Progress in the study of the Devonian–Carboniferous boundary sections in the Berchogur Depression (Mugodzhary Mountains, western Kazakhstan)

The Devonian–Carboniferous boundary beds in the Berchogur (Birshogyr) sections in the Mugodzhary (Mugalzhary) Mountains in western Kazakhstan, known to contain various fossil groups, including ammonoids, conodonts, foraminifers, corals, crinoids, and trilobites, are re-examined. The siliciclastic-carbonate succession of the Zhangana Formation reveals the presence of several ammonoid, conodont, and foraminiferal zones. The succession contains the Acutimitoceras ammonoid Genozone, equivalent to the level of the Stockum ammonoid fauna of Germany, with the conodont Siphonodella sulcata appearing within the Genozone. The same beds show mass occurrences of the foraminifer Tournayellina pseudobeata. The study of the Berchogur sections began in the 1980s; these sections are among very few successions globally with ammonoids of the Acutimitoceras Genozone in association with conodonts and foraminifers. At that time, several outcrops in a small area in the upper reaches of Burtybai (Zhangansai) Creek and two boreholes drilled near the sections were studied, and new taxa of ammonoids, foraminifers, ostracods, conodonts, algae, and spores were described. New excavations in 2018–2020, in conjunction with a search for a new definition of the Devonian–Carboniferous boundary, provided abundant new information on these taxa, on the lithology, and on crinoids, trilobites, and corals. The exact position of marker fossils and lithological changes are documented in several sections along Burtybai Creek allowing an amended correlation with sections of the D–C boundary beds in Western Europe.


Introduction
The Berchogur (Birshogyr) sections in the Mugodzhary (Mugalzhary) Mountains (western Kazakhstan) ( Fig. 1) are among a few sections of the Devonian-Carboniferous (D-C) boundary beds with a variety of fossil remains including ammonoids, conodonts, foraminifers, corals, brachiopods, ostracods, trilobites, and other groups (Maslov 1987, etc.). Hence, these sections have been the focus of D-C boundary research for a number of years. Although many authors (Barskov et al. 1984(Barskov et al. , 1986(Barskov et al. , 1988Kusina 1985;Maslov 1987;Vorozhbitov et al. 1992; Barskov and Kusina 1996) recognized their high importance for D-C boundary studies, no field research was undertaken on these sections between 1984 and 2018. This was due to the seemingly stable situation with this boundary following its ratification as a GSSP in the La Serre section in France at the supposed level of the first appearance datum (FAD) of the conodont Siphonodella sulcata (Huddle, 1934) (Paproth et al. 1991). Since S. sulcata was subsequently discovered in the La Serre GSSP section below the ratified level, the definition of the boundary and the GSSP section have been under re-consideration (see Kaiser 2009;Becker et al. 2016;Kaiser et al. 2019 for references). Several international teams are currently working on both these tasks. It is likely that the composite level corresponding to the base of the Protognathodus kockeli Zone, the beginning of radiation, and the end of major regression (top of Hangenberg Sandstone), will be the primary new boundary markers (Corradini et al. 2011Becker et al. 2016;Kaiser et al. 2017;Spalletta et al. 2017). For the South Urals in Russia and western Kazakhstan, it means that the composite nature of the proposed boundary definition is the best option because it can be applied even where Pr. kockeli (Bischoff, 1957) is absent. This is important because 58º40′26.01′′ E), Loc. 7 (Locality 7; Member 4, loose, 48º34′09.7′′ N, 58º40′22.3′′ E). Note that BK-1 is nearly the same as BK-1 of Barskov et al. (1984Barskov et al. ( , 1988, Trench BK-3 was extended downwards in 2019 after Barskov et al. (1984Barskov et al. ( , 1988, whereas "BK-2" and "BS-2" of Barskov et al. (1988) were not re-examined and are not mapped here. Locality 5 of 2019 is very close to Locality 3 of 2018. Locality 7 of 2019 is presumably very close to the ammonoid locality in Member 4 of Kusina (1985); Locality 4 is not shown, as it is near the top of BK-3 and would not have been visible at this scale. Boreholes BS-1 and BS-2 are after Barskov et al. (1988) protognathids are extremely rare in the Uralian and Northern Eurasian D-C sections. Our current research in western Kazakhstan aims to contribute to the study of D-C successions by re-examination of the sections in the Berchogur (Birshogyr) Depression (Mugodzhary Mountains), from which the Acutimitoceras ammonoid fauna and the conodont Siphonodella sulcata were previously reported (Barskov et al. 1984(Barskov et al. , 1988Maslov 1987). These sections were studied by several teams of researchers in the 1950s and 1980s (see references below), but the precise position of the localities and the fossiliferous levels needed updating because of the disruption of previous topography by the extensive limestone quarrying in the eastern area of the Berchogur Depression (Alabas quarries), construction of new roads, changing of the course of local waterways, and destruction of several published outcrops. We have re-examined and resampled several remaining sections in the upper reaches of Burtybai Creek in the Berchogur Depression, recording the position of fossil occurrences in the sections as precisely as possible.

Geological background
The Berchogur sections of the D-C boundary beds are studied in western Kazakhstan (Aktobe Region, Shalkar District), in the Berchogur Depression (Syncline) between the Main Mugodzhary Mountains and Zhan-Gana Range (Ivanov et al. 1986). The D-C boundary beds crop out throughout the Berchogur Depression (Petrenko 1946;Balashova 1953), but the most complete sections are known in its western and eastern wings. The best studied sections are located at Burtybai Creek (= Zhangansai Creek), right tributary of the Shyuldak (= Chuuldak or Shuldak) River (Fig. 1). These sections are on the left bank of Burtybai Creek in its meridional course, north of the road connecting the old and new Alabas quarries, 5 km east of the village of Alabas and 15 km northeast of Berchogur (Birshogyr) railway station. Here, fossil occurrences are found in the Zhangana Formation, studied in seven localities (including four trenches excavated in 1981, 1983, and 2019) and two boreholes (drilled in 1983) (Barskov et al. 1984(Barskov et al. , 1988Nikolaeva and Mustapaeva 2018) (Fig. 1).
The Berchogur Depression is one of the prominent structures of the South Urals, a meridional syncline (ca. 30 × 60 km) with Middle and Upper Devonian and lower Carboniferous sediments in the middle surrounded by thick Lower-Middle Devonian volcanic series of the Mugodzhary, Kurkuduk, and Milgashin formations (Petrenko 1946;Ivanov et al. 1986;Maslov 1987). The lower Carboniferous beds include the Zhangana, Berchogur, and Karabulak formations, represented by 900 m-thick siliciclastics and carbonate-siliciclastics with coal and coalified shale beds in the upper horizons. Tectonically, the Berchogur Depression is part of the Western Mugodzhary (Puchkov 2000(Puchkov , 2010, an area frequently interpreted as a continuation of the Magnitogorsk Megasynclinorium, whereas some reconstructions (Zonenshain et al. 1990;Ivanov et al. 2014) interpret this area as occurring near the western border of the Mugodzhary Microcontinent separating the latter from the Palaeoasian Ocean. However, the discussion of the tectonics of this area is beyond the scope of this paper.
