrates that there is bounded to the north and west by the mountains of  Paleomagnetic and palynologic analyses of Albian to Santonian strata 833 the Mongolian Altai, Khangai and Khentei ranges plex than that, and records a prolonged paleoenviron- (Figure 1; Jerzykiewicz & Russell, 1991), and to the mental and climatic shift that occurred in response to south and southeast by the Lang Shan and Yin Shan the final tectonic evolution of the region. In the Early ranges, respectively (Jerzykiewicz & Russell, 1991; Cretaceous (Aptian), large perennial lakes formed in Jerzykiewicz, 1995). The larger part of the Gobi Basin broad subsiding basins, but by the Albian continued lies within the borders of Mongolia, and has been block-faulting in the region began to fracture these referred to as Outer Mongolia, but the southeastern basins into numerous small-scale depocenters, and portion (which is not shown in Figure 1) lies in fluviolacustrine conditions with smaller ephemeral northern China, and is named Inner Mongolia. The lakes prevailed over a broad region (Jerzykiewicz, interior of the basin itself is broken up by a series of 1995). An erosional unconformity separates these large faults that divide it into a series of broadly lower units from the overlying Upper Cretaceous northwest–southeast trending fault-block mountains, red beds, which range in age from the Turonian or horsts, with intervening shallow sub-basins or (Jerzykiewicz & Russell, 1991) or possibly Ceno- grabens. The tectonic history of the Gobi Basin sig- manian (Shuvalov, 1982; Samoilov & Benjamini, nificantly post-dates the initial collision between the 1996) through to the Maastrichtian. This unconform- Serbian and North China cratons that occurred as ity developed as a result of the partitioning of the Gobi early as the Permian. The Gobi Basin, as a structural Basin into the Cenozoic graben system (Sun et al., entity, was created in the Jurassic by complex regional 1989). This began in the Late Cretaceous and lasted tectonism related to subduction of the western Pacific into the early Tertiary as the Gobi Basin was broken plate beneath the margin of the eastern Asia continent up into the structurally distinct Shiregin Gashun, (Hsu¨ , 1989). In the Early Cretaceous, the onset of the Ulan Nur, Nemegt and Bayan Mandahu grabens Yanshanian Orogeny, which lasted until the early (Figure 1). The lower series of Upper Cretaceous Tertiary, marked a shift from a compressional to an sediments is dominated by a semi-arid depositional extensional tectonic regime (Li et al., 1995; Vincent & environment with aeolian and intermittent lacustrine Allen, 1999) that strongly influenced sedimentation in sedimentation (Lefeld, 1971; Gradzinski et al., 1977; the Gobi Basin throughout the Cretaceous. The old- Eberth, 1993; Jerzykiewsicz et al., 1993; Jerzykiewicz, est rocks deposited in the basin are Early through 1989, 1995, 1996). A paleoenvironmental shift has Middle Jurassic metasedimentary and volcanic rocks been noted from widespread lacustrine or fluviolacus- that are found exposed within and at the margins of trine sedimentation in the Early Cretaceous, which the mountain uplifts. In some areas these units are lasted in some areas until as late as Baynshirenian time interfingered with coarse-grained clastics that were (Cenomanian to early Campanian?; see Figure 2, F shed off the adjacent eroding highlands (Jerzykiewicz and G), to a semi-arid environment in Barungoyotian & Russell, 1991). This was followed by a significant time (middle Campanian to middle Maastrichtian; period of non-deposition, uplift and erosion that Jerzykiewicz & Russell, 1991). This shift is marked at formed an angular unconformity that can be identified the Bayn Shireh and Djadokhta formational contact across the region (Shuvalov, 1975). This unconform- by a caliche paleosol and hardpan indicative of a ity is in turn overlain by a sequence of faulted Upper substantial period of non-deposition in a semi-arid Jurassic through Lower Cretaceous volcanic and clas- climate (Jerzykiewicz & Russell, 1991). The overlying tic rocks. These units are separated from the overlying Barun Goyot Formation contains abundant evidence Upper Cretaceous sediments by another unconform- of an aeolian origin (Gradzinski & Jerzykiewicz, ity or level of non-deposition (Morris, 1936), just as 1974a, b). A similar, Campanian-aged, climatic the stratigraphically highest Cretaceous units are un- change also occurred in the Western Interior Basin of conformably overlain by units of late Paleocene age. North America, as similar indicators of increasing Nowhere in the region is the stratigraphic record aridity have been observed in the upper part of the complete because from the Early Jurassic until the Belly River Group in Canada (Jerzykiewicz & Sweet, Recent, intervals of localized syndepositional block 1988). In Asia this change in climate and sedimen- faulting occurred across a complex mosaic of grabens tation coincides with a considerable change in dino- within which relatively thin and often discontinuous saurian assemblages in the Gobi Desert and surround- sheets of sediment accumulated. ing regions (Jerzykiewicz & Russell, 1991). A similar An oft-repeated and very prevalent misconception change is also recorded in the continental and marine in the literature is that the Upper Cretaceous through vertebrate assemblages of the Western Interior of Paleocene sediments are dominantly lacustrine in ori- North America in the early to middle Campanian gin (Shuvalov, 1982; Tumanova, 1987). In fact the (Lillegraven & McKenna, 1986). This may reflect a stratigraphic record of the Gobi Basin is more com- Northern Hemisphere or even global climatic shift 834 J. F. Hicks et al. (Jerzykiewicz et al., 1993). If this is true, then Matsukawa et al. (1997) logged a series of 38 the Baynshirenian age may extend into the early sections in the Choir Basin, including the Khuren Campanian (Jerzykiewicz & Russell, 1991). The Dukh locality. These authors divided the Lower Nemegt Formation of Maastrichtian age, which con- Cretaceous of the Choir Basin into three formations: formably overlies the Barun Goyot Formation, reflects Sharilin, Tsagaantsav and Shinekhudag. They applied the onset of a wetter and more humid climate with the name Shinekhudag Formation to the uppermost deposition in broad alluvial meandering systems of the three stratigraphic units at Khuren Dukh. The (Gradzinski, 1970), although the presence of caliche Khuren Dukh locality lies on the western margin of paleosols and the lack of organic-rich detritus indi- the basin and spans the lower and middle members of cates that a relatively dry soil moisture regime pre- the Shinekhudag Formation as defined by Matsukawa vailed at least seasonally (Jerzykiewicz & Russell, et al. (1997). The Shinekhudag may be an informal 1991). Recent work by Jerzykiewicz (1998) sum- assignation, but it is one that is used in the marizes the spectrum of sedimentary paleoenviron- literature (cf. Shuvalov, 1994). It is derived from the ments that have been described from the Cretaceous Shinkhundukian zone of Jerzykiewicz & Russell of central Asia, and makes a convincing comparison (1991, p. 348, Figure 2, composite stratigraphic col- between the Late Cretaceous sedimentary record of umn of Mesozoic strata in the Gobi Basin). This name the Gobi Basin and the modern analogue of the assignation underlines the difficulties that arise when a Okavango region of Botswana. Both regions are formation name is derived from the name of a bio- located in continental extensional tectonic regimes stratigraphic zone (see discussion of ‘svita’, below). and show striking changes from arid to semi-arid The Shinkhundukian zone is approximately Valangin- conditions. ian to Hauterivian in age (Jerzykiewicz & Russell, 1991; Matsukawa et al., 1997), but its namesake, the Shinekhudag Formation, appears to be considerably 3. Geology of the eastern Gobi region younger, as our palynologic study shows. Nowhere in the Gobi Basin is a complete Cretaceous On the basis of the vertebrates, invertebrates and section preserved, and the distribution of the for- palynological remains found at the Khuren Dukh mations varies markedly across the eastern part of the locality, the Shinekhudag Formation has been region. The wide geographic separation of exposures assigned an Early Cretaceous age, known in this makes lithologic correlation difficult and there are region as the Khukhtekian (Jerzykiewicz & Russell, often contradictions in the literature regarding the 1991), named after the Khukhtek ‘svita’, which is well names and ages of some of the formations. This study represented in the Undursil Sum and Shaadangin deals primarily with two formations: the Bayn Shireh Gobi regions of the northern part of the eastern Gobi and the Shinekhudag. Neither of these stratigraphic (Minzhin, 1994). The term ‘svita’ as a stratigraphic units has been formally named in accordance with classification was introduced to the region by Russian widely accepted procedures such as those outlined in workers and subsequently approved by the Inter- the North American Commission on Stratigraphic departmental Stratigraphic Committee of the USSR Nomenclature (NACSN, 1983), but they are cited in (Mezhdedomstviennyi Stratigraficheskiy Komitet the literature. The name Bayn Shireh Formation is SSSR, 1988). The term ‘svita’ combines both litho- attributed to Martinson (1982; see Jerzykiewicz & stratigraphy and biostratigraphy, and has been errone- Russell, 1991), and the name Shinekhudag Formation ously equated with the western term ‘formation’, but was most recently used by Shuvalov (1994) and it has since been redefined and is based on a described Matsukawa et al. (1997) for the strata at Khuren and published reference section. The term ‘svita’ has Dukh. Other names that have been applied to the been somewhat discredited and most modern authors strata at the Khuren Dukh locality (Figure 1), such as