First occurrences of selected foraminifers and their interpreted phylogenies in the uppermost Arundian Substage of the South Cumbria Shelf, northern England, allows the distinction of four biostratigraphic events. Event A1 is characterised by the occurrence of occluded Nodosarchaediscus, Consobrinellopsis, and Archaediscus krestovnikovi and is assigned to the ‘upper’ Cf4δ subzone. Event A2 is characterised by the first occurrence of Pojarkovella ketmenica and Archaediscus at concavus stage (including the first A. pauxillus) and is assigned to the base of the Cf5α subzone. Event A3 is characterised by Koskinotextularia aff. cribriformis, Koskinotextularia sp., Archaediscus moelleri, Endothyranopsis compressa, and Omphalotis minima and is assigned to the base of the Cf5β subzone. Event A4 is characterised by the first occurrence of Koskinotextularia cribriformis, Koskinotextularia bradyi, Koskinotextularia obliqua, and Pojarkovella nibelis morphotype 2. The base of the Cf5α subzone likely coincides with the base of the Livian in Belgium and Tulian in the Russian Platform, although the occurrence of hiatuses and facies barren in foraminifers leads to uncertainty in correlation. The Cf5β subzone is correlated with that part of the Livian and Russian Platform succession with preserved foraminifers and thus can be used for worldwide correlations. However, formally, it cannot be confirmed if this horizon is an isochronous level (due to hiatuses and hostile facies), and it is necessary to locate a slightly higher level of potential correlation that better displays the absence or presence of the successions, such as event A4.

Subdivision of long chronostratigraphical stages is a debatable issue—some authors prefer retaining long-established stages, now known to be of great length in chronometric time, for reasons of nomenclatural stability. Hence, any finer-scale chronostratigraphic subdivision would be based around precise worldwide correlation of regional substages or the establishment of formal global substages. The Viséan is the longest stage in the Carboniferous (Poty et al., 2014; Lucas, 2021; Pointon et al., 2021), between 15.8 Myr to 16.39 Myr in duration (sensu Davydov et al., 2012 and Aretz et al., 2020, respectively), and the second longest stage in the Phanerozoic. In Europe, the most important regional substages of the Viséan were defined in the Russian Platform or East European Platform (EEP), Ukraine (from the Donetz Basin), Belgium (Namur-Dinant Basin) and Britain. Numerous substages were also defined in different Russian regions (e.g., Kuznetsk Basin, Timan, Kolyma-Omolon; see Alekseev et al., 2022, fig. 2), although currently the most widely used are those mostly based on the Russian Platform or EEP, and those from the Western and Eastern slopes of the Urals. Even in the Western slope there are sub-regional stages for the southwestern outcrops (Reitlinger et al., 1996; Alekseev, 2008; Alekseev et al., 2022; Kulagina, 2022).

Poty et al. (2014) proposed to subdivide the Viséan worldwide into three new stages, using the Belgian substages, Moliniacian, Livian, and Warnantian, replacing the Viséan as a stage. Informally, this tripartite subdivision is known as lower, middle, and upper Viséan in Western Europe, which contrasts with the bipartite subdivision of the Viséan in Russia, lower and upper Viséan (Fig. 1). Owing to this dual lithostratigraphic usage, the studied interval here will be informally referred to as the mid-Viséan. The bases for the Belgian substages correspond approximately with the bases of the Chadian, Holkerian, and Asbian substages in Britain and with the Radaevkian, Tulian, and Aleksinian of the EEP (Fig. 1).

The base of the Moliniacian corresponds with the base of the Viséan, since it is the formally defined Viséan GSSP (Devuyst et al., 2003), and coincides with the Radaevkian, Glubokinsky, and the redefined base of the Chadian (sensu Riley, 1994; Fig. 1).

The base of the Warnantian possibly corresponds to the base of the upper Viséan in Western Europe, although this subdivision and its lateral equivalents are rather problematic. In Western Europe, the Asbian Substage, defined in northern England, is classically correlated with the base of the Warnantian Substage in Belgium, a correlation that is not reliable. Firstly, the base of the Warnantian is located in the Thon-Samsom Member, which is sandwiched between hostile facies for foraminifers (stromatolitic evaporates), and thus, assemblages are extremely poor (Poty & Hance, 2006b). Therefore, in terms of faunal records, the scenario for the base of the Warnantian is to situate it between two hiatuses. Secondly, the base of the Asbian stratotype section at Little Asby Scar (east Cumbria) is a purely lithological boundary, possibly misidentified (Waters et al., 2021). Thirdly, important macrofossil occurrences have not been re-found in these key boundary beds in Britain, and the basal foraminiferal assemblage based on the Cf6 or Asperodiscus (= Neoarchaediscus) Zone (Conil et al., 1977a, 1980) occurs from much older levels (Cózar et al., 2022b). Consequently, there is no guide for the base of the Asbian as a chronostratigraphical unit, as it is currently defined, and its base should be moved down to coincide with the start of a suitable foraminiferal zone (Fig. 1). Fourthly, there is no clear basal Asbian or Warnantian equivalence with the Russian or Ukrainian substages, although commonly these are correlated with the base of the Aleksinian Substage (e.g., Brenckle, 2004; Kabanov et al., 2016; Aretz et al., 2020), but the foraminiferal markers recorded in the Aleksinian Substage are usually recorded in the late Asbian of Western Europe (e.g., Cózar et al., 2022b), whereas in Ukraine, it may be correlated with the midparts of the Donetsky Stage (e.g., Davydov et al., 2010; Fig. 1).

Therefore, currently, the most promising level for consistent inter-regional correlation is lower in the Viséan, at the mid-Viséan, at around the Holkerian-Livian-Tulian-Styl’sky basal boundaries (Fig. 1), which in general terms, present more uniform foraminiferal assemblages. Nevertheless, the Holkerian, as was originally defined by George et al. (1976), and the Livian, as defined by Conil et al. (1977a), and likewise the Tulian, Kurtymsky, and Styl’sky also present some problems.

The most similar faunistic succession to that recorded in northern England is generally the neighbouring Belgian zonal scale, although the precise characterisation of the Belgian Livian substage (Fig. 1) is hampered by the presence of dolomitised intervals, stromatolitic beds barren of fauna, and the paucity of faunal records at the base of the Livian (Poty et al., 2002). According to the chronometric age model of Pointon et al. (2021), using U-Pb dates from bentonites (altered volcanic ashes), there is some 1.25-Myr duration between the base of the Lives Formation (in the Livian stratotype section) to the first occurrence of Pojarkovella nibelis (Durkina, 1959; 14.3 m above the formation base), implying a large faunal gap. The base of the Lives Formation (including the Banc d’Or de Bachant bentonite) is formally considered the base of the Livian (Poty & Hance, 2006a). Furthermore, the base of the MFZ12 foraminiferal zone (Poty et al., 2006, renamed from Cf5 Zone of Conil et al., 1977a; Fig. 1) is located at the first occurrence of P. nibelis, and hence, the underlying 14.3 m of carbonates and bentonites should be formally included in the MFZ11 Zone. Significantly, this faunal gap in the basal Lives Formation also includes the top part of the underlying Moliniacian (Poty et al., 2014), indicating no direct correlation is possible between lithostratigraphical units and foraminiferal zones defined in Britain and Belgium for this interval.

In northern England, the South Cumbria Shelf (SCS) is a key platform region where rich foraminiferal assemblages are found in Viséan carbonates, which despite some local issues of dolomitization and clastic input, form a more or less complete faunal succession. Hence, this region is of major significance for the inter-comparison with platform successions elsewhere in Europe. In addition, this region contains the Holkerian stratotype section at Barker Scar (Fig. 2).