There are two main regions with fossiliferous D-C boundary interval sections exposed in the Aktobe Region (Mugodzhary Mountains), in the east and west of the Berchogur Depression. Balashova (1945aBalashova ( , b, 1953Balashova ( , 1956 described the stratigraphy and fossils of these sections, which she began to study in 1939 (see Petrenko 1946, p. 2) and described ammonoids from the western regions of the depression: (1) Aktan and Kust-Kara dry creeks; (2) middle reaches of Sarysai Creek; (3) west of the Zhan-Gana Range, Tungulyk-Bulak Gully; and (4) Kabaksai Creek (Fig. 1). Balashova (1945a, p. 189) and Balashova in Petrenko (1946) considered the ammonoid assemblage containing mainly species of a genus that she identified as "Imitoceras" (most of these species are currently assigned to Acutimitoceras, while some need further revision and are referred to as "Imitoceras", see the discussion below) to indicate a lower Gattendorfia ammonoid fauna. She drew the D-C boundary 10 m below the limestones with ammonoids, at the base of limestones containing the brachiopod Paulonia ranovensis (Peetz, 1893), at the level coinciding with the top of the limestones with the brachiopod Retzia ulentica Nalivkin, 1937. Rozman (1960) studied in detail the D-C boundary beds in the Berchogur Depression, assembled a collection of ammonoids, brachiopods, and foraminifers, and placed the D-C boundary considerably lower, at the base of the limestones with the brachiopod Ptychomaletoechia panderi (Semenov and Moeller, 1864), at the level corresponding to the base of the Zhangana Formation. In 1981, a research group led by B.I. Bogoslovsky reexamined one of Rozman's sections in the upper reaches of Burtybai Creek (Locality 3, Fig. 1, Trench BK-3 in Barskov et al. 1984Barskov et al. , 1988 and collected ca. 200 specimens of ammonoids later described by Kusina (Barskov et al. 1984;Kusina 1985;Bogoslovsky 1988). In the 1980s, trench BK-3 was essentially a pit that only exposed Member 3. In 1979 and 1981, a team led by N.M. Kochetkova, studied the sections along the middle reaches of Burtybai Creek (Kulagina 1982;Kochetkova and Kulagina in Maslov, 1987). These were the Burtybai section and "Outcrop 8". The Burtybai section extended from 2.5 km west of the mouth of Burtybai Creek to the valley in its upper reaches (currently crossed by a new road and not easily identifiable). Deposits of the Burtybai section were subdivided into 11 lithological units. "Outcrop 8" was located ca. 3 km south-west of the mouth of Burtybai Creek (confluence with the Shyuldak River. The thickness of "Outcrop 8" is 5.3 m and it includes six beds (Kochetkova and Kulagina in Maslov, 1987). Bed 1 (2.5 m-thick) of "Outcrop 8" correlates with the uppermost part of "Unit VIII" of the Burtybai section (Beds with the ostracods Armenites asiaticus-Selebratina subtarchanica -Bairdia turgida) and Beds 2-6 of "Outcrop 8" were described as "Unit IX" in the Burtybai section (upper part of the Zhangana Formation with ostracods of the Pseudoleperditia tuberculifera-Coryellina alba-Cribroconcha primaris assemblage, the foraminifer Tournayellina, and the calcareous alga Garwoodia) (Maslov 1987). In 1983, a team led by I.S. Barskov and colleagues from the Aktobe Geological Survey focussed on the section in the upper reaches of Burtybai Creek, excavated two trenches (BK-1 and BK-2), and drilled two boreholes (BS-1 and BS-2) (Barskov et al. 1984(Barskov et al. , 1988 (Fig. 1). They referred to these sections as the "Berchogur sections" and recognized a different set of members in the Zhangana Formation (From 1 to 4), which we follow here. It was established that the limestones with ammonoids (Member 3) are underlain (Member 2) and overlain (Member 4) by shales with lenses and interbeds of argillaceous limestones. A second ammonoid-bearing horizon (here "Locality 7") was discovered on the slope ca. 18.5 m above Trench BK-3, and in Borehole BS-2. Foraminifers (E.A. Reitlinger and E.I. Kulagina) and ostracods (N.M. Kochetkova) were reexamined based on the joint collections of all the teams that had worked in the region and were published as collective research (Simakov et al. 1985;Maslov 1987), where several foraminiferal and ostracod-based units were recognized. Barskov et al. (1984, p. 213) proposed that the D-C boundary in Berchogur can be drawn "within the lineage of the earliest Acutimitoceras. This boundary may be translated at least in term of conodonts (base of the Siphonodella sulcata Zone)…". This level is within Member 3, at the same horizon as ammonoid Sample 3/5 of Barskov et al. (1984), and also at the level of Member 3c (this paper, Figs. 2, 3). Barskov et al. (1984Barskov et al. ( , 1988 referred to this earlier finding, but also considered that samples BS-2/51 from the depth of 29.0 m contained "specimens of Pseudopolygnathus similar to P. fusiformis". Barskov et al. (1984Barskov et al. ( , 1988) went on to suggest that this level can therefore correspond to the interval of the Siphonodella sulcata Zone, and already be Carboniferous, as it would agree with the decision of the International Working Group referred to by Paproth and Streel (1985). Therefore, Barskov et al. (1988, p. 178) suggested that the conodont-based boundary as the "boundary between the B. costatus and S. sulcata zones in the Berchogur section can be expected within Member 2b in the argillaceous series 70-cm thick above the base of the "sandstone-like" limestone of Member 2a". However, subsequent searches have not confirmed these expectations, and the base of the S. sulcata Zone is currently drawn at the base of Member 3 (which is slightly below the FOD of S. sulcata in Member 3c (Sample BK-3/5a) in Trench BK-3 and in Member 3b (Sample 19/12-3) in Trench BK-1, as it is likely that the earliest entry of S. sulcata in Member 3b is not its evolutionary first appearance) (Fig. 3).
In 2018-2019, we re-excavated Trench BK-1 of Barskov et al. (1984Barskov et al. ( , 1988 and excavated two more new trenches in the upper reaches of Burtybai Creek. One of these new trenches is an expansion of Trench BK-3 of Barskov et al. (1984Barskov et al. ( , 1988Figs. 2-3). The presented research re-examined a series of parallel sections of the upper part of the Zhangana Formation (members 2 and 3 in Barskov et al. 1984Barskov et al. , 1988 in the meridional portion of Burtybai Creek that were referred to as the "Berchogur section" by Barskov et al. (1984Barskov et al. ( , 1988 and Simakov et al. (1985). This interval of the succession overlies the Zhangana limestones (algal-foraminiferal wackestones of Member 1 of Barskov et al. (1984Barskov et al. ( , 1988 and Unit VIII in the Burtybai section of Maslov (1987)) and contains two major parts. The lower part is mainly shale with interbeds of limestone and is referred to as Members 2a and 2b exposed in trenches BK-1, BK-3 (Member 2a is not exposed in BK-3) (see Figs. 2, 3), BK-4 ( Fig. 3), and boreholes BS-1, BS-2 (Fig. 4). The upper part is Member 3 represented by peloid wackestones forming a good marker horizon, 0.45 cm thick that yielded ammonoids of the Acutimitoceras Genozone, the conodont Siphonodella sulcata, and the foraminifer Tournayellina pseudobeata. Limestones of Member 3 are overlain by shales of Member 4, which are studied in isolated outcrops. The total thickness of the succession is ca. 35-40 m.

Lithology and depositional settings
The depositional settings were interpreted based on the study of thin sections and microfossils (other than conodonts and foraminifers) and by manual isolation from the residue of several samples from Members 2a and 2b boiled in fresh water with no acid added, and then washed and settled and separated using a light microscope. Importantly, all samples throughout the section lacked any visible siliciclastics or silt-or sand-size particles. Previously published reports of "arenaceous shale" (e.g., Maslov 1987) were based on field observations of some "grains" in shales and limestones, which represent aggregations of calcareous bioclasts, mainly algae. Below, the results of new observations are included in the description of the recognized lithological units, in stratigraphic order: Member 1. (Figs. 3,4) (> 10 m thick). Light-grey, pinkishgrey, and greenish-grey, medium-to thick-bedded limestone with foraminifers, stromatoporoids, rugose corals, ostracods, shallow-water conodonts of the Polygnathus-Icriodus biofacies (Barskov et al. 1988). Microfacies: algal-foraminiferal wackestone.
A bed of gravelstone (Bed 5 in Fig. 3), 1-5 mm thick, is present in the upper part of the unit and is immediately followed by algal grainstone (limestone) of Unit 2a 2 (0.5-1 m thick, Bed 6) terminating the member. The upper bedding surface of this limestone is slightly irregular, brownish, and represents a hardground suggesting a possible gap in sedimentation. Bedding surfaces of limestones and shale of Bed 6 contain accumulations of small coalified plant remains. Limestones contain algae, foraminifers, ostracods, and spores (Barskov et al. 1988).