distance themselves from it (Gradzinski et al., 1977; ‘Khukhtek Formation’ are similarly either informal or Jerzykiewicz & Russell, 1991). Here we use the term inappropriate because they are based on the Russian ‘formation’ but acknowledge that often these for- stratigraphic concept, the ‘svita’, which is not the mations are not as rigorously defined as is required in same as the internationally recognized concept of the formal nomenclature (e.g., NACSN, 1983). formation (see Jerzykiewicz & Russell, 1991). The Rocks of Khukhtekian age have been correlated strata sampled at the Khuren Dukh locality were with the middle portion of a coeval (Aptian–Albian- described (but not named) by Novodvorskaya (1974) aged) lacustrine formation called the Manlay Lake and also by Shuvalov (1974); Shuvalov (1994, Table deposits that are found in the southeastern Gobi 31, p. 150) used the name Shinekhudag for these (Krassilov, 1980; Lopatin, 1980; Jerzykiewicz & strata, as did Matsukawa et al. (1997). Russell, 1991). The Manlay Lake deposits have been  Paleomagnetic and palynologic analyses of Albian to Santonian strata 835 biostratigraphically correlated to the Dushilin For- date from the 19th Century, and are summarized in mation in the western Gobi, which has in turn been Rozhdestvensky (1977). Almost nothing was known correlated to the Khulsyngol Formation in the north- of the Gobi Basin until intensive paleontological western (Mongol-Altai) Gobi; the latter contains plant exploration was begun there in the 1920s by the megafossils that can be correlated to Aptian–Albian Central Asiatic Expedition of the American Museum sediments in Japan and China (Krassilov, 1982; see of Natural History (1921–1930; Andrews, 1932). Figure 2, A). The Shinekhudag Formation, which Their remarkable discoveries of hitherto unknown is named from exposures in the Choir Basin dinosaur species and the discovery of the first dinosaur (Matsukawa et al., 1997) where it is divided into three nests and eggs paved the way for subsequent ex- members that are of fluvial or fluviolacustrine origin, peditions such as the Sino–Swedish Expedition consists of a sequence of cross-bedded sandstone, (1927–1935) to central and northern China, including conglomerate, thin coal beds and fine-grained units. Inner Mongolia. After World War II the Mongolian The lower Upper Cretaceous Bayn Shireh For- Paleontological Expedition of the USSR Academy of mation of Martinson (1982) is well preserved in the Sciences (1946–1949) conducted extensive exca- southern and eastern Gobi, but only small fragments vations and recovered entire skeletons of Upper are found in the Shiregin Gashun Basin and in the Cretaceous dinosaurs primarily in the Nemegt Basin western (or Trans-Altai) Gobi (for a complete list of (Efremov, 1954). The Soviet–Chinese Expedition the localities of the Bayn Shireh Formation, see (1959–1960) focused on Inner Mongolia in China, Jerzykiewicz & Russell, 1991, p. 365, 366). In con- as did the later Sino–Canadian Dinosaur Project trast, the overlying Cretaceous formations, the (1986–1990) and the recent Sino–Belgium Dinosaur Nemegt and Barun Goyot, are exposed only in Expedition (1995–1997; Dong, 1997). The Polish– the western and southern Gobi and not at all in the Mongolian Expedition (1963–1971) ranged widely eastern part. The contacts between the Bayn Shireh across Outer Mongolia (Kielan-Jaworowska & and these stratigraphically higher Cretaceous for- Barsbold, 1972), as did the joint Soviet–Mongolian mations can only be observed in rare isolated outcrop. Expedition (1968–present; Kramarenko, 1974), and For example, the formational contact between the the recent American Museum of Natural History– Bayn Shireh and the Barun Goyot is exposed at Mongolian Academy of Sciences Expedition (1990– Khermeen Tsav in the western Nemegt Basin but present). Since 1990 there has been a resurgence in nowhere else in the western Gobi (Gradzinski & interest in the region, especially since the new discov- Jerzykiewicz, 1972, 1974a). The most complete sec- eries reported by the Sino–Canadian Dinosaur Project tions of the Bayn Shireh Formation, which is up to (Currie, 1993) and the joint American Museum of 300 m thick, are found in the eastern Gobi and consist Natural History–Mongolian Academy of Sciences of fine-grained, often cross-stratified, sandstone, inter- Expedition (Norell et al., 1994; Novacek et al., 1994; bedded with claystone and concretionary, intrafor- Dashzeveg et al., 1995; for overview, see Novacek, mational conglomerates (Sochava, 1975; Martinson, 1996; Norell, 1997). 1982). Although it is considered to be at least partially The Mongolia–Japan Joint Paleontological Expedi- lacustrine by Samoilov & Benjamini (1996), large tion (1992–present), with whom the authors of this scale cross-stratification in many of the sandstone paper are affiliated, has been conducting research layers at the stratotype locality are more indicative of a across a wide area of Mongolia, most notably meandering fluvial system. The upper boundary of the the south-central Gobi locality of Tugrikin-Shireh Bayn Shireh Formation is marked by a horizon of (Fastovsky et al., 1997) and the exposures in the large concretions and concretionary conglomerate eastern Gobi that are the subject of this paper. (Sochava, 1975; Jerzykiewicz & Russell, 1991). Over- lying the Bayn Shireh Formation is either the 4.1. Paleontology, Khuren Dukh locality, Shinekhudag Djadokhta or Barun Goyot Formation, depending on Formation locality. As has already been noted, sediments in these overlying formations mark a significant change in the The Khuren Dukh locality (Figure 1; Shuvalov, 1974; paleoenvironment from primarily fluvial to a semi-arid Novodvorskaya, 1974) lies 60 km south-southwest depositional system with aeolian dunes, interdune of the town of Choyr (or Choir) in southeastern channels and ephemeral lakes. Mongolia. Fossils were first discovered at this locality in 1970 by the Soviet–Mongolian Expedition, who 4. Previous work noted the ‘‘skeletal occurrence of iguanodontids, tur- The earliest geologic work and the first paleontologi- tles and ganoid fishes’’ (Barsbold et al., 1971, p. 273; cal discoveries from the broad Central Asia region Novodvorskaya, 1974). Contra Novodvorskaya 836 J. F. Hicks et al. (1974), Bakhurina & Unwin (1995) and Hicks & saurs and other reptile skeletons have been reported Brinkman (1997), the strata at the Khuren Dukh from some 12 horizons in the sequence (Matsukawa locality are attributed to the Shinekhudag Formation et al., 1997). (Matsukawa et al., 1997) and not to the Khukhtek and The locality has been assigned an Aptian–Albian Dzun Bayan formations (Shuvalov, 1975) as pre- age on the basis of an invertebrate fossil assemblage viously reported, although the formations are prob- identified by G. G. Martinson and pollen identified by ably correlative (see Figure 2, A and B). The Dzun G. M. Bratseva (Shuvalov, 1974). The molluscan Bayan Formation is currently recognized only in the assemblage includes species of freshwater gastropods areas around the settlements of Sayn Shand and Dzun and pelecypods that are characteristic of Aptian– Bayan in the southern part of the Eastern Gobi Aymag Albian deposits in China (Yakushina, 1964), and of (Minzhin, 1994). other areas of Mongolia believed to be of the same Detailed sections of the Khuren Dukh locality with age. The vertebrate fossil assemblage (see Matsukawa descriptions of the bone-bearing layers have been et al., 1997, Table 1) includes Harpimimus okladnikovi published by Novodvorskaya (1974), Shuvalov (Barsbold & Perle, 1984), a new species of ornitho- (1974) and Matsukawa et al. (1997). At Khuren pod, Altirhinus kurzanovi (Norman, 1998; based on Dukh, the Shinekhudag consists of a 100- to 130-m- material previously referred to Iguanodon orientalis thick basin-fill sequence of poorly lithified and poorly Norman, 1996; Norman & Kurzanov, 1997), sorted sandstone and thin beds of gravel in the lower Psittacosaurus mongoliensis, an ornithocheirid ptero- part, overlain by mudstone of fluviolacustrine origin saur and champsosaurs (Novodvorskaya, 1974; (see also Bakhurina & Unwin, 1995). The sedimen- Rozhdestvensky, 1977, p. 109; Jerzykiewicz & Russell, tary sequence rests on a basement surface of granite of 1991; Bakhurina & Unwin, 1995; see Figure 2, B). Early Cretaceous age and Precambrian metamorphic Recent work by the Mongolia–Japan Joint Paleonto- rocks (Matsukawa et al., 1997). In ascending order, logical Expedition at Khuren Dukh has produced a the Cretaceous deposits of the Choir Basin consist of new macrobaenid turtle (Narmandakh, 1997). Fossil the Sharilin, Tsagaantsav and Shinekhudag for- spores and pollen recovered from several mudstone mations, and can reach 300 m in thickness in drill layers at Khuren Dukh indicate an Aptian–Albian age holes in the central part of the basin. The Khuren (Bratseva & Novodvorskaya, 1975; Nichols et al., Dukh locality lies on the western edge of the out- 1997; see Figure 2, B) based on comparisons to crop of the Shinekhudag Formation, which covers the palynomorph biozonation of sediments of Early the western and central part of the Choir Basin Cretaceous age in Siberia (Kotova, 1964, 1968, 1970; (Matsukawa et al., 1997). To the east, the Markova, 1971; Ivanova & Markova, 1981); but see Shinekhudag lies conformably on the Tsagaantsav also discussion of more recent palynological analyses Formation, but in the vicinity of Khuren Dukh it lies later in this report. unconformably on the basement rocks. The strati- According to Matsukawa et al. (1997), ostracods graphic section studied and described here corre- from the middle member of the Shinekhudag sponds to the lower and middle members of the Formation include species that are common in the Shinekhudag Formation. The lower member consists Hauterivian to Barremian stages. This led Matsukawa of a white arkosic sandstone, occasionally cross- et al. (1997) to attribute