This section was originally characterised by the first occurrence of foraminifers assigned to the Cf5 Zone in bed K (George et al., 1976; Ramsbottom, 1981), above 25 m of dolostones, dolomitic limestones, and foraminifer-poor clastic-rich units and limestone beds (Fig. 3). From Barker Scar, Cózar et al. (2022a) questioned the positioning of the stratotype level, since typical foraminifers attributed to the Cf5 Zone occur in limestones below bed K, so repositioning the base of Cf5 at a horizon 14 m lower in the section. The search for a better Holkerian stratotype section is ongoing, but any new selected base should be located at the base of the Cf5β foraminiferal subzone as discussed in Cózar et al. (2023; Fig. 3). Currently, the formal base of the Holkerian is still in bed K at Barker Scar (Fig. 3, right columns), and thus, the strata analysed here currently belong to the uppermost Arundian. Hounslow et al. (2022) have revised the upper Arundian and lower Holkerian lithostratigraphy in the SCS, of which, five stratigraphic sections within the upper half of the Raven’s Member (upper Dalton Formation) are detailed here: Barker Scar, White Scar, Grubbins Wood, Blackstone Point, and Low Frith (Figs. 26). Lithostratigraphical and geographical details of the sections can be found in Hounslow et al. (2022).

Comparatively little attention has been paid to the foraminifers in the mid-Viséan, as most biostratigraphic studies give generalised lists of foraminifers, in some cases accompanied by illustrations (but not always). In addition, many sections are affected by problems of dolomitization and faunistic gaps due to hostile facies or hiatus, and thus, only the first occurrence of some taxa were published, whereas the complete study of possible phylogenies and precise relative position of each species have been never undertaken for this interval. This fact has led to a poor knowledge of the foraminiferal assemblages about the mid-Viséan, and particularly for the concept of some species. The more or less continuous succession in the SCS provides an exceptional succession to study the phylogenetic lineages of several genera that suffered different evolutionary processes around the mid-Viséan boundary, so providing data for establishment of potential chronostratigraphic horizons.

Our aim is to determine the first occurrences of foraminifers in these five sections from the uppermost Arundian, in the type region (where the Holkerian Substage is defined), in order to better understand their potential phylogenetic lineages and so help define foraminiferal assemblages suitable for international correlations. This approach provides improved knowledge of the basal Holkerian and its international equivalents leading to an improved chronostratigraphic worldwide correlation in the mid-Viséan.

The Arundian Substage, as defined by George et al. (1976), contains the Cf4β, Cf4γ, and Cf4δ foraminiferal subzones defined by Conil et al. (1977a, 1980, 1991), which provide an informal subdivision into lower, middle, and upper Arundian, respectively (Fig. 1). These subzones correspond to the MFZ10 (lower half of Cf4β) and MFZ11 (upper half of Cf4β and the entire Cf4γ–Cf4δ) of Poty et al. (2006; Fig. 1). The MFZ11 combines several subzones, which, as shown by Cózar et al. (2020), can be recognised in many sections, and hence the zonal amalgamation into MFZ11, if used more widely, would be a loss of stratigraphic definition. These subzones include: the MFZ11α (upper half of Cf4β), MFZ11β (Cf4γ), and MFZ11γ (Cf4δ) (Fig. 1).

The Dalton Formation of south Cumbria mostly contains assemblages of the Cf4δ subzone, and this subzone is also recorded in the underlying Red Hill Limestone Formation (Hounslow et al., 2022). Some of the more representative foraminifers were illustrated in Hounslow et al. (2022, figs. 15–16) and Cózar et al. (2022a, fig. 3). Assemblages are composed of common species of Ammarchaediscus, Archaediscus at involutus stage (evolutionary stage sensu Conil et al., 1980), Brunsia, Conilidiscus, Eoparastaffella, Eosinopsis, Glomodiscus, Lapparentidiscus, Omphalotis, Plectogyranopsis, Paralysella, Pseudolituotubella, and Uralodiscus. The Cf4δ subzone is biostratigraphically characterised by the first occurrences of Cribranopsis, Endospiroplectammina syzranica (Lipina, 1948), Lituotubella glomospiroidesRauzer-Chernousova, 1948a, Latiendothyranopsis solida (Conil & Lys, 1964), Omphalotis aff. minima (Rauzer-Chernousova and Reitlinger in Rauzer-Chernousova et al., 1936) sensu Cózar et al. (2022a), large Plectogyranopsis (P. moraviae Conil & Longerstaey in Conil et al., 1980, and P. settlensis Conil & Longerstaey in Conil et al., 1980), as well as primitive and questionable forms of Nodosarchaediscus (e.g., Nodosarchaediscus? cornua (Conil & Lys, 1964), N.? viaeVachard, 1977, N.? sp. 1 sensu Cózar et al., 2022a).

At the base of the upper Raven’s Member of the Dalton Formation, Endothyranopsis aff. compressa (Rauzer-Chernousova & Reitlinger in Rauzer-Chernousova et al., 1936) sensu Cózar et al. (2022a) occurs with more evolved/occluded species of Nodosarchaediscus [N. exiguus (Bozorgnia, 1973), N. demaneti (Conil & Lys, 1964), N. hirta (Conil & Lys, 1964) and N. tchaboksarensis (Bozorgnia, 1973)]. In the upper part of the Cf4δ subzone, Consobrinellopsis and the most occluded forms of Nodosarchaediscus occur [N. pirleti (Bozorgnia, 1973), N. rostratus (Bozorgnia, 1973), N. saleei (Conil & Lys, 1964), N. tchalussensis (Bozorgnia, 1973) and N. conili (Bozorgnia, 1973); Figs. 36]. These taxa suggest a potential subdivision of the Cf4δ subzone into lower, middle, and upper units (Hounslow et al., 2022; Fig. 1). A similar progression in the occlusion of Nodosarchaediscus in the upper Moliniacian of Belgium was observed by Laloux (1987).

Four biostratigraphic events with important new first appearances are defined in the sections below the classical base of the Holkerian, A1 to A4 (Table 1). Event A1 in Grubbins Wood and Barker Scar is recognised about 2 m below the mid-clastic unit of the Raven’s Member (Figs. 34), whereas at White Scar it occurs at the level of these clastics (Fig. 5) and in Low Frith and Blackstone Point, it is developed about 2 m above (Fig. 6). Event A1 is characterised by the first occurrences of highly occluded Nodosarchaediscus, Consobrinellopsis, and Archaediscus krestovnikovi. It is interpreted to be the base of the ‘upper’ Cf4δ, and event A1 is always located close to this horizon of clastic input.

Event A2 is characterised by the first occurrences of Archaediscus at concavus stage (oblique sections and rare A. pauxillus also) and Pojarkovella ketmenica (very rare). This level is recognised in the middle of bed B at Barker Scar (Fig. 3), and c. 4–5 m above the mid-clastic interval in the Raven’s Member in Grubbins Wood, White Scar, and Blackstone Point (Figs. 46). Interestingly, this event also coincides with the earliest Lituotubella magna in the Grubbins Wood section (Fig. 4). This horizon is considered the base of the Cf5α subzone (Cózar et al., 2020). Unfortunately, the rarity of Archaediscus specimens does not allow precise species identifications. This fact implies that the boundaries selected in Cózar et al. (2022a) were imprecisely located due to poor preservation in the then studied thin-sections, and thus, scarcity in foraminifers. The new material has revealed that the base of the Cf5α subzone lies within bed B at Barker Scar (not at the base of the overlying bed C, as previously suggested; Fig. 3).

Event A3 is the most widely recognizable event with more consistent taxa. The base of this event is primarily characterised by primitive Koskinotextularia (K. aff. cribriformis), Endothyranopsis compressa, Omphalotis minima, and common Archaediscus pauxillus. Less than 1 m above, a wide variety of taxa first occurs, such as Pojarkovella occidentalis, Archaediscus moelleri, Pojarkovella pura, and Ugurus intermedius. More rarely, Pojarkovella nibelis morphotype 1 also occurs, although it is more common higher up. Pojarkovella ketmenica is rare in older horizons and becomes a common taxon from the base of A3. Similarly, more species of Archaediscus at concavus stage are recorded from A3. In higher levels, other species of Pojarkovella occur, as well as more typically, Lituotubella magna, and the first Endostaffella fucoides occur. This event is interpreted as the base of the classical Cf5 (or MFZ12) zone, now considered as the Cf5β or MFZ12β subzones (Cózar et al., 2020), and is situated at the base of bed C at Barker Scar.