Unit 2a 2 corresponds to the so-called "sandstonelike limestone" (Maslov 1987). The bioclast composition was studied in samples BK-1/11 (Trench 1) (Fig. 3) and BS-1/25 (Borehole BS-1) (Fig. 4). This grainstone is mainly composed of remains of calcareous dasycladaceans and Menselina. The residue also contains occasional crinoidal columnals, fragmented and juvenile brachiopod shells, remains of echinoids, ophiuroids, and plates of machaeridians. The assemblage is less diverse than in the underlying Unit 2a 1 , but still marine, although indicating a shallower depth. It is likely that the basin somewhere had algal meadows producing a very large amount of carbonate bioclasts. At the same time the residue contains some altered mica plates suggesting an influx of some clastic components from the land. Interestingly, the assemblage contained one spirally coiled microconchid tube, which might indicate a slight decrease in salinity (Gierlowski-Kordesch and Cassle 2015).

Member 2b.
(2-2.5 m) Dark-brown, greenish-grey to black shale, yellowish brown where weathered. In the middle of the member there is a very thin (1-3 cm) interbed of marly limestone with abundant bivalve imprints (Bed 8, Sample BK-14) (Fig. 3). Within Bed 10 there is a layer with numerous carbonate concretions and abundant fossils, including tabulate and rugose corals, brachiopods, bryozoans, and crinoidal columnals.
Two samples were studied from Trench BK-2 from the year 1983 (BK-2/5 and BK-2/6). Shales of this unit contain smaller and larger (up to 3.5 × 3 × 1.5 cm) carbonate concretions. Some concretions contain moulds and, in some cases, fragments (up to 1 cm) of turriform gastropod shells. The shale contains many fragments of brachiopod shells and ferruginous moulds of juvenile brachiopods. There are hundreds or thousands of small plates of the cirripedian Pabulum Whyte (1976) described from the upper Viséan of Yorkshire and Scotland, crinoidal columnals, ostracod tests, branching bryozoans, and asteroid plates, the latter with somewhat different morphology to the plates from Unit 2a 1 . Ostracod tests and bivalve moulds, and echinoid, ophiuroid, holothurian, and machaeridian remains are rare. The residues contained fragments of orthoconic non-ammonoid cephalopod shells, and one small conical fish tooth. No identifiable remains of calcareous algae or Menselina were found. This composition of bioclasts clearly indicates a fast deepening of the basin during accumulation of the sediments of Member 2b and increased salinity, while the general taxonomic composition of the biota remained largely unchanged compared to Member 2a. Again, almost no fish remains were found. This suggests that the eustatic sea-level fall event between Members 2a and 2b was not long and did not result in a fundamental change in the environment. It is likely to be related to a significant depth increase, with the depth of the seafloor below the photic zone.
In summary, Member 2a reflects a single short-term stage in the evolution of the basin, showing a significant shallowing trend. Deposits of Member 2b were accumulated after a short gap in deposition in a considerably deeper basin, although retaining the marine biota of previous times.
Beds 9-11 are represented by greenish-blue shale and are one unit. Before our excavations "Bed 9" and "Bed 11" were covered intervals separated by an exposed interval of "Bed 10" (Barskov et al. 1988, text-fig. 2). Therefore, we use bed nos. 9 and 11 for continuity. It would be more logical to have a boundary between beds 10 and 11 lower, at the level within Bed 10 with numerous carbonate concretions and abundant fossils. However, at this stage we focus on maintaining correspondence with previous studies, especially because in 2019 we used the same samples nos. as in Barskov et al. (1988). A detailed conodont study including processing samples of 5 kg throughout the section is in progress and will be published in a separate paper.
There is a 10-cm thick bed of shale between 3b and 3c.
In summary, sediments of Member 3 were accumulated below the fair-weather wave base in an oxic environment, suggested by the presence of benthic organisms and traces of bioturbation.
Member 4. (18.5 m) Dark grey to black calcareous shale, sometimes with thin (1 cm) beds of argillaceous limestone or marl and microbioclastic wackestone. Thin sections show numerous thin-walled ostracods, foraminifers, algae, and echinoids fragments (Fig. 6C, E). Fossils are found at the base and in the top part and include ammonoids, nautiloids, brachiopods, corals, conularians, and crinoids. Conodonts are mainly found at the base of the member (Barskov et al. 1988). In borehole BS-2, Member 4 is overlain by sandstones with shale interbeds of the Berchogur Formation, probably of continental origin. This member is studied in several poorly exposed intervals above 3c.

Conodonts
In 2019 we resampled the Berchogur section (carbonates and shale members) for conodonts (trenches BK-1, BK-3, BK-4). Trench BK-2 (of Barskov et al. 1984Barskov et al. , 1988 has not been resampled. The conodont processing is ongoing, but we can present here our preliminary results. The studied samples contained 36 conodont specimens of which 18 are platform elements and the rest are ramiform ones. Seven conodont genera are identified: Polygnathus, Neopolygnathus, Bispathodus, Siphonodella, Mehlina, Icriodus, and Apatognathus. The most important find was a discovery in Unit 3b in Trench BK-1 (Sample BK1-19-12/3) of Siphonodella (Eosiphonodella) sulcata, which allowed this member to be assigned to the sulcata Zone, as it was established previously in Trench BK-3 (Barskov et al. 1988) (Fig. 7).
In this paper we also add the distribution of conodonts based on the sampling of 1981-1983 from three trenches  Birina, 1948 in peloidal wackestone, longitudinal section, Sample 19BK3b (1); b Tournayellina pseudobeata Reitlinger and Kulagina in Maslov, 1987, median sections, Sample 19BK3a (1); c, e Bioclastic peloidal wackestone, Locality 7, Sample BK3-4a (1); c with Tournayellina pseudobeata (T), thin-walled ostracods (O), and a single echinoid spine (E); e with Garwoodia gregaria (Nicholson, 1888), d Mudstone with crinoids and juvenile shell of ammonoid, Sample 19BK3c (1); f Peloidal wackestone with bioclasts of bryozoans (B), crinoids (C), and ostracods (O), Sample 19BK3b (2); g Peloidal wackestone with rare foraminifers, the same thin section as in A with lower magnification; h Garwoodia gregaria Nicholson, 1888, Sample 19BK3a (2); i Fine-grained bioclastic peloidal wackestone, same thin section as in B with lower magnification; scale bar for a-b 0.2 mm; scale bars for c-i are shown in mm in each image 1 3 and two boreholes based on ca. 1 kg samples. The conodonts are extremely rare, discovered only in a few samples. Some conodont elements from these samples were already illustrated in Maslov (1987). Conodont collection no. 243 is housed at the Department of Paleontology, Lomonosov Moscow State University (MSU). The analysis of the conodont distribution in the Berchogur sections allows the recognition of the following conodont zones (Fig. 8 We have no local record of the praesulcata Zone in Members 1 and 2a that fall in the interval of the B. ultimus and S. (Eo.) praesulcata s.l. zones. Considering the conodont data, we provisionally correlate Members 1 and 2a with the Wocklum Limestone of the Rhenish Massif (cf. recent review by Becker et al. 2016 andKaiser et al. 2016) (Fig. 8). The