an older age to the bedded, that is interbedded with units of conglomer- Shinekhudag. Our study supports the interpretation of ate and thin beds of mudstone. The middle member Shuvalov (1974), however, and on the basis of the consists of a black laminated mudstone that occurs at molluscan, vertebrate and palynologic assemblages the top of the section. The coarse conglomeratic described above, we assign the Khuren Dukh locality deposits were transported into the basin from a west- to the Aptian–Albian (Khukhtekian age; K1/4–5 bio- ern highland source in response to periodic tectonic stratigraphic zone; Jerzykiewicz & Russell, 1991; see pulses and subsidence in the basin. These coarse Figure 2, C). But, as will be discussed below, our lithofacies grade south and eastward from their source palynological analysis further refines this age. in the Ikhe-Nartinskii massif into finer grained sand- stone in the more distal parts of the basin. The middle 4.2. Paleontology, Bayn Shireh locality, Bayn Shireh mudstone member of the Shinekhudag reflects a Formation gradual slowing of subsidence and a gentle paleoslope The Bayn Shireh locality (Figure 1) was established in across the basin from the western source areas. The 1948 with the discovery of an ankylosaurid skeleton abundance of fossil turtles, insects and mollusks in by the Mongolian Paleontological Expedition these units indicates fluviolacustrine and swamp (Maleev, 1952). The units of variegated mudstone environments. Abundant vertebrate remains of dino- and interbedded concretionary sandstone that make  Paleomagnetic and palynologic analyses of Albian to Santonian strata 837 up the sediments at this locality were first described by Formation, contains non-terrestrial vertebrates in- Vasiliev et al. (1959). The original skeleton was found cluding hybodont sharks, rays and plesiosaurs (Currie in a red mudstone that most probably corresponds to & Eberth, 1993). Similar assemblages from the Upper the lower or middle portion of the stratigraphic col- Cretaceous coastal plain of North America are umn at Bayn Shireh. This specimen was assigned to believed to have inhabited estuarine and fluvial a new genus and species, Talarurus plicatospineus environments, which implies that during late (Maleev, 1952; re-described by Tumanova, 1987), Baynshirenian time, large rivers with direct connec- and is among the oldest of the Ankylosauridae family tions to the sea drained at least part of the eastern known from Mongolia (see also Maleev, 1954, 1956), Gobi region. although Shamosaurus scutatus of the Aptian–Albian Baynshirenian age vertebrate assemblages are dis- aged ‘Khukhtekskaya Svita’ remains the earliest and tinctive because of their variety and the abundance of most primitive ankylosaurid from Mongolia (Coombs turtles (Jerzykiewicz & Russell, 1991). No micro- & Maryanska, 1990; Barsbold, 1997). vertebrates have been described and the known dino- The most complete sections of the formation are saurian assemblage is not overly diverse, although it found in the eastern Gobi region, but it is also known includes a wide range of therizinosaurids (formerly from the exposures in the south-central and western segnosaurs; Jerzykiewicz & Russell, 1991; Russell & Gobi (Jerzykiewicz & Russell, 1991). The strati- Dong, 1993), the tyrannosaurid Alectrosaurus (Perle, graphic interval was assigned a Baynshirenian age, 1977) and primitive hadrosaurids (Maryanska & which was based on the association of fossils and Osmolska, 1975, 1981). New specimens of hadro- lithologies that were attributed to the ‘Bainshirein- saurs from the Bayn Shireh Formation at Baishin Tsav skaya Svita’ (Gradzinski et al., 1977; see Figure 2, F). (and those previously collected by the joint Soviet– The stratotype locality was revisited by the Polish– Mongolian Expedition from this site in the eastern Mongolian Expedition of 1963 (Kielan-Jaworowska & Gobi) have recently been attributed to the lambeosau- Dovchin, 1968), and a number of fragmentary dino- rine Bactrosaurus johnsoni (Tsogtbataar, 1997). Recent saur skeletons, tortoise shells, and fish remains were collections by the Mongolia–Japan Joint Paleontologi- recovered. cal Expedition at the Bayn Shireh locality have in- Sochava (1975) has described the detailed stratigra- cluded a variety of turtles, ankylosaurs, probable phy of the Bayn Shireh locality. In this area of the juvenile hadrosaur material and an ornithomimid. In eastern Gobi, three formations have been defined addition, team members have uncovered dinosaur (Sochava, 1975). The lower conglomerate and eggs and nests for the first time in the Bayn Shireh sandstone red beds are assigned to the Saynshand Formation, some of which have been attributed to Formation. The middle sedimentary sequence is the family Dendroolithidae (Watabe et al., 1997; called the Bayn Shireh Formation and is composed of Ariunchimeg, 1997). Collections from a second, fairly gray sandstone and conglomerate, with relatively thick new locality in the Bayn Shireh Formation, called units of red-brown mudstone in the upper part. The Burkhant, which is located approximately 8 km NW uppermost units of calcareous mudstone, sandstone of the stratotype locality, are far less diverse, but and conglomerate are assigned to the Dzhibkhalant include an indeterminate sauropod and a large Formation. indeterminate dromaeosaurid. In 1968 the Soviet–Mongolian Paleontological When considered together, the faunal assemblages Expedition began a long-term series of ongoing inves- from both the Bayn Shireh and the Burkhant localities tigations in Mongolia. From this work, ostracods and are consistent with those assigned to a Baynshirenian freshwater mollusks recovered from the lower part age, which has been cited in the literature as ranging of the Bayn Shireh Formation (Barsbold, 1972; from the late Cenomanian to early Santonian (i.e., the Sochava, 1975) have been compared to similar species ‘Bainshireinskaya Svita’; Martinson, 1975; Gradzinski from Cretaceous sediments found interbedded with et al., 1977; see Figure 2, F) and more recently from Cenomanian to Santonian marine strata in the the Turonian to lower Campanian (interval K2; Fergana and Aral Sea regions (Yakushina, 1964; Jerzykiewicz & Russell, 1991; see Figure 2, G). Martinson, 1982; Makulbekov & Kurzanov, 1986; see Figure 2, E). It is interesting to note that although the 4.3. Paleomagnetism Bayn Shireh Formation at the stratotype locality is considered to be entirely terrestrial, just 200 km to the Except for a preliminary abstract (Hicks & Brinkman, southeast, across the border with China, the Upper 1997), little magnetostratigraphic work from the Cretaceous Iren Dabasu Formation, which has been Cretaceous of Mongolia has been reported. Paleo- correlated with the upper part of the Bayn Shireh magnetic samples of Cretaceous sedimentary rocks 838 J. F. Hicks et al. Table 1. Due to the poor resolution of the available maps, all localities were placed geographically using a GPS system. This table lists either the starting point for each measured stratigraphic section or the paleomagnetic sites referred to in the text. Locality name Global Positioning Satellite (GPS) Location 1 Khuren Dukh Section 4549 50.34 N, 10826 41.48 E, altitude 1176 m, standard error 28 m 2 Khuren Dukh Hawaii (01) 4550 21.43 N, 10826 48.29 E, altitude 1116 m, standard error 5.2 m 3 Khuren Dukh Hawaii (02) 4550 20.43 N, 10826 44.43 E, altitude 1202 m, standard error 5.5 m 4 Bayn Shireh Section 4416 11.33 N, 10955 17.26 E, altitude 866 m, standard error 12.7 m 5 Burkhant Section 4420 23.70 N, 10951 32.23 E, altitude 809 m, standard error 12.7 m have been collected by other recent expeditions in the Burkhant, as well as from one additional locality Gobi Basin (Novacek et al., 1994) and paleomagnetic informally designated as Khuren Dukh Hawaii. The research has been carried out on Cambrian-aged only topographic maps available for the region are sections from western Mongolia (Evans et al., 1996). Russian military maps that are at a small scale and Pruner (1987, 1992) made a paleomagnetic analysis have low resolution. The localities sampled as part of of two suites of Cretaceous rock samples, and this study were mapped using a Global Positioning published a paleopole position and paleolatitude for Satellite (GPS) instrument without differential correc- Mongolia. These determinations were made primarily tion that yields statistical three dimensional accuracies from Cretaceous-aged andesitic basalt collected from of between 2 and 28 m. The longitude and latitude the northern margin of the upper part of the Nilgin of all four localities studied are summarized in Table Basin, to the south of the Chulut Tsagan Del ore 1. At the Khuren Dukh locality, a magnetostrati- deposit (labeled A in Figure 1). Cretaceous volcanics, graphic section comprising four paleomagnetic sites basalt and andesite form extensive flows across the was collected over a 94-m interval logged through the Nilgin Basin and overlie a conglomerate of Cretaceous Shinekhudag Formation. Two paleomagnetic sites age that contains pebbles that are traceable to a were also sampled from the adjacent locality in the Jurassic granitic source. The second suite of samples same formation (Khuren Dukh Hawaii, GPS location was obtained from an unnamed locality of Cretaceous in Table 1). At the Bayn Shireh locality, a 92-m sedimentary rocks, the location of which is shown in section composed of 13 paleomagnetic sites was Figure 1 (labeled B). Both suites gave a reversed collected through the Bayn Shireh Formation. At paleomagnetic direction with a mean of 187.7 (dec- Burkhant (GPS location in Table 1), a short 16-m lination),
61.1 (inclination) and a paleopole pos- section comprising three sites was collected ition of 83N, 222E (Pruner, 1987). Insufficient from a relatively small exposure of the Bayn Shireh information is available at present to determine which Formation. reversal this is. The inferred paleolatitude of the Mongolian region in the Cretaceous is shown in Figure 1. With additional data, Pruner (1992) recal- 5.2. Methodology culated the mean paleomagnetic directions (182.9,