Event A4 is recorded in the middle of bed C at Barker Scar and is primarily characterised by the first occurrence of evolved Koskinotextularia, such as K. cribriformis, followed 1–2 m above by K. bradyi and K. obliqua. In this interval, Pojarkovella nibelis morphotype 2 is commonly recorded, although in White Scar (Fig. 5) this species is recorded 0.3 m below the first K. cribriformis. This part of the succession is also characterised by the first occurrences of Holkeria topleyensis, H. aff. avonensis, Endostaffella delicata, and Vissarionovella holkeriana, although these taxa are always rare.

Higher up in the succession, in bed K from Barker Scar, the formally defined base of the Holkerian (George et al., 1976), are more evolved species of Holkeria (H. avonensis, H. daggeri), although it cannot be tested yet if these occurrences are consistent in other sections or if these first occurrences are an artefact due to strong dolomitization and hostile facies in beds D to J in the section. However, in White Scar Quarry, beds equivalent to I–J are well-preserved limestones (Fig. 5), and no Holkeria has been recorded, which might suggest its low potential for correlation. Other records of Holkeria (Strank, 1982), are located always far above the base of this substage, such as 34 m above the base of the Fawes Wood Limestone Formation in Beckermonds Scar Borehole, or even younger in the Garsdale Limestone Formation in the ‘late’ Holkerian in Raydale Borehole and River Clough (incorrectly attributed to the Fawes Wood Limestone by Strank, 1982) in the North Askrigg area. A similar record is in the upper Holkerian in the Wood Dale Formation in Derbyshire, as well as in upper parts of the Holkerian in the South Askrigg block (Waters et al., 2017). Other records of Holkeria are in the Asbian.

In the middle part of Barker Scar at bed T, the occurrences of Pseudoendothyra, ‘MillerellaexcavataConil & Lys, 1964, Globoendothyra globulus (von Eichwald, 1860), and Archaediscus at concavus transitional to angulatus stage seem to be an important horizon for correlation, which was defined as the base of Cf5β2 by Cózar et al. (2022a), a level which has been also recognised in Grubbins Wood above a covered part of the section. In these younger levels, evolved species of Eostaffella are also recorded, such as E. parastruveiRauzer-Chernousova, 1948d, and E. mosquensisVissarionova, 1948.

Eleven genera show similar patterns of first occurrences in the studied sections and can be used in defining chronostratigraphic horizons for the upper Arundian. These include Koskinotextularia, Archaediscus, Endothyranopsis, Omphalotis, Pojarkovella, Lituotubella, Endostaffella, Holkeria, Vissarionovella, Ugurus, and Cribrospira. The occurrence of typical taxa from the underlying early Viséan beds is a notable feature, having been recorded in all the five sections studied. These latter taxa include most species of Glomodiscus (except for G. rigensConil & Lys, 1964, which extends up to the late Viséan), Conilidiscus, Uralodiscus (except for U. adindaniiBrenckle & Marchant, 1987, which also extends up to the late Viséan), Lapparentidiscus, Lysella, Paralysella, Pseudolituotubella, Chernobaculites, Septabrunsiina, Spinolaxina, and most species of Eoparastaffella.

Koskinotextularia

Possibly the most important lineage for recognition of the mid-Viséan boundary, with the clearest taxonomic features, is that recorded in the genus Koskinotextularia, with biserial pairs of chambers and a cribrate aperture, arising from Consobrinellopsis, with a simple basal aperture (Fig. 7.1). These are represented in the upper part of the Cf4δ subzone in the SCS (Cózar et al., 2020; Hounslow et al., 2022). Two primitive species first occur, Koskinotextularia aff. cribriformis sensu Cózar et al., 2022a (from the base of Cf5β), and rarely, K. sp. A (in slightly younger levels; Fig. 7.2, 7.5). They are characterised by a rudimentary wall (with coarse grains) and cribrate aperture (with a few holes), as well as an irregular arrangement of the biserial pairs of chambers. Koskinotextularia sp. A shows a higher number of chambers, although it is less common, and is only recorded in the Grubbins Wood section, where it occurs close to the top of the section (Fig. 4). Koskinotextularia sp. A resembles an immature specimen of K. strictum (Conil & Lys, 1964), where the final distorted pairs are not observed, although only a holotype was illustrated. The species is not endemic to Britain, because it is recorded from the mid-Viséan in the Brabant Massif (e.g, Bless et al., 1976, pl. 8, fig. 1). These primitive forms evolved to more typical and well-formed species of Koskinotextularia a few metres above, where successively, K. cribriformisEickhoff, 1968, and K. bradyi (von Möller, 1879) occur (this latter species is characterised by the thickening of the wall; Figs. 7.37.4). Other species arose slightly later, such as K. obliqua (Conil & Lys, 1964), which usually show two distorted final pairs of chambers (Fig. 7.6). The possible lineage is: Consobrinellopsis lipinae (Conil & Lys, 1964; ‘upper’ Cf4δ) → Koskinotextularia aff. cribriformis (base Cf5α) → K. cribriformis (close to the base Cf5β) → K. bradyi (close to the base Cf5β). Koskinotextularia obliqua probably arose from K. sp. A, which share similar morphologies, but the latter presents more rudimentary aspects (Fig. 7). The evolution of the family Palaeotextulariidae in Western Europe contrasts notably with that in the EEP, Urals, and Kazakhstan. In the latter regions, genera with a fibrous wall (e.g., Palaeotextularia, Cribrostomum), first occur nearly at the same time as those genera without a fibrous wall (e.g., Consobrinellopsis and Koskinotextularia) during the Tulian (e.g., Makhlina et al., 1993; Brenckle & Milkina, 2003; Kulagina, 2022), whereas the occurrence of those taxa with a fibrous wall in Western Europe is not recorded until the Asbian or Warnantian.

Archaediscus

The archaediscids are one of the main foraminiferal families used to subdivide the Viséan (Conil et al., 1980), and some taxa are also important for recognition of the mid-Viséan. In the underlying levels of the studied succession, the Archaediscus at involutus stage (cf. Conil et al., 1980) are predominant (Fig. 8), whereas those in the Cf4δ subzone show a predominance of Archaediscus at involutus transitional to the concavus stage. The common sigmoidal Archaediscus teresConil & Lys, 1964 (Figs. 9.19.2), a transitional form between the involutus and concavus stages, evolved into Archaediscus pauxillusShlykova, 1951 (Figs. 9.49.5), at concavus stage (with the base of the lumen in the final whorl undulose). Conil et al. (1980) and Brenckle & Grelecky (1993) suggested that this species might be the juvenile of other species, such as A. moelleriRauzer-Chernousova, 1948c, A. convexus Grozdilova & Lebedeva in Grozdilova, 1953, or the juvenarium of A. koktjubensisRauzer-Chernousova, 1948d. For the dimensions of A. pauxillus, its relationship with A. koktjubensis does not seem plausible, whereas A. convexus has been first recorded from younger strata. However, this species has been recorded together with Archaediscus moelleri (Figs. 9.69.9) in White Scar Quarry (Fig. 5), but in older levels in the other sections. Archaediscus moelleri seems to be an evolved species (more rapidly expanding coiling leaving high-open lumina, thickened fibrous wall in the inner whorls), with a moderately-sized test (larger than A. pauxillus) and a larger proloculum. These features, although subtle, can be observed from the juveniles of A. moelleri, and are slightly different from those assigned here to A. pauxillus (compare Figs. 9.49.5 with 9.69.7). The concavus sutures are variable; in A. teres, being present in only 2 or 3 whorls in older levels of the early Viséan, whereas the specimens which occur in the mid-Viséan, present 4 or 4.5 whorls with concavus sutures. In such cases, the species is rather difficult to distinguish from the first A. pauxillus, a fact which questions the possible presence of this taxon in older levels (Figs. 9.3, 10.1–10.2). This earlier occurrence is in agreement with Brenckle & Milkina (2003), who documented A. pauxillus from rocks assigned to the Bobrikian. Although there are Archaediscus at involutus stage with sigmoidal coiling and a macrospheric proloculum, transitional forms between the involutus and concavus stages have not been recorded, and thus, it is assumed that A. moelleri arose directly from A. pauxillus (Fig. 9). We follow the recommendation of Brenckle & Grelecky (1993) to keep A. pauxillus independent from A. moelleri, mostly due to this small difference in the stratigraphic range of both species and the slightly more “evolved” characters in A. moelleri.