endemic species Neognathodus mugodzaricus (Gagiev, Kononova and Pazukhin in Maslov, 1987) (samples BK1-19-6/3; Fig. 7 Famennian-Tournaisian conodonts from the Berchogur (Birshogyr) sections; a, b Siphonodella (Eosiphonodella) sulcata (Huddle, 1934); a BK-3/5a; also in Barskov et al. 1984, pl. 6, fig. 7a; also in Maslov, 1987, pl. 27, fig. 12a Member 3, Bed 4, Sample 5a; lower Tournaisian, sulcata Zone; b BK-3/5a; Member 3, Bed 4, Sample 5a; lower Tournaisian, sulcata Zone; c, e Icriodus costatus (Thomas, 1949): c BS-2/54; also in Barskov et al. 1984, pl. 7, fig. 2a;Maslov, 1987, pl. 25, fig. 9; Member 2a 2 , Sample 54; uppermost Famennian, Lower-Middle praesulcata zones (top praesulcata Zone sensu Kaiser et al. 2009); e BS-2/54; Sample 54; uppermost Famennian, Lower-Middle praesulcata zones (top praesulcata Zone sensu Kaiser et al. 2009); d Neopolygnathus communis (Branson and Mehl, 1934b); BK-2/10; also in Maslov 1987, pl. 26, fig. 1a; Member 3, Sample 10; lower Tournaisian, sulcata Zone; f, g Polygnathus inornatus inornatus Branson, 1934; f BS-2/54; also in Maslov 1987, pl. 27, fig. 1a; Member 2a 2 , Sample 54; uppermost Famennian, Lower-Middle praesulcata zones (top praesulcata Zone sensu Kaiser et al. 2009); g BS-1/24; Member 2a 2 , Sample 24; uppermost Famennian, Lower-Middle praesulcata zones (top praesulcata Zone sensu Kaiser et al. 2009); h-k Siphonodella cf. praesulcata Sandberg in Sandberg et al., 1972; h BS-1/7; also in Maslov, 1987, pl. 27, fig. 6a; Member 2b, Sample 7; uppermost Famennian, Middle praesulcata Zone (costatus-kockeli Interregnum); i BS-1/12; also in Maslov, 1987, pl. 27, fig. 7a; Member 2b, Sample 12; uppermost Famennian, Upper praesulcata Zone (costatus-kockeli Interregnum); j BS-1/10; also in Barskov et al. 1984, pl. 7 (Maslov 1987, p. 95, pl. 26, figs. 4-8) and has not been recorded from any other region. This species belongs to the N. communis group, which is classified in the genus Neopolygnathus by Vorontzova in Barskov et al. (1991). The presence in Sample BK-19-1/1 (top of Member 1) of Mehlina strigosa and Apatognathus varians cipitis Kononova in Bushmina and Kononova, 1981 supports the assignment of this part of the section to the uppermost Famennian (Bushmina and Kononova 1981;Ziegler and Sandberg 1984 Fig. 4), whereas in other regions its appearance defines the Lower praesulcata Zone (s.l., see discussion in Söte et al. 2017). The assemblage of Member 2b also contains Pseudopolygnathus cf. fusiformis (Branson and Mehl, 1934b) (see Barskov et al. 1984), Polygnathus inornatus, and Neopolygnathus mugodzaricus (BK-1, Sample 17). Our specimens of S. (Eo). praesulcata are not very similar to specimens of S. (Eo.) praesulcata illustrated by Sandberg et al. (1972). Specimens from Berchogur appear to be more derived in the praesulcata/sulcata lineage, as can be judged from Flajs and Feist's (1988, p. 74) summary of differences between these two species based on the angle of the carina curvature. In specimens from Member 2b, this angle is less than 12°, hence we identified them as S. (Eo.) praesulcata. The specimen illustrated in Fig. 6H, with a curvature of ca. 10°, is an intermediate form. This specimen appears to be similar  Fig. 6i-k the angles appear to be less than 12° as in 6H, however, this is an artefact of the photography, i.e., the specimens are slightly inclined, so the carinae appear to be more strongly curved. All these specimens possess a pseudokeel and can therefore be assigned to S. (Eo.) praesulcata. The Last Appearance Datum (LAD) of Icriodus costatus, Mehlina strigosa, and Apatognathus varians cipitis at the top of Member 2a probably correlates with the main conodont extinction defining the base of the costatus-kockeli Interregnum of Kaiser et al. (2009). The lower part of Member 2b is interpreted as the probable record of a transgression, which can probably be correlated with the top part of the Hangenberg Sandstone (see Becker et al. 2016). 3. The Upper praesulcata Zone correlates with the interval of the Lower Protognathodus fauna of Ziegler (1962), the Upper praesulcata Zone of Ziegler and Sandberg (1990) and Corradini (2003), the kockeli Zone sensu Kaiser et al. (2016), and the lower part of the Pr. kockeli Zone sensu Corradini et al. (2017) and Spalletta et al. (2017). The entry of S. (Eo.) cf. praesulcata in Member 2b (Fig. 7hk) is not the FAD of the zonal index species characterizing the Lower praesulcata Zone because the Berchogur specimens have a more derived morphology. The assemblage also contains Polygnathus inornatus (Fig. 7f, g), which can also indicate the upper part of the praesulcata Zone. In the Berchogur sections there are no records of Protognathodus kockeli. However, this level (= kockeli Zone) is known to contain associations of Bispathodus, whereas one of these, B. spinulicostatus (Branson, 1934) (Fig. 7o, p) frequently occurs in this zone . By correlation, largely based on sea-level changes, the level of the FAD of Pr. kockeli could be expected in the uppermost beds of Member 2b (dark shale unit above the limestone with bivalves) (Fig. 3). 4. The sulcata Zone is recognized based on the presence of the index species Siphonodella (Eosiphonodella) sulcata (Fig. 7a, b) and several representatives of the genus Bispathodus (Bispathodus aculeatus aculeatus (Branson and Mehl, 1934a) (Fig. 7q-s), B. aculeatus anteposicornis (Scott, 1961), B. spinulicostatus and B. stabilis vulgaris (Dzik, 2006) (Fig. 7t)). This zone also contains Neopolygnathus communis (Branson and Mehl, 1934b) (Fig. 7d). It is noteworthy that the first appearance of S. (Eo.) sulcata in this section is fixed ca. 20 cm above the base of the limestone beds of Member 3 (Unit 3b) in Trench BK-1 (Sample BK-19-12/3) and in Unit 3c in Trench BK-3 (the latter based on Barskov et al. 1984). The assemblage contains Pseudopolygnathus cf. fusiformis, which occurs in the uppermost Famennian and basal Tournaisian. This occurrence is at approximately the same level as the first ammonoids that we found in 2018-2019 and agrees with Barskov et al.'s (1988) conclusion that the sulcata Zone can only be reliably identified in Member 3. The level of the base of the sulcata Zone within Member 2b (as suggested in Maslov 1987) is not upheld in this paper, as we so far have no evidence of that. Considering the lithology (change from clay to wackestones with algae and rare foraminifers and then to wackestones with ammonoids) and the presence of Acutimitoceras and Siphonodella, Member 3 can be interpreted as a transgressive phase of the evolution of the basin, and at least its upper part (3c) can be considered an equivalent of the Stockum Limestone of the Rhenish Massif (Clausen et al. 1994) and basal Hastière Formation of the Ardennes (Becker et al. 2016, etc.). The genus Bispathodus is represented by Bispathodus aculeatus aculeatus found in the lower part of Member 3 (Sample BK1-19-12/1). This subspecies appeared at the base of the Middle expansa Zone and continued to the lower Tournaisian (Ziegler and Sandberg 1984;Corradini et al. 2017). Polygnathids are represented by Neopolygnathus communis (Branson and Mehl, 1934b) (Sample BK1-19-12/1) (Fig. 3). Neopolygnathus communis ranges from the lower Famennian to the lower Tournaisian . It is possible that there is a short hiatus between Members 2b and 3, as Yu. Gatovsky observed in 2019 a ferruginous crust 1-2 cm thick between clays of Member 2b and limestones of Member 3, whereas the basal Member 3a is slightly brecciated.