62.6) and the paleopole position (86.9N, In the field, four separately oriented hand samples 252.8E). These revised directions clearly show that were collected for paleomagnetic analysis from each Mongolia and the North China Block share an iden- stratigraphic level or site. Samples collected from tical pole wander path from the late Paleozoic coarser-grained units required the use of potassium onwards. The Cretaceous paleorotation and paleolati- silicate solution (as a hardening agent) and plastic tudinal differences of about 15 between the Siberian containers. Three samples from each site were then Platform and the Mongolia/North China Block are dry-sanded into cubes approximately 2 cm square, in presumably related to the final closure of the ocean preparation for measurement in a CTF DRM-430 basin between them at some time in the Cretaceous cryogenic magnetometer inside a shielded room (Pruner, 1992). (background field <200 nT) located in the paleomag- netic laboratory at Scripps Institution of Oceanogra- 5. Paleomagnetic analysis phy. Samples in plastic containers were opened and rewrapped in aluminium foil prior to thermal treat- 5.1. Localities studied ment. After measurement of the natural remnant A suite of paleomagnetic samples was collected from magnetization (NRM) at room temperature, the three main localities: Khuren Dukh, Bayn Shireh and samples were thermally demagnetized in a series of  Paleomagnetic and palynologic analyses of Albian to Santonian strata 839 closely spaced temperature steps in a zero magnetic were considered to have been completely overprinted field inside a specially constructed shielded oven, and were discarded from subsequent analysis. Such a modeled after a type developed at Lamont-Doherty specimen, BSI12B, is shown in Figure 3E, F. The Geological Observatory. The samples were demagnet- vector end-points track toward a normal direction, ized in between 14 and 19 steps beginning at 125C, following a circuitous path from temperature steps increasing in steps of 25 to a maximum temperature 125 to about 475, but then the directions become of between 450 and 550. Some of the stronger random, and move widely both in direction and samples were taken to higher temperatures in the intensity between steps. Some specimens were ran- later stages of the analysis when their behavior to dom in orientation, and were discarded from further thermal demagnetization was better known. study. For the sites that comprised at least three The paleomagnetic data analysis was carried out statistically significant directions obtained by either using software developed in-house at Scripps Insti- PCA or Fisher analysis, we calculated a site mean tution. First the directional data were plotted on direction using standard Fisher statistics (Fisher, vector end-point (Zijderveld) and equal-area diagrams 1953). The samples from some sites displayed poor (Figure 3). When these plots were analysed visually, consistency and the site mean from such sites was the largest majority of the specimens were found to discarded if the CSD was larger than 35. The site display a simple quasi-linear characteristic component means from acceptably consistent levels were then that trended straight towards the origin. This consist- converted to virtual geomagnetic poles (VGPs). The ent direction was usually attained after the 225 to VGP latitude of each site mean is a good guide to 250 temperature step (Figure 3A, B), but in some polarity (a positive value being normal and negative cases it was not attained until the 350 and higher being reversed). In the VGP diagrams, the VGP temperature steps (Figure 3C). Over the thickness of latitude of each site mean is plotted as filled squares the Bayn Shireh Formation, these simple quasi-linear against height in section (e.g., Figure 4B). The samples were highly consistent, as is shown in Figure individual sample data based on lines calculated by 3A, B. The two illustrated samples are from the lower PCA are plotted as open diamonds, and those based and upper part of the section, respectively, some 60 m on Fisher analysis of clusters are plotted as open apart, but show almost identical magnetic behavior. triangles. Using principal component analysis (PCA; Kirshvink, 1980) best-fit lines were calculated from a minimum of five (and usually many more) consecutive demag- 6. Paleomagnetic results netization steps, and were considered acceptable if 6.1. Khuren Dukh they had a maximum angle of deviation (MAD) of less than 25. At the Khuren Dukh locality (Figure 4A), a 94-m In a few specimens the end-points tended to cluster section was logged primarily through the lower mem- or hover in one orientation rather than trend to the ber of the Shinekhudag Formation, which dips 20 to origin. This is clearly shown in Figure 3D, where the the ESE at this locality, and consists of fine to very overprint is completely removed by the 275 to 300 coarse, often cross-stratified, sandstone and conglom- step. In successive temperature steps up to about erate interbedded with thin layers of mudstone and 475, the specimen adopted a characteristic reversed shale. Samples from all four paleomagnetic sites at this component that ‘hovered’ in a stable direction and did locality were collected from sandstone. Three samples not systematically decrease in intensity. After the 475 were analysed from each of the four sites, but only the step the data became essentially random in direction samples from the two sites in the middle of the section and intensity. In cases such as this we discarded the were sufficiently consistent in direction to allow the low (NRM to 250) and high (500 to 575) tempera- calculation of site mean directions (Figure 4B). Even ture data and selected a set of at least five consecutive within these two normal polarity sites there was con- demagnetization points that clustered around a stable siderable variability, as is shown in Figure 4C where direction that best defined the characteristic compo- the six samples from the two significant sites are nent, and then calculated a mean from these using plotted along with the two mean site directions, but Fisher statistics (Fisher, 1953). This mean was con- the direction is consistent with a mid-latitude North- sidered acceptable if the circular standard deviation ern Hemisphere sampling locality. The declination of (CSD) was less than 35. the site means lies to the east of the present day field In a very few cases, the specimen moved towards a (calculated from the current International Geomag- stable direction over a number of temperature steps, netic Reference Field, IGRF, declination 356.2; but failed to reach that stable direction. These samples inclination 65.2, Figure 4C). The individual samples  Paleomagnetic and palynologic analyses of Albian to Santonian strata 841 Figure 4. Khuren Dukh paleomagnetic section. A, lithostratigraphy of the section, showing the stratigraphic position of the paleomagnetic sample sites. B, VGP plot of the magnetostratigraphic section. VGP latitudes are plotted versus height in section. Sample means from PCA (best-fit lines) are shown as open diamonds, Fisher averages of clusters are shown as open triangles. Filled squares are statistically significant site means. C, equal-area plot of the sample mean directions (open heavy circles) shown with the corresponding site means (filled diamond) and the present day field direction (declination 356.2, inclination 65.2, shown as a crossed circle). analysed from this section displayed a normal positive 6.2. Khuren Dukh Hawaii (down) inclination, and some variability in decli- nation, but the site means are very consistently Two paleomagnetic sample sites were collected from oriented to the northeast at 35 (Figure 4C). the middle member of the Shinekhudag Formation at Figure 3. Representative samples selected from the Bayn Shireh section, showing: A, B vector end-point diagram of two quasi-linear demagnetization paths from the lower (A) and upper (B) part of the Bayn Shireh section, clearly showing excellent consistency in magnetic direction. Each demagnetization step is plotted as a pair of the vector’s cartesian components (North, East and Down). Solid circles, N and E; open squares, N and Down; C, vector end-point diagram showing removal of an overprint at the 350 step, and subsequent northeast direction, consistent with A and B; D, ‘hovering’ behavior where the sample mean was calculated by Fisher average of a cluster of points from the 300 to 475 temperature steps (directions are random by the 500 step); E, circuitous demagnetization path that fails to reach a stable direction and is essentially random by the 500 step; F, corresponding equal-area plot of E (open/black circles, upper/lower hemisphere). 