The discoidal Archaediscus krestovnikoviRauzer-Chernousova, 1948d (showing undulose and concavus sutures; Figs. 11.711.8), arose in older levels (Figs. 36), traditionally considered to be derived from A. koktjubensis (Figs. 11.311.4). Archaediscus koktjubensis shows some inner whorls in a well-defined sigmoidal coiling, whereas in A. krestovnikovi, this sigmoidal coiling is only incipient with a high degree of axial rotation (Brenckle & Grelecky, 1993, plate 1, figs. 11, 18). In addition, we only observed undulose or concavus sutures in A. krestovnkovi. From our point of view, both species arose from Archaediscus pusillusRauzer-Chernousova, 1948c (Figs. 11.111.2), which shows a more imperfect sigmoidal to oscillant inner coiling. Brenckle & Grelecky (1993) recommended consideration of A. pusillus as a synonym of A. koktjubensis because the inner whorl and proloculum is not well-observed in the holotype of A. pusillus, whereas other characters are rather similar in both species. We interpret that the difference in the coiling of the inner whorls in the studied material is sufficient to keep A. pusillus, characterised by this incomplete sigmoidal to oscillant coiling, as an independent species from the typical well-formed sigmoidal inner whorls in A. koktjubensis. Some intermediate forms occur between A. koktjubensis and A. krestovnikovi, in which the occurrence of some whorls with concavus sutures is frequent (Figs. 11.511.6), as well as a variable thickness for the microgranular layer. Since the revision of the Archaediscidae by Brenckle et al. (1987), some authors have followed the practice that those forms with a more marked microgranular layer should be included in the genus Paraarchaediscus, such as Paraarchaediscus koktjubensis (e.g, Kulagina, 2022), whereas, supposedly, the microgranular layer should not be present in Archaediscus krestovnikovi. Possibly, until the first occurrence of the eosigmoilinids in the Serpukhovian, all the Archaediscus sensu lato, possess a visible microgranular layer, which certainly, is progressively reduced, but that observations of this character varies with the occurrence of micrite infillings. The reduction of the microgranular layer during the evolution of the archaediscids in the Viséan and Serpukhovian is notable in Archaediscus krestovnikovi because the holotype of the species came from Serpukhovian rocks (Reitlinger in Brenckle & Grelecky, 1993)—a period where the microgranular layer is negligible in most species. Specimens with a well-marked microgranular layer form approximately 99% of the assemblages in the mid-Viséan, although specimens with a negligible microgranular layer are first recorded from the Cf4γ subzone. These occurrences fall to around 60% during the Asbian, less than 40% during the Brigantian, and less than 20% during the Serpukhovian. The problem has been discussed in Cózar & Somerville (2020) and Cózar et al. (2022b). Owing to the impossibility of establishing a robust differentiation based on the microgranular layer thickness, for pragmatical purposes, all the specimens are included under the genus Archaediscus.

Transitional forms to A. krestovnikovi display the inner whorls at a clear concavus stage, but the final whorls show a well-marked microgranular layer, at involutus stage, and thick fibrous layer in the area of the suture. The last two or three whorls in A. krestovnikovi tend to be planispiral, partly evolute and with prominent undulose sutures, but they do not develop typical concavus sutures (cf. Conil et al., 1980). Archaediscus krestovnikovi with all concavus sutures are only recorded in much younger beds, although it can be questioned if they should be included in this species or in another. At levels close to the first occurrence of A. krestovnikovi, the first oblique sections of Archaediscus at concavus stage are also recorded (Figs. 10.110.2), although species identification was not possible. In younger levels of the succession, there are other species of the genus, typically at concavus stage, that usually first occur, such as Archaediscus convexus (Fig. 10.5); A. aff. globosusConil & Lys, 1964; A. inflexusConil & Lys, 1964; A. krestovnikovi amplaConil & Lys, 1964 (Fig. 10.4); A. itinerariusShlykova, 1951; and A. krestovnikovi ovataConil & Lys, 1964 (Fig. 10.3). However, these taxa, as well as numerous oblique sections of Archaediscus at concavus stage, are also of small size (less than 300–400 μm). The only evolution within the genus is recorded in intermediate levels of the Park Limestone Formation, where moderate-sized species become more common [e.g., Archaediscus grandiculusShlykova, 1951 (Fig. 10.6), A. subangustusConil & Lys, 1964, and A. krestovnikovi with all the whorls at concavus stage]. This fact coincides with the occurrence of Archaediscus at concavus transitional angulatus stage. Interestingly, this event seems to be coeval with the first occurrence of Pseudoendothyra in the successions in Barker Scar (Cózar et al., 2022a), within the lower-mid Holkerian.

Endothyranopsis

The primitive forms corresponding to the subgenus Endothyranopsis (Eosinopsis), are widely represented in the Chadian and Arundian of the SCS (Hounslow et al., 2022; and unpublished data of the authors), with up to four species of the genus being recorded. Endothyranopsis (Eosinopsis) solidaHance et al., 2011 (Figs. 12.112.2) is morphologically the most similar species to Endothyranopsis (Endothyranopsis), due to its more regular coiling, although more distorted in the inner whorls. This species might be the ancestral form to the first Endothyranopsis s.s., E. aff. compressa sensu Cózar et al., 2022a, which is a species of moderate size (<600 μm), planispiral coiling, and septa and wall relatively thin (Figs. 12.412.5). Endothyranopsis aff. compressa is recorded from the ‘middle’ Cf4δ subzone in the region (Hounslow et al., 2022). There are other species of Eosinopsis (E. sp. A; Fig. 12.3) which show a marked thickening of the wall and septa and resemble the typical E. compressa, although they are rather scarce, and the marked sutures and distortions relate them (most likely) to Latiendothyranopsis. Endothyranopsis aff. compressa is not a completely new species, and it can be found in the literature, although misidentified (e.g., Conil & Naum, 1976, pl. VIII, fig. 101, as Eostaffella sp.; Bless et al., 1976, pl. 8, fig. 14, as Loeblichia sp.). The thickening of the septa clearly distinguishes this genus from primitive Eostaffella, and the more perfect planispiral coiling in the inner whorls from Eosinopsis. Most of the Endothyranopsis ex gr. compressa documented in Cózar et al. (2020) belong to E. aff. compressa.

However, between 1 and 1.5 m above the base of bed C (at Barker Scar, Fig. 3), Endothyranopsis compressa (Figs. 12.612.7) is first recorded, and also at similar levels in other studied sections (Figs. 46). This species is a classical marker for the Tulian in the Russian literature (e.g, Lipina & Reitlinger, 1971). However, in Belgium, its occurrence is high in the MFZ12 Zone, and using the old notation in the region, from the V2bγ (Conil & Naum, 1976; see Fig. 1), which might imply a horizon more than 48 m above the base of the Livian, in the Awirs Member. In higher levels of the Park Limestone Formation, the septa of E. compressa become thicker, and the tests reach larger diameters. These are considered as the first transitional forms to E. crassa (Brady, 1876), and are recorded in the upper parts of the Holkerian, but more frequently found in the early Asbian (Cózar et al., 2022a).

The phylogeny in the group suggests: Eosinopsis solida (possibly base of the Viséan) → Endothyranopsis aff. compressa (‘middle’ Cf4δ) → E. compressa (close to base of Cf5α; Fig. 12) → E. compressa trans. crassa (late Holkerian).