Foraminifers
Foraminifers of the Zhangana Formation were first studied by Reitlinger (1961) based on Rozman's (1960Rozman's ( , 1962 collections from the Berchogur Depression. More than 20 years later, Reitlinger and Kulagina (in Simakov et al. 1985;Barskov et al. 1984;Maslov 1987) published the results of their study of foraminifers from the Burtybai and Berchogur sections based on material collected in 1978-1983. Foraminifers of boreholes BS-1, BS-2 and trenches BK-1, BK-2, and BK-3 were studied by Reitlinger, and those of "Outcrop 8" were studied by Kulagina, Pazukhin and Kochetkova (Maslov 1987, Table 5). Based on all these data, the following foraminiferal beds were recognized from bottom to top: (1) Beds with unilocular foraminifers and rare tournayellids (89.5 m, units I-IV, Kushelgian Regional Substage), In the correlation scheme in Maslov (1987, text- fig. 7, insert), Beds 3-5 in "Outcrop 8" (total thickness ca. 1.5 m) were correlated with the upper part of Member 2b and Member 3 of the Berchogur sections and the "Beds with T. pseudobeata" were recognized in Beds 4-6, 0.8 m thick. Later the base of the Beds with T. pseudobeata in "Outcrop 8" was drawn at Bed 1 = upper part of Unit VIII, approximately m below its top based on the first appearance of the zonal index (Kulagina 2013). This boundary corresponds to a level in the lower part of Member 2a in the Berchogur sections (Maslov 1987).
In 2018, Mustapayeva collected foraminiferal samples from natural outcrops along the upper reaches of Burtybai Creek and Locality 3, approximately at places where in 2019 we re-excavated Trenches BK-1 and BK-3. In 2019 we received new material from re-excavated trenches BK-1 and BK-3, re-examined thin sections from the collection of Reitlinger (Geological Institute, Russian Academy of Sciences, Moscow) from samples collected by Kononova from Borehole BS-1 in 1983.
Quasiendothyra kobeitusana substricta Zone. This unit is defined by the presence of the zonal subspecies Quasiendothyra (Quasiendothyra) kobeitusana substricta Conil and Lys, 1964 and the upper boundary is drawn by its disappearance. We recognize the Quasiendothyra kobeitusana substricta Zone, rather than the Beds with Quasiendothyra kobeitusana substricta-Endoglomospiranella imminuta, because Endoglomospiranella imminuta is found in the upper part of Unit VIII in "Outcrop 8" of Maslov (1987), where Q. (Q.) kobeitusana substricta is not found and is correlated with Unit 2a 2 in the Berchogur sections. E. imminuta apparently continues higher up the section, compared to Q. (Q.) kobeitusana substricta.
Only the uppermost part of the Quasiendothyra kobeitusana substricta Zone is recognized in Borehole BS-1 where it corresponds to Members 1 and 2a (about 3 m).
Tournayellina pseudobeata Acme Zone. The lower boundary of this unit is recognized based on the disappearance of the index subspecies of the previous unit and most species of the underlying beds and on the mass occurrence of T. pseudobeata.

Ammonoids
Ammonoids collected in 1981 and 1983 by several teams of researchers were identified by Kusina (in Barskov et al. 1984;Kusina, 1985) and discussed by Bogoslovsky (1986) and Bogoslovsky in Maslov (1987). The taxonomy of the species-rich genus Acutimitoceras is still debatable. It comprises a variety of species with and without constrictions, with a long or short ventral lobe, and with or without a carinate venter. Becker (1996) proposed several subgenera within Acutimitoceras, apart from the nominal subgenus (type species A. acutum Schindewolf, 1923). He named Stockumites, with the type species Imitoceras intermedium Schindewolf, 1923 and Streeliceras (type species Imitoceras heterolobatum Vöhringer, 1960), and considered Sulcimitoceras Kusina, 1985 as a subgenus of Acutimitoceras Librovitch, 1957. Korn andKlug (2002) considered Stockumites Becker as a junior synonym of Acutimitoceras and the genus Streeliceras Becker as a junior synonym of Nicimitoceras Korn, 1993(type species Imitoceras subacre Vöhringer, 1960. Korn and Klug (2002) considered Sulcimitoceras Kusina, 1985 as a separate genus. The range of morphological variation in species currently included in Stockumites is quite wide and is likely to warrant further subdivision. The two Berchogur species A. dzhanganense and A. alabasense, show some affinity with the type species A. (Stockumites) intermedium, but their cross sections and sutures are quite different (see Korn 1994, text-figs. 59B, C;Nikolaeva, 2020, text-figs. 3 and 5), and it is possible that these two species could be assigned to a different subgenus; we do not make a final decision on their subgeneric assignment in this paper. We tentatively assign a single loosely collected specimen identified as Imitoceras (Acutimitoceras) carinatum Schmidt, 1924by Kusina (1985 fig. 3b, 1i), to the genus ?Nicimitoceras, taking into consideration that the early stages are noticeably less evolute than in Streeliceras heterolobatum (see Korn 1994, text- fig. 59E). In addition, the ventral lobe in the Berchogur specimen (Kusina, 1985, text- fig. 3b) is not shortened, as in Streeliceras. At the same time, the early whorls of this specimen (Kusina 1985, text-fig. 1i) are more evolute than in N. subacre (Vöhringer 1960, p. text-fig. 2;Korn 1994, text- fig. 59C). Therefore, the generic affinity of this specimen remains problematic. In the description, Kusina (1985) noted that this specimen develops a carinate venter on the sixth whorl (later than in the type material of carinatum Schmidt) and has a wider shell; the Berchogur specimen is likely to represent a new species, but more material is needed for a positive identification.
The ammonoid assemblage of Member 3c belongs to the Acutimitoceras Genozone based on the occurrence of the index genus. In addition to Acutimitoceras (prionoceratids with evolute inner whorls), the assemblage contains species with involute or semiinvolute inner whorls, "I." bertchogurense, and N.? (or A.?) cf. carinatum; the latter is similar to specimens from the prorsum Zone in its type region in Germany (Korn 1994;Becker et al. 2016). It should be also noted that A. dzhanganense is similar to specimens commonly assigned to A. (Stockumites) subbilobatum (even though A. dzhanganense has no constrictions), which are found in the prorsum Zone in Franconia, in the Rhenish Massif, and in Morocco ). It is not clear whether the German and Moroccan specimens attributed to A. (S.) subbilobatum represent the same species. Therefore, the assemblage from Member 3 probably correlates with that of the prorsum Zone. Becker (1996) proposed to consider the size of the shells developing a carinate venter, suggesting that the development of the oxyconic morphotypes could be shifted to earlier ontogenetic stages in the evolution of the genus Acutimitoceras. For instance, a previously figured specimen of Acutimitoceras sp. from Berchogur (Acutimitoceras sp., Barskov et al. 1984, pl. 1, figs. 8a-b, "Imitoceras (A.)" sp., Kusina 1985, pl. III, fig. 8), developed an oxyconic shell at 15 mm shell diameter, whereas in A. acutum from the eponymous zone in its type region in Germany, such shell morphology is present at the shell diameter of 6-8 mm, in the fourth to fifth (see Korn 1994, text- fig. 50B) whorl. More similar to the Berchogur specimen is A. (A.) oxynotum Bartzsch and Weyer, 1996 from Thuringia, which becomes oxyconic at ca. 15 mm in diameter or even somewhat later (paratype MB.C.1976). Both forms can be distinguished by different types of shell constrictions (very short and only ventral in the Berchogur specimen). Unfortunately, the precise level of occurrence of A. (Acutimitoceras) sp. at Berchogur is not clear (Kusina 1985;Bartzsch and Weyer 1996, p. 53); the specimen was not collected in situ. The ammonoid fauna of Berchogur can be tentatively correlated with that of the lower part of the Wangyou Formation in China in the Muhua sections (Ruan 1981;Qie et al. 2015), containing A. wangyuense Sun andShen, 1965. Becker andWeyer (2004, pp. 29-30) discussed A. wangyuense and excluded it from Acutimitoceras. It is most likely an early karagandoceratid species and was cited as ?Bartzschiceras wangyuense. It was also re-emphasized (following Bartzsch and Weyer 1996, p. 93) that Sun and Chen's original specimen and the supposed later wangyuense specimen of Ruan (1981) are not conspecific. In China, Acutimitoceras sp. were found in the same formation in the S. sulcata Zone, or at the base of the S. duplicata Zone (Member C in the Gedongguan section (Youjiang Basin, Changshun County, Guizhou) (Qie et al. 2015, text-fig. 4). Similar shell morphology is observed in ammonoids from Faunule 1 (Bed RTB 10) of the Gara Bou Tlidat section, in the lower part of the Fezzou Formation in Morocco (Becker et al. 2018) but this level also includes gattendorfiids and is younger than the Stockum Limestone (Fig. 12). The position of the ammonoid fauna in Berchogur in relation to the conodont zones is currently under investigation (see "Discussion").