842 J. F. Hicks et al. Figure 5. Bayn Shireh paleomagnetic section. Caption as for Figure 4, except for C. C, equal-area plot of the site mean directions (filled circles), shown with the overall section mean direction (heavy open diamond) and associated alpha 95 (dashed circle), and present day field direction shown as a crossed circle. the nearby Khuren Dukh Hawaii locality (Table 1), 6.3. Bayn Shireh which is located in an area of limited exposure and low relief where bedding is tilted at 15 to the east. The At Bayn Shireh (Table 1; Figure 5A) a 92-m section paleomagnetic samples from site KDH01 were col- of 13 paleomagnetic sites was sampled at an average lected from an organic-rich, dark brown mudstone site spacing of approximately 7.1 m. The section is laminated with macerated plant fragments and con- well exposed and suitable for paleomagnetic analysis taining palynomorphs. Samples from the second site as the sediments are moderately fine grained (usually were collected from a nearby light gray, fine- to caliche-rich mudstone or siltstone) with only a few predominately coarse-grained friable sandstone. The interbedded sandstone beds and thin conglomeratic samples from site KDH01 were consistent and uni- intervals (Figure 5A). Of the 13 sites analysed, 11 formly normal in polarity. The site mean lies close to were statistically significant. The mean VGP direc- the expected dipole direction for a site at this latitude, tions for each of those sites are shown in Figure 5B as but some 20 to the northeast and some 10 more filled squares. All the sites are normal in polarity. In shallow (Figure 6). The samples from the second site, Figure 5C the directions of the 11 site means are KDH02, were inconsistent and uninterpretable, plotted as black filled circles, although the consistency probably due to their unsuitable coarse sandstone and density of these data indicate that the majority all lithology. plotted in essentially the same place and cannot be  Paleomagnetic and palynologic analyses of Albian to Santonian strata 843 istics, however. They all lie in the same broad time interval, they are consistent with previously published data, they are different from the present day field, they are consistent over a wide geographic region, a variety of bedding attitudes and they span a number of discrete sedimentary basins. 6.4. Burkhant At Burkhant (Table 1, Figure 7A) a 16-m section of three paleomagnetic sites was logged and sampled. The samples were collected from highly fractured mudstone or fine sandstone layers that were interbed- ded with coarser, cross-bedded sandstone and con- glomerate. The section at this locality is capped by an additional 2–3 m of highly weathered mudstone and fine to very coarse, trough cross-bedded sandstone and conglomerate that was not logged or sampled. Of the three sites analyzed, only one, a normal polarity site, was statistically significant. The mean VGP lati- tude of that site (calculated from three samples) is plotted as a filled square in Figure 7B. The mean VGP sample directions from the adjacent sites are plotted on that diagram as well, but the Fisher mean for these directions exceeded our criteria of a CSD angle of no more than 35, so no mean site direction was calcu- lated. The three sample mean directions that make up the statistically significant site are shown on an equal area plot in Figure 7C, along with the calculated site mean which is shown as an open diamond symbol (declination 3.5, inclination 58.2). Consistent with the nearby Bayn Shireh section, the measured direc- tion is to the east of the present field, and at the same inclination. Figure 6. Equal-area plot of the samples that make up the single significant site at the Khuren Dukh Hawaii 7. Palynologic analyses locality. Previous palynological studies of samples from Khuren Dukh (Bratseva & Novodvorskaya, 1975; differentiated. The section mean, or the calculated Nichols et al., 1997) concluded that the age of these Fisher average of all the site means, is plotted with an deposits is Aptian–Albian. Our new interpretation of open diamond symbol. The calculated Fisher average the data from these previous studies supports an of all eleven significant sites is declination 4.7, incli- Albian age determination, however. More impor- nation 64.0, with an alpha 95 of 9.6 (shown as a tantly, results of analyses of new samples collected by dotted circle in Figure 5C). This orientation is some members of our expedition at Khuren Dukh provide 8.5 east of the present field and at approximately the new evidence supporting an Albian age—probably same inclination (Figure 5C). middle to late Albian—for the Shinekhudag For- The limited nature of the exposures and low angle mation at Khuren Dukh. Seven samples collected at of dips in the eastern Gobi did not allow us to conduct or near the localities discussed here yielded assem- a formal fold test to check that the characteristic blages of fossil pollen, spores and algal cysts. A split remnant magnetization is definitely Cretaceous. The of the organic-rich sample from the Khuren Dukh magnetic directions we have obtained from the geo- Hawaii locality (KDH01) yielded a low-diversity graphically separate Bayn Shireh, Burkhant and assemblage in which the numerically dominant forms Khuren Dukh localities share a number of character- present are several species of bisaccate gymnosperm 844 J. F. Hicks et al. Figure 7. Burkhant paleomagnetic section. Caption as for Figure 4, except for C. C, equal-area plot of the sample mean directions (filled circles), shown with the corresponding site mean direction (heavy open diamond), and present day field direction shown as a crossed circle. pollen of pinaceous affinity referable to the genera 4A) provide data that restrict the age of the Pityosporites, Pinuspollenites, Piceapollis and Podocar- Shinekhudag Formation to Albian. pidites; gymnosperm pollen of the genera Cycadopites A sample from unit 8 of the Khuren Dukh and Inaperturopollenites is present also. Less common measured section (sample collected close to paleo- are trilete fern spores (species of Cyathidites, Leptole- magnetic sample KD104; see Figure 4A) is more pidites and Lycopodiumsporites) and chlorophycean al- useful for palynostratigraphy than those from the gal cysts. The algal cysts are indicative of deposition in Khuren Dukh Hawaii locality. Similar to those from a freshwater paleoenvironment. Two other samples the other samples, the assemblage from this sample is collected close to the Khuren Dukh Hawaii paleomag- dominated by bisaccate gymnosperm pollen, es- netic sample site yielded closely similar assemblages, pecially species of the genera Pinuspollenites, Piceapollis adding Cedripites, Cerebropollenites and Corollina to the and Podocarpidites. Also present are species of the list of gymnosperm pollen genera. These assemblages gymnosperm genera Inaperturopollenites, Cerebropol- are consistent with a late Early Cretaceous (Aptian– lenites, Corollina and Cycadopites. The assemblage also Albian) age, and are similar to the assemblages re- includes monolete and trilete fern spores and algal ported by Bratseva & Novodvorskaya (1975) from the coenobia, the latter being indicative of a freshwater Khuren Dukh locality. No species were recovered lacustrine environment. More important for age de- from our samples that serve to narrow this rather termination is the presence of tricolpate angiosperm broad age determination, but other samples collected pollen. The assemblage from this and other samples at or near the Khuren Dukh measured section (Figure from Khuren Dukh is illustrated in Figure 8. Figure 8. Palynomorphs from the Shinekhudag Formation at Khuren Dukh. 1a, b, reticulate tricolpate pollen cf. Tricolpites vulgaris (Pierce 1961) Srivastava 1969 (two levels of focus on the same specimen); 2, psilate tricolpate pollen; 3–5, Asteropollis asteroides Hedlund & Norris 1968; 6, Corollina sp.; 7, Cycadopites sp.; 8, Laevigatosporites sp.; 9, Cyathidites sp.; 10, trilete spore cf. Deltoidospora; 11, Pityosporites sp.; 12a, b, Podocarpidites sp. (two levels of focus on the same specimen); 13, Pityosporites sp. cf. Pinuspollenites; 14, Inaperturopollenites sp.; 15, Cerebropollenites sp.; 16, Piceapollenites sp.; 17, Botryococcus sp. (algal coenobium); 18, Pediastrum sp. (algal coenobium). All specimens shown at the same magnification except 16, which is 50% of the others. 846 J. F. Hicks et al. The sample from unit 8 yielded two biostratigraphi- Novodvorskaya’s (1975) record of this species sup- cally very important species. A specimen of reticulate ports our palynologic evidence for an Albian age at tricolpate pollen of the genus Tricolpites and a second Khuren Dukh. species of psilate tricolpate pollen. Singh (1975) docu- Hua (1991) reported Asteropollis pollen from the mented that the lowest stratigraphic occurrence of Eren Basin of Inner Mongolia, China, and concluded reticulate tricolpate pollen in North America is in that it indicated an Aptian age for the Saihan Tal rocks of middle Albian age. Although tricolpate an- Formation, but he misinterpreted the data of Doyle & giosperm pollen has been reported from older Creta- Robbins (1977) that he cited to support that con- ceous rocks elsewhere in the world (see review by clusion. As published by Doyle & Robbins (1977, Batten, 1996), an analysis by Brenner (1976) shows p. 57, fig. 4), the age of the biostratigraphic subzone in that tricolpate pollen is not present at middle latitudes which Asteropollis asteroides, Stephanocolpites fredericks- below the middle Albian. Brenner’s analysis was for burgensis and tricolpate pollen including species of the Western Hemisphere, but we assume that the Tricolpites occur is middle to early late Albian in age. trend in angiosperm migration patterns in middle Thus, a middle to middle–late Albian age is indicated Cretaceous time is equally valid for the Asian conti- for both the Saihan Tal Formation in the Eren Basin nent. Maps published by Smith et al. (1981) show that and the Shinekhudag Formation at Khuren Dukh. central Asia and North America were at about the same paleolatitude in Albian time. The presence 8. Conclusions of tricolpate pollen in combination with other angiosperm pollen previously reported (reviewed The chronostratigraphy of the Mongolian Cretaceous below) indicates that the Shinekhudag Formation at sedimentary sequence has been difficult to deduce Khuren Dukh is middle to late Albian in age. with any precision. The nature of the intracratonic As reported by Nichols et al. (1997) and Nichols graben style of terrestrial sedimentation has meant et al. (in press), three samples previously collected at that the sedimentary sequences are relatively thin Khuren Dukh yielded assemblages of palynomorphs and discontinuous and punctuated with significant similar in composition to those discussed above. They hiatuses and unconformities. Regional biostrati- are dominated by pollen of various gymnosperm graphic