Omphalotis

During the Chadian and Arundian the genus Omphalotis is common in the SCS [O. frequentata (Ganelina, 1956), O. chariessa (Conil & Lys, 1964), and O. exilis (Rauzer-Chernousova, 1948b)]. Also, in the Cf4δ subzone, Omphalotis aff. minima sensu Cózar et al., 2022a (Figs. 13.313.4) has been distinguished. This species resembles the nominal species, but it shows smaller dimensions: smaller diameter, thinner wall, lower basal secondary deposits, and lower number of whorls and chambers per whorl. Apparently, it could be considered as an immature form. However, it occurs consistently through the Cf4δ subzone, from the basal levels, whereas the larger O. minima (Figs. 13.513.6) has only been recorded in much higher levels (Figs. 34). The ancestral form for O. aff. minima is interpreted to be Omphalotis frequentata (first occurring from the Chadian), which shows more distorted coiling (Fig. 13.1). Records of Omphalotis minima in the early Viséan, in some cases, seem to be a misidentification of O. chariessa or with O. aff. minima (e.g., Groessens et al., 1982, figs. 122–123, respectively), whereas in other cases, specimens look like the nominal species (e.g., Brenckle et al., 2009, pl. 7, figs. 10–11). Furthermore, it is not possible to validate the occurrence of those O. minima when it was not illustrated, such as in Brazhnikova et al. (1967), where the species is mentioned from the base of the Viséan, although this stratigraphic range does not seem to be very plausible. Brazhnikova & Vdovenko (1973, pl. 24, figs. 1–2) illustrated two specimens from the C1Vd2 (= Bobrikian), of which the second specimen is more properly O. aff. minima, whereas the first specimen is questionable, and resembles O. minima? recorded in the SCS. Some specimens in the SCS have been identified as Omphalotis minima?, although they have not been recorded in the sections spanning the mid-Viséan boundary, but in Coastguard, Grubbins Wood 2, and Low Frith sections (Figs. 13.713.12). These Omphalotis minima? are of large size and show rather developed basal secondary deposits and the typical differentiated wall of Omphalotis. This latter feature allows distinction from the large Omphalotis chariessa (compare Fig. 13.2 with Brenckle, 2005, pl. 12, fig. 3). In addition, these specimens show a slightly more distorted coiling than the nomimal species. Their stratigraphic range is also unusual because they are restricted to the top of the Cf4γ up to the base of the ‘middle’ Cf4δ subzones but are not present in most of the ‘middle’ Cf4δ and the ‘upper’ Cf4δ subzone. Thus, the role of this O. minima? in the phylogeny of the genus is unknown; it does not seem to be related to O. minima, but certainly its occurrence and misidentification could induce mistakes in the stratigraphic range of O. minima. Possibly it is a new species which did not extend into the mid-Viséan.

Pojarkovella

The Pojarkovella stock is also well represented in the SCS, although with uneven distribution. This genus is interpreted to arise from the common Eoparastaffella species in the Cf4δ subzone, possibly E. iniqua Postoyalko in Postoyalko & Garan, 1972 or E. modesta (Hance et al., 2011). The former species is common through the Cf4γ–δ in the SCS, whereas E. modesta is rarely recorded from the ‘lower’ and ‘middle’ part of the Cf4δ subzone in the Coastguard Quarry, Barker Scar, and White Scar sections (Figs. 3, 5). Eoparastaffella iniqua is rather abundant and seem to be the ancestral form of the stock, and it could have an intermediate form, E. modesta, a species with strong similarities to the genus Pojarkovella (Figs. 14.114.2). In fact, E. modesta was originally defined as a Pojarkovella species by Hance et al. (2011). The first representative, interpreted herein as being included in the genus Pojarkovella, is P. ketmenicaSimonova & Zub, 1975, a small species with some typical septa as in Pojarkovella, involute, and moderate to low pseudochomata (Figs. 14.314.4). Its descendant is P. occidentalis Vachard & Cózar in Vachard et al., 2016 (also small, but evolute in the final whorl; Figs. 14.514.6), which arose generally from similar levels in Barker Scar and White Scar sections (Figs. 3, 5) around the mid-Viséan boundary. However, rarely, P. ketmenica is recorded in older rocks, from the event A2 in Blackstone Point and Grubbins Wood sections (Figs. 4, 6). Its earlier occurrence in some sections cannot be explained by facies control, because similar packstone facies are also common in the studied section for coeval intervals. In the case of the Barker Scar section, it is argued that their low diversity is due to the stronger dolomitization of those beds. However, its occurrence in some sections and not in others seem to be a result of random sampling and low abundances. The following species in the same lineage seem to be P. pura Simonova in Simonova & Zub, 1975 (moderate to large size and involute; Figs. 14.714.8), and P. nibelis (moderate size and evolute; Figs. 14.914.12). As Cózar (2002) proposed, P. honesta Simonova in Simonova & Zub, 1975, seems to be a junior synonym of P. nibelis. Both species present similar parameters, and only subtle differences, in that the axial section in P. honesta is slightly wider, a feature that does not seem to justify its consideration as an independent valid species because both axial sections of P. honesta illustrated in the type material by Simonova & Zub (1975) are slightly oblique sections. Simonova & Zub (1975) distinguished P. honesta from P. nibelis also by their size, and consequently, lesser number of chambers and whorls, with P. honesta restricted to specimens <600 µm and P. nibelis >600 µm in diameter. The larger size confers to the latter species a more evolute appearance, and thus, narrower axial section and more pronounced umbilici. However, this diagnosis of P. nibelis does not correspond to the original definition of Durkina (1959), which is a reinterpretation by Simonova & Zub (1975). Possibly, the most logical step, should be to rename the taxon described by the latter authors as ‘P. nibelis’, as a new species. However, to keep nomenclatural stability, they will be referred herein as morphotype 1 (<600 µm; Figs. 14.914.10) and morphotype 2 (>600 µm; Figs. 14.1114.12). Unfortunately, this inconsistent use of the species is commonly observed in several biostratigraphical studies in Britain and Belgium, such as: Conil & Lys (1967, pl. 4, figs. 37–39) who illustrated three Quasiendothyra nibelis that correspond to the morphotype 1; Conil & Lys (1968, pl. VIII, fig. 121) who illustrated a specimen of morphotype 2; Austin et al. (1973, pl. II, fig. 27) who illustrated a specimen that corresponds to the morphotype 2; Kimpe et al. (1978) pl. 10, figs. 57, 59 belong to morphotype 1, whereas pl. 10 fig. 58 fits with the diagnosis of morphotype 2; Conil et al. (1981) pl. III, figs. 43–44 are morphotype 2, and pl. III, fig. 45 to morphotype 1; Strank (1981) illustrated many specimens from different ages and sections, but taking into consideration only the specimens illustrated from the Barker Scar stratotype, pl. III, fig. 6 as Nibelia aff. nibelis and pl. III, fig. 9 as N. nibelis (Nibelia = Pojarkovella), and both belong to morphotype 1, whereas pl. VII, fig. 18, also as N. nibelis corresponds to morphotype 2. The latter author also illustrated specimens from other sections, where, that in pl. V, figs. 5, 13 and pl. VI, fig. 21 correspond also to morphotype 2. This inconsistent use of the species questions their use as primary markers, until the taxonomy of the genus is clarified.

A few metres above the occurrence of P. nibelis morph. 2, wider, larger and more evolute species are recorded (e.g., P. mutabilisSimonova & Zub, 1975, P. erigentisSimonova & Zub, 1975, P. ovoides Simonova in Simonova & Zub, 1975, and P. eostaffelloides Simonova in Simonova & Zub, 1975), but they are rare, and their occurrence and distribution do not follow a consistent trend (Figs. 35). Thus, a complete lineage can be proposed: Eoparastaffella modesta (‘lower’ Cf4δ) → Pojarkovella ketmenica (base Cf5α, more frequently base of Cf5β) → P. occidentalis (base of Cf5β) → P. nibelis morph. 1 (above base of Cf5β) → P. nibelis morph. 2 (mostly base of A4) → P. mutabilis (advanced A4; Fig. 14), whereas the other species mentioned above seem to arise directly from P. ketmenica and P. occidentalis.

The occurrence of primitive forms before P. nibelis, in strata assigned originally to the Cf4δ subzone, is known in Western Europe from Conil (1980), strata that were later assigned to the Cf5α subzone by Cózar et al. (2020). Similarly, in China, primitive species have been assigned to the MFZ11B subzone (Hance et al., 2011; Liu et al., 2015), and Pojarkovella sp. is recorded about 120 m below P. nibelis in the Tarim Basin (Brenckle, 2004). Their occurrence in Iran from the lower Viséan is not confirmed, because Pojarkovella wushiensis? (Li, 1991 emend. Brenckle, 2004) documented by Brenckle et al. (2009, pl. 7, fig. 9) is considered herein as Endothyranopsis (Eosinopsis) solida, whereas Zandkarimi et al. (2016) documented P. wushiensis together with P. nibelis. However, primitive Pojarkovella sp. are first recorded in Iran a few metres below P. nibelis, although misidentified within P. cf. pennarroyensisCózar, 2002 (see Zandkarimi et al., 2019). In some cases, the first occurrence of Pojarkovella s.l. was considered the base of the Holkerian or the MFZ12 zone (e.g., Waters et al., 2017; Zandkarimi et al., 2019).