The present ca. 180 specimens were mostly collected at the top of Member 3, either in trenches BK-1 and BK-3 and Locality 5 (Fig. 1) or as loose specimens on the surface. Besides fragments of indeterminable syringoporids s.l., the corals are mostly represented by small solitary and some, larger, dissepimented rugose corals. A short description of this fauna is provided here.
Proheterelasma omaliusi De Koninck, 1872 (Fig. 13f) is represented by a single solitary ceratoid coral, 4-5 mm in diameter and with 22 major septa grouped in four bundles joining the axial end of the cardinal columella.
Caninia cf. cornucopiae Michelin, 1840 in Michelin 1840-1847 ( Fig. 13n-q) is trochoid or trocho-ceratoid, 8-20 mm-high, 4-15 mm in diameter and with 22-26 major septa in mature stages. It displays thickened septa in the cardinal quadrants (convex side of the corals) and conspicuous septal furrows on the outer surface. The major septa are usually withdrawn in adult stages, but the counter septum is commonly elongated and reaches the axis where it can form a rudimentary axial structure with the upturned tabulae. The cardinal septum in the well-developed cardinal fossula is short. The minor septa appear as crests on the external wall. Dissepiments occur only in corals > 10 mm in diameter. These corals display a wide morphological variability but the presence of thickened cardinal quadrants, cardinal fossula, and elongated counter septa are relatively stable characters. We refer this population to C. cornucopiae with caution, as the material is poorly preserved, hampering the proper observation of internal characters, and because their size is relatively small. Similar corals occurring in the Sultsiferovyi (sulcifer) Horizon and Karakingir Beds (Upper Famennian) of Mount Aktas were described as Caninia cornucopiae by Ulitina (1975), whereas Volkova (1941) reported similar specimens under the name Laccophyllum fossulatum from the beds in 1941 assigned to the "lower Tournaisian" of the Shiderty River, Karaganda region of Kazakhstan. If Volkova's material could be re-examined, these specimens could be attributed to the species Canina fossulatum (Volkova, 1941); the illustration does not allow suitable comparison.
Rylstonia brevisepta Hudson, 1942 (Fig. 13i-l) is a small (< 10 mm in diameter, 10-15 mm-high, 18-22 major septa) ceratoid solitary coral with frequent constrictions and episodes of rejuvenescence. Its axial structure made of thickened axial plate and radial lamellae is prominent in the calices. Its dimensions and narrow dissepimentarium make it comparable to specimens of R. brevisepta described by Hudson (1942) (as a variety of the late Tournaisian species R. benecompacta Hudson and Platt, 1927) from the upper Courceyan of the British Isles. The columellate corals from the upper Kassinian Regional Substage (lower Tournaisian) of the Kyzylzhal Mountains of Kazakhstan, described as Rylstonia cystosa Keller, 1959, are probably a junior synonym of Hudson's (1942) species as well as Litvophyllum karakingiria Keller, 1959 andpossibly Dibunophyllum urunsaica Keller, 1959, also from the upper Kassinian, which differs from R. cystosa only in the development of few transeptal dissepiments.
Conilophyllum priscum (Münster, 1840) (Fig. 13b-e) is a cylindrical coral 10-14 mm in diameter, having weakly thickened 24-30 major septa. The minor septa are short and commonly contraclinant. The dissepimentarium comprises rare interseptal or transeptal dissepiments. The cardinal fossula is inconspicuous and the tabulae are flat, mesa-shaped. The outer wall is festooned. Despite the rarity of dissepiments, the material can be attributed with confidence to the Hastarian (lower Tournaisian) species Conilophyllum priscum of Western Europe, which commonly displays a wide range of variation (Poty and Boland 1994). The species was described as Kassinella longiseptata Keller, 1959 from the upper Kassinian Regional Substage (lower Tournaisian), reported from Mount Kyzylzhal in the Zhezkazgan (formerly Dzhezkazgan) area and identified as Guerichiphyllum kazakhstanicum Ulitina, 1975 from the Sultsiferovyi (sulcifer) Regional Substage (Upper Famennian) of Mount Aktas (Ulitina 1975). Both should be regarded as junior synonyms of Conilophyllum priscum.
"Permia" cavernula Hudson, 1943 (Fig. 13r-s) is ceratoid to patellate, up to 10 mm in diameter, having 24 major septa. In the early growth stages, the septa are long and rhopaloid and form an axial column with a narrow lumen. In later stages, the septa withdraw and form an aulos. Fedorowski (1971) synonymized Permia with Aulophyllum Milne-Edwards and Haime, 1850; hence a new generic name is required for these species formerly assigned to 'Permia'. Ulitina (1975) figured Amplexocarinia muralis from the Sultsiferovyi (sulcifer) Regional Substage (Upper Famennian) of Mount Aktas, that could be conspecific with the present material.
A last very eroded fragment of cylindrical coral (Fig. 13m) displays only the thickened tabular part of the major septa divided in four quadrants by four open fossulae. This single specimen is possibly a uralinid (or keyserlingophyllid) coral, but its preservation precludes a more precise attribution.
Of these eight coral taxa, only Molophyllum sp. has an Upper Devonian affinity as it was described from the uppermost Famennian deposits of the Omolon (Onoprienko 1979;Shilo et al. 1984). Conilophyllum priscum also occurs in the uppermost Famennian of Omolon (Poty 1999) but only in the lowermost Tournaisian (Hastarian) of Western Europe, where it is the guide-taxa of the RC1α subzone of Poty et al. (2006). Cyathaxonia cornu, Caninia cornucopiae, and uralinids typically occur in lower Tournaisian strata in Eurasia (Poty et al. 2006) whereas "Permia" cavernicula was collected from the lower Courceyan shales of Ireland, attributed to the RC2 zone. Columellate Caninia cornucopiae, Rylstonia brevisepta, and Proheterelasma omaliusi appear in the lower upper Tournaisian (Ivorian Substage of NW Europe) RC3α zone, but early occurrences in Central Asia cannot be excluded.
In summary, most of the coral fauna recovered from the Berchogur trenches point to a (lower or basal upper) Tournaisian age, with the notable exception of the possibly Devonian Molophyllum sp.