zonation of the sequence has gone a long way species. One of them (sample KD1–4p), from the towards dating individual formations or depositional lower part of the Khuren Dukh section, also contained basins, whether by using palynoflora, megaflora, mol- the biostratigraphically important angiosperm pollen lusks or vertebrates (Jerzykiewicz & Russell, 1991), species Asteropollis asteroides Hedlund & Norris, 1968. but interbasinal correlations remain difficult. Mag- There are numerous records of this species in the netostratigraphy is a dating method that can refine the Albian of North America (Hedlund & Norris, 1968; chronology in such sedimentary and tectonic regimes, Srivastava, 1975; Wingate, 1980; Nichols & Jacobson, especially where isotopic ages are not available. Our 1982; Ward, 1986), and on this basis Nichols et al. (in paleomagnetic results demonstrate that the Creta- press) concluded that the age of the strata at Khuren ceous sedimentary rocks in the eastern Gobi preserve Dukh is Albian. That conclusion is strengthened by a measurable magnetic orientation that is in close the discovery reported here of tricolpate pollen in the accordance with that expected for this age and latitude section at Khuren Dukh sampled for paleomagnetic from the magnetic analysis of igneous rocks of analysis. approximately the same age (Pruner, 1987, 1992). In There is additional supporting evidence for our the same region of the eastern Gobi, Pruner (1987) conclusion based on a reevaluation of the study by measured a reversed magnetic direction from a Bratseva & Novodvorskaya (1975). They did not Cretaceous sandstone (Figure 1) that is antipodal to report either Asteropollis or any species of tricolpate our normal directions and is strong proof that these pollen from Khuren Dukh, but they did report are primary depositional remnant directions, and not another biostratigraphically important species of a later pervasive overprint. angiosperm pollen not observed in our samples, Steph- All of the sections and localities studied by us are anocolpites aff. fredericksburgensis. The species Steph- normal in polarity, and on the basis of their biostrati- anocolpites fredericksburgensis Hedlund & Norris 1968 graphic age estimates (Figure 2), they can be convinc- is well known in deposits of Albian age in North ingly correlated to chron 34 normal (C34n), the America (Hedlund & Norris, 1968; Doyle & Robbins, Cretaceous Long Normal that ranges in age from 1977; Wingate, 1980) and has also been reported approximately 121 to 83.5 Ma (Obradovich, 1993; from the Albian of Asia, in Siberia (Chlonova, Cande & Kent, 1995). The previous biostratigraphic 1985) and China (Yu, 1985). Thus, Bratseva & age estimates for the Shinekhudag Formation (Figure  Paleomagnetic and palynologic analyses of Albian to Santonian strata 847 2, B) place this sequence somewhere within the 22.5 and Mr T. Sonoda, for their assistance in the field; million year span of the Aptian and Albian Stages of the Mongolian members of the 1996 expedition, the Lower Cretaceous. In this case magnetostratigra- preparators, Enkhbat and Otgonjargal, drivers, phy can do little to refine this broad chronologic Bat-Orkhon, Ganbaatar, Ganulzii and Tseveen, and interpretation. Palynostratigraphy indicates, however, cook Tsetsegee, for their respective contributions. that the time span encompasses only the Albian part Grateful acknowledgment is made to: Dr K. Ishii, of chron 34 normal, which ranges in age from 112 to Director, Hayashibara Museum of Natural Sciences 98.5 Ma, and perhaps only the middle to late Albian, for his overall guidance, and to Mr K. Hayashibara, an interval of only about 7 million years (Figure 2, D). President of the Hayashibara Company Limited, for In the case of the Bayn Shireh Formation, the his generous support of this research. Acknowledg- biostratigraphic age estimates range from the Ceno- ment is also made to the donors of the Petroleum manian to the early Campanian (Figure 2, E–G). The Research Fund, administered by the American top of C34n, is believed to lie at or near to the Chemical Society, for partial support of this research Santonian/Campanian stage boundary (Lillegraven, with Grant A.C.S.-P.R.F. 30442-AC8, to Dr L. 1991; Obradovich, 1993; Harland et al., 1989). Both Tauxe, Scripps Institution of Oceanography, Paleo- of the studied exposures of the Bayn Shireh For- magnetic Laboratory. Our thanks to Dr J. Gee and Mr mation are entirely normal in polarity, which may S. Didonna for technical assistance in the Scripps suggest that the Baynshirenian age attributed to these laboratory; to Dr I. MacInnes and to Dr L. McNeil of units is no younger than the latest Santonian (Figure Yale University for their assistance on early drafts of 2, H), contra Jerzykiewicz & Russell (1991), but in this manuscript. This paper constitutes Hayashibara accordance with the estimates of Martinson (1975) Museum of Natural Sciences (Okayama, Japan) and Gradzinski et al. (1977; Figure 2, F). As was Contribution Number 3. pointed out earlier, however, even the best exposures in the Gobi usually contain only part of a sedimentary References sequence. Thus, the possibility of finding a reversal Andrews, R. C. 1932. The new conquest of Central Asia. Natural near the top of a more complete section of the Bayn History of Central Asia 1, 678 pp. (American Museum of Natural Shireh Formation remains plausible. Regardless of History, New York). this seeming ambiguity, the results of our study clearly Ariunchimeg, Y. 1997. Results of studies of dinosaur eggs. Abstracts of Report Meeting, Mongolia–Japan Joint Paleontological Expedition, indicate that there is enormous potential for further p. 13 (Mongolian Paleontological Center, Ulaan Baatar, magnetostratigraphic work in the Mesozoic of the Mongolia). Gobi Basin, especially in the younger, post-Santonian Bakhurina, N. H. & Unwin, D. M. 1995. A survey of pterosaurs from the Jurassic and Cretaceous of the former Soviet Union and or post-early Campanian sediments of the central and Mongolia. Historical Biology 10, 197–245. western Gobi region, where some of the most signifi- Barsbold, R. 1972. Biostratigraphy and fresh-water molluscs from cant Cretaceous and Tertiary vertebrate discoveries the Upper Cretaceous of Mongolia, 1–88. (Izdatelstvo Nauka, Moscow) [In Russian] have been made. Extensive magnetostratigraphic work Barsbold, R. 1997. Mongolian dinosaurs. In The encyclopedia of was conducted at some of these younger fossil verte- dinosaurs (eds Currie, P. J. & Padian, K.), pp. 447–450 brate localities during the summer of 1997 by the (Academic Press, San Diego). Barsbold, R. & Perle, A. 1984. The first record of a primitive Mongolian–Japan Joint Paleontological Expedition, ornithomimosaur from the Cretaceous of Mongolia. Paleontologi- the results of which will be the subject of a subsequent cal Journal 18, 118–120. report. Barsbold, R., Veronin, U. I. & Zhegallo, V. I. 1971. The work of the Soviet–Mongolian Paleontological Expedition in 1969—1970. Paleontological Journal 5, 272–276. Batten, D. J. 1996. Upper Jurassic and Cretaceous miospores. Acknowledgements In Palynology: principles and applications (eds Jansonius, J. & McGregor, D. C.), pp. 807–830 (American Association of The authors thank the following people for their Stratigraphic Palynologists, Dallas). contribution to this project: Dr R. Barsbold, Director Bratseva, G. M. & Novodvorskaya, I. M. 1975. Spores and pollen of the Mongolian Paleontological Center, Academy from Lower Cretaceous deposits of a site Khuren Dukh, MNR (Mongolian People’s Republic). In Fossil fauna and flora of of Sciences of Mongolia, for his expertise and Mongolia. Trudy Sovmestnaya Sovetsko–Mongoliskaya Paleontolog- all-encompassing assistance on this project; Mr K. icheskaya Ekspeditsiya 2, 205–209. [In Russian] Tsogbaatar, Director of the Paleontological Museum, Brenner, G. J. 1976. Middle Cretaceous floral provinces and early migrations of angiosperms. In Origin and early evolution of an- Academy of Sciences of Mongolia, for his oversight giosperms (ed. Beck, C. B.), pp. 23–47 (Columbia University and management of the fieldwork; Mongolian paleon- Press, New York). tologists Dr Y. Aruinchimeg and Ms P. Narmandakh, Burnham, D. A., Derstler, K. L. & Linster, C. J. 1997. A new specimen of Velociraptor (Dinosauria: Theropoda) from the Two as well as the Japanese members of the 1996 expedi- Medicine Formation of Montana. In Dinofest International (eds tion from the Hayashibara Museum, Mr S. Suzuki Wolberg, D. L., Stump, E. & Rosenberg, G. D.), Proceedings of 848 J. F. Hicks et al. a Symposium sponsored by Arizona State University, pp. 73–75 Harland, W. B., Armstrong, R. L., Cox, A. V., Craig, L. E., Smith, (Academy of Natural Sciences, Philadelphia). A. G. & Smith, D. G. 1990. A geologic time scale 1989, 263 pp. Cande, S. C. & Kent, D. V. 1995. Revised calibration of (Cambridge University Press, Cambridge). the geomagnetic polarity time scale for the Late Cretaceous Hedlund, R. W. & Norris, G. 1968. Spores and pollen grains from and Cenozoic. Journal of Geophysical Research 100 (B4), 6093– Fredericksburgian (Albian) strata, Marshall County, Oklahoma. 