Lituotubella

Another genus whose species were studied in relation to mid-Viséan divisions is Lituotubella. The genus is derived from Pseudolituotubella, which is common in parts of the Tournaisian and early Viséan, possibly arising from P. tenuissima (Vdovenko, 1954; Fig. 15.1). In SCS, the first species of Lituotubella is L. glomospiroides (Fig. 15.2), which is recorded from the base of the Cf4δ subzone (e.g., Blackstone Point section; Hounslow et al., 2022), and extends into the late Viséan. However, at levels with Koskinotextularia cribriformis and Pojarkovella nibelis morph. 2, Lituotubella magnaRauzer-Chernousova 1948a is first recorded in Barker Scar and White Scar (Figs. 3, 5), whereas in Grubbins Wood the species first occurs in slightly older levels (Fig. 15.4). The only unusual record is observed in Grubbins Wood, where there is a partial section of the coiled part close to event A2 (Fig. 4), which according to its dimensions (Fig. 15.3), should correspond to L. magna, and thus its stratigraphic range could be extended down to the Cf5α subzone (although exceptionally). The first occurrence of the genus was erroneously placed at the base of the middle Viséan Livian or Holkerian in Western Europe (e.g., Conil et al., 1980, 1991; Poty et al., 2006), although it is common in the Cf4δ subzone in the SCS (Hounslow et al., 2022). There is little information about the genus below the Tulian Substage in Russia, although in Ukraine, the genus occurs from the upper part of Horizon XIII (e.g., Brazhnikova et al., 1967), which is coeval with the Tulian, and Lituotubella glomospiroides from the C1Vc (Brazhnikova & Vdovenko, 1973), which would be part of the Glubokinsky (Fig. 1). Vdovenko (1970) described three new species of Lituotubella, one in open nomeclature from the Dukuchaevsky (Tournaisian) and L. eoglomospiroides and L. curta lata from the Glubokinsky and Sukhonsky (Fig. 1), but they are immature specimens and more closely related to Pseudolituotubella. Nevertheless, Vdovenko (2001) and Davydov et al. (2010) highlighted that the species first occurring at the base of the Tulian is Lituotubella magna.

Endostaffella

Other auxilliary (secondary) markers, although of low reliability, are the Endostaffella species. In Grubbins Wood, E. fucoidesRozovskaya, 1963 (Fig. 16.1), occurs in the mid Cf5α subzone, whereas E. delicataRozovskaya, 1963 (Figs. 16.216.3) arose at the base of the Cf5β, although both species are not recorded in White Scar Quarry and only recorded from higher in the Cf5β subzone in Barker Scar (Cózar et al., 2022a). Brenckle (2005) considered E. fucoides as synonymous with E. parva (von Möller, 1879). Herein, E. parva is preserved for specimens with more skew-coiled inner whorls and usually a more prominent juvenarium compared to the final planispiral whorls. It is certainly a subtle difference that could be considered as insufficient to keep both species as valid. However, forms attributed to E. parva in the SCS have been only recorded in much younger levels, in the upper Holkerian (Cózar et al., 2022a), which suggest a higher degree in the evolution of the genus. This larger contrast is observed in the old zonation from Ukraine, where Endostaffella parva was considered a marker for the Serpukhovian (e.g., Poletaev et al., 1991), whereas Endostaffella first occurs from the C1Ve2 (Styl’sky Substage; Fig. 1; Vdovenko, 2001). The ancestral form of the genus is not clear, and it could be a species of Medioendothyra, Mediocris, or another similar genus. Rauzer-Chernousova (1948c) considered E. parva as part of the family Eostaffellidae, similar to other more recent authors (van Ginkel, 2010), Loeblichiidae for Conil et al. (1980), and Endostaffellidae for Rozovskaya (1963) and Rauzer-Chernousova et al. (1996). However, there is a notable difference in the definition of the genus. The above-mentioned authors considered the possession of weak secondary deposits (i.e., pseudochomata) whereas for Brenckle (2005) and Cózar et al. (2016) the genus Endostaffella s.s. does not contain secondary deposits, and so some of the species containing pseudochomata have been transferred to the genera Praeplectostaffella, Praeostaffellina, and Endostaffellopsis (Cózar et al., 2008, 2016).

Only a few authors have considered this genus as a possible marker for the mid-Viséan (e.g., Conil et al., 1980; Vdovenko, 2001), a fact that our study seems to confirm. However, there is no robust biostratigraphic analysis of the genus for it to be considered as a possible guide, since it is rarely documented from the lower part of the early Viséan (e.g., Groessens et al., 1982). Van Ginkel (2010) also considered some primitive forms recorded from the lower Viséan as included in the genus, such as Endostaffella barzassiensis (Lebedeva, 1954) and E. zakharovi (Bogush & Juferev, 1960), although Brenckle (2005) questioned the inclusion of these species in the genus Endostaffella and discussed their possible inclusion in Medioendothyra or Planoendothyra. Herein, those species are not included in Endostaffella because they do not contain the typical numbers of chambers, the dense coiling, or the possible presence of lateral secondary deposits. Specimens of Endostaffella in the studied sections from the SCS are rather scarce (Figs. 36).

Holkeria

Another auxilliary marker considered for the Holkerian, particularly in Britain, is Holkeria (e.g., Conil et al., 1980; Fewtrell et al., 1981; Strank, 1982; Riley, 1993). Three species of Holkeria were described by Strank (1982), recorded high in the stratotype section at Barker Scar, which are only present from bed K, the horizon that traditionally marks the base of the Holkerian (Cózar et al., 2022a). It is noteworthy that Holkeria is absent from older levels in Barker Scar (Fig. 3), although the strong dolomitization from bed F upwards, as well as the thick clastic-rich intervals, both hamper knowledge of the assemblages from this interval (c. 7 m thick interval, from bed F to bed K; Fig. 3). However, the most primitive species, Holkeria topleyensisStrank, 1982 (Fig. 17.2), is recorded at the same level as Pojarkovella nibelis morph. 2 in Grubbins Wood and White Scar (Figs. 4, 5), as well as Holkeria aff. avonesis (Conil & Longerstaey in Conil et al., 1980; Fig. 17.3). Within the Holkeria species, owing to their more primitive structures and the degree of evolution of the final whorl, the possible lineage is (Fig. 17): H. topleyensisH. aff. avonensisH. avonensisH. daggeriStrank, 1982. Strank (1982) described H. daggeri as having “sharply raised final chambers,” but in the same assemblages, we have recorded two to three final chambers trending to uncoiling, but also others, which demonstrate that the species developed an uncoiled part of the test (Fig. 17.3). While the ancestor of the genus is unknown, Strank (1982) proposed a Tournayellidae, due to the occurrence of tournayellid chambers, whereas Conil et al. (1980) suggested it could be a Rhodesinella (= Cribrospira; see Figs. 18.318.4). Interestingly, although a questionable specimen of Cribrospira has been recorded in SCS above event A4, the shape of the septa and chambers do not seem to suggest a link with Holkeria. In addition, usually, the primitive Cribrospira are also first recorded in the strata equivalent to the mid-Viséan (e.g., Lipina & Reitlinger, 1971; Vdovenko, 2001). Although the tournayellid character of the chambers is observed in some specimens, the shape of the test and wall suggest that the genus might arise also from small species of Plectogyranopsis, such as P. foeda (Conil & Lys, 1964). This latter species shows a more similar distortion of the coiling, a trend to a final uncoiled whorl, irregular septa, and slightly coarse microgranular wall (Fig. 17.1), which resemble primitive characters observed in H. topleyensis.