Crinoids
Numerous well-preserved crinoid columnals and pluricolumnals are found in the D-C boundary beds of the Berchogur section (mostly at the top of Bed 3c and loose). The collection is housed at the TsNIGR Museum, St. Petersburg, no. 12735. The assemblage is characterized by the family Floricyclidae ( Fig. 14a-e, h). This family was established by Moore and Jeffords (1968) based on crinoid stems from various Mississippian and Pennsylvanian localities of North America. It represents a morphologically well-defined taxonomic group with a clearly outlined pentalobate lumen surrounded by a perilumen, a wide areola, and a coarse crenularium. Later, members of this family were identified from the (mainly lower) Carboniferous of many regions of Kazakhstan, South Tien-Shan, South China, Mongolia, Transbaikal, Altai, South Urals, and Poland (Dubatolova and Shao 1959;Stukalina 1973;Yeltyschewa and Poljarnaja 1975;Dubatolova 1976;Głuchowsky 1981, etc.). In Kazakhstan, the earliest occurrences of this family are from the Sultsiferovyi (sulcifer) Regional Substage of the Upper Famennian and, in addition, they are common in the Simorinian Regional Substage, and also in Tournaisian and Viséan beds (Sisova 1979(Sisova , 1983Chernova and Stukalina 1989). In Transbaikal, members of Floricyclidae are also recorded in the uppermost Famennian (Kotikhinian Regional Substage) with a brachiopod assemblage including Sphenospira julii (Dehée, 1929), the index species of the uppermost Famennian Sphenospira julii-Spinocarinifera nigra Zone (Kurilenko et al. 2002;Kurilenko and Kulkov 2008), although the crinoid assemblage of this interval is mixed, as in addition to the earliest occurrences of the typical Carboniferous species it contains the latest occurrences of Late Devonian taxa. Species of Floricyclidae are also identified in the lower Tournaisian beds of Eastern Transbaikal containing a lower Tournaisian crinoid assemblage. In the Berchogur sections, we identified representatives of the genera Floricyclus Moore andJeffords, 1968 (F. kazangapicus Sisova, 1983 (Fig. 14a-e); Floricyclus sp.) and Lamprosterigma Moore andJeffords, 1968 (L. cf. welleri (Moore andJeffords, 1968)) (Fig. 14h). Floricyclus has an axial canal with long club-like lobes. Lamprosterigma is distinguished by a wide axial canal with short, weakly differentiated lobes. Chernova and Stukalina (1989) showed a trend of successive changes in morphological elements depending on the position of columnals in a stem. Distally, the lumen lobes become narrower, the areola increases, and the crenularium decreases. This trend is also observed in the material from Berchogur (Fig. 14a-c). The species Floricyclus kazangapicus is typical of the Tournaisian of South Kazakhstan (upper part of the Kassinian Regional Substage) and the lower part of the Rusakovian Regional Substage) (Sisova 1983;Chernova and Stukalina Fig. 14 Crinoids from the Berchogur (Birshogyr) sections; a-e Floricyclus kazangapicus Sisova, 1983: a, b articular facets of the proximal columnals (Loc. 5); c articular facet of a distal columnal (Member 3); d lateral view of a distal pluricolumnal (Member 3); e lateral view of a proximal pluricolumnal (Loc. 5); f, i Baryschyr sp. (Loc. 5): f articular facet of a columnal, i lateral view of a pluricolumnal: g, j, k Cyclocaudex cf. aptus Moore and Jeffords, 1968: g, j articular facets of the columnals (Loc. 5, loose), k lateral view of a pluricolumnal (Loc. 6); h Lamprosterigma cf. welleri (Moore and Jeffords, 1968): articular facet of a columnal (loosely collected in Trench BK-3, Member 3); l-n Platycrinites? cf. subtuberosus Stukalina, 1973 (loosely collected in Trench BK-3, Member 3). l, m articular facets of the columnals; n side view of a columnal; o-s Gilbertsocrinus? circumvallatus (Yeltyschewa in Dubatolova and Shao, 1959): o articular facet of a proximal columnal (Member 3); s articular facets of a distal columnal (Member 4); p-r lateral views of the proximal pluricolumnals (Member 3); t, u Taranshicrinus sp.: t articular facet of a columnal; u side view of a columnal (Loc. 5); v Goniostathmus sp.: lateral view of a pluricolumnal (Member 3) ◂ 1989) and Floricyclus cf. kazangapicus is known from Eastern Transbaikal (Pavlovian Regional Substage) (Kurilenko et al. 2002). In Transbaikal, deposits of the Pavlovian Regional Substage contain an abundant lower Tournaisian fauna of bryozoans, brachiopods, and crinoids (Kurilenko et al. 2002;Kurilenko and Kulkov 2008). The genus Lamprosterigma is found in the lower Carboniferous of Kazakhstan and North America. Lamprosterigma welleri (Moore and Jeffords, 1968) was described from the lower Mississippian of Kentucky (Brodhead Formation, Borden Group, Osagean) and from the Viséan of Holy Cross Mountains, Poland (Moore and Jeffords 1968;Głuchowsky 1981;Chernova and Stukalina 1989).
Columnals of the family Platycrinitidae Austin and Austin, 1842 (Fig. 14l-n) are easily identified. They are bilaterally symmetrical, subdivided by a fulcral ridge into two halves with an areola in each (Stukalina 1973;Dubatolova 1976;Kurilenko 1989;Kurilenko et al. 2002, etc.). The genus Platycrinites Miller, 1821 was established based on calices; hence, when it is identified from columnals, a question mark is traditionally placed next to the generic name. Members of the family are widespread in the Carboniferous of many regions of Kazakhstan, Central Asia, Russia, and North America. The species from Berchogur, similar to Platy crinites? subtuberosus, is distinguished from typical specimens of this species in the presence of columnals of two orders and their smaller size. In Transbaikal, Platycrinites? subtuberosus is typically found in the D-C boundary beds in association with Gilbertsocrinus? circumvallatus, which is known from the Tournaisian of Western Siberia, Kuznetsk Basin, southern Transcaucasia, Altai, Kazakhstan, Far East, Mongolia, and South China (Stukalina 1973;Dubatolova 1976, etc.). The assemblage also includes Goniostathmus sp. (Fig. 14v), a genus also known from the Upper Famennian of Central Kazakhstan and Lower Mississippian of Iowa (Burlington Limestone, Osagean) (Moore and Jeffords 1968).
Columnals of Cyclocaudex, Floricyclus, Gilbertsocrinus and Platycrinites (mostly identified to genus) have been previously recorded from the Tournaisian (sulcata Conodont Zone of Pilton Beds, Fremington Quay, North Devon (England)) (Głuchowsky and Racki 2005) in an assemblage similar to those of Poland, Czech Republic, North America, and Kazakhstan. In summary, the crinoid assemblage from the Berchogur sections supports an early Tournaisian age for the host deposits, as indicated by the conodonts, ammonoids, and corals.
Trilobites collected in 2019 come from Member 3 (Trench BK-3) of the Berchogur sections. Previously, a few trilobites were found in the member and were identified as Archegonus (Waribole) sp. (Maslov 1987, p. 48;Barskov et al. 1988, p. 175) (this collection is presumed to be lost). Additional trilobites that were not identified were recovered from the yellowish-grey limestone with Tournayellina pseudobeata (Bed 4, Unit IX, "Outcrop 8"), which is located 3 km south of the Burtybai section of Maslov (1987, p. 27).