6095. Pollen et Spores 10, 129–159. Chlonova, A. F. 1985. Fundamental stratigraphic events for the Hicks, J. & Brinkman, D. 1997. Preliminary paleomagnetic results type Cretaceous angiosperm pollen. In Palynostratigraphy of the from the Cretaceous of the eastern Gobi, Mongolia. Abstract of Mesozoic and Cainozoic of Siberia (eds Volkova, V. S. & Chlonova, Report Meeting, Mongolia–Japan Joint Paleontological Expedition, A. F.), Akademia Nauk SSSR, Sibirskoye Otdeleniye, Trudy 620, p. 12 (Mongolian Paleontological Center, Ulaan Baatar, 13–17. [In Russian] Mongolia). Coombs, W. P. Jr & Maryanska, T. 1990. Ankylosauria. In The Hsu¨ , K. J. 1989. Origin of sedimentary basins of China. In Chinese Dinosauria (eds Weishampel, D. B., Dodson, P. & Osmolska, sedimentary basins (ed. Zhu, X.), pp. 207–227 (Elsevier, H.), pp. 456–483 (University of California Press, Berkeley). Amsterdam). Currie, P. J. (ed.) 1993. Results from the Sino-Canadian Dinosaur Hua, R. 1991. Early Cretaceous angiospermous pollen from Eren Basin, Project. Canadian Journal of Earth Sciences 30, 1997–2272. Nei Mongol, 65 pp. (Geological Publishing House, Beijing). [In Currie, P. J. & Eberth, D. A. 1993. Paleontology, sedimentology Chinese with English summary] and paleoecology of the Iren Dabasu Formation (Upper Creta- Ivanova, E. A. & Markova, L. G. 1981. Palynological character of ceous), Inner Mongolia, People’s Republic of China. Cretaceous the Cretaceous deposits. Hauterivian–Barremian; Aptian–Albian. Research 14, 127–144. In Pollens and spores of the western Siberia. Jurassic-Paleogene. Dashzeveg, D., Novacek, M., Norell, M., Clark, J., Chiappe, L., Trudy Vsesoiuznogo Neftianogo Nauchno-issledovatelskogo Davidson, A., McKenna, M., Dingus, L., Swisher, C. & Geologorazvedochnogo Instituta (VNIGRI) 177, 282–290. Altangeral, P. 1995. Extraordinary preservation in a new verte- Jerzykiewicz, T. 1989. Sino-Canadian Dinosaur Project Expedition brate assemblage from the Late Cretaceous of Mongolia. Nature successful in Inner Mongolia. GEOS 18 (4), 1–6. 374, 446–449. Jerzykiewicz, T. 1995. Cretaceous vertebrate-bearing strata of the Dong, Z. 1997. Hunting dinosaurs in China. In Dinofest Gobi and Ordos basins—a demise of the Central Asian lacustrine International (eds Wolberg, D. L., Stump, E. & Rosenberg, dinosaur habitat. In Environmental and tectonic history of East and G. D.), Proceedings of a Symposium sponsored by Arizona South Asia (eds Chang, K. H. & Park, S. O.), Proceedings of State University, pp. 259–264 (Academy of Natural Sciences, the 15th International Symposium of Kyungpook National Philadelphia). University, Taegu, pp. 233–256. Doyle, J. A. & Robbins, E. I. 1977. Angiosperm pollen zonation of Jerzykiewicz, T. 1996. Late Cretaceous dinosaurian habitats of the continental Cretaceous of the Atlantic Coastal Plain and its western Canada and Central-Asia—a comparison from a geologi- application to deep wells in the Salisbury Embayment. Palynology cal standpoint. In Cretaceous stratigraphy and paleoenvironments 1, 43–78. (ed. Sahni, A.), L. Rama Rao Volume, Geological Society of India, Eberth, D. A. 1993. Depositional environments and facies tran- Memoir 37, 63–83. sitions of dinosaur bearing Upper Cretaceous redbeds at Bayan Jerzykiewicz, T. 1998. Okavango Oasis, Kalahari Desert: a Mandahu (Inner Mongolia, People’s Republic of China). contemporary analogue for the Late Cretaceous vertebrate Canadian Journal of Earth Sciences 30, 2196–2213. habitat of the Gobi Basin, Mongolia. Geoscience Canada 25 (1), Efremov, I. A. 1954. The paleontological investigations in the 15–26. Mongolian People’s Republic (the preliminary results of the Jerzykiewicz, T. & Russell, D. A. 1991. Late Mesozoic stratigraphy expeditions in 1946, 1947 and 1949). Trudy Mongolskoy Komissii and vertebrates of the Gobi Basin. Cretaceous Research 12, 345–377. Academiya Nauk SSSR 59, 3–32. [In Russian] Jerzykiewicz, T. & Sweet, A. R. 1988. Sedimentological and paly- Evans, D. A., Zhuravlev, A. Y., Budney, C. J. & Kirschvink, J. L. nological evidence of regional climatic changes in the Campanian 1996. Paleomagnetism of the Bayan Gol Formation, western to Paleocene sediments of the Rocky Mountain Foothills, Mongolia. Geological Magazine 133, 487–496. Canada. Sedimentary Geology 59, 29–76. Fastovsky, D. E., Badamgarav, D., Ishimoto, H., Watabe, M. & Jerzykiewicz, T., Currie, P. J., Eberth, D. A., Johnston, P. A., Weishampel, D. B. 1997. The paleoenvironments of Tugrikin- Koster, E. H. & Zheng, J.-J. 1993. Djadokhta Formation cor- Shireh (Gobi Desert, Mongolia) and aspects of the taphonomy relative strata in Chinese Inner Mongolia: an overview of the and paleoecology of Protoceratops (Dinosauria: Ornithischichia). stratigraphy, sedimentary geology, and paleontology and com- Palaios 12, 59–70. parisons with the type locality in the pre-Altai Gobi. Canadian Fisher, R. A. 1953. Dispersion on a sphere. Proceedings of the Royal Journal of Earth Sciences 30, 2180–2195. Society of London A 217, 295–305. Karczewska, J. & Ziembinska-Tworzydlo, M. 1983. Age of Gradzinski, R. 1970. Sedimentation of dinosaur-bearing Upper the Upper Cretaceous Nemegt Formation (Mongolia) on Cretaceous deposits of the Nemegt Basin, Gobi Desert. Palaeon- charophytan evidence. Acta Palaeontologica Polonica 28, tologica Polonica 21, 147–229. 137–146. Gradzinski, R. & Jerzykiewicz, T. 1972. Additional geographical Kielan-Jaworaska, Z. & Barsbold, R. 1972. Narrative of the Polish– and geological data from the Polish–Mongolian paleontological Mongolian Expedition 1967–1971. Palaeontologica Polonica 27, expedition. Palaeontologica Polonica 27, 17–30. 5–13. Gradzinski, R. & Jerzykiewicz, T. 1974a. Dinosaur and Kielan-Jaworaska, Z. & Dovchin, N. 1968. Narrative of the Polish– mammal-bearing aeolian and associated deposits of the Upper Mongolian Expeditions 1963–1965. Palaeontologica Polonica 19, Cretaceous in the Gobi Desert (Mongolia). Sedimentary Geology 7–30. 12, 249–278. Kirschvink, J. L. 1980. The least-squares line and plane and Gradzinski, R. & Jerzykiewicz, T. 1974b. Sedimentation of the the analysis of paleomagnetic data. Royal Astronomical Society, Barun Goyot Formation. Palaeontologica Polonica 30, 111–146. Geophysical Journal 62, 699–718. Gradzinski, R., Kazmierczak, J. & Lefeld, J. 1968. Geographical Kotova, I. Z. 1964. Age of continental deposits of Gusinoozer and geological data from the Polish–Mongolian Paleontological basins and specialties of composition of Early Cretaceous flora of Expeditions. Palaeontologica Polonica 19, 33–82. Transbaikalia. Izvestiya Akademiya Nauk SSSR, Seriya Geologiya Gradzinski, R., Kielan-Jaworowska, Z. & Maryanska, T. 1977. 8, 84–93. [In Russian] Upper Cretaceous Djadokhta, Barun Goyot, and Nemegt for- Kotova, I. Z. 1968. On age of coal deposits of Eastern Transbaikalia mations of Mongolia, including remarks on previous subdivision. (Bukachachinskaya Basin). Izvestiya Akademiya Nauk SSSR, Acta Geologica Polonica 27, 281–317. Seriya Geologiya 11, 95–103. [In Russian]  Paleomagnetic and palynologic analyses of Albian to Santonian strata 849 Kotova, I. Z. 1970. Palynological basis of age of Jurassic and Lower Maryanska, T. & Osmolska, H. 1975. Protoceratopsidae (Dino- Cretaceous deposits of Transbaikalia. Sovetskaya Geologiya 7, sauria) of Asia. Paleontologica Polonica 33, 133–181. 19–30. [In Russian] Maryanska, T. & Osmolska, H. 1981. Cranial anatomy of Saurolo- Kramarenko, N. N. 1974. On the work of the Joint Soviet– phus augustirostris with comments on the Asian Hadrosauridae Mongolian paleontological expedition during the years 1969– (Dinosauria). Palaeontologica Polonica 42, 5–24. 1972. Sovmestnaya Sovetsko–Mongolskaya Paleontologicheskaya Matsukawa, M., Nagata, H., Taketani, Y., Khanda, Y., Khosbajar, Ekspeditsia, Trudy 1, 9–18. [In Russian] P., Badamgarav, D. & Obata, I. 1997. Dinosaur bearing Lower Krassilov, V. 1980. The fossil plants of Manlay. In Rannemelovoe Cretaceous deposits in the Choir Basin, S.E. Mongolia— ozero Manlay, Sovetsko–Mongolskaya Paleontologicheskaya Eks- stratigraphy and sedimentary environments. Journal of the Geo- peditsia (eds Kalugina, N. S., Barsbold, R., Luwsandansan, B., logical Society of the Philippines 52 (3–4), 99–114. Tatarinov, L. P., Trofimov, B. A., Reshetov, V. Yu., Shishkin, Mezhdedomstviennyi Stratigraficheskiy Komitet SSSR 1988. M. A. & Kalugina, N. S.) Trudy 13, 40–42. [In Russian] Stratigraficheskiy Kodeks SSSR. [Stratigraphic Code of the USSR]. Krassilov, V. 1982. Early Cretaceous flora of Mongolia. Palaeon- (Ministerstvo Geologii SSSR, Akademiya Nauk SSSR, tolographica B 181, 1–43. Leningrad). [In Russian] Krassilov, V. & Makulbekov, N. M. 1995. Maastrichtian aquatic Minzhin, C. 1994. Mongolian stratigraphic dictionary, 171 pp. (Min- plants from Mongolia. Paleontological Journal 29 (2A), 119–140. istry of Geology and Mineral Resources of Mongolia, Institute of Lefeld, J. 1971. Geology of the Djadokhta Formation at Bayn Dzak Geology and Mineral Resources, Biostratigraphy Sector, Ulaan (Mongolia). Palaeontologica Polonica 25, 101–127. Baatar). Li, S., Yang, S. & Jerzykiewicz, T. 1995. Upper Triassic-Jurassic Morris, F. K. 1936. Central Asia in Cretaceous time. Geological foreland sequence of the Ordos Basin in China. In Stratigraphic Society of America, Bulletin 47, 1477–1534. evolution of foreland basins (eds Dorobek, B. L. & Ross, G. M.), NACSN (North American Commission on Stratigraphic Nomen- SEPM (Society for Sedimentary Geology) Special Publication 52, clature) 1983. North American Stratigraphic Code. American 23–241. Association of Petroleum Geologists, Bulletin 67, 841–875. Lillegraven, J. A. 1991. Stratigraphic placement of the Santonian- Narmandakh, P. 1997. New genus of turtle from Lower Cretaceous Campanian boundary (Upper Cretaceous) in the North deposit of Mongolia, Batoremys. Abstract of Report Meeting, American Gulf Coastal Plain and Western Interior, with impli- Mongolia–Japan Joint Paleontological Expedition, p. 10 (Mongolian cations to global geochronology. Cretaceous Research 12, 115–136. Paleontological Center, Ulaan Baatar, Mongolia). Lillegraven, J. A. & McKenna, M. C. 1986. Fossil mammals from Nichols, D. J., Watabe, M., Ichinnorov, N. & Ariunchimeg, Y. the ‘‘Mesaverde’’ Formation (Late Cretaceous, Judithian) of 1997. Preliminary report on palynology of the Cretaceous of the the Bighorn and Wind River Basins, Wyoming, with definitions Gobi Desert. Abstract of Report Meeting, Mongolia–Japan Joint of Late Cretaceous North American Land-Mammal ‘‘Ages’’. Paleontological Expedition, p. 16 (Mongolian Paleontological American Museum Novitates 2840, 1–68. Center, Ulaan Baatar, Mongolia). Lopatin, V. M. 1980. Stratigrafia nizhnego mela Shavokhinskoy Nichols, D. J., Watabe, M., Ichinnorov, N. & Ariunchimeg, Ya. (In