The most evolved species of the genus Holkeria (H. avonensis and H. daggeri; Figs. 17.417.5), in contrast, seem to be the only markers for the “classical Holkerian,” as they are recorded from beds K and L in Barker Scar (Ramsbottom, 1981; Cózar et al., 2022a). Nevertheless, the genus is not particularly common in any section assigned to the Holkerian in Britain, being absent in most sections. This fact prevents use of those species as potential markers for the recognition of the Holkerian in a traditional sense, nor are they useful for worldwide correlations, because the genus is not recorded in the ∼600 km distant Belgian successions.

Vissarionovella

A classical marker in the British literature for the Holkerian is Dainella(?) holkeriana Conil & Longerstaey in Conil et al., 1980. The species was questioned by Cózar & Vachard (2001) as a “non-typical Dainella, Lysella nor Bessiella,” and as suggested by those authors, the occurrence of a marked tectum is closer to the genus Vissarionovella. Worldwide, the genus Vissarionovella has its first occurrence from the mid-Viséan (Cózar & Vachard, 2001), and becomes common in the late Viséan (e.g., Hance et al., 2011; Kulagina, 2022). Specimens recorded in the SCS show this differentiated tectum, and thus are assigned to Vissarionovella (Figs. 18.518.6). No other species of the genus has been recorded in the studied successions. However, V. tujmasensis (Vissarionova, 1948) is commonly described from France, China, Russia, and Britain in the mid-Viséan (e.g, Vachard, 1977; Austin et al., 1973 as Dainella sp.; see Hance et al., 2011, and references therein). Nevertheless, there are also many questionable specimens attributed to Bessiella and Dainella from the so-called “V2a” (= Cf4δ sensu Conil et al., 1980; Fig. 1) from Belgium which need to be revised because the characteristic differentiation of the wall is observed (e.g., Conil & Lys, 1964, figs. 564, 778, 781; Bless et al. 1976, pl. 9, fig. 23). On the other hand, Hance et al. (2011) proposed that true Vissarionovella only occurs in the late Viséan. Admittedly, the genus is much more abundant in younger levels (e.g., Kulagina, 2022), but specimens of the genus have also been recorded from the Cf4γ–δ in France (Vachard et al., 2018) and even mentioned from the latest Tournaisian in China (Devuyst et al., 2003).

Owing to the more marked skew-coiled whorls and irregular coiling, the species seems to be derived from Bessiella legrandi Conil & Hance in Groessens et al., 1982 (common in older rocks). Vissarionovella holkeriana is recorded from the base of event A4, but is absent in Barker Scar from those levels, and only recorded from much younger levels in bed N (Cózar et al., 2022a). With the questioned systematics of the genus and its poorly known stratigraphic range, consideration of it as a reliable marker is unwise. Thus, neither the genus nor species can be considered as primary markers for the recognition of the mid-Viséan, but only as auxilliary markers, and their precise stratigraphic ranges need to be further investigated.

Ugurus

This genus is usually interpreted to be derived from Omphalotis or considered as a subgenus of Omphalotis (Vachard & Le Coze, 2022) as they display very similar wall structures. Ugurus intermediusVachard et al., 2018 (Fig. 18.1), is recorded very close to the base of events A3 and A4 (Figs. 3, 5). Its stratigraphic range in southern France was restricted to middle Viséan rocks, which seems to coincide with its stratigraphic range in Britain. This species seems to be one of the most primitive, due to the poor thickening/division of the septa, compared to U. uchtovensis (Durkina, 1959) and U. mirificus (Rauzer-Chernousova, 1948b), which could suggest that U. intermedius arose directly from Omphalotis. However, a specimen of U. mirificus has been recorded from the “lower” Cf4δ which discards this possible phylogeny (Fig. 18.2). The occurrence of those species, and the precision of their stratigraphic ranges, allows us to discard the phylogeny proposed by Vachard & Le Coze (2022, fig. 4) for the bradyinoid lineages, because Ugurus is the oldest genus, whereas Holkeria occurs later. Furthermore, Bibradya, is situated by those authors in an intermediate position, yet is only recorded from the late Asbian (see Cózar & Somerville, 2020; Cózar et al., 2022b).

Cribrospira

Other classical markers previously proposed for the mid-Viséan division are the primitive Cribrospira (Lipina & Reitlinger, 1971; Vdovenko, 2001), although the concept of the genus has changed slightly, since previously, it included species attributed to Rhodesinella and to Holkeria. In the SCS, rare specimens of Cribrospira? pansaConil & Lys, 1965 (Fig. 18.3) are recorded in White Scar Quarry (Fig. 5), close to the mid-Viséan boundary, as well as in much younger levels (Cózar et al., 2022a). It is noteworthy that the species was described only from the V2b subzone by Conil & Lys (1964), although in Conil et al. (1980), the genus (Rhodesina preoccupied = Rhodesinella) was considered as first occurring in the uppermost early Viséan. In the SCS, clear specimens have not been recorded in older levels, only a questionable specimen (identified as Cribrospira? sp.; Fig. 18.4), which resembles the Rhodesina sp. illustrated from the late Viséan by Conil et al. (1980, pl. 30, fig. 15). Taking into consideration the scarcity of specimens, it is difficult to validate if the primitive Cribrospira can be used as a solid marker.

The ‘upper’ Cf4δ (A1 to A2 interval) is not a robust subdivision as is desirable, and outside the SCS, it could be a valid subdivision for the rest of Britain, although the rarity of the taxa to define the base of the A1 event make this difficult. Owing to these problems in the SCS, its potential for an international correlation is not too promising; it needs further investigation since not enough attention has been paid to the evolution of nodes in the genus Nodosarchaediscus, and for some authors (Brenckle & Grelecky, 1993) they are interpreted as cements. The preservation of the nodes and wall under cathodoluminescence is variable, a fact with no clear explanation, and some hypotheses have been suggested: reduction of porosity in the nodes, different mineralogical phases between nodes and wall, absence of activators and inhibitors in the nodes, and removilization and homogenization of the activators in the nodes (Laloux, 1987). However, to consider the nodes as simply cement contrasts with worldwide biostratigraphic use of the genera with nodes. In practice, it would be nearly impossible to explain the same evolution of the genus based on the progressive occlusion of the lumen, as has been clearly observed in coeval strata of Belgium and Britain.

The Cf5α subzone (A2 to A3 interval) needs some reconsideration, because, in contrast to the earlier study (Cózar et al., 2020), two of the key markers proposed, have now been demonstrated to occur at older levels in the SCS. The first occurrence of Endothyranopsis s.s., is in the ‘mid’ Cf4δ, and Archaediscus krestovnikovi has been recorded from the ‘upper’ Cf4δ. This latter taxon is frequently used in Russian sections as a marker for the Tulian (Stepanov & Donakova, 1982; Vdovenko et al., 1990; Makhlina et al., 1993; Makhlina, 1996), although an older occurrence was proposed by Vdovenko et al. (1990) and Reitlinger et al. (1996) as a species from the Bobrikian Substage. Similarly, Vdovenko et al. (1990) questioned the possible occurrence of Endothyranopsis compressa in the Bobrikian. On the other hand, some rare species of Archaediscus at concavus stage (e.g., A. pauxillus) from the Bobrikian have been questioned (Brenckle & Milkina, 2003), and the Pojarkovella species have not been used for biostratigraphy in the Russian or Ukrainian zonal schemes, where they usually occur in the upper part of the Tulian or equivalents (e.g., Kulagina, 2022). The basal Tulian unit in the Moscow Basin (the C1tl1) is an alluvial lithostratigraphical unit devoid of foraminifers (e.g., Makhlina et al., 1993), so possible ranges into the lowest Tulian are unknown. A similar problem exists in the SW Urals, where the basal Tulian unconformably overlies Tournaisian rocks, so part of the Tulian might be missing (Kulagina, 2022; Fig. 1).