Our present collection contains two rock samples with trilobites. The first contains an isolated pygidium, the second is a small rock fragment with exuviae of a pygidium, cranidium, and possibly of a free cheek (Fig. 15). In our opinion they are identical to those that were assigned by Balashova (1956) to Phillipsia carringtonensis and Cyrtosymbole (Waribole) baiburensis, which also were found as aggregated group of remains in the Kust-Kara locality (south of Burtybai Creek). Cranidia of all these taxa have a moderately narrow glabella with a very weakly pronounced constriction and a smoothly rounded front margin. The glabella surface, mainly centrally, is covered with small tubercles, which in the studied cranidium were only preserved in the posterior right corner of the glabella (Fig. 15d). An important character is a very wide pre-glabella field with relatively straight front margin. In addition, the fixigenae are also relatively wide. The pygidia are also similar in our specimen from the aggregation of remains; the lateral border is very wide, and the pleural ribs end far from the edge of the pygidium. The number of rings of the axis is 10-11 (Fig. 15a), whereas in Balashova's specimens it is 11-12, and there are 8-9 pleural ribs. Hahn and Hahn's (1972, p. 344) revision states that the cranidium identified by Balashova as Cyrtosymbole (Waribole) baiburensis, most likely belongs to Cummingella carringtonensis, but later Hahn and Hahn (2008, p. 80) excluded Balashova's specimens from C. carringtonensis. It is possible that the cranidium recognized by Balashova (1956) as the new species Phillipsia coronata also belongs to the above species. She described two small incomplete cranidia (ca. 3 mm long) that come from a different locality in the upper reaches of the Shyuldak River. Based on morphology, especially the position of furrows on the glabella, its shape, and raised cornea layer, it is possible to interpret Ph. coronata as young specimens of C. carringtonensis. The generic assignment of Ph. coronata remains debatable: Osmólska (1970, p. 94) tentatively assigned this species to the genus Piltonia, whereas Hahn and Hahn (1972, p. 391) assigned it to the subgenus Phillipsia (Phillipsia). The analysis of Hahn and Hahn's (2008) catalogue shows that the Berchogur trilobite is most similar to early Tournaisian species of the subgenus Liobolina (Liobolina) Richter and Richter, 1951. In having a considerable wide preglabellar field, they are most similar to the common species Liobolina (Liobolina) nebulosa Richter andRichter, 1951 (Hahn andHahn 2008, text-figs. 220-222), from the lower Tournaisian (Hangenberg Limestone), of the duplicata/sandbergi conodont zone of Germany, Austria, and England. However, in that species, the glabella is smooth, in contrast to the Berchogur taxon, which displays a prominently ornamented cranidium. More specimens are needed to identify the Berchogur trilobites more precisely, but until that, this taxon can be provisionally identified as Liobolina (Liobolina) coronata (Balashova, 1956).  (Balashova, 1956); Member 3, lower Tournaisian: a pygidium, specimen PIN, unnumbered; b limestone fragment with group of exuviae of trilobites; c pygidium, specimen PIN, unnumbered; d cranidium, specimen PIN, unnumbered; scale bar 5 mm

Discussion
Renewed interest in the Berchogur sections was a result of searching for and testing of new criteria to correlate the proposed levels for the D-C boundary in regions where it is not possible to use the FAD of the conodont Protognathodus kockeli. This conodont species is common in Western Europe, where it enters slightly above the main regressive phase represented by the "Hangenberg Sandstone Event" (Middle Crisis Interval), i.e., at the base of the Lower Stockum Limestone = base of the "Upper Crisis Interval" ). This level is above the Hangenberg Black Shale (= main part of the Lower Crisis Interval) separated from it by the Hangenberg Shale/Sandstone (= Regression Interval). The search for the new proposed boundary level, identified by the end of the "Hangenberg Sandstone regressive event" and near the FAD of Pr. kockeli is a current task of the working group of the Devonian-Carboniferous boundary (see references in Becker et al. 2016 andKaiser et al. 2016). Protognathodus kockeli, like all species of the genus Protognathodus, is extremely rare in the Middle and South Urals and Kazakhstan. In this situation, correlations might be possible by identifying coeval faunas among other groups of organisms and correlating them and finding traces of transgressive-regressive episodes. Previous studies focussed on the first appearance datum of the conodont S. (Eo). sulcata (see Barskov et al. 1984), which was then considered as a potential boundary marker, but was later devalued due to its facies dependence (absence in shallow-water intervals and extreme rarity in pelagic Stockum Limestone facies) and an earlier occurrence in La Serre (Kaiser 2009). The mostly shallow-water facies of the Berchogur sections contains a diversity of fossils allowing the interpretation of the biota and environments of the South Urals-western Kazakhstan basins at the D-C boundary. The base of the section is represented by shallow-water algal-foraminiferal limestones with stromatoporoids (lower Zhangana Formation, Member 1) overlain by an alternation of shales and marly limestones (Member 2), then by three beds of limestone (Member 3, 45 cm thick in total in Trench BK-3), in the upper part with ammonoids, mostly of the genus Acutimitoceras, and then by shale with rare interbeds of limestone (Member 4). It is not possible to document the occurrence of Pr. kockeli in this section, but both the S.  Barskov et al. 1984;Simakov et al. 1985; although later excavations did not confirm the occurrence of two separate ammonoid levels). Main ammonoid, coral and crinoid occurrences are located at the top of Bed 3c. This level possibly correlates with the sulcata FOs at Seiler, Müssenberg and Hasselbachtal in the Rhenish Massif (Paproth and Streel 1985;Clausen et al. 1989;Bless et al. 1993;Kaiser et al. 2006;Becker et al. 2016;etc.). The correlation of the FO of S. (Eo.) praesulcata with the FO of Pr. kuehni (Upper Protognathodus Fauna of Alberti et al. 1974) was proven in some sections, for example at Seiler (Clausen et al. 1989;compare Kaiser et al. 2019). In the Berchogur Depression, no ammonoids are found within the praesulcata zones (Members 1 and 2, B. ultimus ultimus Zone to ckl), which is characterized by shallowing upwards within Member 2a, a sharp deepening at the base of Member 2b (most likely the transgressive Lower Crisis Interval and Hangenberg Black Shale equivalent). We draw the base of the S. (Eo.) sulcata Zone in Trench BK-1 and Trench BK-3 at the base of Member 3. The presence of carbonates (mudstones) with ammonoids in Bed 3c represents the continuation of a transgressive trend after the end of a regressive episode in the upper part of Member 2b (Simakov et al. 1985). An important element in understanding this succession is the occurrence of reworked material forming the gravelstone horizon (between samples 5 and 6) in Member 2a below the limestones forming Unit 2a 2 . Limestones of Members 3a-b, in addition to conodonts, contain foraminifers (Tournayellina and Paracaligelloides) together with the alga Garwoodia (trenches BK-1 and BK-3). The latter suggests a shallow-water depositional environment. Considering that the Acutimitoceras fauna is found only within Member 3c, it is possible to suggest a quick succession of changes in bathymetry. This agrees with the data from the European successions, where a small-scale regression (Hangenberg Sandstone regression) is presumed near the FAD of S. (Eo.) sulcata. It is presently known that the first appearance of the conodont S. (Eo.) cf. praesulcata is in Member 2b approximately one metre below Member 3, and the interval of ammonoids in Unit 3c shows the presence of S. (Eo.) sulcata in its upper part (Barskov et al. 1984(Barskov et al. , 1988Maslov 1987; our own observation). By correlation based on sea-level changes, the FAD of Pr. kockeli could be expected in the uppermost beds of Member 2b (dark shale unit above the limestone with neritic faunas), and we draw the base of the sulcata zone at the base of Member 3 slightly below the FO of S. (Eo.) sulcata. In any case, the upper part of Member 3c can be correlated with the upper part of the prorsum Zone, as S. (Eo.) sulcata enters within the prorsum Zone in Western Europe (Korn and Weyer 2003;Clausen and Korn 2008). It is also possible that beds of Member 4 correlate with higher horizons of the lower Tournaisian, but so far ammonoid data for Member 4 (a series of shale and marly limestones with another level with Acutimitoceras) are inconclusive. This higher level could correspond to the Gattendorfia Zone, but so far, no gattendorfiids were found in this section, possibly due to the very poor exposure of this level. In this respect, it would be important to know the precise level of A. (Acutimitoceras), which is so far only known from Gattendorfia levels of Germany. The correlation of Unit 3c with the upper Stockum Limestone (Tournaisian) is supported by the presence of basal Tournaisian corals and crinoids. It is evident that the succession shows small-scale depositional cycles that can be roughly correlated with those in Western Europe, as episodes dominated by siliciclastic deposition alternated with short events of carbonate accumulation.

Conclusions
The re-excavation and re-examination of the Berchogur sections revealed an interesting history of evolution of a shallow-water mixed terrigenous-carbonate basin showing cycles of limestone and shale deposition around the Devonian-Carboniferous boundary in Western Kazakhstan. The integrated approach allows its correlation with the up-todate stratigraphic schemes of the Rhenish Massif and other regions and the rough locality of the future D-C boundary in the absence of conodont markers, including Pr. kockeli. This agrees with the currently prevalent approach to search for a composite set of criteria denoting the D-C boundary, rather than a single species that may or may not be present in remote successions. Based on the identification of a regressive event that most probably records the "Hangenberg Sandstone event", and with the clear distinction of Devonian and Carboniferous faunal associations (foraminifers, conodonts, ammonoids, etc.), the future D-C boundary is likely to be within the upper part of Member 2b, somewhat below the base of Member 3, as this is the end of the regressive episode and beginning of a new transgression. The current D-C boundary is located within Member 3c and is drawn by the FO of S. (Eo.) sulcata. Member 3c contains ammonoids similar to those of the A. prorsum Zone of the Rhenish Stockum Limestone, which ranges from the Pr. kockeli to the basal S. (Eo.) sulcata zones. Crinoids and corals, which are published for the first time from Member 3, show clear lower Tournaisian affinities suggesting that the new Mississippian benthic faunas appeared at Berchogur as early as in the S. (Eo.) sulcata Zone. This pattern agrees with observations in other regions (Denayer and Webb 2017).