vpadiny yugo-vostochnoy Mongolii [The stratigraphy of the press). Preliminary report on the palynology of the Cretaceous of Lower Cretaceous of Shavokhta Depression of southeast Mon- the Gobi Desert, Mongolia. Proceedings of the Ninth International golia]. In Rannemelovoe ozero Manlay, Sovmestnaya Sovetsko– Palynological Congress (American Association of Stratigraphic Mongolskaya Paleontologicheskaya Ekspeditsia (eds Barsbold, R., Palynologists, Dallas). Luwsandansan, B., Tatarinov, L. P., Trofimov, B. A., Reshetov, Nichols, D. J. & Jacobson, S. R. 1982. Palynostratigraphic frame- V. Yu., Shishkin, M. A. & Kalugina, N. S.) Trudy 13, 6–19. [In work for the Cretaceous (Albian–Maastrichtian) of the overthrust Russian] belt of Utah and Wyoming. Palynology 6, 119–147. Makulbekov, N. M. & Kurzanov, S. M. 1986. The biogeographical Norell, M. A. 1997. Central Asiatic expeditions. In The encyclopedia relationships of the Late Cretaceous biota of Mongolia. In of dinosaurs (eds Currie, P. J. & Padian, K.), pp. 100–104. Problemy paleografii Azii, Sovmestnaya Sovetsko–Mongolskaya (Academic Press, San Diego). Paleontologicheskaya Ekspeditsia (eds Rozanov, A. Y., Tatarinov, Norell, M. A., Clark, J. M., Dashzeveg, D., Barsbold, R., Chiappe, L. P., Luwsandansan, B., Afanasheva, G. A., Barsbold, R., L. M., Davidson, A. R., McKenna, M. C., Perle, A. & Novacek, Morosova, I. P., Novitskaya, L. J., Rasnitsyn, A. P., Reshetov, M. J. 1994. A theropod dinosaur embryo and the affinities of the V. Yu., Sysoev, V. A., Trofimov, B. A. & Rozanov, A. Yu.) Flaming Cliff dinosaur eggs. Science 266, 779–782. Trudy 29, 106–112. [In Russian] Norman, D. B. 1996. On Mongolian ornithopods (Dinosauria: Maleev, E. A. 1952. A new family of armoured dinosaurs from the Ornithischia). 1. Iguanodon orientalis Rozhdestvensky 1952. Zoo- Upper Cretaceous of Mongolia. Doklady Academiya Nauk SSSR logical Journal of the Linnean Society 116, 303–315. 87, 131–134. [In Russian] Norman, D. B. 1998. On Asian ornithopods (Dinosauria: Ornithis- Maleev, E. A. 1954. The Upper Cretaceous armoured dinosaurs. chia). 3. A new species of iguanodontid dinosaur. Zoological Trudy Paleontologicheskaya Institut, Academiya Nauk SSSR 48, Journal of the Linnean Society 122, 291–348. 142–170. [In Russian] Norman, D. B. & Kurzanov, S. M. 1997. New ornithopod dino- Maleev, E. A. 1956. The armoured dinosaurs of Mongolia, Part I. saurs from Asia and their evolution (abstract). Journal of Verte- Trudy Paleontologicheskaya Institut, Academiya Nauk SSSR 62, brate Paleontology 17, Supplement to Number 3, 67A. 51–91. [In Russian] Novacek, M. 1996. Dinosaurs of the Flaming Cliffs, 367 pp. Markova, L. G. 1971. History of development of Early Cretaceous (Doubleday, New York). flora of western Siberian lowland (on the basis of palynology). Novacek, M. J., Norell, M., McKenna, M. C. & Clark, J. 1994. Trudy, Siberian Institute of Geology, Geophysics and Mineral Fossils of the Flaming Cliffs. Scientific American 271, 60–69. Resources (SNIIGGIMS) 82, 1–98. [In Russian] Novodvorskaya, I. M. 1974. Taphonomy of locality of Lower Martinson, G. G. 1975. To the question about principles of Cretaceous vertebrates: Khuren Dukh. In Fauna and biostratigra- stratigraphy and correlation of Mesozoic continental deposits. In phy of Mesozoic and Cenozoic of Mongolia, Transactions of Soviet Stratigraphy of the Mesozoic deposits of Mongolia, Joint Soviet- and Mongolia Joint Paleontological Expedition (eds Kramarenko, Mongolian Paleontological Expedition, Transactions 13, 7–24 N. N., Luwsandansan, B., Voronin, Yu. I., Barsbold, R., (Izdatelstvo Nauka, Moscow). Rozhdestvensky, A. K., Trofimov, B. A. & Reshetov, V. Yu.) 1, Martinson, G. G. 1982. The Upper Cretaceous mollusks of 305–313 (Nauka, Leningrad). Mongolia. In Sovmestnaya Sovetsko–Mongolskaya Paleontolog- Obradovich, J. D. 1993. A Cretaceous time scale. In Evolution of icheskaya Ekspeditsia (eds Nevesskaya, L. A., Tatarinov, L. P., the Western Interior Basin (eds Caldwell, W. G. E. & Kauffman, Luwsandansan, B., Barsbold, R., Trofimov, B. A., Reshetov, E. G.), Geological Association of Canada, Special Paper 39, 379– V. Yu. & Nevesskaya, L. A.) Trudy 17, 5–76. [In Russian] 396. 850 J. F. Hicks et al. Perle, A. 1977. On the first discovery of Alectrosaurus from the Late Singh, C. 1975. Stratigraphic significance of early angiosperm Cretaceous of Mongolia. Problems of Mongolian Geology 3, 104– pollen in the mid-Cretaceous of Alberta. In The Cretaceous System 113. [In Russian] in the Western Interior of North America (ed. Caldwell, W. G. E.), Pruner, P. 1987. Paleomagnetism and paleogeography of Mongolia Geological Association of Canada, Special Paper 13, 365–389. in the Cretaceous, Permian and Carboniferous—preliminary Smith, A. G., Hurley, A. M. & Briden, J. C. 1981. Phanerozoic data. Tectonophysics 139, 155–167. paleocontinental world maps, 102 pp. (Cambridge University Press, Pruner, P. 1992. Paleomagnetism and paleogeography of Mongolia Cambridge). from the Carboniferous to the Cretaceous—final report. Physics of Sochava, A. V. 1975. Stratigraphy and lithology of the Upper the Earth and Planetary Interiors 70, 169–177. Cretaceous sediments in southern Mongolia. In Stratigraphy of Rozhdestvensky, A. K. 1952. A new representative of the Mesozoic sediments of Mongolia, Transactions of Joint Soviet– duck-billed dinosaurs from the Upper Cretaceous deposits of Mongolian Scientific Research and Geological Expedition (ed. Mongolia. Doklady Akademii Nauk SSSR (Report of the Academy Martinson, G. G.), 13, 113–182. of Sciences of USSR) 86, 405–408. [In Russian] Srivastava, S. K. 1975. Microspores from the Fredericksburg Rozhdestvensky, A. K. 1977. The study of dinosaurs in Asia. In Group (Albian) of the southern United States. Paleobiologie Jurij Alexandrovich Orlov Memorial Number (eds Singh, S. N. & Continentale 6 (2), 1–119. Singh, S. K.), Palaeontological Society of India 20, 102–119. Sun, Z., Xie, Q. & Yang, J. 1989. Ordos Basin—a typical example Russell, D. A. & Chamney, T. P. 1967. Notes on the biostratigra- of an unstable cratonic interior superimposed basin. In phy of dinosaurian and microfossil faunas in the Edmonton Chinese sedimentary basins (ed. Zhu, X.), pp. 63–75 (Elsevier, Formation (Cretaceous), Alberta. Natural History Museum of Amsterdam). Canada, Natural History Paper 35, 21 pp. Tsogtbataar, K. 1997. Preliminary results of study on Mongolian Russell, D. A. & Dong, Z. 1993. The affinities of a new therapod hadrosaurids. Abstract of Report Meeting, Mongolia–Japan Joint from the Alxa Desert, Inner Mongolia, People’s Republic of Paleontological Expedition, p. 15 (Mongolian Paleontological China. In Results from the Sino-Canadian Dinosaur Project (ed. Center, Ulaan Baatar, Mongolia). Currie, P. J.), Canadian Journal of Earth Sciences 30, 2107–2127. Tumanova, T. A. 1987. The armoured dinosaurs of Mongolia. Samoilov, V. S. & Benjamini, C. 1996. Geochemical features of Joint Soviet–Mongolian Paleontological Transactions 32, 1–80. [In dinosaur remains from the Gobi Desert, South Mongolia. Palaios Russian] 11, 519–531. Vasiliev, V. G., Volkhonin, V. C., Grishin, G. L., Ivanov, A. K., Shuvalov, V. F. 1974. On geological structure and age of localities Marinov, I. A. & Mokshancev, K. B. 1959. Geological structure of Khobur and Khuren Dukh. Transactions of Joint Soviet and the People’s Republic of Mongolia (stratigraphy and tectonics), Mongolian Paleontological Expedition 1, 296–304. 492 pp. (Gostoptekhizdat, Leningrad). [In Russian] Shuvalov, V. F. 1975. Mesozoic stratigraphy of central Mongolia. Vincent, S. J. & Allen, M. B. 1999. Evolution of the Minle and In Stratigrafia mezozoiskikh otlozenii Mongolii, Sovmestnaya Chaoshui Basins, China: implications for Mesozoic strike-slip Sovetsko–Mongolskaya Nauchno-Issledovatelskaya Geologischeskaya basin formation in Central Asia. Geological Society of America, Ekspeditisia (eds Martinson, G. G., Zaitsev, N. S., Luwsan- Bulletin 111, 725–742. dansan, B., Menner, V. V., Pavlova, T. G., Peive, A. V., Ward, J. V. 1986. Early Cretaceous angiosperm pollen from the Timofeev, P. P., Tumortogoo, O. & Yanshin, A. L.), Trudy 13, Cheyenne and Kiowa Formations (Albian) of Kansas, U.S.A. 50–112. [In Russian] Palaeontographica, Abteilung B 202, 1–81. Shuvalov, V. F. 1982. Paleogeography and history of Mongolian Watabe, M., Ariunchimeg, Y. & Brinkman, D. 1997. Dinosaur egg lake systems in Jurassic and Cretaceous. In Mesozoic lake basins of nests and their sedimentary environments in the Bayn Shire Mongolia (ed. Martinson, G. G.), pp. 18–80 (Nauka, Leningrad). locality (Late Cretaceous), eastern Gobi. Abstract of Report Meet- Shuvalov, V. F. 1994. The paleogeography of Mongolian lakes of ing, Mongolia–Japan Joint Paleontological Expedition, p. 11 the Mesozoic. In Limnology and paleolimnology of Mongolia (eds (Mongolian Paleontological Center, Ulaan Baatar, Mongolia). Sevastyanov, D. V., Shuvalov, V. F. & Neustrueva, I. Yu.), Wingate, F. H. 1980. Plant microfossils from the Denton Shale pp. 148–181 (Nauka, St. Petersburg). [In Russian] Member of the Bokchito Formation (Lower Cretaceous, Albian) Shuvalov, V. F. & Chkhikvadze, V. M. 1979. On stratigraphical and in southern Oklahoma. Oklahoma Geological Survey, Bulletin 130, systematical position of some freshwater turtles from new Creta- 93 pp. ceous localities in Mongolia. In Fauna Mezozoya i Kainozoya Yakushina, A. A. 1964. On some Cretaceous fresh-water molluscs Mongolii, Sovmestnaya Sovetsko–Mongolskaya Paleontologicheskaya of South Primoriya. In Stratigraphy and paleontology of Mesozoic Ekspeditsia (eds Tatarinov, L. P., Barsbold, R., Vorobieva, E. I., and Cenozoic deposits in Eastern Siberia and Far East, pp. 280–293 Luwsandansan, B., Tatarinov, L. P., Trofimov, B. A., Reshetov, (Nauka, Leningrad). V. Yu., Rodendorf, B. B. & Shishkin, M. A.), Trudy 8, 58–76. [In Yu, J., Han, X. & Wu, Y. 1985. Cretaceous spore-pollen in Jiangxi Russian] Province. Jiangxi Baieji Baozi Huafen 7, 1–200. [In Chinese]