Two zonal schemes are used for the Moscow Basin (the Endothyranopsis compressaArchaediscus krestovnikovi Zone, as in Stepanov & Donakova, 1982) and for the Urals (Endothyranopsis compressaParaarchaediscus koktjubensis Zone, as in Alekseev, 2008), equivalent to the older Endothyranopsis compressa Zone of Lipina & Reitlinger (1971) and Kulagina et al. (2003). However, the main markers are similar: Endothyranopsis compressa, Globoendothyra s.s. (or Globoendothyra globulus), Eostaffella mosquensisVissarionova, 1948, Parastaffella (= Pseudoendothyra herein), Lituotubella (for some authors, L. magna), primitive Cribrospira, Endostaffella, Vissariotaxis exilis (Vissarionova, 1948), Omphalotis minima (rarely O. omphalota (Rauzer-Chernoussova & Reitlinger in Rauzer-Chernousova et al., 1936) is also included), Archaediscus moelleri, rather diversified species of Archaediscus or Paraarchaediscus, and the Palaeotextulariidae (including Consobrinellopsis, Koskinotextularia, Cribrostomum and Koskinobigenerina; Lipina & Reitlinger, 1971; Vdovenko et al., 1990; Makhlina et al., 1993; Vdovenko, 2001; Brenckle & Milkina, 2003; Kulagina, 2022). The Cf5α subzone possibly corresponds to the interval lacking foraminifers at the base of the Tulian (C1tl1), and so the Cf5β should correspond to the Tulian part with foraminifers, the C1tl2 and C1tl3 (Fig. 1). This conclusion is based on the occurrence of Endothyranopsis compressa, Archaediscus moelleri, A. pauxillus (and rather diversified Archaediscus from those basal levels), Koskinotextularia, Omphalotis minima, the most common occurrence of Lituotubella magna, and rare Endostaffella.

On the other hand, in the succession in the SW Urals, the substages defined in this region are correlated with those of the EEP, where the base of the succession was correlated with the Cf5α (cf. Kulagina, 2022, fig. 3). However, we interpret that the base of the Sikasya River section (Tulian interval = Kurtymsky of Sinitsyna et al., 1984 for this region) should be assigned to the Cf5β subzone due to the occurrence of well-formed Endothyranopsis compressa from the basal samples, and thus it is difficult to establish precisely if this latter subzone is complete or if there is a hiatus at its base.

In the Belgian succession, Cózar et al. (2020) and Pointon et al. (2021) inferred that the basal mid-Viséan hiatus might represent only the upper part of the Cf4δ subzone, whereas the Banc d’Or de Bachant and the lower 14 m of the Haut-le-Wastia Member of the Livian stratotype (barren in foraminifers), might represent the Cf5α subzone or MFZ12α. Hence, the Cf5β or MFZ12β in Belgium might be complete or almost complete (Fig. 1), although this is a hypothesis that is difficult to validate. Unfortunately, bentonites studied in White Scar Quarry only contain older detrital zircons (Hounslow et al., 2022) and are not useful for comparison with those studied by Pointon et al. (2021). The foraminifers have been never illustrated in the Livian stratotype, which also complicates any comparison. In addition, due to the problematic taxonomy in Pojarkovella, and the varied primitive Koskinotextularia species from northern Belgium, it cannot be confirmed that the identification of Pojarkovella species (occurring from 14.3 m above the base of the Haut-le-Wastia Member) nor the Koskinotextularia species (its first occurrence is not precisely positioned in the section) was correct. According to Conil et al. (1977b), Koskinotextularia first occurs from the so-called V2bα (Fig. 1), but it was not illustrated. The only specimens illustrated from lower levels are those from the Brabant Massif, which correspond to primitive Koskinotextularia and Pojarkovella nibelis morphotype 1, from reworked breccias (e.g., Kimpe et al., 1978). It would be desirable for an updated foraminiferal succession from the Namur-Dinant Basin to be achieved, in order to establish precise correlations, and to confirm which subzones are complete in the Livian, or alternatively, to find other sections in Britain yielding productive bentonites allowing convergence of zonal and chronometric scales between these areas.

A clearer chronostratigraphic correlation between the regional substages is at the base of the Cf5α subzone, whereas a cruder correlation of isochronous horizons by means of foraminiferal first appearances can be established at the base of the Cf5β subzone (Fig. 1). This latter possibility suggests that the current formal base of the Holkerian, Livian, and Tulian should be modified to be 14.3 m above the current base of the Livian and at the base of the C1tl2 unit for the Tulian. The case for the currently defined Holkerian (base of bed K, Fig. 3) is weak, because its base is an arbitary lithological change which cannot be precisely correlated regionally (Cózar et al., 2022a, 2023; Hounslow et al., 2022). If the base of the Cf5β is used to define the base of the Holkerian, it is necessary to move the boundary down 14 m, from bed K to the base of bed C at Barker Scar, representing event A3 (Figs. 1, 3).

The base of Cf5β is a distinctive boundary that is useful for overall international correlation, but due to the uncertainties at the base of the Tulian (in the EEP and Urals) and Livian (in Belgium), it cannot be formally confirmed that this boundary corresponds to a biostratigraphic isochronous horizon. This is particularly important for sections in Belgium, when elsewhere the basal few metres of the middle Viséan (here understood as the Cf5β) usually contain a mixture of middle Viséan and early Viséan foraminifers. This interval with typically early Viséan fauna still preserved during the middle Viséan is observed in basins from England, Ukraine, Russia, Kazakhstan, and China (e.g., Austin et al., 1973; Poletaev et al., 1991; Reitlinger et al., 1996; Brenckle & Milkina, 2003; Devuyst et al., 2003, respectively). This opens the question of the time duration of the hostile stromatolites and mudstones at the base of the Livian in Belgium. To solve this problem from a palaeontological perspective, it is necessary to recognise other horizons of correlation, such as event A4, as seen in the SCS, and in so doing to address which parts are represented. This approach provides a better chance of finding isochronous events that are recognised inter-regionally at the mid-Viséan boundary.

The analysis of sections yielding the uppermost Arundian strata in the South Cumbria Shelf allow the precise study of classical mid-Viséan foraminiferal markers and the recognition of four main occurrence events.

First occurrences of species of Koskinotextularia, Archaediscus, Endothyranopsis, Omphalotis, and Pojarkovella provide robust primary markers. In contrast, species of Lituotubella, Endostaffella, Vissarionovella, Ugurus, Cribrospira, and Holkeria are too scarce, and so their first occurrences are insufficiently robust and thus should be considered as auxilliary markers.

Event A1 is characterised by the occurrence of occluded Nodosarchaediscus, Consobrinellopsis, and the first Archaediscus krestovnikovi. This event is recognised as the base of the ‘upper’ Cf4δ subzone. However, the rarity of specimens makes this event not as consistent as would be desirable, being located within an interval of 2 m above and below the mid-clastic unit in the Raven’s Member in the SCS sections.

Event A2 is characterised by the first occurrence of Pojarkovella ketmenica, Archaediscus at concavus stage (including the first A. pauxillus), and exceptionally, Lituotubella magna. The event is assigned to the base of the Cf5α subzone.

Event A3 is characterised by Koskinotextularia aff. Cribriformis, K. sp. A, Archaediscus moelleri, Pojarkovella occidentalis, Endothyranopsis compressa, Omphalotis minima, and common A. pauxillus. The event is also recognised with secondary markers using the first occurrences of Pojarkovella pura, P. nibelis morphotype 1, common P. ketmenica, Endostaffella fucoides, usually Lituotubella magna, common Archaediscus at concavus stage, and rare Ugurus intermedius. This event is recognised as the base of the Cf5β subzone.

Event A4 is characterised by the first occurrence of Koskinotextularia cribriformis, K. bradyi, Pojarkovella nibelis morphotype 2 (exceptionally, from the top of previous subzone), with auxilliary markers provided by Koskinotextularia obliqua, Holkeria, Vissarionovella, Cribrospira? Pansa, and Endostaffella delicata.

The base of the Cf5α subzone might coincide with the base of the Livian in Belgium and Tulian in the Russian Platform, although the occurrence of hiatuses and hostile facies barren in foraminifers do not allow confirmation of this. The Cf5β is correlated with the part of the Livian and Russian Platform with preserved foraminifers (14.3 m above the base of the Haut-le-Bastia Member and base of C1tl2 unit, respectively) and can be used for worldwide correlations. Formally, it cannot be confirmed that this horizon is an isochronous level (due to hiatuses and hostile facies), and it is necessary to search for another potential level of correlation, slightly higher, to illuminate the absence or presence of events, such as event A4 in other successions.

We would like to thank E. Kulagina and Q. Sheng for their constructive comments. MWH was part funded by NERC (grant NE/P00170X/1).